Skip to contentSkip to conversation

ethdev: introduce Rx buffer split

RFC In review Slava Ovsiienko 10 participants 172 messages Aug 17, 2020 – Oct 19, 2020 Network drivers Page 2 of 3

lore
Share & tools
lore
Acked-by on v6 1/6
#89 lore raw
In reply to Viacheslav Ovsiienko
Quoted 124 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The following structure is introduced to specify the Rx packet segment:

struct rte_eth_rxseg {
struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specifiy Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuraion, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

If the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array:

- the first network buffer will be allocated from the memory pool, specified in the first segment description element, the second network buffer - from the pool in the second segment description element and so on. If there is no enough elements to describe the buffer for entire packet of maximal length the pool from the last valid element will be used to allocate the buffers from for the rest of segments

- the offsets from the segment description elements will provide the data offset from the buffer beginning except the first mbuf - for this one the offset is added to the RTE_PKTMBUF_HEADROOM to get actual offset from the buffer beginning. If there is no enough elements to describe the buffer for entire packet of maximal length the offsets for the rest of segment will be supposed to be zero.

- the data length being received to each segment is limited by the length specified in the segment description element. The data receiving starts with filling up the first mbuf data buffer, if the specified maximal segment length is reached and there are data remaining (packet is longer than buffer in the first mbuf) the following data will be pushed to the next segment up to its own maximal length. If the first two segments is not enough to store all the packet remaining data the next (third) segment will be engaged and so on. If the length in the segment description element is zero the actual buffer size will be deduced from the appropriate memory pool properties. If there is no enough elements to describe the buffer for entire packet of maximal length the buffer size will be deduced from the pool of the last valid element for the remaining segments.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

Quoted 1 line
Quoted 322 lines
---
 doc/guides/nics/features.rst           |  15 +++++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 +++
 lib/librte_ethdev/rte_ethdev.c         | 111 +++++++++++++++++++++++++--------
 lib/librte_ethdev/rte_ethdev.h         |  62 +++++++++++++++++-
 5 files changed, 171 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +15 -0
index dd8c955..832ea3b 100644
@@ -185,6 +185,21 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[provides]   eth_dev_ops**: ``rxq_info_get:buffer_split``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index 584e720..232cd54 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index bcc0fc2..bcc2479 100644
@@ -60,6 +60,12 @@ New Features
   Added the FEC API which provides functions for query FEC capabilities and
   current FEC mode from device. Also, API for configuring FEC mode is also provided.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -253,6 +259,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +86 -25
index 892c246..96ecb91 100644
@@ -105,6 +105,9 @@ struct rte_eth_xstats_name_off {
 #define RTE_RX_OFFLOAD_BIT2STR(_name)  \
        { DEV_RX_OFFLOAD_##_name, #_name }
 
+#define RTE_ETH_RX_OFFLOAD_BIT2STR(_name)      \
+       { RTE_ETH_RX_OFFLOAD_##_name, #_name }
+
 static const struct {
        uint64_t offload;
        const char *name;
@@ -128,9 +131,11 @@ struct rte_eth_xstats_name_off {
        RTE_RX_OFFLOAD_BIT2STR(SCTP_CKSUM),
        RTE_RX_OFFLOAD_BIT2STR(OUTER_UDP_CKSUM),
        RTE_RX_OFFLOAD_BIT2STR(RSS_HASH),
+       RTE_ETH_RX_OFFLOAD_BIT2STR(BUFFER_SPLIT),
 };
 
 #undef RTE_RX_OFFLOAD_BIT2STR
+#undef RTE_ETH_RX_OFFLOAD_BIT2STR
 
 #define RTE_TX_OFFLOAD_BIT2STR(_name)  \
        { DEV_TX_OFFLOAD_##_name, #_name }
@@ -1784,38 +1789,94 @@ struct rte_eth_dev *
                return -EINVAL;
        }
 
-       if (mp == NULL) {
-               RTE_ETHDEV_LOG(ERR, "Invalid null mempool pointer\n");
-               return -EINVAL;
-       }
-
        RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_queue_setup, -ENOTSUP);
 
-       /*
-        * Check the size of the mbuf data buffer.
-        * This value must be provided in the private data of the memory pool.
-        * First check that the memory pool has a valid private data.
-        */
        ret = rte_eth_dev_info_get(port_id, &dev_info);
        if (ret != 0)
                return ret;
 
-       if (mp->private_data_size < sizeof(struct rte_pktmbuf_pool_private)) {
-               RTE_ETHDEV_LOG(ERR, "%s private_data_size %d < %d\n",
-                       mp->name, (int)mp->private_data_size,
-                       (int)sizeof(struct rte_pktmbuf_pool_private));
-               return -ENOSPC;
-       }
-       mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+       if (mp) {
+               /* Single pool configuration check. */
+               if (rx_conf->rx_seg || rx_conf->rx_nseg) {
+                       RTE_ETHDEV_LOG(ERR,
+                                      "Ambiguous segment configuration\n");
+                       return -EINVAL;
+               }
+               /*
+                * Check the size of the mbuf data buffer, this value
+                * must be provided in the private data of the memory pool.
+                * First check that the memory pool(s) has a valid private data.
+                */
+               if (mp->private_data_size <
+                               sizeof(struct rte_pktmbuf_pool_private)) {
+                       RTE_ETHDEV_LOG(ERR, "%s private_data_size %u < %u\n",
+                               mp->name, mp->private_data_size,
+                               (unsigned int)
+                               sizeof(struct rte_pktmbuf_pool_private));
+                       return -ENOSPC;
+               }
+               mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+               if (mbp_buf_size < dev_info.min_rx_bufsize +
+                                  RTE_PKTMBUF_HEADROOM) {
+                       RTE_ETHDEV_LOG(ERR,
+                                      "%s mbuf_data_room_size %u < %u"
+                                      " (RTE_PKTMBUF_HEADROOM=%u +"
+                                      " min_rx_bufsize(dev)=%u)\n",
+                                      mp->name, mbp_buf_size,
+                                      RTE_PKTMBUF_HEADROOM +
+                                      dev_info.min_rx_bufsize,
+                                      RTE_PKTMBUF_HEADROOM,
+                                      dev_info.min_rx_bufsize);
+                       return -EINVAL;
+               }
+       } else {
+               const struct rte_eth_rxseg *rx_seg = rx_conf->rx_seg;
+               uint16_t n_seg = rx_conf->rx_nseg;
+               uint16_t seg_idx;
 
-       if (mbp_buf_size < dev_info.min_rx_bufsize + RTE_PKTMBUF_HEADROOM) {
-               RTE_ETHDEV_LOG(ERR,
-                       "%s mbuf_data_room_size %d < %d (RTE_PKTMBUF_HEADROOM=%d + min_rx_bufsize(dev)=%d)\n",
-                       mp->name, (int)mbp_buf_size,
-                       (int)(RTE_PKTMBUF_HEADROOM + dev_info.min_rx_bufsize),
-                       (int)RTE_PKTMBUF_HEADROOM,
-                       (int)dev_info.min_rx_bufsize);
-               return -EINVAL;
+               /* Extended multi-segment configuration check. */
+               if (!rx_conf->rx_seg || !rx_conf->rx_nseg) {
+                       RTE_ETHDEV_LOG(ERR,
+                                      "Memory pool is null and no"
+                                      " extended configuration provided\n");
+                       return -EINVAL;
+               }
+               /*
+                * Check the sizes and offsets against buffer sizes
+                * for each segment specified in extended configuration.
+                */
+               for (seg_idx = 0; seg_idx < n_seg; seg_idx++) {
+                       struct rte_mempool *mpl = rx_seg[seg_idx].mp;
+                       uint32_t length = rx_seg[seg_idx].length;
+                       uint32_t offset = rx_seg[seg_idx].offset;
+                       uint32_t head_room = seg_idx ? 0 : RTE_PKTMBUF_HEADROOM;
+
+                       if (mpl == NULL) {
+                               RTE_ETHDEV_LOG(ERR, "null mempool pointer\n");
+                               return -EINVAL;
+                       }
+                       if (mpl->private_data_size <
+                               sizeof(struct rte_pktmbuf_pool_private)) {
+                               RTE_ETHDEV_LOG(ERR,
+                                              "%s private_data_size %u < %u\n",
+                                              mpl->name,
+                                              mpl->private_data_size,
+                                              (unsigned int)sizeof(struct
+                                              rte_pktmbuf_pool_private));
+                               return -ENOSPC;
+                       }
+                       mbp_buf_size = rte_pktmbuf_data_room_size(mpl);
+                       length = length ? length : (mbp_buf_size - head_room);
+                       if (mbp_buf_size < length + offset + head_room) {
+                               RTE_ETHDEV_LOG(ERR,
+                                       "%s mbuf_data_room_size %u < %u"
+                                       " (segment length=%u +"
+                                       " segment offset=%u)\n",
+                                       mpl->name, mbp_buf_size,
+                                       length + offset, length, offset);
+                               return -EINVAL;
+                       }
+               }
        }
 
        /* Use default specified by driver, if nb_rx_desc is zero */
lib/librte_ethdev/rte_ethdev.h +61 -2
index 5bcfbb8..e019f4a 100644
@@ -970,6 +970,16 @@ struct rte_eth_txmode {
 };
 
 /**
+ * A structure used to configure an RX packet segment to split.
+ */
+struct rte_eth_rxseg {
+       struct rte_mempool *mp; /**< Memory pool to allocate segment from. */
+       uint16_t length; /**< Segment data length, configures split point. */
+       uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
+       uint32_t reserved; /**< Reserved field. */
+};
+
+/**
  * A structure used to configure an RX ring of an Ethernet port.
  */
 struct rte_eth_rxconf {
@@ -977,6 +987,43 @@ struct rte_eth_rxconf {
        uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
        uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
        uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
+       uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
+       /**
+        * Points to the array of segment descriptions. Each array element
+        * describes the properties for each segment in the receiving
+        * buffer.
+        *
+        * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
+        * the PMD will split the received packets into multiple segments
+        * according to the specification in the description array:
+        *
+        * - the first network buffer will be allocated from the memory pool,
+        *   specified in the first array element, the second buffer, from the
+        *   pool in the second element, and so on.
+        *
+        * - the offsets from the segment description elements specify
+        *   the data offset from the buffer beginning except the first mbuf.
+        *   For this one the offset is added with RTE_PKTMBUF_HEADROOM.
+        *
+        * - the lengthes in the elements define the maximal data amount
+        *   being received to each segment. The receiving starts with filling
+        *   up the first mbuf data buffer up to specified length. If the
+        *   there are data remaining (packet is longer than buffer in the first
+        *   mbuf) the following data will be pushed to the next segment
+        *   up to its own length, and so on.
+        *
+        * - If the length in the segment description element is zero
+        *   the actual buffer size will be deduced from the appropriate
+        *   memory pool properties.
+        *
+        * - if there is not enough elements to describe the buffer for entire
+        *   packet of maximal length the following parameters will be used
+        *   for the all remaining segments:
+        *     - pool from the last valid element
+        *     - the buffer size from this pool
+        *     - zero offset
+        */
+       struct rte_eth_rxseg *rx_seg;
        /**
         * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
         * Only offloads set on rx_queue_offload_capa or rx_offload_capa
@@ -1260,6 +1307,7 @@ struct rte_eth_conf {
 #define DEV_RX_OFFLOAD_SCTP_CKSUM      0x00020000
 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM  0x00040000
 #define DEV_RX_OFFLOAD_RSS_HASH                0x00080000
+#define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000
 
 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
                                 DEV_RX_OFFLOAD_UDP_CKSUM | \
@@ -2027,9 +2075,21 @@ int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue,
  *   No need to repeat any bit in rx_conf->offloads which has already been
  *   enabled in rte_eth_dev_configure() at port level. An offloading enabled
  *   at port level can't be disabled at queue level.
+ *   The configuration structure also contains the pointer to the array
+ *   of the receiving buffer segment descriptions, see rx_seg and rx_nseg
+ *   fields, this extended configuration might be used by split offloads like
+ *   RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mp_pool is not NULL,
+ *   the extended configuration fields must be set to NULL and zero.
  * @param mb_pool
  *   The pointer to the memory pool from which to allocate *rte_mbuf* network
- *   memory buffers to populate each descriptor of the receive ring.
+ *   memory buffers to populate each descriptor of the receive ring. There are
+ *   two options to provide Rx buffer configuration:
+ *   - single pool:
+ *     mb_pool is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is 0.
+ *   - multiple segments description:
+ *     mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0.
+ *     Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads.
+ *
  * @return
  *   - 0: Success, receive queue correctly set up.
  *   - -EIO: if device is removed.
--

1.8.3.1

#90 lore raw
In reply to Viacheslav Ovsiienko
Quoted 245 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The following structure is introduced to specify the Rx packet segment:

struct rte_eth_rxseg {
struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specifiy Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuraion, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

If the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array:

- the first network buffer will be allocated from the memory pool, specified in the first segment description element, the second network buffer - from the pool in the second segment description element and so on. If there is no enough elements to describe the buffer for entire packet of maximal length the pool from the last valid element will be used to allocate the buffers from for the rest of segments

- the offsets from the segment description elements will provide the data offset from the buffer beginning except the first mbuf - for this one the offset is added to the RTE_PKTMBUF_HEADROOM to get actual offset from the buffer beginning. If there is no enough elements to describe the buffer for entire packet of maximal length the offsets for the rest of segment will be supposed to be zero.

- the data length being received to each segment is limited by the length specified in the segment description element. The data receiving starts with filling up the first mbuf data buffer, if the specified maximal segment length is reached and there are data remaining (packet is longer than buffer in the first mbuf) the following data will be pushed to the next segment up to its own maximal length. If the first two segments is not enough to store all the packet remaining data the next (third) segment will be engaged and so on. If the length in the segment description element is zero the actual buffer size will be deduced from the appropriate memory pool properties. If there is no enough elements to describe the buffer for entire packet of maximal length the buffer size will be deduced from the pool of the last valid element for the remaining segments.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---
 doc/guides/nics/features.rst           |  15 +++++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 +++
 lib/librte_ethdev/rte_ethdev.c         | 111 +++++++++++++++++++++++++--------
 lib/librte_ethdev/rte_ethdev.h         |  62 +++++++++++++++++-
 5 files changed, 171 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +15 -0
index dd8c955..832ea3b 100644
@@ -185,6 +185,21 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[provides]   eth_dev_ops**: ``rxq_info_get:buffer_split``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index 584e720..232cd54 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index bcc0fc2..bcc2479 100644
@@ -60,6 +60,12 @@ New Features
   Added the FEC API which provides functions for query FEC capabilities and
   current FEC mode from device. Also, API for configuring FEC mode is also provided.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -253,6 +259,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +5 -10
index 892c246..96ecb91 100644
@@ -105,6 +105,9 @@ struct rte_eth_xstats_name_off {
 #define RTE_RX_OFFLOAD_BIT2STR(_name)	\
 	{ DEV_RX_OFFLOAD_##_name, #_name }
 
+#define RTE_ETH_RX_OFFLOAD_BIT2STR(_name)	\
+	{ RTE_ETH_RX_OFFLOAD_##_name, #_name }
+
 static const struct {
 	uint64_t offload;
 	const char *name;
@@ -128,9 +131,11 @@ struct rte_eth_xstats_name_off {
 	RTE_RX_OFFLOAD_BIT2STR(SCTP_CKSUM),
 	RTE_RX_OFFLOAD_BIT2STR(OUTER_UDP_CKSUM),
 	RTE_RX_OFFLOAD_BIT2STR(RSS_HASH),
+	RTE_ETH_RX_OFFLOAD_BIT2STR(BUFFER_SPLIT),
 };
 
 #undef RTE_RX_OFFLOAD_BIT2STR
+#undef RTE_ETH_RX_OFFLOAD_BIT2STR
 
 #define RTE_TX_OFFLOAD_BIT2STR(_name)	\
 	{ DEV_TX_OFFLOAD_##_name, #_name }
@@ -1784,38 +1789,94 @@ struct rte_eth_dev *
 		return -EINVAL;
 	}
 
-	if (mp == NULL) {
-		RTE_ETHDEV_LOG(ERR, "Invalid null mempool pointer\n");
-		return -EINVAL;
-	}
-
 	RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_queue_setup, -ENOTSUP);
 
-	/*
-	 * Check the size of the mbuf data buffer.
-	 * This value must be provided in the private data of the memory pool.
-	 * First check that the memory pool has a valid private data.
-	 */
 	ret = rte_eth_dev_info_get(port_id, &dev_info);
 	if (ret != 0)

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1798

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +1 -7
 		return ret;
 
-	if (mp->private_data_size < sizeof(struct rte_pktmbuf_pool_private)) {
-		RTE_ETHDEV_LOG(ERR, "%s private_data_size %d < %d\n",
-			mp->name, (int)mp->private_data_size,
-			(int)sizeof(struct rte_pktmbuf_pool_private));
-		return -ENOSPC;
-	}
-	mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+	if (mp) {

Please, compare vs NULL [1]

[1] https://doc.dpdk.org/guides/contributing/coding_style.html#null-pointers

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1800

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.c +2 -0
+		/* Single pool configuration check. */
+		if (rx_conf->rx_seg || rx_conf->rx_nseg) {

Please, compare vs NULL and 0. IMHO, rx_nsegs check is sufficient. If it is 0, nobody cares what is in rx_seg.

Quoted 14 lines
lib/librte_ethdev/rte_ethdev.c +14 -0
+			RTE_ETHDEV_LOG(ERR,
+				       "Ambiguous segment configuration\n");
+			return -EINVAL;
+		}
+		/*
+		 * Check the size of the mbuf data buffer, this value
+		 * must be provided in the private data of the memory pool.
+		 * First check that the memory pool(s) has a valid private data.
+		 */
+		if (mp->private_data_size <
+				sizeof(struct rte_pktmbuf_pool_private)) {
+			RTE_ETHDEV_LOG(ERR, "%s private_data_size %u < %u\n",
+				mp->name, mp->private_data_size,
+				(unsigned int)

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1824

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +10 -0
+				sizeof(struct rte_pktmbuf_pool_private));
+			return -ENOSPC;
+		}
+		mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+		if (mbp_buf_size < dev_info.min_rx_bufsize +
+				   RTE_PKTMBUF_HEADROOM) {
+			RTE_ETHDEV_LOG(ERR,
+				       "%s mbuf_data_room_size %u < %u"
+				       " (RTE_PKTMBUF_HEADROOM=%u +"
+				       " min_rx_bufsize(dev)=%u)\n",

Do not split format string. It is not a problem that it is long.

Quoted 12 lines
lib/librte_ethdev/rte_ethdev.c +11 -0
+				       mp->name, mbp_buf_size,
+				       RTE_PKTMBUF_HEADROOM +
+				       dev_info.min_rx_bufsize,
+				       RTE_PKTMBUF_HEADROOM,
+				       dev_info.min_rx_bufsize);
+			return -EINVAL;
+		}
+	} else {
+		const struct rte_eth_rxseg *rx_seg = rx_conf->rx_seg;
+		uint16_t n_seg = rx_conf->rx_nseg;
+		uint16_t seg_idx;
 

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1838

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +2 -8
-	if (mbp_buf_size < dev_info.min_rx_bufsize + RTE_PKTMBUF_HEADROOM) {
-		RTE_ETHDEV_LOG(ERR,
-			"%s mbuf_data_room_size %d < %d (RTE_PKTMBUF_HEADROOM=%d + min_rx_bufsize(dev)=%d)\n",
-			mp->name, (int)mbp_buf_size,
-			(int)(RTE_PKTMBUF_HEADROOM + dev_info.min_rx_bufsize),
-			(int)RTE_PKTMBUF_HEADROOM,
-			(int)dev_info.min_rx_bufsize);
-		return -EINVAL;
+		/* Extended multi-segment configuration check. */
+		if (!rx_conf->rx_seg || !rx_conf->rx_nseg) {

Please, compare vs NULL and 0

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1841

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.c +3 -0
+			RTE_ETHDEV_LOG(ERR,
+				       "Memory pool is null and no"
+				       " extended configuration provided\n");

Do not split format string. It is not a problem that it is long.

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1852

Quoted 11 lines
lib/librte_ethdev/rte_ethdev.c +11 -0
+			return -EINVAL;
+		}
+		/*
+		 * Check the sizes and offsets against buffer sizes
+		 * for each segment specified in extended configuration.
+		 */
+		for (seg_idx = 0; seg_idx < n_seg; seg_idx++) {
+			struct rte_mempool *mpl = rx_seg[seg_idx].mp;
+			uint32_t length = rx_seg[seg_idx].length;
+			uint32_t offset = rx_seg[seg_idx].offset;
+			uint32_t head_room = seg_idx ? 0 : RTE_PKTMBUF_HEADROOM;

Why? Shouldn't it be in offset? IMHO too many offsets this way.

Quoted 7 lines
lib/librte_ethdev/rte_ethdev.c +7 -0
+
+			if (mpl == NULL) {
+				RTE_ETHDEV_LOG(ERR, "null mempool pointer\n");
+				return -EINVAL;
+			}
+			if (mpl->private_data_size <
+				sizeof(struct rte_pktmbuf_pool_private)) {

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1869

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +10 -0
+				RTE_ETHDEV_LOG(ERR,
+					       "%s private_data_size %u < %u\n",
+					       mpl->name,
+					       mpl->private_data_size,
+					       (unsigned int)sizeof(struct
+					       rte_pktmbuf_pool_private));
+				return -ENOSPC;
+			}
+			mbp_buf_size = rte_pktmbuf_data_room_size(mpl);
+			length = length ? length : (mbp_buf_size - head_room);

Compare length with 0.
What does ensure that mbp_buf_size is greater or equal to head_room size?

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1874

Quoted 5 lines
lib/librte_ethdev/rte_ethdev.c +5 -0
+			if (mbp_buf_size < length + offset + head_room) {
+				RTE_ETHDEV_LOG(ERR,
+					"%s mbuf_data_room_size %u < %u"
+					" (segment length=%u +"
+					" segment offset=%u)\n",

Do not split format string. It is not a problem that it is long.

Quoted 13 lines
lib/librte_ethdev/rte_ethdev.c +5 -0
+					mpl->name, mbp_buf_size,
+					length + offset, length, offset);
+				return -EINVAL;
+			}
+		}
 	}
 
 	/* Use default specified by driver, if nb_rx_desc is zero */
lib/librte_ethdev/rte_ethdev.h +0 -0
index 5bcfbb8..e019f4a 100644

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 973

Quoted 4 lines
lib/librte_ethdev/rte_ethdev.h +1 -0
@@ -970,6 +970,16 @@ struct rte_eth_txmode {
 };
 
 /**
+ * A structure used to configure an RX packet segment to split.

RX -> Rx

Quoted 88 lines
lib/librte_ethdev/rte_ethdev.h +60 -1
+ */
+struct rte_eth_rxseg {
+	struct rte_mempool *mp; /**< Memory pool to allocate segment from. */
+	uint16_t length; /**< Segment data length, configures split point. */
+	uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
+	uint32_t reserved; /**< Reserved field. */
+};
+
+/**
  * A structure used to configure an RX ring of an Ethernet port.
  */
 struct rte_eth_rxconf {
@@ -977,6 +987,43 @@ struct rte_eth_rxconf {
 	uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
 	uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
 	uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
+	uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
+	/**
+	 * Points to the array of segment descriptions. Each array element
+	 * describes the properties for each segment in the receiving
+	 * buffer.
+	 *
+	 * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
+	 * the PMD will split the received packets into multiple segments
+	 * according to the specification in the description array:
+	 *
+	 * - the first network buffer will be allocated from the memory pool,
+	 *   specified in the first array element, the second buffer, from the
+	 *   pool in the second element, and so on.
+	 *
+	 * - the offsets from the segment description elements specify
+	 *   the data offset from the buffer beginning except the first mbuf.
+	 *   For this one the offset is added with RTE_PKTMBUF_HEADROOM.
+	 *
+	 * - the lengthes in the elements define the maximal data amount
+	 *   being received to each segment. The receiving starts with filling
+	 *   up the first mbuf data buffer up to specified length. If the
+	 *   there are data remaining (packet is longer than buffer in the first
+	 *   mbuf) the following data will be pushed to the next segment
+	 *   up to its own length, and so on.
+	 *
+	 * - If the length in the segment description element is zero
+	 *   the actual buffer size will be deduced from the appropriate
+	 *   memory pool properties.
+	 *
+	 * - if there is not enough elements to describe the buffer for entire
+	 *   packet of maximal length the following parameters will be used
+	 *   for the all remaining segments:
+	 *     - pool from the last valid element
+	 *     - the buffer size from this pool
+	 *     - zero offset
+	 */
+	struct rte_eth_rxseg *rx_seg;
 	/**
 	 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
 	 * Only offloads set on rx_queue_offload_capa or rx_offload_capa
@@ -1260,6 +1307,7 @@ struct rte_eth_conf {
 #define DEV_RX_OFFLOAD_SCTP_CKSUM	0x00020000
 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM  0x00040000
 #define DEV_RX_OFFLOAD_RSS_HASH		0x00080000
+#define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000
 
 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
 				 DEV_RX_OFFLOAD_UDP_CKSUM | \
@@ -2027,9 +2075,21 @@ int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue,
  *   No need to repeat any bit in rx_conf->offloads which has already been
  *   enabled in rte_eth_dev_configure() at port level. An offloading enabled
  *   at port level can't be disabled at queue level.
+ *   The configuration structure also contains the pointer to the array
+ *   of the receiving buffer segment descriptions, see rx_seg and rx_nseg
+ *   fields, this extended configuration might be used by split offloads like
+ *   RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mp_pool is not NULL,
+ *   the extended configuration fields must be set to NULL and zero.
  * @param mb_pool
  *   The pointer to the memory pool from which to allocate *rte_mbuf* network
- *   memory buffers to populate each descriptor of the receive ring.
+ *   memory buffers to populate each descriptor of the receive ring. There are
+ *   two options to provide Rx buffer configuration:
+ *   - single pool:
+ *     mb_pool is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is 0.
+ *   - multiple segments description:
+ *     mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0.
+ *     Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads.
+ *
  * @return
  *   - 0: Success, receive queue correctly set up.
  *   - -EIO: if device is removed.
#91 lore raw

15/10/2020 12:11, Andrew Rybchenko:

Quoted 6 lines · Viacheslav Ovsiienko

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1800

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.c +2 -0
+		/* Single pool configuration check. */
+		if (rx_conf->rx_seg || rx_conf->rx_nseg) {

Please, compare vs NULL and 0. IMHO, rx_nsegs check is sufficient. If it is 0, nobody cares what is in rx_seg.

Yes the pointer should not be a criteria.
Having more than zero items is enough to check.

[...]

Quoted 6 lines

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1824

Quoted 4 lines
lib/librte_ethdev/rte_ethdev.c +4 -0
+			RTE_ETHDEV_LOG(ERR,
+				       "%s mbuf_data_room_size %u < %u"
+				       " (RTE_PKTMBUF_HEADROOM=%u +"
+				       " min_rx_bufsize(dev)=%u)\n",

Do not split format string. It is not a problem that it is long.

The benefit of keeping format string on the same line is for "grepping" the source code. But after a format specifier, I think we can split. Who is grepping "< %u (RTE_PKTMBUF_HEADROOM" ?
I would just change the split on the second line after the %u.

#92 lore raw

PATCH v6 2/6 app/testpmd: add multiple pools per core creation

In reply to Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

---
 app/test-pmd/bpf_cmd.c                |  4 +--
 app/test-pmd/cmdline.c                |  2 +-
 app/test-pmd/config.c                 |  6 ++--
 app/test-pmd/parameters.c             | 24 +++++++++----
 app/test-pmd/testpmd.c                | 63 +++++++++++++++++++----------------
 app/test-pmd/testpmd.h                | 24 ++++++++++---
 doc/guides/testpmd_app_ug/run_app.rst |  7 ++--
 7 files changed, 83 insertions(+), 47 deletions(-)
app/test-pmd/bpf_cmd.c +2 -2
index 16e3c3b..0a1a178 100644
@@ -69,7 +69,7 @@ struct cmd_bpf_ld_result {
 
 	*flags = RTE_BPF_ETH_F_NONE;
 	arg->type = RTE_BPF_ARG_PTR;
-	arg->size = mbuf_data_size;
+	arg->size = mbuf_data_size[0];
 
 	for (i = 0; str[i] != 0; i++) {
 		v = toupper(str[i]);
@@ -78,7 +78,7 @@ struct cmd_bpf_ld_result {
 		else if (v == 'M') {
 			arg->type = RTE_BPF_ARG_PTR_MBUF;
 			arg->size = sizeof(struct rte_mbuf);
-			arg->buf_size = mbuf_data_size;
+			arg->buf_size = mbuf_data_size[0];
 		} else if (v == '-')
 			continue;
 		else
app/test-pmd/cmdline.c +1 -1
index 273fb1a..a585cf0 100644
@@ -2907,7 +2907,7 @@ struct cmd_setup_rxtx_queue {
 		if (!numa_support || socket_id == NUMA_NO_CONFIG)
 			socket_id = port->socket_id;
 
-		mp = mbuf_pool_find(socket_id);
+		mp = mbuf_pool_find(socket_id, 0);
 		if (mp == NULL) {
 			printf("Failed to setup RX queue: "
 				"No mempool allocation"
app/test-pmd/config.c +3 -3
index d4be694..5f501f6 100644
@@ -690,7 +690,7 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 	printf("\nConnect to socket: %u", port->socket_id);
 
 	if (port_numa[port_id] != NUMA_NO_CONFIG) {
-		mp = mbuf_pool_find(port_numa[port_id]);
+		mp = mbuf_pool_find(port_numa[port_id], 0);
 		if (mp)
 			printf("\nmemory allocation on the socket: %d",
 							port_numa[port_id]);
@@ -3352,9 +3352,9 @@ struct igb_ring_desc_16_bytes {
 	 */
 	tx_pkt_len = 0;
 	for (i = 0; i < nb_segs; i++) {
-		if (seg_lengths[i] > (unsigned) mbuf_data_size) {
+		if (seg_lengths[i] > mbuf_data_size[0]) {
 			printf("length[%u]=%u > mbuf_data_size=%u - give up\n",
-			       i, seg_lengths[i], (unsigned) mbuf_data_size);
+			       i, seg_lengths[i], mbuf_data_size[0]);
 			return;
 		}
 		tx_pkt_len = (uint16_t)(tx_pkt_len + seg_lengths[i]);
app/test-pmd/parameters.c +18 -6
index 15ce8c1..4db4987 100644
@@ -106,7 +106,9 @@
 	       "(flag: 1 for RX; 2 for TX; 3 for RX and TX).\n");
 	printf("  --socket-num=N: set socket from which all memory is allocated "
 	       "in NUMA mode.\n");
-	printf("  --mbuf-size=N: set the data size of mbuf to N bytes.\n");
+	printf("  --mbuf-size=N,[N1[,..Nn]: set the data size of mbuf to "
+	       "N bytes. If multiple numbers are specified the extra pools "
+	       "will be created to receive with packet split features\n");
 	printf("  --total-num-mbufs=N: set the number of mbufs to be allocated "
 	       "in mbuf pools.\n");
 	printf("  --max-pkt-len=N: set the maximum size of packet to N bytes.\n");
@@ -892,12 +894,22 @@
 				}
 			}
 			if (!strcmp(lgopts[opt_idx].name, "mbuf-size")) {
-				n = atoi(optarg);
-				if (n > 0 && n <= 0xFFFF)
-					mbuf_data_size = (uint16_t) n;
-				else
+				unsigned int mb_sz[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs, i;
+
+				nb_segs = parse_item_list(optarg, "mbuf-size",
+					MAX_SEGS_BUFFER_SPLIT, mb_sz, 0);
+				if (nb_segs <= 0)
 					rte_exit(EXIT_FAILURE,
-						 "mbuf-size should be > 0 and < 65536\n");
+						 "bad mbuf-size\n");
+				for (i = 0; i < nb_segs; i++) {
+					if (mb_sz[i] <= 0 || mb_sz[i] > 0xFFFF)
+						rte_exit(EXIT_FAILURE,
+							 "mbuf-size should be "
+							 "> 0 and < 65536\n");
+					mbuf_data_size[i] = (uint16_t) mb_sz[i];
+				}
+				mbuf_data_size_n = nb_segs;
 			}
 			if (!strcmp(lgopts[opt_idx].name, "total-num-mbufs")) {
 				n = atoi(optarg);
app/test-pmd/testpmd.c +35 -28

145 lines not on this page · raw message

app/test-pmd/testpmd.h +19 -5
index 9a29d7a..b42d710 100644
@@ -42,6 +42,13 @@
  */
 #define RTE_MAX_SEGS_PER_PKT 255 /**< nb_segs is a 8-bit unsigned char. */
 
+/*
+ * The maximum number of segments per packet is used to configure
+ * buffer split feature, also specifies the maximum amount of
+ * optional Rx pools to allocate mbufs to split.
+ */
+#define MAX_SEGS_BUFFER_SPLIT 8 /**< nb_segs is a 8-bit unsigned char. */
+
 #define MAX_PKT_BURST 512
 #define DEF_PKT_BURST 32
 
@@ -393,7 +400,9 @@ struct queue_stats_mappings {
 extern uint8_t dcb_config;
 extern uint8_t dcb_test;
 
-extern uint16_t mbuf_data_size; /**< Mbuf data space size. */
+extern uint32_t mbuf_data_size_n;
+extern uint16_t mbuf_data_size[MAX_SEGS_BUFFER_SPLIT];
+/**< Mbuf data space size. */
 extern uint32_t param_total_num_mbufs;
 
 extern uint16_t stats_period;
@@ -605,17 +614,22 @@ struct mplsoudp_decap_conf {
 
 /* Mbuf Pools */
 static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+		    int name_size, unsigned int idx)
 {
-	snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	if (!idx)
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	else
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+			 sock_id, idx);
 }
 
 static inline struct rte_mempool *
-mbuf_pool_find(unsigned int sock_id)
+mbuf_pool_find(unsigned int sock_id, unsigned int idx)
 {
 	char pool_name[RTE_MEMPOOL_NAMESIZE];
 
-	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name));
+	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name), idx);
 	return rte_mempool_lookup((const char *)pool_name);
 }
 
doc/guides/testpmd_app_ug/run_app.rst +5 -2
index ec085c2..1eb0a10 100644
@@ -107,9 +107,12 @@ The command line options are:
     Set the socket from which all memory is allocated in NUMA mode,
     where 0 <= N < number of sockets on the board.
 
-*   ``--mbuf-size=N``
+*   ``--mbuf-size=N[,N1[,...Nn]``
 
-    Set the data size of the mbufs used to N bytes, where N < 65536. The default value is 2048.
+    Set the data size of the mbufs used to N bytes, where N < 65536.
+    The default value is 2048. If multiple mbuf-size values are specified the
+    extra memory pools will be created for allocating mbufs to receive packets
+    with buffer splittling features.
 
 *   ``--total-num-mbufs=N``
 
Signature · 1 line
1.8.3.1
#93 lore raw

PATCH v6 3/6 app/testpmd: add buffer split offload configuration

In reply to Viacheslav Ovsiienko

This patch add support for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT providing per queue configuration for this offload.

---
 app/test-pmd/cmdline.c | 21 +++++++++++----------
 app/test-pmd/config.c  |  9 +++++++++
 2 files changed, 20 insertions(+), 10 deletions(-)
app/test-pmd/cmdline.c +11 -10
index a585cf0..fa71039 100644
@@ -883,16 +883,16 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"port config <port_id> rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"     Enable or disable a per port Rx offloading"
 			" on all Rx queues of a port\n\n"
 
 			"port (port_id) rxq (queue_id) rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"    Enable or disable a per queue Rx offloading"
 			" only on a specific Rx queue\n\n"
 
@@ -18417,7 +18417,8 @@ struct cmd_config_per_port_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc#rss_hash");
+			   "scatter#buffer_split#timestamp#security#"
+			   "keep_crc#rss_hash");
 cmdline_parse_token_string_t cmd_config_per_port_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_port_rx_offload_result,
@@ -18497,8 +18498,8 @@ struct cmd_config_per_port_rx_offload_result {
 	.help_str = "port config <port_id> rx_offload vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc|rss_hash "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc|rss_hash on|off",
 	.tokens = {
 		(void *)&cmd_config_per_port_rx_offload_result_port,
 		(void *)&cmd_config_per_port_rx_offload_result_config,
@@ -18547,7 +18548,7 @@ struct cmd_config_per_queue_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc");
+			   "scatter#buffer_split#timestamp#security#keep_crc");
 cmdline_parse_token_string_t cmd_config_per_queue_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_queue_rx_offload_result,
@@ -18603,8 +18604,8 @@ struct cmd_config_per_queue_rx_offload_result {
 		    "vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc on|off",
 	.tokens = {
 		(void *)&cmd_config_per_queue_rx_offload_result_port,
 		(void *)&cmd_config_per_queue_rx_offload_result_port_id,
app/test-pmd/config.c +9 -0
index 5f501f6..7126d91 100644
@@ -1092,6 +1092,15 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 			printf("off\n");
 	}
 
+	if (dev_info.rx_offload_capa & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+		printf("RX offload buffer split:       ");
+		if (ports[port_id].dev_conf.rxmode.offloads &
+		    RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT)
+			printf("on\n");
+		else
+			printf("off\n");
+	}
+
 	if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_VLAN_INSERT) {
 		printf("VLAN insert:                   ");
 		if (ports[port_id].dev_conf.txmode.offloads &
Signature · 1 line
1.8.3.1
#94 lore raw

PATCH v6 4/6 app/testpmd: add rxpkts commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxpkts=X[,Y]

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command:

testpmd> set rxpkts (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only). Optionally the multiple memory pools can be specified with --mbuf-size command line parameter and the mbufs to receive will be allocated sequentially from these extra memory pools.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 48 ++++++++++++++++++++++-
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  7 ++++
 app/test-pmd/testpmd.h                      | 11 +++++-
 doc/guides/testpmd_app_ug/run_app.rst       |  9 +++++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 21 +++++++++-
 7 files changed, 165 insertions(+), 7 deletions(-)
app/test-pmd/cmdline.c +58 -3
index fa71039..d8dba54 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxpkts (x[,y]*)\n"
+			"    Set the length of each segment to scatter"
+			" packets on receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set txpkts (x[,y]*)\n"
 			"    Set the length of each segment of TXONLY"
 			" and optionally CSUM packets.\n\n"
@@ -3889,6 +3895,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxpkts_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxpkts;
+	cmdline_fixed_string_t seg_lengths;
+};
+
+static void
+cmd_set_rxpkts_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxpkts_result *res;
+	unsigned int seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_lengths, "segment lengths",
+				  MAX_SEGS_BUFFER_SPLIT, seg_lengths, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_segments(seg_lengths, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxpkts_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxpkts_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 rxpkts, "rxpkts");
+cmdline_parse_token_string_t cmd_set_rxpkts_lengths =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 seg_lengths, NULL);
+
+cmdline_parse_inst_t cmd_set_rxpkts = {
+	.f = cmd_set_rxpkts_parsed,
+	.data = NULL,
+	.help_str = "set rxpkts <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxpkts_keyword,
+		(void *)&cmd_set_rxpkts_name,
+		(void *)&cmd_set_rxpkts_lengths,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF TXONLY PACKETS *** */
 
 struct cmd_set_txpkts_result {
@@ -7517,6 +7569,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxpkts"))
+		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
 		show_tx_pkt_segments();
 	else if (!strcmp(res->what, "txtimes"))
@@ -7529,12 +7583,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19807,6 +19861,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
 	(cmdline_parse_inst_t *)&cmd_set_txtimes,
app/test-pmd/config.c +46 -2

62 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index 4db4987..e4e3635 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
 	printf("  --txonly-multi-flow: generate multiple flows in txonly mode\n");
@@ -662,6 +663,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
 		{ "disable-link-check",		0, 0, 0 },
@@ -1272,6 +1274,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
+				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs;
+
+				nb_segs = parse_item_list
+						(optarg, "rxpkt segments",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_len, 0);
+				if (nb_segs > 0)
+					set_rx_pkt_segments(seg_len, nb_segs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxpkts\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "txpkts")) {
 				unsigned seg_lengths[RTE_MAX_SEGS_PER_PKT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +7 -0
index 7e6ef80..f88c1e2 100644
@@ -210,6 +210,13 @@ struct fwd_engine * fwd_engines[] = {
 uint8_t f_quit;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 uint16_t tx_pkt_length = TXONLY_DEF_PACKET_LEN; /**< TXONLY packet length. */
app/test-pmd/testpmd.h +10 -1
index b42d710..8e5ba6a 100644
@@ -420,6 +420,13 @@ struct queue_stats_mappings {
 extern struct rte_fdir_conf fdir_conf;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 #define TXONLY_DEF_PACKET_LEN 64
@@ -816,7 +823,9 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_record_core_cycles(uint8_t on_off);
 void set_record_burst_stats(uint8_t on_off);
 void set_verbose_level(uint16_t vb_level);
-void set_tx_pkt_segments(unsigned *seg_lengths, unsigned nb_segs);
+void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
+void show_rx_pkt_segments(void);
+void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
 void show_tx_pkt_times(void);
doc/guides/testpmd_app_ug/run_app.rst +9 -0
index 1eb0a10..463b76c 100644
@@ -361,6 +361,15 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxpkts=X[,Y]``
+
+    Set the length of segments to scatter packets on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+    Optionally the multiple memory pools can be specified with --mbuf-size
+    command line parameter and the mbufs to receive will be allocated
+    sequentially from these extra memory pools.
+
 *   ``--txpkts=X[,Y]``
 
     Set TX segment sizes or total packet length. Valid for ``tx-only``
doc/guides/testpmd_app_ug/testpmd_funcs.rst +20 -1
index 795c739..ff88762 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxpkts``: Packets to RX split configuration.
+
 * ``txpkts``: Packets to TX configuration.
 
 * ``txtimes``: Burst time pattern for Tx only mode.
@@ -774,6 +776,23 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxpkts
+~~~~~~~~~~
+
+Set the length of segments to scatter packets on receiving if split
+feature is engaged. Affects only the queues configured with split offloads
+(currently BUFFER_SPLIT is supported only). Optionally the multiple memory
+pools can be specified with --mbuf-size command line parameter and the mbufs
+to receive will be allocated sequentially from these extra memory pools (the
+mbuf for the first segment is allocated from the first pool, the second one
+from the second pool, and so on, if segment number is greater then pool's the
+mbuf for remaining segments will be allocated from the last valid pool).
+
+   testpmd> set rxpkts (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. Zero value
+means to use the corresponding memory pool data buffer size.
+
 set txpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#95 lore raw

PATCH v6 5/6 app/testpmd: add rxoffs commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxoffs=X[,Y]

Sets the offsets of packet segments from the beginning of the receiving buffer if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command, providing the same:

testpmd> set rxoffs (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 44 +++++++++++++++++++++
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  2 +
 app/test-pmd/testpmd.h                      |  4 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  6 +++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 17 +++++++-
 7 files changed, 145 insertions(+), 4 deletions(-)
app/test-pmd/cmdline.c +58 -3
index d8dba54..7182bba 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxoffs (x[,y]*)\n"
+			"    Set the offset of each packet segment on"
+			" receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set rxpkts (x[,y]*)\n"
 			"    Set the length of each segment to scatter"
 			" packets on receiving if split feature is engaged."
@@ -3895,6 +3901,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT OFFSETS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxoffs_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxoffs;
+	cmdline_fixed_string_t seg_offsets;
+};
+
+static void
+cmd_set_rxoffs_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxoffs_result *res;
+	unsigned int seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_offsets, "segment offsets",
+				  MAX_SEGS_BUFFER_SPLIT, seg_offsets, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_offsets(seg_offsets, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxoffs_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxoffs_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 rxoffs, "rxoffs");
+cmdline_parse_token_string_t cmd_set_rxoffs_offsets =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 seg_offsets, NULL);
+
+cmdline_parse_inst_t cmd_set_rxoffs = {
+	.f = cmd_set_rxoffs_parsed,
+	.data = NULL,
+	.help_str = "set rxoffs <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxoffs_keyword,
+		(void *)&cmd_set_rxoffs_name,
+		(void *)&cmd_set_rxoffs_offsets,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
 
 struct cmd_set_rxpkts_result {
@@ -7569,6 +7621,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxoffs"))
+		show_rx_pkt_offsets();
 	else if (!strcmp(res->what, "rxpkts"))
 		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
@@ -7583,12 +7637,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxoffs#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19861,6 +19915,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxoffs,
 	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
app/test-pmd/config.c +44 -0

50 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index e4e3635..2298ba5 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxoffs=X[,Y]*: set RX segment offsets for split.\n");
 	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
@@ -663,6 +664,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxoffs",			1, 0, 0 },
 		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
@@ -1274,6 +1276,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxoffs")) {
+				unsigned int seg_off[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_offs;
+
+				nb_offs = parse_item_list
+						(optarg, "rxpkt offsets",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_off, 0);
+				if (nb_offs > 0)
+					set_rx_pkt_offsets(seg_off, nb_offs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxoffs\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
 				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +2 -0
index f88c1e2..580178d 100644
@@ -215,6 +215,8 @@ struct fwd_engine * fwd_engines[] = {
  */
 uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
app/test-pmd/testpmd.h +4 -0
index 8e5ba6a..fc56b60 100644
@@ -425,6 +425,8 @@ struct queue_stats_mappings {
  */
 extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+extern uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
@@ -825,6 +827,8 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_verbose_level(uint16_t vb_level);
 void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_rx_pkt_segments(void);
+void set_rx_pkt_offsets(unsigned int *seg_offsets, unsigned int nb_offs);
+void show_rx_pkt_offsets(void);
 void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
doc/guides/testpmd_app_ug/run_app.rst +6 -0
index 463b76c..9b0a84a 100644
@@ -361,6 +361,12 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxoffs=X[,Y]``
+
+    Set the offsets of packet segments on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+
 *   ``--rxpkts=X[,Y]``
 
     Set the length of segments to scatter packets on receiving if split
doc/guides/testpmd_app_ug/testpmd_funcs.rst +16 -1
index ff88762..c99d887 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxoffs``: Packet offsets for RX split.
+
 * ``rxpkts``: Packets to RX split configuration.
 
 * ``txpkts``: Packets to TX configuration.
@@ -776,6 +778,19 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxoffs
+~~~~~~~~~~
+
+Set the offsets of segments relating to the data buffer beginning on receiving
+if split feature is engaged. Affects only the queues configured with split
+offloads (currently BUFFER_SPLIT is supported only).
+
+   testpmd> set rxoffs (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. If the list
+of offsets is shorter than the list of segments the zero offsets will be used
+for the remaining segments.
+
 set rxpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#96 lore raw

PATCH v6 6/6 app/testpmd: add extended Rx queue setup

In reply to Viacheslav Ovsiienko

If Rx queue is configured with split feature the extended setup with specified segment sizes and pool will be performed.

---
 app/test-pmd/cmdline.c | 12 ++++++------
 app/test-pmd/testpmd.c | 48 ++++++++++++++++++++++++++++++++++++++++++++++--
 app/test-pmd/testpmd.h |  5 +++++
 3 files changed, 57 insertions(+), 8 deletions(-)
app/test-pmd/cmdline.c +6 -6
index 7182bba..204221f 100644
@@ -2927,12 +2927,12 @@ struct cmd_setup_rxtx_queue {
 				rxring_numa[res->portid]);
 			return;
 		}
-		ret = rte_eth_rx_queue_setup(res->portid,
-					     res->qid,
-					     port->nb_rx_desc[res->qid],
-					     socket_id,
-					     &port->rx_conf[res->qid],
-					     mp);
+		ret = rx_queue_setup(res->portid,
+				     res->qid,
+				     port->nb_rx_desc[res->qid],
+				     socket_id,
+				     &port->rx_conf[res->qid],
+				     mp);
 		if (ret)
 			printf("Failed to setup RX queue\n");
 	} else {
app/test-pmd/testpmd.c +46 -2

66 lines not on this page · raw message

app/test-pmd/testpmd.h +5 -0
index fc56b60..af654ea 100644
@@ -876,6 +876,11 @@ void port_rss_reta_info(portid_t port_id,
 
 void set_vf_traffic(portid_t port_id, uint8_t is_rx, uint16_t vf, uint8_t on);
 
+int
+rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
+	       uint16_t nb_rx_desc, unsigned int socket_id,
+	       struct rte_eth_rxconf *rx_conf, struct rte_mempool *mp);
+
 int set_queue_rate_limit(portid_t port_id, uint16_t queue_idx, uint16_t rate);
 int set_vf_rate_limit(portid_t port_id, uint16_t vf, uint16_t rate,
 				uint64_t q_msk);
Signature · 1 line
1.8.3.1
#97 lore raw

PATCH v2 1/1 eal/rte_malloc: add alloc_size() attribute to allocation functions

In reply to Slava Ovsiienko

By using the alloc_size() attribute the compiler can optimize better and detect errors at compile time.

For example, Gcc will fail one of the invalid allocation examples in app/test/test_malloc.c because the allocation is outside the limits of memory.

---
v2 - rebase onto correct branch (main)

 app/test/test_malloc.c              |  5 ++++-
 lib/librte_eal/include/rte_common.h | 12 ++++++++++++
 lib/librte_eal/include/rte_malloc.h | 24 ++++++++++++++++--------
 3 files changed, 32 insertions(+), 9 deletions(-)
app/test/test_malloc.c +4 -1
index 71b3cfdde5cf..fdf77b4f6a14 100644
@@ -846,6 +846,9 @@ test_malloc_bad_params(void)
 	if (bad_ptr != NULL)
 		goto err_return;
 
+#if defined(RTE_CC_GCC) || defined(RTE_CC_CLANG)
+	/* this test can not be built, will get trapped at compile time! */
+#else
 	/* rte_malloc expected to return null with size will cause overflow */
 	align = RTE_CACHE_LINE_SIZE;
 	size = (size_t)-8;
@@ -857,7 +860,7 @@ test_malloc_bad_params(void)
 	bad_ptr = rte_realloc(NULL, size, align);
 	if (bad_ptr != NULL)
 		goto err_return;
-
+#endif
 	return 0;
 
 err_return:
lib/librte_eal/include/rte_common.h +12 -0
index 2920255fc1e3..e63ef0f1de5e 100644
@@ -134,6 +134,18 @@ typedef uint16_t unaligned_uint16_t;
 	__attribute__((format(printf, format_index, first_arg)))
 #endif
 
+/**
+ * Tells compiler that the function returns a value that points to
+ * memory, where the size is given by the one or two arguments.
+ * Used by compiler to validate object size.
+ */
+#if defined(RTE_CC_GCC) || defined(RTE_CC_CLANG)
+#define __rte_alloc_size(...) \
+	__attribute__((alloc_size(__VA_ARGS__)))
+#else
+#define __rte_alloc_size(...)
+#endif
+
 #define RTE_PRIORITY_LOG 101
 #define RTE_PRIORITY_BUS 110
 #define RTE_PRIORITY_CLASS 120
lib/librte_eal/include/rte_malloc.h +16 -8
index 42ca05182f8e..3af64f87618e 100644
@@ -54,7 +54,8 @@ struct rte_malloc_socket_stats {
  *   - Otherwise, the pointer to the allocated object.
  */
 void *
-rte_malloc(const char *type, size_t size, unsigned align);
+rte_malloc(const char *type, size_t size, unsigned align)
+	__rte_alloc_size(2);
 
 /**
  * Allocate zero'ed memory from the heap.
@@ -80,7 +81,8 @@ rte_malloc(const char *type, size_t size, unsigned align);
  *   - Otherwise, the pointer to the allocated object.
  */
 void *
-rte_zmalloc(const char *type, size_t size, unsigned align);
+rte_zmalloc(const char *type, size_t size, unsigned align)
+	__rte_alloc_size(2);
 
 /**
  * Replacement function for calloc(), using huge-page memory. Memory area is
@@ -106,7 +108,8 @@ rte_zmalloc(const char *type, size_t size, unsigned align);
  *   - Otherwise, the pointer to the allocated object.
  */
 void *
-rte_calloc(const char *type, size_t num, size_t size, unsigned align);
+rte_calloc(const char *type, size_t num, size_t size, unsigned align)
+	__rte_alloc_size(2, 3);
 
 /**
  * Replacement function for realloc(), using huge-page memory. Reserved area
@@ -129,7 +132,8 @@ rte_calloc(const char *type, size_t num, size_t size, unsigned align);
  *   - Otherwise, the pointer to the reallocated memory.
  */
 void *
-rte_realloc(void *ptr, size_t size, unsigned int align);
+rte_realloc(void *ptr, size_t size, unsigned int align)
+	__rte_alloc_size(2);
 
 /**
  * Replacement function for realloc(), using huge-page memory. Reserved area
@@ -155,7 +159,8 @@ rte_realloc(void *ptr, size_t size, unsigned int align);
  */
 __rte_experimental
 void *
-rte_realloc_socket(void *ptr, size_t size, unsigned int align, int socket);
+rte_realloc_socket(void *ptr, size_t size, unsigned int align, int socket)
+	__rte_alloc_size(2, 3);
 
 /**
  * This function allocates memory from the huge-page area of memory. The memory
@@ -181,7 +186,8 @@ rte_realloc_socket(void *ptr, size_t size, unsigned int align, int socket);
  *   - Otherwise, the pointer to the allocated object.
  */
 void *
-rte_malloc_socket(const char *type, size_t size, unsigned align, int socket);
+rte_malloc_socket(const char *type, size_t size, unsigned align, int socket)
+	__rte_alloc_size(2);
 
 /**
  * Allocate zero'ed memory from the heap.
@@ -209,7 +215,8 @@ rte_malloc_socket(const char *type, size_t size, unsigned align, int socket);
  *   - Otherwise, the pointer to the allocated object.
  */
 void *
-rte_zmalloc_socket(const char *type, size_t size, unsigned align, int socket);
+rte_zmalloc_socket(const char *type, size_t size, unsigned align, int socket)
+	__rte_alloc_size(2);
 
 /**
  * Replacement function for calloc(), using huge-page memory. Memory area is
@@ -237,7 +244,8 @@ rte_zmalloc_socket(const char *type, size_t size, unsigned align, int socket);
  *   - Otherwise, the pointer to the allocated object.
  */
 void *
-rte_calloc_socket(const char *type, size_t num, size_t size, unsigned align, int socket);
+rte_calloc_socket(const char *type, size_t num, size_t size, unsigned align, int socket)
+	__rte_alloc_size(2, 3);
 
 /**
  * Frees the memory space pointed to by the provided pointer.
Signature · 1 line
2.27.0
#98 lore raw

Re: [dpdk-dev] [PATCH v2] eal/rte_malloc: add alloc_size() attribute to allocation functions

15/10/2020 02:55, Stephen Hemminger:

Quoted 10 lines

By using the alloc_size() attribute the compiler can optimize better and detect errors at compile time.

For example, Gcc will fail one of the invalid allocation examples in app/test/test_malloc.c because the allocation is outside the limits of memory.

---
v2 - rebase onto correct branch (main)

[...]

lib/librte_eal/include/rte_common.h typedef uint16_t unaligned_uint16_t; · line 147

Quoted 11 lines
lib/librte_eal/include/rte_common.h +11 -0
+/**
+ * Tells compiler that the function returns a value that points to
+ * memory, where the size is given by the one or two arguments.
+ * Used by compiler to validate object size.
+ */
+#if defined(RTE_CC_GCC) || defined(RTE_CC_CLANG)
+#define __rte_alloc_size(...) \
+	__attribute__((alloc_size(__VA_ARGS__)))
+#else
+#define __rte_alloc_size(...)
+#endif

Not sure about the rebase: RTE_CC_GCC and RTE_CC_CLANG don't exist.

#99 lore raw

Re: [dpdk-dev] [PATCH v2] eal/rte_malloc: add alloc_size() attribute to allocation functions

19/10/2020 16:13, Thomas Monjalon:

Quoted 12 lines

15/10/2020 02:55, Stephen Hemminger:

Quoted 10 lines

By using the alloc_size() attribute the compiler can optimize better and detect errors at compile time.

For example, Gcc will fail one of the invalid allocation examples in app/test/test_malloc.c because the allocation is outside the limits of memory.

---
v2 - rebase onto correct branch (main)

[...]

Quoted 13 lines

lib/librte_eal/include/rte_common.h typedef uint16_t unaligned_uint16_t; · line 147

Quoted 11 lines
lib/librte_eal/include/rte_common.h +11 -0
+/**
+ * Tells compiler that the function returns a value that points to
+ * memory, where the size is given by the one or two arguments.
+ * Used by compiler to validate object size.
+ */
+#if defined(RTE_CC_GCC) || defined(RTE_CC_CLANG)
+#define __rte_alloc_size(...) \
+	__attribute__((alloc_size(__VA_ARGS__)))
+#else
+#define __rte_alloc_size(...)
+#endif

Not sure about the rebase: RTE_CC_GCC and RTE_CC_CLANG don't exist.

Actually no, I forgot my 7-month old patch introducing them.

Applied, thanks

#100 lore raw

v7 cover ethdev: introduce Rx buffer split

In reply to Slava Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---

v1: http://patches.dpdk.org/patch/79594/
v2: http://patches.dpdk.org/patch/79893/
- add feature support to mlx5 PMD

v3: http://patches.dpdk.org/patch/80389/
- rte_eth_rx_queue_setup_ex is renamed to rte_eth_rxseg_queue_setup
- DEV_RX_OFFLOAD_BUFFER_SPLIT is renamed to RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT
- commit message update
- documentaion provided
- release notes update
- minor bug fixes in testpmd related part

v4: http://patches.dpdk.org/patch/80401/
- common part of rx_queue_setup/rxseg_queue_setup

v5: http://patches.dpdk.org/patch/80609/
- refactored to approach of providing split configuration in the rte_eth_rxconf structure instead of introducing the new API routine
- added support for rxoffs command to testpmd to provide segment offsets for complete testing of split configurations
- patchset is split into two parts - PMD part will be presented as separate series

v6: http://patches.dpdk.org/patch/80785/
- wordy comments rephrased
- typos fixed
- rte_eth_rx_queue_setup configuration check isolated for two main options
- the rest of comments addressed

v7: - description as union of feature dedicated segment split description
- the split limitations reported in the rte_eth_dev_info->rx_seg_capa
- comments made less wordy

Viacheslav Ovsiienko (6):
ethdev: introduce Rx buffer split app/testpmd: add multiple pools per core creation app/testpmd: add buffer split offload configuration app/testpmd: add rxpkts commands and parameters app/testpmd: add rxoffs commands and parameters app/testpmd: add extended Rx queue setup

 app/test-pmd/bpf_cmd.c                      |   4 +-
 app/test-pmd/cmdline.c                      | 151 +++++++++++++++++++++++----
 app/test-pmd/config.c                       | 107 +++++++++++++++++++-
 app/test-pmd/parameters.c                   |  54 ++++++++--
 app/test-pmd/testpmd.c                      | 120 ++++++++++++++++------
 app/test-pmd/testpmd.h                      |  44 ++++++--
 doc/guides/nics/features.rst                |  15 +++
 doc/guides/rel_notes/deprecation.rst        |   5 -
 doc/guides/rel_notes/release_20_11.rst      |   9 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  22 +++-
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  36 ++++++-
 lib/librte_ethdev/rte_ethdev.c              | 152 +++++++++++++++++++++++-----
 lib/librte_ethdev/rte_ethdev.h              |  87 +++++++++++++++-
 13 files changed, 703 insertions(+), 103 deletions(-)
Signature · 1 line
1.8.3.1
#101 lore raw

PATCH v7 1/6 ethdev: introduce Rx buffer split

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---
 doc/guides/nics/features.rst           |  15 ++++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 ++
 lib/librte_ethdev/rte_ethdev.c         | 152 +++++++++++++++++++++++++++------
 lib/librte_ethdev/rte_ethdev.h         |  87 ++++++++++++++++++-
 5 files changed, 237 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +15 -0
index dd8c955..832ea3b 100644
@@ -185,6 +185,21 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[provides]   eth_dev_ops**: ``rxq_info_get:buffer_split``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index 584e720..232cd54 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index bcc0fc2..bcc2479 100644
@@ -60,6 +60,12 @@ New Features
   Added the FEC API which provides functions for query FEC capabilities and
   current FEC mode from device. Also, API for configuring FEC mode is also provided.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -253,6 +259,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +127 -25

186 lines not on this page · raw message

lib/librte_ethdev/rte_ethdev.h +86 -1

125 lines not on this page · raw message

Signature · 1 line
1.8.3.1
Acked-by on v7 1/6
#102 lore raw
Quoted 98 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

Quoted 1 line

LGTM,

Quoted 409 lines
---
 doc/guides/nics/features.rst           |  15 ++++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 ++
 lib/librte_ethdev/rte_ethdev.c         | 152 +++++++++++++++++++++++++++------
 lib/librte_ethdev/rte_ethdev.h         |  87 ++++++++++++++++++-
 5 files changed, 237 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +15 -0
index dd8c955..832ea3b 100644
@@ -185,6 +185,21 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[provides]   eth_dev_ops**: ``rxq_info_get:buffer_split``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index 584e720..232cd54 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index bcc0fc2..bcc2479 100644
@@ -60,6 +60,12 @@ New Features
   Added the FEC API which provides functions for query FEC capabilities and
   current FEC mode from device. Also, API for configuring FEC mode is also provided.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -253,6 +259,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +127 -25
index 892c246..37635c3 100644
@@ -105,6 +105,9 @@ struct rte_eth_xstats_name_off {
 #define RTE_RX_OFFLOAD_BIT2STR(_name)  \
        { DEV_RX_OFFLOAD_##_name, #_name }
 
+#define RTE_ETH_RX_OFFLOAD_BIT2STR(_name)      \
+       { RTE_ETH_RX_OFFLOAD_##_name, #_name }
+
 static const struct {
        uint64_t offload;
        const char *name;
@@ -128,9 +131,11 @@ struct rte_eth_xstats_name_off {
        RTE_RX_OFFLOAD_BIT2STR(SCTP_CKSUM),
        RTE_RX_OFFLOAD_BIT2STR(OUTER_UDP_CKSUM),
        RTE_RX_OFFLOAD_BIT2STR(RSS_HASH),
+       RTE_ETH_RX_OFFLOAD_BIT2STR(BUFFER_SPLIT),
 };
 
 #undef RTE_RX_OFFLOAD_BIT2STR
+#undef RTE_ETH_RX_OFFLOAD_BIT2STR
 
 #define RTE_TX_OFFLOAD_BIT2STR(_name)  \
        { DEV_TX_OFFLOAD_##_name, #_name }
@@ -1763,6 +1768,77 @@ struct rte_eth_dev *
        return ret;
 }
 
+static int
+rte_eth_rx_queue_check_split(const struct rte_eth_rxseg *rx_seg,
+                            uint16_t n_seg, uint32_t *mbp_buf_size,
+                            const struct rte_eth_dev_info *dev_info)
+{
+       const struct rte_eth_rxseg_capa *seg_capa = &dev_info->rx_seg_capa;
+       struct rte_mempool *mp_first;
+       uint32_t offset_mask;
+       uint16_t seg_idx;
+
+       if (n_seg > seg_capa->max_seg) {
+               RTE_ETHDEV_LOG(ERR,
+                              "Requested Rx segments %u exceed supported %u\n",
+                              n_seg, seg_capa->max_seg);
+               return -EINVAL;
+       }
+       /*
+        * Check the sizes and offsets against buffer sizes
+        * for each segment specified in extended configuration.
+        */
+       mp_first = rx_seg[0].split.mp;
+       offset_mask = (1u << seg_capa->offset_align_log2) - 1;
+       for (seg_idx = 0; seg_idx < n_seg; seg_idx++) {
+               struct rte_mempool *mpl = rx_seg[seg_idx].split.mp;
+               uint32_t length = rx_seg[seg_idx].split.length;
+               uint32_t offset = rx_seg[seg_idx].split.offset;
+
+               if (mpl == NULL) {
+                       RTE_ETHDEV_LOG(ERR, "null mempool pointer\n");
+                       return -EINVAL;
+               }
+               if (seg_idx != 0 && mp_first != mpl &&
+                   seg_capa->multi_pools == 0) {
+                       RTE_ETHDEV_LOG(ERR, "Receiving to multiple pools is not supported\n");
+                       return -ENOTSUP;
+               }
+               if (offset != 0) {
+                       if (seg_capa->offset_allowed == 0) {
+                               RTE_ETHDEV_LOG(ERR, "Rx segmentation with offset is not supported\n");
+                               return -ENOTSUP;
+                       }
+                       if (offset & offset_mask) {
+                               RTE_ETHDEV_LOG(ERR, "Rx segmentat invalid offset alignment %u, %u\n",
+                                              offset,
+                                              seg_capa->offset_align_log2);
+                               return -EINVAL;
+                       }
+               }
+               if (mpl->private_data_size <
+                       sizeof(struct rte_pktmbuf_pool_private)) {
+                       RTE_ETHDEV_LOG(ERR,
+                                      "%s private_data_size %u < %u\n",
+                                      mpl->name, mpl->private_data_size,
+                                      (unsigned int)sizeof
+                                       (struct rte_pktmbuf_pool_private));
+                       return -ENOSPC;
+               }
+               offset += seg_idx != 0 ? 0 : RTE_PKTMBUF_HEADROOM;
+               *mbp_buf_size = rte_pktmbuf_data_room_size(mpl);
+               length = length != 0 ? length : *mbp_buf_size;
+               if (*mbp_buf_size < length + offset) {
+                       RTE_ETHDEV_LOG(ERR,
+                                      "%s mbuf_data_room_size %u < %u (segment length=%u + segment offset=%u)\n",
+                                      mpl->name, *mbp_buf_size,
+                                      length + offset, length, offset);
+                       return -EINVAL;
+               }
+       }
+       return 0;
+}
+
 int
 rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
                       uint16_t nb_rx_desc, unsigned int socket_id,
@@ -1784,38 +1860,64 @@ struct rte_eth_dev *
                return -EINVAL;
        }
 
-       if (mp == NULL) {
-               RTE_ETHDEV_LOG(ERR, "Invalid null mempool pointer\n");
-               return -EINVAL;
-       }
-
        RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_queue_setup, -ENOTSUP);
 
-       /*
-        * Check the size of the mbuf data buffer.
-        * This value must be provided in the private data of the memory pool.
-        * First check that the memory pool has a valid private data.
-        */
        ret = rte_eth_dev_info_get(port_id, &dev_info);
        if (ret != 0)
                return ret;
 
-       if (mp->private_data_size < sizeof(struct rte_pktmbuf_pool_private)) {
-               RTE_ETHDEV_LOG(ERR, "%s private_data_size %d < %d\n",
-                       mp->name, (int)mp->private_data_size,
-                       (int)sizeof(struct rte_pktmbuf_pool_private));
-               return -ENOSPC;
-       }
-       mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+       if (mp != NULL) {
+               /* Single pool configuration check. */
+               if (rx_conf->rx_nseg != 0) {
+                       RTE_ETHDEV_LOG(ERR,
+                                      "Ambiguous segment configuration\n");
+                       return -EINVAL;
+               }
+               /*
+                * Check the size of the mbuf data buffer, this value
+                * must be provided in the private data of the memory pool.
+                * First check that the memory pool(s) has a valid private data.
+                */
+               if (mp->private_data_size <
+                               sizeof(struct rte_pktmbuf_pool_private)) {
+                       RTE_ETHDEV_LOG(ERR, "%s private_data_size %u < %u\n",
+                               mp->name, mp->private_data_size,
+                               (unsigned int)
+                               sizeof(struct rte_pktmbuf_pool_private));
+                       return -ENOSPC;
+               }
+               mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+               if (mbp_buf_size < dev_info.min_rx_bufsize +
+                                  RTE_PKTMBUF_HEADROOM) {
+                       RTE_ETHDEV_LOG(ERR,
+                                      "%s mbuf_data_room_size %u < %u (RTE_PKTMBUF_HEADROOM=%u + min_rx_bufsize(dev)=%u)\n",
+                                      mp->name, mbp_buf_size,
+                                      RTE_PKTMBUF_HEADROOM +
+                                      dev_info.min_rx_bufsize,
+                                      RTE_PKTMBUF_HEADROOM,
+                                      dev_info.min_rx_bufsize);
+                       return -EINVAL;
+               }
+       } else {
+               const struct rte_eth_rxseg *rx_seg = rx_conf->rx_seg;
+               uint16_t n_seg = rx_conf->rx_nseg;
 
-       if (mbp_buf_size < dev_info.min_rx_bufsize + RTE_PKTMBUF_HEADROOM) {
-               RTE_ETHDEV_LOG(ERR,
-                       "%s mbuf_data_room_size %d < %d (RTE_PKTMBUF_HEADROOM=%d + min_rx_bufsize(dev)=%d)\n",
-                       mp->name, (int)mbp_buf_size,
-                       (int)(RTE_PKTMBUF_HEADROOM + dev_info.min_rx_bufsize),
-                       (int)RTE_PKTMBUF_HEADROOM,
-                       (int)dev_info.min_rx_bufsize);
-               return -EINVAL;
+               /* Extended multi-segment configuration check. */
+               if (rx_conf->rx_seg == NULL || rx_conf->rx_nseg == 0) {
+                       RTE_ETHDEV_LOG(ERR,
+                                      "Memory pool is null and no extended configuration provided\n");
+                       return -EINVAL;
+               }
+               if (rx_conf->offloads & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+                       ret = rte_eth_rx_queue_check_split(rx_seg, n_seg,
+                                                          &mbp_buf_size,
+                                                          &dev_info);
+                       if (ret != 0)
+                               return ret;
+               } else {
+                       RTE_ETHDEV_LOG(ERR, "No Rx segmentation offload configured\n");
+                       return -EINVAL;
+               }
        }
 
        /* Use default specified by driver, if nb_rx_desc is zero */
lib/librte_ethdev/rte_ethdev.h +86 -2
index 5bcfbb8..2d387c7 100644
@@ -970,6 +970,27 @@ struct rte_eth_txmode {
 };
 
 /**
+ * A structure used to configure an Rx packet segment to split.
+ */
+struct rte_eth_rxseg_split {
+       struct rte_mempool *mp; /**< Memory pool to allocate segment from. */
+       uint16_t length; /**< Segment data length, configures split point. */
+       uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
+       uint32_t reserved; /**< Reserved field. */
+};
+
+/**
+ * A common structure used to describe Rx packet segment properties.
+ */
+struct rte_eth_rxseg {
+       union {
+               /* The settings for buffer split offload. */
+               struct rte_eth_rxseg_split split;
+               /* The other features settings should be added here. */
+       };
+};
+
+/**
  * A structure used to configure an RX ring of an Ethernet port.
  */
 struct rte_eth_rxconf {
@@ -977,6 +998,46 @@ struct rte_eth_rxconf {
        uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
        uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
        uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
+       uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
+       /**
+        * Points to the array of segment descriptions. Each array element
+        * describes the properties for each segment in the receiving
+        * buffer according to feature descripting structure.
+        *
+        * The supported capabilties of receiving segmentation is reported
+        * in rte_eth_dev_info ->rx_seg_capa field.
+        *
+        * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
+        * the PMD will split the received packets into multiple segments
+        * according to the specification in the description array:
+        *
+        * - the first network buffer will be allocated from the memory pool,
+        *   specified in the first array element, the second buffer, from the
+        *   pool in the second element, and so on.
+        *
+        * - the offsets from the segment description elements specify
+        *   the data offset from the buffer beginning except the first mbuf.
+        *   For this one the offset is added with RTE_PKTMBUF_HEADROOM.
+        *
+        * - the lengths in the elements define the maximal data amount
+        *   being received to each segment. The receiving starts with filling
+        *   up the first mbuf data buffer up to specified length. If the
+        *   there are data remaining (packet is longer than buffer in the first
+        *   mbuf) the following data will be pushed to the next segment
+        *   up to its own length, and so on.
+        *
+        * - If the length in the segment description element is zero
+        *   the actual buffer size will be deduced from the appropriate
+        *   memory pool properties.
+        *
+        * - if there is not enough elements to describe the buffer for entire
+        *   packet of maximal length the following parameters will be used
+        *   for the all remaining segments:
+        *     - pool from the last valid element
+        *     - the buffer size from this pool
+        *     - zero offset
+        */
+       struct rte_eth_rxseg *rx_seg;
        /**
         * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
         * Only offloads set on rx_queue_offload_capa or rx_offload_capa
@@ -1260,6 +1321,7 @@ struct rte_eth_conf {
 #define DEV_RX_OFFLOAD_SCTP_CKSUM      0x00020000
 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM  0x00040000
 #define DEV_RX_OFFLOAD_RSS_HASH                0x00080000
+#define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000
 
 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
                                 DEV_RX_OFFLOAD_UDP_CKSUM | \
@@ -1376,6 +1438,16 @@ struct rte_eth_switch_info {
 };
 
 /**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */
+struct rte_eth_rxseg_capa {
+       uint16_t max_seg; /**< Maximum amount of segments to split. */
+       uint8_t multi_pools; /**< Supports receiving to multiple pools.*/
+       uint8_t offset_allowed; /**< Supports buffer offsets. */
+       uint8_t offset_align_log2; /**< Required offset alignment. */
+};
+
+/**
  * Ethernet device information
  */
 
@@ -1403,6 +1475,7 @@ struct rte_eth_dev_info {
        /** Maximum number of hash MAC addresses for MTA and UTA. */
        uint16_t max_vfs; /**< Maximum number of VFs. */
        uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */
+       struct rte_eth_rxseg_capa rx_seg_capa; /**< Segmentation capability.*/
        uint64_t rx_offload_capa;
        /**< All RX offload capabilities including all per-queue ones */
        uint64_t tx_offload_capa;
@@ -2027,9 +2100,21 @@ int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue,
  *   No need to repeat any bit in rx_conf->offloads which has already been
  *   enabled in rte_eth_dev_configure() at port level. An offloading enabled
  *   at port level can't be disabled at queue level.
+ *   The configuration structure also contains the pointer to the array
+ *   of the receiving buffer segment descriptions, see rx_seg and rx_nseg
+ *   fields, this extended configuration might be used by split offloads like
+ *   RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mp_pool is not NULL,
+ *   the extended configuration fields must be set to NULL and zero.
  * @param mb_pool
  *   The pointer to the memory pool from which to allocate *rte_mbuf* network
- *   memory buffers to populate each descriptor of the receive ring.
+ *   memory buffers to populate each descriptor of the receive ring. There are
+ *   two options to provide Rx buffer configuration:
+ *   - single pool:
+ *     mb_pool is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is 0.
+ *   - multiple segments description:
+ *     mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0.
+ *     Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads.
+ *
  * @return
  *   - 0: Success, receive queue correctly set up.
  *   - -EIO: if device is removed.
--

1.8.3.1

#103 lore raw
In reply to Viacheslav Ovsiienko

15/10/2020 22:17, Viacheslav Ovsiienko:

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1448

Quoted 9 lines

/**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */

lib/librte_ethdev/rte_ethdev.h +6 -0
+struct rte_eth_rxseg_capa {
+	uint16_t max_seg; /**< Maximum amount of segments to split. */
+	uint8_t multi_pools; /**< Supports receiving to multiple pools.*/
+	uint8_t offset_allowed; /**< Supports buffer offsets. */
+	uint8_t offset_align_log2; /**< Required offset alignment. */
+};

Why not having bit fields for these last three?

#104 lore raw
Quoted 13 lines · Thomas Monjalon

15/10/2020 22:17, Viacheslav Ovsiienko:

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1448

Quoted 9 lines

/**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */

lib/librte_ethdev/rte_ethdev.h +6 -0
+struct rte_eth_rxseg_capa {
+     uint16_t max_seg; /**< Maximum amount of segments to split. */
+     uint8_t multi_pools; /**< Supports receiving to multiple pools.*/
+     uint8_t offset_allowed; /**< Supports buffer offsets. */
+     uint8_t offset_align_log2; /**< Required offset alignment. */
+};

Why not having bit fields for these last three?

+1

Quoted 3 lines
#105 lore raw

PATCH v7 2/6 app/testpmd: add multiple pools per core creation

In reply to Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

---
 app/test-pmd/bpf_cmd.c                |  4 +--
 app/test-pmd/cmdline.c                |  2 +-
 app/test-pmd/config.c                 |  6 ++--
 app/test-pmd/parameters.c             | 24 +++++++++----
 app/test-pmd/testpmd.c                | 63 +++++++++++++++++++----------------
 app/test-pmd/testpmd.h                | 24 ++++++++++---
 doc/guides/testpmd_app_ug/run_app.rst |  7 ++--
 7 files changed, 83 insertions(+), 47 deletions(-)
app/test-pmd/bpf_cmd.c +2 -2
index 16e3c3b..0a1a178 100644
@@ -69,7 +69,7 @@ struct cmd_bpf_ld_result {
 
 	*flags = RTE_BPF_ETH_F_NONE;
 	arg->type = RTE_BPF_ARG_PTR;
-	arg->size = mbuf_data_size;
+	arg->size = mbuf_data_size[0];
 
 	for (i = 0; str[i] != 0; i++) {
 		v = toupper(str[i]);
@@ -78,7 +78,7 @@ struct cmd_bpf_ld_result {
 		else if (v == 'M') {
 			arg->type = RTE_BPF_ARG_PTR_MBUF;
 			arg->size = sizeof(struct rte_mbuf);
-			arg->buf_size = mbuf_data_size;
+			arg->buf_size = mbuf_data_size[0];
 		} else if (v == '-')
 			continue;
 		else
app/test-pmd/cmdline.c +1 -1
index 273fb1a..a585cf0 100644
@@ -2907,7 +2907,7 @@ struct cmd_setup_rxtx_queue {
 		if (!numa_support || socket_id == NUMA_NO_CONFIG)
 			socket_id = port->socket_id;
 
-		mp = mbuf_pool_find(socket_id);
+		mp = mbuf_pool_find(socket_id, 0);
 		if (mp == NULL) {
 			printf("Failed to setup RX queue: "
 				"No mempool allocation"
app/test-pmd/config.c +3 -3
index d4be694..5f501f6 100644
@@ -690,7 +690,7 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 	printf("\nConnect to socket: %u", port->socket_id);
 
 	if (port_numa[port_id] != NUMA_NO_CONFIG) {
-		mp = mbuf_pool_find(port_numa[port_id]);
+		mp = mbuf_pool_find(port_numa[port_id], 0);
 		if (mp)
 			printf("\nmemory allocation on the socket: %d",
 							port_numa[port_id]);
@@ -3352,9 +3352,9 @@ struct igb_ring_desc_16_bytes {
 	 */
 	tx_pkt_len = 0;
 	for (i = 0; i < nb_segs; i++) {
-		if (seg_lengths[i] > (unsigned) mbuf_data_size) {
+		if (seg_lengths[i] > mbuf_data_size[0]) {
 			printf("length[%u]=%u > mbuf_data_size=%u - give up\n",
-			       i, seg_lengths[i], (unsigned) mbuf_data_size);
+			       i, seg_lengths[i], mbuf_data_size[0]);
 			return;
 		}
 		tx_pkt_len = (uint16_t)(tx_pkt_len + seg_lengths[i]);
app/test-pmd/parameters.c +18 -6
index 15ce8c1..4db4987 100644
@@ -106,7 +106,9 @@
 	       "(flag: 1 for RX; 2 for TX; 3 for RX and TX).\n");
 	printf("  --socket-num=N: set socket from which all memory is allocated "
 	       "in NUMA mode.\n");
-	printf("  --mbuf-size=N: set the data size of mbuf to N bytes.\n");
+	printf("  --mbuf-size=N,[N1[,..Nn]: set the data size of mbuf to "
+	       "N bytes. If multiple numbers are specified the extra pools "
+	       "will be created to receive with packet split features\n");
 	printf("  --total-num-mbufs=N: set the number of mbufs to be allocated "
 	       "in mbuf pools.\n");
 	printf("  --max-pkt-len=N: set the maximum size of packet to N bytes.\n");
@@ -892,12 +894,22 @@
 				}
 			}
 			if (!strcmp(lgopts[opt_idx].name, "mbuf-size")) {
-				n = atoi(optarg);
-				if (n > 0 && n <= 0xFFFF)
-					mbuf_data_size = (uint16_t) n;
-				else
+				unsigned int mb_sz[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs, i;
+
+				nb_segs = parse_item_list(optarg, "mbuf-size",
+					MAX_SEGS_BUFFER_SPLIT, mb_sz, 0);
+				if (nb_segs <= 0)
 					rte_exit(EXIT_FAILURE,
-						 "mbuf-size should be > 0 and < 65536\n");
+						 "bad mbuf-size\n");
+				for (i = 0; i < nb_segs; i++) {
+					if (mb_sz[i] <= 0 || mb_sz[i] > 0xFFFF)
+						rte_exit(EXIT_FAILURE,
+							 "mbuf-size should be "
+							 "> 0 and < 65536\n");
+					mbuf_data_size[i] = (uint16_t) mb_sz[i];
+				}
+				mbuf_data_size_n = nb_segs;
 			}
 			if (!strcmp(lgopts[opt_idx].name, "total-num-mbufs")) {
 				n = atoi(optarg);
app/test-pmd/testpmd.c +35 -28

145 lines not on this page · raw message

app/test-pmd/testpmd.h +19 -5
index 9a29d7a..b42d710 100644
@@ -42,6 +42,13 @@
  */
 #define RTE_MAX_SEGS_PER_PKT 255 /**< nb_segs is a 8-bit unsigned char. */
 
+/*
+ * The maximum number of segments per packet is used to configure
+ * buffer split feature, also specifies the maximum amount of
+ * optional Rx pools to allocate mbufs to split.
+ */
+#define MAX_SEGS_BUFFER_SPLIT 8 /**< nb_segs is a 8-bit unsigned char. */
+
 #define MAX_PKT_BURST 512
 #define DEF_PKT_BURST 32
 
@@ -393,7 +400,9 @@ struct queue_stats_mappings {
 extern uint8_t dcb_config;
 extern uint8_t dcb_test;
 
-extern uint16_t mbuf_data_size; /**< Mbuf data space size. */
+extern uint32_t mbuf_data_size_n;
+extern uint16_t mbuf_data_size[MAX_SEGS_BUFFER_SPLIT];
+/**< Mbuf data space size. */
 extern uint32_t param_total_num_mbufs;
 
 extern uint16_t stats_period;
@@ -605,17 +614,22 @@ struct mplsoudp_decap_conf {
 
 /* Mbuf Pools */
 static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+		    int name_size, unsigned int idx)
 {
-	snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	if (!idx)
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	else
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+			 sock_id, idx);
 }
 
 static inline struct rte_mempool *
-mbuf_pool_find(unsigned int sock_id)
+mbuf_pool_find(unsigned int sock_id, unsigned int idx)
 {
 	char pool_name[RTE_MEMPOOL_NAMESIZE];
 
-	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name));
+	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name), idx);
 	return rte_mempool_lookup((const char *)pool_name);
 }
 
doc/guides/testpmd_app_ug/run_app.rst +5 -2
index ec085c2..1eb0a10 100644
@@ -107,9 +107,12 @@ The command line options are:
     Set the socket from which all memory is allocated in NUMA mode,
     where 0 <= N < number of sockets on the board.
 
-*   ``--mbuf-size=N``
+*   ``--mbuf-size=N[,N1[,...Nn]``
 
-    Set the data size of the mbufs used to N bytes, where N < 65536. The default value is 2048.
+    Set the data size of the mbufs used to N bytes, where N < 65536.
+    The default value is 2048. If multiple mbuf-size values are specified the
+    extra memory pools will be created for allocating mbufs to receive packets
+    with buffer splittling features.
 
 *   ``--total-num-mbufs=N``
 
Signature · 1 line
1.8.3.1
#106 lore raw

PATCH v7 3/6 app/testpmd: add buffer split offload configuration

In reply to Viacheslav Ovsiienko

This patch add support for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT providing per queue configuration for this offload.

---
 app/test-pmd/cmdline.c | 21 +++++++++++----------
 app/test-pmd/config.c  |  9 +++++++++
 2 files changed, 20 insertions(+), 10 deletions(-)
app/test-pmd/cmdline.c +11 -10
index a585cf0..fa71039 100644
@@ -883,16 +883,16 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"port config <port_id> rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"     Enable or disable a per port Rx offloading"
 			" on all Rx queues of a port\n\n"
 
 			"port (port_id) rxq (queue_id) rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"    Enable or disable a per queue Rx offloading"
 			" only on a specific Rx queue\n\n"
 
@@ -18417,7 +18417,8 @@ struct cmd_config_per_port_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc#rss_hash");
+			   "scatter#buffer_split#timestamp#security#"
+			   "keep_crc#rss_hash");
 cmdline_parse_token_string_t cmd_config_per_port_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_port_rx_offload_result,
@@ -18497,8 +18498,8 @@ struct cmd_config_per_port_rx_offload_result {
 	.help_str = "port config <port_id> rx_offload vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc|rss_hash "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc|rss_hash on|off",
 	.tokens = {
 		(void *)&cmd_config_per_port_rx_offload_result_port,
 		(void *)&cmd_config_per_port_rx_offload_result_config,
@@ -18547,7 +18548,7 @@ struct cmd_config_per_queue_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc");
+			   "scatter#buffer_split#timestamp#security#keep_crc");
 cmdline_parse_token_string_t cmd_config_per_queue_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_queue_rx_offload_result,
@@ -18603,8 +18604,8 @@ struct cmd_config_per_queue_rx_offload_result {
 		    "vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc on|off",
 	.tokens = {
 		(void *)&cmd_config_per_queue_rx_offload_result_port,
 		(void *)&cmd_config_per_queue_rx_offload_result_port_id,
app/test-pmd/config.c +9 -0
index 5f501f6..7126d91 100644
@@ -1092,6 +1092,15 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 			printf("off\n");
 	}
 
+	if (dev_info.rx_offload_capa & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+		printf("RX offload buffer split:       ");
+		if (ports[port_id].dev_conf.rxmode.offloads &
+		    RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT)
+			printf("on\n");
+		else
+			printf("off\n");
+	}
+
 	if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_VLAN_INSERT) {
 		printf("VLAN insert:                   ");
 		if (ports[port_id].dev_conf.txmode.offloads &
Signature · 1 line
1.8.3.1
#107 lore raw

PATCH v7 4/6 app/testpmd: add rxpkts commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxpkts=X[,Y]

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command:

testpmd> set rxpkts (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only). Optionally the multiple memory pools can be specified with --mbuf-size command line parameter and the mbufs to receive will be allocated sequentially from these extra memory pools.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 48 ++++++++++++++++++++++-
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  7 ++++
 app/test-pmd/testpmd.h                      | 11 +++++-
 doc/guides/testpmd_app_ug/run_app.rst       |  9 +++++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 21 +++++++++-
 7 files changed, 165 insertions(+), 7 deletions(-)
app/test-pmd/cmdline.c +58 -3
index fa71039..d8dba54 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxpkts (x[,y]*)\n"
+			"    Set the length of each segment to scatter"
+			" packets on receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set txpkts (x[,y]*)\n"
 			"    Set the length of each segment of TXONLY"
 			" and optionally CSUM packets.\n\n"
@@ -3889,6 +3895,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxpkts_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxpkts;
+	cmdline_fixed_string_t seg_lengths;
+};
+
+static void
+cmd_set_rxpkts_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxpkts_result *res;
+	unsigned int seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_lengths, "segment lengths",
+				  MAX_SEGS_BUFFER_SPLIT, seg_lengths, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_segments(seg_lengths, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxpkts_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxpkts_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 rxpkts, "rxpkts");
+cmdline_parse_token_string_t cmd_set_rxpkts_lengths =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 seg_lengths, NULL);
+
+cmdline_parse_inst_t cmd_set_rxpkts = {
+	.f = cmd_set_rxpkts_parsed,
+	.data = NULL,
+	.help_str = "set rxpkts <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxpkts_keyword,
+		(void *)&cmd_set_rxpkts_name,
+		(void *)&cmd_set_rxpkts_lengths,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF TXONLY PACKETS *** */
 
 struct cmd_set_txpkts_result {
@@ -7517,6 +7569,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxpkts"))
+		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
 		show_tx_pkt_segments();
 	else if (!strcmp(res->what, "txtimes"))
@@ -7529,12 +7583,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19807,6 +19861,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
 	(cmdline_parse_inst_t *)&cmd_set_txtimes,
app/test-pmd/config.c +46 -2

62 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index 4db4987..e4e3635 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
 	printf("  --txonly-multi-flow: generate multiple flows in txonly mode\n");
@@ -662,6 +663,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
 		{ "disable-link-check",		0, 0, 0 },
@@ -1272,6 +1274,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
+				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs;
+
+				nb_segs = parse_item_list
+						(optarg, "rxpkt segments",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_len, 0);
+				if (nb_segs > 0)
+					set_rx_pkt_segments(seg_len, nb_segs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxpkts\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "txpkts")) {
 				unsigned seg_lengths[RTE_MAX_SEGS_PER_PKT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +7 -0
index 7e6ef80..f88c1e2 100644
@@ -210,6 +210,13 @@ struct fwd_engine * fwd_engines[] = {
 uint8_t f_quit;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 uint16_t tx_pkt_length = TXONLY_DEF_PACKET_LEN; /**< TXONLY packet length. */
app/test-pmd/testpmd.h +10 -1
index b42d710..8e5ba6a 100644
@@ -420,6 +420,13 @@ struct queue_stats_mappings {
 extern struct rte_fdir_conf fdir_conf;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 #define TXONLY_DEF_PACKET_LEN 64
@@ -816,7 +823,9 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_record_core_cycles(uint8_t on_off);
 void set_record_burst_stats(uint8_t on_off);
 void set_verbose_level(uint16_t vb_level);
-void set_tx_pkt_segments(unsigned *seg_lengths, unsigned nb_segs);
+void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
+void show_rx_pkt_segments(void);
+void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
 void show_tx_pkt_times(void);
doc/guides/testpmd_app_ug/run_app.rst +9 -0
index 1eb0a10..463b76c 100644
@@ -361,6 +361,15 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxpkts=X[,Y]``
+
+    Set the length of segments to scatter packets on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+    Optionally the multiple memory pools can be specified with --mbuf-size
+    command line parameter and the mbufs to receive will be allocated
+    sequentially from these extra memory pools.
+
 *   ``--txpkts=X[,Y]``
 
     Set TX segment sizes or total packet length. Valid for ``tx-only``
doc/guides/testpmd_app_ug/testpmd_funcs.rst +20 -1
index 795c739..ff88762 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxpkts``: Packets to RX split configuration.
+
 * ``txpkts``: Packets to TX configuration.
 
 * ``txtimes``: Burst time pattern for Tx only mode.
@@ -774,6 +776,23 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxpkts
+~~~~~~~~~~
+
+Set the length of segments to scatter packets on receiving if split
+feature is engaged. Affects only the queues configured with split offloads
+(currently BUFFER_SPLIT is supported only). Optionally the multiple memory
+pools can be specified with --mbuf-size command line parameter and the mbufs
+to receive will be allocated sequentially from these extra memory pools (the
+mbuf for the first segment is allocated from the first pool, the second one
+from the second pool, and so on, if segment number is greater then pool's the
+mbuf for remaining segments will be allocated from the last valid pool).
+
+   testpmd> set rxpkts (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. Zero value
+means to use the corresponding memory pool data buffer size.
+
 set txpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#108 lore raw

PATCH v7 5/6 app/testpmd: add rxoffs commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxoffs=X[,Y]

Sets the offsets of packet segments from the beginning of the receiving buffer if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command, providing the same:

testpmd> set rxoffs (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 44 +++++++++++++++++++++
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  2 +
 app/test-pmd/testpmd.h                      |  4 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  6 +++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 17 +++++++-
 7 files changed, 145 insertions(+), 4 deletions(-)
app/test-pmd/cmdline.c +58 -3
index d8dba54..7182bba 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxoffs (x[,y]*)\n"
+			"    Set the offset of each packet segment on"
+			" receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set rxpkts (x[,y]*)\n"
 			"    Set the length of each segment to scatter"
 			" packets on receiving if split feature is engaged."
@@ -3895,6 +3901,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT OFFSETS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxoffs_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxoffs;
+	cmdline_fixed_string_t seg_offsets;
+};
+
+static void
+cmd_set_rxoffs_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxoffs_result *res;
+	unsigned int seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_offsets, "segment offsets",
+				  MAX_SEGS_BUFFER_SPLIT, seg_offsets, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_offsets(seg_offsets, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxoffs_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxoffs_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 rxoffs, "rxoffs");
+cmdline_parse_token_string_t cmd_set_rxoffs_offsets =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 seg_offsets, NULL);
+
+cmdline_parse_inst_t cmd_set_rxoffs = {
+	.f = cmd_set_rxoffs_parsed,
+	.data = NULL,
+	.help_str = "set rxoffs <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxoffs_keyword,
+		(void *)&cmd_set_rxoffs_name,
+		(void *)&cmd_set_rxoffs_offsets,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
 
 struct cmd_set_rxpkts_result {
@@ -7569,6 +7621,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxoffs"))
+		show_rx_pkt_offsets();
 	else if (!strcmp(res->what, "rxpkts"))
 		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
@@ -7583,12 +7637,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxoffs#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19861,6 +19915,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxoffs,
 	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
app/test-pmd/config.c +44 -0

50 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index e4e3635..2298ba5 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxoffs=X[,Y]*: set RX segment offsets for split.\n");
 	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
@@ -663,6 +664,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxoffs",			1, 0, 0 },
 		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
@@ -1274,6 +1276,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxoffs")) {
+				unsigned int seg_off[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_offs;
+
+				nb_offs = parse_item_list
+						(optarg, "rxpkt offsets",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_off, 0);
+				if (nb_offs > 0)
+					set_rx_pkt_offsets(seg_off, nb_offs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxoffs\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
 				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +2 -0
index f88c1e2..580178d 100644
@@ -215,6 +215,8 @@ struct fwd_engine * fwd_engines[] = {
  */
 uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
app/test-pmd/testpmd.h +4 -0
index 8e5ba6a..fc56b60 100644
@@ -425,6 +425,8 @@ struct queue_stats_mappings {
  */
 extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+extern uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
@@ -825,6 +827,8 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_verbose_level(uint16_t vb_level);
 void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_rx_pkt_segments(void);
+void set_rx_pkt_offsets(unsigned int *seg_offsets, unsigned int nb_offs);
+void show_rx_pkt_offsets(void);
 void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
doc/guides/testpmd_app_ug/run_app.rst +6 -0
index 463b76c..9b0a84a 100644
@@ -361,6 +361,12 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxoffs=X[,Y]``
+
+    Set the offsets of packet segments on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+
 *   ``--rxpkts=X[,Y]``
 
     Set the length of segments to scatter packets on receiving if split
doc/guides/testpmd_app_ug/testpmd_funcs.rst +16 -1
index ff88762..c99d887 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxoffs``: Packet offsets for RX split.
+
 * ``rxpkts``: Packets to RX split configuration.
 
 * ``txpkts``: Packets to TX configuration.
@@ -776,6 +778,19 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxoffs
+~~~~~~~~~~
+
+Set the offsets of segments relating to the data buffer beginning on receiving
+if split feature is engaged. Affects only the queues configured with split
+offloads (currently BUFFER_SPLIT is supported only).
+
+   testpmd> set rxoffs (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. If the list
+of offsets is shorter than the list of segments the zero offsets will be used
+for the remaining segments.
+
 set rxpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#109 lore raw

PATCH v7 6/6 app/testpmd: add extended Rx queue setup

In reply to Viacheslav Ovsiienko

If Rx queue is configured with split feature the extended setup with specified segment sizes and pool will be performed.

---
 app/test-pmd/cmdline.c | 12 ++++++------
 app/test-pmd/testpmd.c | 48 ++++++++++++++++++++++++++++++++++++++++++++++--
 app/test-pmd/testpmd.h |  5 +++++
 3 files changed, 57 insertions(+), 8 deletions(-)
app/test-pmd/cmdline.c +6 -6
index 7182bba..204221f 100644
@@ -2927,12 +2927,12 @@ struct cmd_setup_rxtx_queue {
 				rxring_numa[res->portid]);
 			return;
 		}
-		ret = rte_eth_rx_queue_setup(res->portid,
-					     res->qid,
-					     port->nb_rx_desc[res->qid],
-					     socket_id,
-					     &port->rx_conf[res->qid],
-					     mp);
+		ret = rx_queue_setup(res->portid,
+				     res->qid,
+				     port->nb_rx_desc[res->qid],
+				     socket_id,
+				     &port->rx_conf[res->qid],
+				     mp);
 		if (ret)
 			printf("Failed to setup RX queue\n");
 	} else {
app/test-pmd/testpmd.c +46 -2

66 lines not on this page · raw message

app/test-pmd/testpmd.h +5 -0
index fc56b60..af654ea 100644
@@ -876,6 +876,11 @@ void port_rss_reta_info(portid_t port_id,
 
 void set_vf_traffic(portid_t port_id, uint8_t is_rx, uint16_t vf, uint8_t on);
 
+int
+rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
+	       uint16_t nb_rx_desc, unsigned int socket_id,
+	       struct rte_eth_rxconf *rx_conf, struct rte_mempool *mp);
+
 int set_queue_rate_limit(portid_t port_id, uint16_t queue_idx, uint16_t rate);
 int set_vf_rate_limit(portid_t port_id, uint16_t vf, uint16_t rate,
 				uint64_t q_msk);
Signature · 1 line
1.8.3.1
#110 lore raw

v8 cover ethdev: introduce Rx buffer split

In reply to Slava Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---

v1: http://patches.dpdk.org/patch/79594/
v2: http://patches.dpdk.org/patch/79893/
- add feature support to mlx5 PMD

v3: http://patches.dpdk.org/patch/80389/
- rte_eth_rx_queue_setup_ex is renamed to rte_eth_rxseg_queue_setup
- DEV_RX_OFFLOAD_BUFFER_SPLIT is renamed to RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT
- commit message update
- documentaion provided
- release notes update
- minor bug fixes in testpmd related part

v4: http://patches.dpdk.org/patch/80401/
- common part of rx_queue_setup/rxseg_queue_setup

v5: http://patches.dpdk.org/patch/80609/
- refactored to approach of providing split configuration in the rte_eth_rxconf structure instead of introducing the new API routine
- added support for rxoffs command to testpmd to provide segment offsets for complete testing of split configurations
- patchset is split into two parts - PMD part will be presented as separate series

v6: http://patches.dpdk.org/patch/80785/
- wordy comments rephrased
- typos fixed
- rte_eth_rx_queue_setup configuration check isolated for two main options
- the rest of comments addressed

v7: http://patches.dpdk.org/patch/80978/
- description as union of feature dedicated segment split description
- the split limitations reported in the rte_eth_dev_info->rx_seg_capa
- comments made less wordy

v8: - bitfield instead of uint8;
- named union to avoid warinings;
- typos fixed*** SUBJECT HERE ***

---
Viacheslav Ovsiienko (6):
ethdev: introduce Rx buffer split app/testpmd: add multiple pools per core creation app/testpmd: add buffer split offload configuration app/testpmd: add rxpkts commands and parameters app/testpmd: add rxoffs commands and parameters app/testpmd: add extended Rx queue setup

 app/test-pmd/bpf_cmd.c                      |   4 +-
 app/test-pmd/cmdline.c                      | 151 +++++++++++++++++++++++----
 app/test-pmd/config.c                       | 107 +++++++++++++++++++-
 app/test-pmd/parameters.c                   |  54 ++++++++--
 app/test-pmd/testpmd.c                      | 120 ++++++++++++++++------
 app/test-pmd/testpmd.h                      |  44 ++++++--
 doc/guides/nics/features.rst                |  15 +++
 doc/guides/rel_notes/deprecation.rst        |   5 -
 doc/guides/rel_notes/release_20_11.rst      |   9 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  22 +++-
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  36 ++++++-
 lib/librte_ethdev/rte_ethdev.c              | 152 +++++++++++++++++++++++-----
 lib/librte_ethdev/rte_ethdev.h              |  88 +++++++++++++++-
 13 files changed, 704 insertions(+), 103 deletions(-)
Signature · 1 line
1.8.3.1
#111 lore raw

PATCH v8 1/6 ethdev: introduce Rx buffer split

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---
 doc/guides/nics/features.rst           |  15 ++++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 ++
 lib/librte_ethdev/rte_ethdev.c         | 152 +++++++++++++++++++++++++++------
 lib/librte_ethdev/rte_ethdev.h         |  88 ++++++++++++++++++-
 5 files changed, 238 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +15 -0
index dd8c955..832ea3b 100644
@@ -185,6 +185,21 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[provides]   eth_dev_ops**: ``rxq_info_get:buffer_split``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index 584e720..232cd54 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index bcc0fc2..bcc2479 100644
@@ -60,6 +60,12 @@ New Features
   Added the FEC API which provides functions for query FEC capabilities and
   current FEC mode from device. Also, API for configuring FEC mode is also provided.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -253,6 +259,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +127 -25

186 lines not on this page · raw message

lib/librte_ethdev/rte_ethdev.h +87 -1

126 lines not on this page · raw message

Signature · 1 line
1.8.3.1
#112 lore raw
Quoted 383 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---
 doc/guides/nics/features.rst           |  15 ++++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 ++
 lib/librte_ethdev/rte_ethdev.c         | 152 +++++++++++++++++++++++++++------
 lib/librte_ethdev/rte_ethdev.h         |  88 ++++++++++++++++++-
 5 files changed, 238 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +15 -0
index dd8c955..832ea3b 100644
@@ -185,6 +185,21 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[provides]   eth_dev_ops**: ``rxq_info_get:buffer_split``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index 584e720..232cd54 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index bcc0fc2..bcc2479 100644
@@ -60,6 +60,12 @@ New Features
   Added the FEC API which provides functions for query FEC capabilities and
   current FEC mode from device. Also, API for configuring FEC mode is also provided.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -253,6 +259,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +127 -25
index 892c246..0eda3f4 100644
@@ -105,6 +105,9 @@ struct rte_eth_xstats_name_off {
 #define RTE_RX_OFFLOAD_BIT2STR(_name)	\
 	{ DEV_RX_OFFLOAD_##_name, #_name }
 
+#define RTE_ETH_RX_OFFLOAD_BIT2STR(_name)	\
+	{ RTE_ETH_RX_OFFLOAD_##_name, #_name }
+
 static const struct {
 	uint64_t offload;
 	const char *name;
@@ -128,9 +131,11 @@ struct rte_eth_xstats_name_off {
 	RTE_RX_OFFLOAD_BIT2STR(SCTP_CKSUM),
 	RTE_RX_OFFLOAD_BIT2STR(OUTER_UDP_CKSUM),
 	RTE_RX_OFFLOAD_BIT2STR(RSS_HASH),
+	RTE_ETH_RX_OFFLOAD_BIT2STR(BUFFER_SPLIT),
 };
 
 #undef RTE_RX_OFFLOAD_BIT2STR
+#undef RTE_ETH_RX_OFFLOAD_BIT2STR
 
 #define RTE_TX_OFFLOAD_BIT2STR(_name)	\
 	{ DEV_TX_OFFLOAD_##_name, #_name }
@@ -1763,6 +1768,77 @@ struct rte_eth_dev *
 	return ret;
 }
 
+static int
+rte_eth_rx_queue_check_split(const struct rte_eth_rxseg *rx_seg,
+			     uint16_t n_seg, uint32_t *mbp_buf_size,
+			     const struct rte_eth_dev_info *dev_info)
+{
+	const struct rte_eth_rxseg_capa *seg_capa = &dev_info->rx_seg_capa;
+	struct rte_mempool *mp_first;
+	uint32_t offset_mask;
+	uint16_t seg_idx;
+
+	if (n_seg > seg_capa->max_seg) {
+		RTE_ETHDEV_LOG(ERR,
+			       "Requested Rx segments %u exceed supported %u\n",
+			       n_seg, seg_capa->max_seg);
+		return -EINVAL;
+	}
+	/*
+	 * Check the sizes and offsets against buffer sizes
+	 * for each segment specified in extended configuration.
+	 */
+	mp_first = rx_seg[0].conf.split.mp;
+	offset_mask = (1u << seg_capa->offset_align_log2) - 1;
+	for (seg_idx = 0; seg_idx < n_seg; seg_idx++) {
+		struct rte_mempool *mpl = rx_seg[seg_idx].conf.split.mp;
+		uint32_t length = rx_seg[seg_idx].conf.split.length;
+		uint32_t offset = rx_seg[seg_idx].conf.split.offset;
+
+		if (mpl == NULL) {
+			RTE_ETHDEV_LOG(ERR, "null mempool pointer\n");
+			return -EINVAL;
+		}
+		if (seg_idx != 0 && mp_first != mpl &&
+		    seg_capa->multi_pools == 0) {
+			RTE_ETHDEV_LOG(ERR, "Receiving to multiple pools is not supported\n");
+			return -ENOTSUP;
+		}
+		if (offset != 0) {
+			if (seg_capa->offset_allowed == 0) {
+				RTE_ETHDEV_LOG(ERR, "Rx segmentation with offset is not supported\n");
+				return -ENOTSUP;
+			}
+			if (offset & offset_mask) {
+				RTE_ETHDEV_LOG(ERR, "Rx segmentat invalid offset alignment %u, %u\n",
+					       offset,
+					       seg_capa->offset_align_log2);
+				return -EINVAL;
+			}
+		}
+		if (mpl->private_data_size <
+			sizeof(struct rte_pktmbuf_pool_private)) {
+			RTE_ETHDEV_LOG(ERR,
+				       "%s private_data_size %u < %u\n",
+				       mpl->name, mpl->private_data_size,
+				       (unsigned int)sizeof
+					(struct rte_pktmbuf_pool_private));
+			return -ENOSPC;
+		}
+		offset += seg_idx != 0 ? 0 : RTE_PKTMBUF_HEADROOM;
+		*mbp_buf_size = rte_pktmbuf_data_room_size(mpl);
+		length = length != 0 ? length : *mbp_buf_size;
+		if (*mbp_buf_size < length + offset) {
+			RTE_ETHDEV_LOG(ERR,
+				       "%s mbuf_data_room_size %u < %u (segment length=%u + segment offset=%u)\n",
+				       mpl->name, *mbp_buf_size,
+				       length + offset, length, offset);
+			return -EINVAL;
+		}
+	}
+	return 0;
+}
+
 int
 rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
 		       uint16_t nb_rx_desc, unsigned int socket_id,
@@ -1784,38 +1860,64 @@ struct rte_eth_dev *
 		return -EINVAL;
 	}
 
-	if (mp == NULL) {
-		RTE_ETHDEV_LOG(ERR, "Invalid null mempool pointer\n");
-		return -EINVAL;
-	}
-
 	RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_queue_setup, -ENOTSUP);
 
-	/*
-	 * Check the size of the mbuf data buffer.
-	 * This value must be provided in the private data of the memory pool.
-	 * First check that the memory pool has a valid private data.
-	 */
 	ret = rte_eth_dev_info_get(port_id, &dev_info);
 	if (ret != 0)
 		return ret;
 
-	if (mp->private_data_size < sizeof(struct rte_pktmbuf_pool_private)) {
-		RTE_ETHDEV_LOG(ERR, "%s private_data_size %d < %d\n",
-			mp->name, (int)mp->private_data_size,
-			(int)sizeof(struct rte_pktmbuf_pool_private));
-		return -ENOSPC;
-	}
-	mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+	if (mp != NULL) {
+		/* Single pool configuration check. */
+		if (rx_conf->rx_nseg != 0) {
+			RTE_ETHDEV_LOG(ERR,
+				       "Ambiguous segment configuration\n");
+			return -EINVAL;
+		}
+		/*
+		 * Check the size of the mbuf data buffer, this value
+		 * must be provided in the private data of the memory pool.
+		 * First check that the memory pool(s) has a valid private data.
+		 */
+		if (mp->private_data_size <
+				sizeof(struct rte_pktmbuf_pool_private)) {
+			RTE_ETHDEV_LOG(ERR, "%s private_data_size %u < %u\n",
+				mp->name, mp->private_data_size,
+				(unsigned int)
+				sizeof(struct rte_pktmbuf_pool_private));
+			return -ENOSPC;
+		}
+		mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+		if (mbp_buf_size < dev_info.min_rx_bufsize +
+				   RTE_PKTMBUF_HEADROOM) {
+			RTE_ETHDEV_LOG(ERR,
+				       "%s mbuf_data_room_size %u < %u (RTE_PKTMBUF_HEADROOM=%u + min_rx_bufsize(dev)=%u)\n",
+				       mp->name, mbp_buf_size,
+				       RTE_PKTMBUF_HEADROOM +
+				       dev_info.min_rx_bufsize,
+				       RTE_PKTMBUF_HEADROOM,
+				       dev_info.min_rx_bufsize);
+			return -EINVAL;
+		}
+	} else {
+		const struct rte_eth_rxseg *rx_seg = rx_conf->rx_seg;
+		uint16_t n_seg = rx_conf->rx_nseg;
 
-	if (mbp_buf_size < dev_info.min_rx_bufsize + RTE_PKTMBUF_HEADROOM) {
-		RTE_ETHDEV_LOG(ERR,
-			"%s mbuf_data_room_size %d < %d (RTE_PKTMBUF_HEADROOM=%d + min_rx_bufsize(dev)=%d)\n",
-			mp->name, (int)mbp_buf_size,
-			(int)(RTE_PKTMBUF_HEADROOM + dev_info.min_rx_bufsize),
-			(int)RTE_PKTMBUF_HEADROOM,
-			(int)dev_info.min_rx_bufsize);
-		return -EINVAL;
+		/* Extended multi-segment configuration check. */
+		if (rx_conf->rx_seg == NULL || rx_conf->rx_nseg == 0) {
+			RTE_ETHDEV_LOG(ERR,
+				       "Memory pool is null and no extended configuration provided\n");
+			return -EINVAL;
+		}
+		if (rx_conf->offloads & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+			ret = rte_eth_rx_queue_check_split(rx_seg, n_seg,
+							   &mbp_buf_size,
+							   &dev_info);
+			if (ret != 0)
+				return ret;
+		} else {
+			RTE_ETHDEV_LOG(ERR, "No Rx segmentation offload configured\n");
+			return -EINVAL;
+		}
 	}
 
 	/* Use default specified by driver, if nb_rx_desc is zero */
lib/librte_ethdev/rte_ethdev.h +4 -0
index 5bcfbb8..6f18bec 100644
@@ -970,6 +970,27 @@ struct rte_eth_txmode {
 };
 
 /**
+ * A structure used to configure an Rx packet segment to split.
+ */
+struct rte_eth_rxseg_split {
+	struct rte_mempool *mp; /**< Memory pool to allocate segment from. */

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 986

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+	uint16_t length; /**< Segment data length, configures split point. */
+	uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
+	uint32_t reserved; /**< Reserved field. */
+};
+
+/**
+ * A common structure used to describe Rx packet segment properties.
+ */
+struct rte_eth_rxseg {
+	union {

Why not just 'union rte_eth_rxseg' ?

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 988

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+		/* The settings for buffer split offload. */
+		struct rte_eth_rxseg_split split;

Pointer to a split table must be here. I.e. struct rte_eth_rxseg_split *split;
Also it must be specified how the array is terminated. We need either a number of define last item condition (mp == NULL ?)

Quoted 45 lines
lib/librte_ethdev/rte_ethdev.h +5 -0
+		/* The other features settings should be added here. */
+	} conf;
+};
 
 
 
+
+/**

* A structure used to configure an RX ring of an Ethernet port. */
struct rte_eth_rxconf {

lib/librte_ethdev/rte_ethdev.h +30 -0
@@ -977,6 +998,46 @@ struct rte_eth_rxconf {
 	uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
 	uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
 	uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
+	uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
+	/**
+	 * Points to the array of segment descriptions. Each array element
+	 * describes the properties for each segment in the receiving
+	 * buffer according to feature descripting structure.
+	 *
+	 * The supported capabilities of receiving segmentation is reported
+	 * in rte_eth_dev_info ->rx_seg_capa field.
+	 *
+	 * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
+	 * the PMD will split the received packets into multiple segments
+	 * according to the specification in the description array:
+	 *
+	 * - the first network buffer will be allocated from the memory pool,
+	 *   specified in the first array element, the second buffer, from the
+	 *   pool in the second element, and so on.
+	 *
+	 * - the offsets from the segment description elements specify
+	 *   the data offset from the buffer beginning except the first mbuf.
+	 *   For this one the offset is added with RTE_PKTMBUF_HEADROOM.
+	 *
+	 * - the lengths in the elements define the maximal data amount
+	 *   being received to each segment. The receiving starts with filling
+	 *   up the first mbuf data buffer up to specified length. If the
+	 *   there are data remaining (packet is longer than buffer in the first
+	 *   mbuf) the following data will be pushed to the next segment
+	 *   up to its own length, and so on.
+	 *
+	 * - If the length in the segment description element is zero
+	 *   the actual buffer size will be deduced from the appropriate

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1040

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+	 *   memory pool properties.
+	 *
+	 * - if there is not enough elements to describe the buffer for entire
+	 *   packet of maximal length the following parameters will be used
+	 *   for the all remaining segments:
+	 *     - pool from the last valid element
+	 *     - the buffer size from this pool
+	 *     - zero offset
+	 */
+	struct rte_eth_rxseg *rx_seg;

It must not be a pointer. It looks really strange this way taking into account that it is a union in fact. Also, why is it put here in the middle of exsiting structure? IMHO it should be added after offlaods.

Quoted 12 lines
lib/librte_ethdev/rte_ethdev.h +1 -0
 	/**
 	 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
 	 * Only offloads set on rx_queue_offload_capa or rx_offload_capa
@@ -1260,6 +1321,7 @@ struct rte_eth_conf {
 #define DEV_RX_OFFLOAD_SCTP_CKSUM	0x00020000
 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM  0x00040000
 #define DEV_RX_OFFLOAD_RSS_HASH		0x00080000
+#define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000
 
 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
 				 DEV_RX_OFFLOAD_UDP_CKSUM | \

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1445

Quoted 8 lines
lib/librte_ethdev/rte_ethdev.h +5 -0
@@ -1376,6 +1438,17 @@ struct rte_eth_switch_info {
 };
 
 /**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */
+__extension__
+struct rte_eth_rxseg_capa {
+	uint16_t max_seg; /**< Maximum amount of segments to split. */

May be 'max_segs' to avoid confusing vs maximum segment length.

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1448

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+	uint16_t multi_pools:1; /**< Supports receiving to multiple pools.*/
+	uint16_t offset_allowed:1; /**< Supports buffer offsets. */
+	uint16_t offset_align_log2:4; /**< Required offset alignment. */

4 bits are even insufficient to specify cache-line alignment. IMHO at least 8 bits are required.

Consider to put 32 width bit-fields at start of the structure. Than, max_segs (16), offset_align_log2 (8), plus reserved (8).

lib/librte_ethdev/rte_ethdev.h struct rte_eth_dev_info { · line 1479

Quoted 11 lines
lib/librte_ethdev/rte_ethdev.h +4 -0
+};
+
+/**
  * Ethernet device information
  */
 
@@ -1403,6 +1476,7 @@ struct rte_eth_dev_info {
 	/** Maximum number of hash MAC addresses for MTA and UTA. */
 	uint16_t max_vfs; /**< Maximum number of VFs. */
 	uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */
+	struct rte_eth_rxseg_capa rx_seg_capa; /**< Segmentation capability.*/

Why is it put in the middle of existing structure?

Quoted 27 lines
lib/librte_ethdev/rte_ethdev.h +13 -1
 	uint64_t rx_offload_capa;
 	/**< All RX offload capabilities including all per-queue ones */
 	uint64_t tx_offload_capa;
@@ -2027,9 +2101,21 @@ int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue,
  *   No need to repeat any bit in rx_conf->offloads which has already been
  *   enabled in rte_eth_dev_configure() at port level. An offloading enabled
  *   at port level can't be disabled at queue level.
+ *   The configuration structure also contains the pointer to the array
+ *   of the receiving buffer segment descriptions, see rx_seg and rx_nseg
+ *   fields, this extended configuration might be used by split offloads like
+ *   RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mp_pool is not NULL,
+ *   the extended configuration fields must be set to NULL and zero.
  * @param mb_pool
  *   The pointer to the memory pool from which to allocate *rte_mbuf* network
- *   memory buffers to populate each descriptor of the receive ring.
+ *   memory buffers to populate each descriptor of the receive ring. There are
+ *   two options to provide Rx buffer configuration:
+ *   - single pool:
+ *     mb_pool is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is 0.
+ *   - multiple segments description:
+ *     mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0.
+ *     Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads.
+ *
  * @return
  *   - 0: Success, receive queue correctly set up.
  *   - -EIO: if device is removed.
Reviewed-by
#113 lore raw
Quoted 98 lines · Andrew Rybchenko
Quoted 97 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

Quoted 3 lines
Quoted 3 lines

With below review notes processed:

Quoted 441 lines
Quoted 283 lines
---
 doc/guides/nics/features.rst           |  15 ++++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 ++
 lib/librte_ethdev/rte_ethdev.c         | 152 +++++++++++++++++++++++++++------
 lib/librte_ethdev/rte_ethdev.h         |  88 ++++++++++++++++++-
 5 files changed, 238 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +15 -0
index dd8c955..832ea3b 100644
@@ -185,6 +185,21 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[provides]   eth_dev_ops**: ``rxq_info_get:buffer_split``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index 584e720..232cd54 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index bcc0fc2..bcc2479 100644
@@ -60,6 +60,12 @@ New Features
   Added the FEC API which provides functions for query FEC capabilities and
   current FEC mode from device. Also, API for configuring FEC mode is also provided.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -253,6 +259,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +127 -25
index 892c246..0eda3f4 100644
@@ -105,6 +105,9 @@ struct rte_eth_xstats_name_off {
 #define RTE_RX_OFFLOAD_BIT2STR(_name)	\
 	{ DEV_RX_OFFLOAD_##_name, #_name }
 
+#define RTE_ETH_RX_OFFLOAD_BIT2STR(_name)	\
+	{ RTE_ETH_RX_OFFLOAD_##_name, #_name }
+
 static const struct {
 	uint64_t offload;
 	const char *name;
@@ -128,9 +131,11 @@ struct rte_eth_xstats_name_off {
 	RTE_RX_OFFLOAD_BIT2STR(SCTP_CKSUM),
 	RTE_RX_OFFLOAD_BIT2STR(OUTER_UDP_CKSUM),
 	RTE_RX_OFFLOAD_BIT2STR(RSS_HASH),
+	RTE_ETH_RX_OFFLOAD_BIT2STR(BUFFER_SPLIT),
 };
 
 #undef RTE_RX_OFFLOAD_BIT2STR
+#undef RTE_ETH_RX_OFFLOAD_BIT2STR
 
 #define RTE_TX_OFFLOAD_BIT2STR(_name)	\
 	{ DEV_TX_OFFLOAD_##_name, #_name }
@@ -1763,6 +1768,77 @@ struct rte_eth_dev *
 	return ret;
 }
 
+static int
+rte_eth_rx_queue_check_split(const struct rte_eth_rxseg *rx_seg,
+			     uint16_t n_seg, uint32_t *mbp_buf_size,
+			     const struct rte_eth_dev_info *dev_info)
+{
+	const struct rte_eth_rxseg_capa *seg_capa = &dev_info->rx_seg_capa;
+	struct rte_mempool *mp_first;
+	uint32_t offset_mask;
+	uint16_t seg_idx;
+
+	if (n_seg > seg_capa->max_seg) {
+		RTE_ETHDEV_LOG(ERR,
+			       "Requested Rx segments %u exceed supported %u\n",
+			       n_seg, seg_capa->max_seg);
+		return -EINVAL;
+	}
+	/*
+	 * Check the sizes and offsets against buffer sizes
+	 * for each segment specified in extended configuration.
+	 */
+	mp_first = rx_seg[0].conf.split.mp;
+	offset_mask = (1u << seg_capa->offset_align_log2) - 1;
+	for (seg_idx = 0; seg_idx < n_seg; seg_idx++) {
+		struct rte_mempool *mpl = rx_seg[seg_idx].conf.split.mp;
+		uint32_t length = rx_seg[seg_idx].conf.split.length;
+		uint32_t offset = rx_seg[seg_idx].conf.split.offset;
+
+		if (mpl == NULL) {
+			RTE_ETHDEV_LOG(ERR, "null mempool pointer\n");
+			return -EINVAL;
+		}
+		if (seg_idx != 0 && mp_first != mpl &&
+		    seg_capa->multi_pools == 0) {
+			RTE_ETHDEV_LOG(ERR, "Receiving to multiple pools is not supported\n");
+			return -ENOTSUP;
+		}
+		if (offset != 0) {
+			if (seg_capa->offset_allowed == 0) {
+				RTE_ETHDEV_LOG(ERR, "Rx segmentation with offset is not supported\n");
+				return -ENOTSUP;
+			}
+			if (offset & offset_mask) {
+				RTE_ETHDEV_LOG(ERR, "Rx segmentat invalid offset alignment %u, %u\n",
+					       offset,
+					       seg_capa->offset_align_log2);
+				return -EINVAL;
+			}
+		}
+		if (mpl->private_data_size <
+			sizeof(struct rte_pktmbuf_pool_private)) {
+			RTE_ETHDEV_LOG(ERR,
+				       "%s private_data_size %u < %u\n",
+				       mpl->name, mpl->private_data_size,
+				       (unsigned int)sizeof
+					(struct rte_pktmbuf_pool_private));
+			return -ENOSPC;
+		}
+		offset += seg_idx != 0 ? 0 : RTE_PKTMBUF_HEADROOM;
+		*mbp_buf_size = rte_pktmbuf_data_room_size(mpl);
+		length = length != 0 ? length : *mbp_buf_size;
+		if (*mbp_buf_size < length + offset) {
+			RTE_ETHDEV_LOG(ERR,
+				       "%s mbuf_data_room_size %u < %u (segment length=%u + segment offset=%u)\n",
+				       mpl->name, *mbp_buf_size,
+				       length + offset, length, offset);
+			return -EINVAL;
+		}
+	}
+	return 0;
+}
+
 int
 rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
 		       uint16_t nb_rx_desc, unsigned int socket_id,
@@ -1784,38 +1860,64 @@ struct rte_eth_dev *
 		return -EINVAL;
 	}
 
-	if (mp == NULL) {
-		RTE_ETHDEV_LOG(ERR, "Invalid null mempool pointer\n");
-		return -EINVAL;
-	}
-
 	RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_queue_setup, -ENOTSUP);
 
-	/*
-	 * Check the size of the mbuf data buffer.
-	 * This value must be provided in the private data of the memory pool.
-	 * First check that the memory pool has a valid private data.
-	 */
 	ret = rte_eth_dev_info_get(port_id, &dev_info);
 	if (ret != 0)
 		return ret;
 
-	if (mp->private_data_size < sizeof(struct rte_pktmbuf_pool_private)) {
-		RTE_ETHDEV_LOG(ERR, "%s private_data_size %d < %d\n",
-			mp->name, (int)mp->private_data_size,
-			(int)sizeof(struct rte_pktmbuf_pool_private));
-		return -ENOSPC;
-	}
-	mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+	if (mp != NULL) {
+		/* Single pool configuration check. */
+		if (rx_conf->rx_nseg != 0) {
+			RTE_ETHDEV_LOG(ERR,
+				       "Ambiguous segment configuration\n");
+			return -EINVAL;
+		}
+		/*
+		 * Check the size of the mbuf data buffer, this value
+		 * must be provided in the private data of the memory pool.
+		 * First check that the memory pool(s) has a valid private data.
+		 */
+		if (mp->private_data_size <
+				sizeof(struct rte_pktmbuf_pool_private)) {
+			RTE_ETHDEV_LOG(ERR, "%s private_data_size %u < %u\n",
+				mp->name, mp->private_data_size,
+				(unsigned int)
+				sizeof(struct rte_pktmbuf_pool_private));
+			return -ENOSPC;
+		}
+		mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+		if (mbp_buf_size < dev_info.min_rx_bufsize +
+				   RTE_PKTMBUF_HEADROOM) {
+			RTE_ETHDEV_LOG(ERR,
+				       "%s mbuf_data_room_size %u < %u (RTE_PKTMBUF_HEADROOM=%u + min_rx_bufsize(dev)=%u)\n",
+				       mp->name, mbp_buf_size,
+				       RTE_PKTMBUF_HEADROOM +
+				       dev_info.min_rx_bufsize,
+				       RTE_PKTMBUF_HEADROOM,
+				       dev_info.min_rx_bufsize);
+			return -EINVAL;
+		}
+	} else {
+		const struct rte_eth_rxseg *rx_seg = rx_conf->rx_seg;
+		uint16_t n_seg = rx_conf->rx_nseg;
 
-	if (mbp_buf_size < dev_info.min_rx_bufsize + RTE_PKTMBUF_HEADROOM) {
-		RTE_ETHDEV_LOG(ERR,
-			"%s mbuf_data_room_size %d < %d (RTE_PKTMBUF_HEADROOM=%d + min_rx_bufsize(dev)=%d)\n",
-			mp->name, (int)mbp_buf_size,
-			(int)(RTE_PKTMBUF_HEADROOM + dev_info.min_rx_bufsize),
-			(int)RTE_PKTMBUF_HEADROOM,
-			(int)dev_info.min_rx_bufsize);
-		return -EINVAL;
+		/* Extended multi-segment configuration check. */
+		if (rx_conf->rx_seg == NULL || rx_conf->rx_nseg == 0) {
+			RTE_ETHDEV_LOG(ERR,
+				       "Memory pool is null and no extended configuration provided\n");
+			return -EINVAL;
+		}
+		if (rx_conf->offloads & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+			ret = rte_eth_rx_queue_check_split(rx_seg, n_seg,
+							   &mbp_buf_size,
+							   &dev_info);
+			if (ret != 0)
+				return ret;
+		} else {
+			RTE_ETHDEV_LOG(ERR, "No Rx segmentation offload configured\n");
+			return -EINVAL;
+		}
 	}
 
 	/* Use default specified by driver, if nb_rx_desc is zero */
lib/librte_ethdev/rte_ethdev.h +4 -0
index 5bcfbb8..6f18bec 100644
@@ -970,6 +970,27 @@ struct rte_eth_txmode {
 };
 
 /**
+ * A structure used to configure an Rx packet segment to split.
+ */
+struct rte_eth_rxseg_split {
+	struct rte_mempool *mp; /**< Memory pool to allocate segment from. */

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 986

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+	uint16_t length; /**< Segment data length, configures split point. */
+	uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
+	uint32_t reserved; /**< Reserved field. */
+};
+
+/**
+ * A common structure used to describe Rx packet segment properties.
+ */
+struct rte_eth_rxseg {
+	union {

Why not just 'union rte_eth_rxseg' ?

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 988

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+		/* The settings for buffer split offload. */
+		struct rte_eth_rxseg_split split;

Pointer to a split table must be here. I.e. struct rte_eth_rxseg_split *split;
Also it must be specified how the array is terminated. We need either a number of define last item condition (mp == NULL ?)

Quoted 45 lines
lib/librte_ethdev/rte_ethdev.h +5 -0
+		/* The other features settings should be added here. */
+	} conf;
+};
 
 
 
+
+/**

* A structure used to configure an RX ring of an Ethernet port. */
struct rte_eth_rxconf {

lib/librte_ethdev/rte_ethdev.h +30 -0
@@ -977,6 +998,46 @@ struct rte_eth_rxconf {
 	uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
 	uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
 	uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
+	uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
+	/**
+	 * Points to the array of segment descriptions. Each array element
+	 * describes the properties for each segment in the receiving
+	 * buffer according to feature descripting structure.
+	 *
+	 * The supported capabilities of receiving segmentation is reported
+	 * in rte_eth_dev_info ->rx_seg_capa field.
+	 *
+	 * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
+	 * the PMD will split the received packets into multiple segments
+	 * according to the specification in the description array:
+	 *
+	 * - the first network buffer will be allocated from the memory pool,
+	 *   specified in the first array element, the second buffer, from the
+	 *   pool in the second element, and so on.
+	 *
+	 * - the offsets from the segment description elements specify
+	 *   the data offset from the buffer beginning except the first mbuf.
+	 *   For this one the offset is added with RTE_PKTMBUF_HEADROOM.
+	 *
+	 * - the lengths in the elements define the maximal data amount
+	 *   being received to each segment. The receiving starts with filling
+	 *   up the first mbuf data buffer up to specified length. If the
+	 *   there are data remaining (packet is longer than buffer in the first
+	 *   mbuf) the following data will be pushed to the next segment
+	 *   up to its own length, and so on.
+	 *
+	 * - If the length in the segment description element is zero
+	 *   the actual buffer size will be deduced from the appropriate

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1040

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+	 *   memory pool properties.
+	 *
+	 * - if there is not enough elements to describe the buffer for entire
+	 *   packet of maximal length the following parameters will be used
+	 *   for the all remaining segments:
+	 *     - pool from the last valid element
+	 *     - the buffer size from this pool
+	 *     - zero offset
+	 */
+	struct rte_eth_rxseg *rx_seg;

It must not be a pointer. It looks really strange this way taking into account that it is a union in fact. Also, why is it put here in the middle of exsiting structure? IMHO it should be added after offlaods.

Quoted 12 lines
lib/librte_ethdev/rte_ethdev.h +1 -0
 	/**
 	 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
 	 * Only offloads set on rx_queue_offload_capa or rx_offload_capa
@@ -1260,6 +1321,7 @@ struct rte_eth_conf {
 #define DEV_RX_OFFLOAD_SCTP_CKSUM	0x00020000
 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM  0x00040000
 #define DEV_RX_OFFLOAD_RSS_HASH		0x00080000
+#define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000
 
 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
 				 DEV_RX_OFFLOAD_UDP_CKSUM | \

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1445

Quoted 8 lines
lib/librte_ethdev/rte_ethdev.h +5 -0
@@ -1376,6 +1438,17 @@ struct rte_eth_switch_info {
 };
 
 /**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */
+__extension__
+struct rte_eth_rxseg_capa {
+	uint16_t max_seg; /**< Maximum amount of segments to split. */

May be 'max_segs' to avoid confusing vs maximum segment length.

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1448

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+	uint16_t multi_pools:1; /**< Supports receiving to multiple pools.*/
+	uint16_t offset_allowed:1; /**< Supports buffer offsets. */
+	uint16_t offset_align_log2:4; /**< Required offset alignment. */

4 bits are even insufficient to specify cache-line alignment. IMHO at least 8 bits are required.

Consider to put 32 width bit-fields at start of the structure. Than, max_segs (16), offset_align_log2 (8), plus reserved (8).

lib/librte_ethdev/rte_ethdev.h struct rte_eth_dev_info { · line 1479

Quoted 11 lines
lib/librte_ethdev/rte_ethdev.h +4 -0
+};
+
+/**
  * Ethernet device information
  */
 
@@ -1403,6 +1476,7 @@ struct rte_eth_dev_info {
 	/** Maximum number of hash MAC addresses for MTA and UTA. */
 	uint16_t max_vfs; /**< Maximum number of VFs. */
 	uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */
+	struct rte_eth_rxseg_capa rx_seg_capa; /**< Segmentation capability.*/

Why is it put in the middle of existing structure?

Quoted 27 lines
lib/librte_ethdev/rte_ethdev.h +13 -1
 	uint64_t rx_offload_capa;
 	/**< All RX offload capabilities including all per-queue ones */
 	uint64_t tx_offload_capa;
@@ -2027,9 +2101,21 @@ int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue,
  *   No need to repeat any bit in rx_conf->offloads which has already been
  *   enabled in rte_eth_dev_configure() at port level. An offloading enabled
  *   at port level can't be disabled at queue level.
+ *   The configuration structure also contains the pointer to the array
+ *   of the receiving buffer segment descriptions, see rx_seg and rx_nseg
+ *   fields, this extended configuration might be used by split offloads like
+ *   RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mp_pool is not NULL,
+ *   the extended configuration fields must be set to NULL and zero.
  * @param mb_pool
  *   The pointer to the memory pool from which to allocate *rte_mbuf* network
- *   memory buffers to populate each descriptor of the receive ring.
+ *   memory buffers to populate each descriptor of the receive ring. There are
+ *   two options to provide Rx buffer configuration:
+ *   - single pool:
+ *     mb_pool is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is 0.
+ *   - multiple segments description:
+ *     mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0.
+ *     Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads.
+ *
  * @return
  *   - 0: Success, receive queue correctly set up.
  *   - -EIO: if device is removed.
#114 lore raw
In reply to Andrew Rybchenko

Hi, Andrew

Quoted 17 lines

-----Original Message-----
From: Andrew Rybchenko <[email protected]>
Sent: Friday, October 16, 2020 11:52
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH v8 1/6] ethdev: introduce Rx buffer split

Quoted 6 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

[snip]

Quoted 10 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 986

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+struct rte_eth_rxseg {
+	union {

Why not just 'union rte_eth_rxseg' ?

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 988

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+		/* The settings for buffer split offload. */
+		struct rte_eth_rxseg_split split;

Pointer to a split table must be here. I.e.
struct rte_eth_rxseg_split *split;

OK, will try to simplify with that, thanks.

Quoted 2 lines

Also it must be specified how the array is terminated.
We need either a number of define last item condition (mp == NULL ?)

We have one, please see: "rte_eth_rxconf->rx_nseg"

Quoted 16 lines
Quoted 13 lines
lib/librte_ethdev/rte_ethdev.h +5 -0
+		/* The other features settings should be added here. */
+	} conf;
+};
 
 
 
+
+/**
  * A structure used to configure an RX ring of an Ethernet port.
  */
 struct rte_eth_rxconf {

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 999

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +0 -0
@@ -977,6 +998,46 @@ struct rte_eth_rxconf {
 	uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
 	uint8_t rx_drop_en; /**< Drop packets if no descriptors are available.

*/

Quoted 6 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1040

Quoted 1 line
lib/librte_ethdev/rte_ethdev.h +1 -0
+	struct rte_eth_rxseg *rx_seg;

It must not be a pointer. It looks really strange this way taking into account that it is a union in fact.
Also, why is it put here in the middle of exsiting structure? IMHO it should be added after offlaods.

OK, agree, will move.

Quoted 14 lines
Quoted 14 lines
lib/librte_ethdev/rte_ethdev.h +1 -1
 	/**
 	 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
 	 * Only offloads set on rx_queue_offload_capa or rx_offload_capa @@
-1260,6 +1321,7 @@ struct rte_eth_conf {
 #define DEV_RX_OFFLOAD_SCTP_CKSUM	0x00020000
 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM  0x00040000
 #define DEV_RX_OFFLOAD_RSS_HASH		0x00080000
+#define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000
 
 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
 				 DEV_RX_OFFLOAD_UDP_CKSUM | \
@@ -1376,6 +1438,17 @@ struct rte_eth_switch_info {  };
 
Quoted 9 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1445

Quoted 6 lines
lib/librte_ethdev/rte_ethdev.h +5 -0
 /**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */
+__extension__
+struct rte_eth_rxseg_capa {
+	uint16_t max_seg; /**< Maximum amount of segments to split. */

May be 'max_segs' to avoid confusing vs maximum segment length.

OK, max_nseg would be more appropriate.

Quoted 6 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1448

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+	uint16_t multi_pools:1; /**< Supports receiving to multiple pools.*/
+	uint16_t offset_allowed:1; /**< Supports buffer offsets. */
+	uint16_t offset_align_log2:4; /**< Required offset alignment. */

4 bits are even insufficient to specify cache-line alignment. IMHO at least 8 bits are required.

4 bits seems to be quite sufficient. It is a log2, tells how many lsbs in offset should be zeroes. 2^15 is 32K, it covers all reasonable alignments for uint16_t type.

Quoted 3 lines

Consider to put 32 width bit-fields at start of the structure. Than, max_segs (16), offset_align_log2 (8), plus reserved (8).

OK, np.

With best regards, Slava

#115 lore raw
Quoted 30 lines · Slava Ovsiienko

Hi, Andrew

Quoted 17 lines

-----Original Message-----
From: Andrew Rybchenko <[email protected]>
Sent: Friday, October 16, 2020 11:52
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH v8 1/6] ethdev: introduce Rx buffer split

Quoted 6 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

[snip]

Quoted 10 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 986

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+struct rte_eth_rxseg {
+	union {

Why not just 'union rte_eth_rxseg' ?

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 988

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+		/* The settings for buffer split offload. */
+		struct rte_eth_rxseg_split split;

Pointer to a split table must be here. I.e.
struct rte_eth_rxseg_split *split;

Quoted 6 lines

OK, will try to simplify with that, thanks.

Quoted 2 lines

Also it must be specified how the array is terminated.
We need either a number of define last item condition (mp == NULL ?)

We have one, please see: "rte_eth_rxconf->rx_nseg"

A bit confusing to have it outside of union far from split but may be acceptable for the experimental API... If there are better ideas - welcome.

Quoted 47 lines
Quoted 16 lines
Quoted 13 lines
lib/librte_ethdev/rte_ethdev.h +5 -0
+		/* The other features settings should be added here. */
+	} conf;
+};
 
 
 
+
+/**
  * A structure used to configure an RX ring of an Ethernet port.
  */
 struct rte_eth_rxconf {

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 999

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +0 -0
@@ -977,6 +998,46 @@ struct rte_eth_rxconf {
 	uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
 	uint8_t rx_drop_en; /**< Drop packets if no descriptors are available.

*/

Quoted 6 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1040

Quoted 1 line
lib/librte_ethdev/rte_ethdev.h +1 -0
+	struct rte_eth_rxseg *rx_seg;

It must not be a pointer. It looks really strange this way taking into account that it is a union in fact.
Also, why is it put here in the middle of exsiting structure? IMHO it should be added after offlaods.

OK, agree, will move.

Quoted 14 lines
Quoted 14 lines
lib/librte_ethdev/rte_ethdev.h +1 -1
 	/**
 	 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
 	 * Only offloads set on rx_queue_offload_capa or rx_offload_capa @@
-1260,6 +1321,7 @@ struct rte_eth_conf {
 #define DEV_RX_OFFLOAD_SCTP_CKSUM	0x00020000
 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM  0x00040000
 #define DEV_RX_OFFLOAD_RSS_HASH		0x00080000
+#define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000
 
 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
 				 DEV_RX_OFFLOAD_UDP_CKSUM | \
@@ -1376,6 +1438,17 @@ struct rte_eth_switch_info {  };
 
Quoted 9 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1445

Quoted 6 lines
lib/librte_ethdev/rte_ethdev.h +5 -0
 /**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */
+__extension__
+struct rte_eth_rxseg_capa {
+	uint16_t max_seg; /**< Maximum amount of segments to split. */

May be 'max_segs' to avoid confusing vs maximum segment length.

Quoted 1 line

OK, max_nseg would be more appropriate.

Agreed.

Quoted 18 lines
Quoted 7 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1448

Quoted 4 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+	uint16_t multi_pools:1; /**< Supports receiving to multiple pools.*/
+	uint16_t offset_allowed:1; /**< Supports buffer offsets. */
+	uint16_t offset_align_log2:4; /**< Required offset alignment. */

4 bits are even insufficient to specify cache-line alignment. IMHO at least 8 bits are required.

4 bits seems to be quite sufficient. It is a log2, tells how many lsbs in offset should be zeroes. 2^15 is 32K, it covers all reasonable alignments for uint16_t type.

Quoted 3 lines

Consider to put 32 width bit-fields at start of the structure. Than, max_segs (16), offset_align_log2 (8), plus reserved (8).

OK, np.

With best regards, Slava

#116 lore raw
Quoted 43 lines

-----Original Message-----
From: Andrew Rybchenko <[email protected]>
Sent: Friday, October 16, 2020 12:28
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [dpdk-dev] [PATCH v8 1/6] ethdev: introduce Rx buffer split

Quoted 32 lines · Slava Ovsiienko

Hi, Andrew

Quoted 17 lines

-----Original Message-----
From: Andrew Rybchenko <[email protected]>
Sent: Friday, October 16, 2020 11:52
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH v8 1/6] ethdev: introduce Rx buffer split

Quoted 6 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

[snip]

Quoted 10 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 986

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+struct rte_eth_rxseg {
+	union {

Why not just 'union rte_eth_rxseg' ?

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 988

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+		/* The settings for buffer split offload. */
+		struct rte_eth_rxseg_split split;

Pointer to a split table must be here. I.e.
struct rte_eth_rxseg_split *split;

OK, will try to simplify with that, thanks.

Quoted 8 lines
Quoted 4 lines
Quoted 2 lines

Also it must be specified how the array is terminated.
We need either a number of define last item condition (mp == NULL ?)

We have one, please see: "rte_eth_rxconf->rx_nseg"

A bit confusing to have it outside of union far from split but may be acceptable for the experimental API...
If there are better ideas - welcome.

It can't be inside union - it does not share the memory layout with descriptors itself.
It can't be inside the descriptor structure either - rx_nseg describes the size of array, not segment property. It is uint16_t rx_nseg and is placed before offload to save some space due to rte_eth_rxconf members alignment. Will mark rte_eth_rxseg_split as experimental, nice clue, thanks.

With best regards, Slava

#117 lore raw
In reply to Slava Ovsiienko

16/10/2020 11:15, Slava Ovsiienko:

Quoted 20 lines

From: Andrew Rybchenko <[email protected]>

Quoted 10 lines · Viacheslav Ovsiienko

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1445

Quoted 6 lines

/**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */

lib/librte_ethdev/rte_ethdev.h +3 -0
+__extension__
+struct rte_eth_rxseg_capa {
+	uint16_t max_seg; /**< Maximum amount of segments to split. */

May be 'max_segs' to avoid confusing vs maximum segment length.

OK, max_nseg would be more appropriate.

Quoted 6 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1448

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+	uint16_t multi_pools:1; /**< Supports receiving to multiple pools.*/
+	uint16_t offset_allowed:1; /**< Supports buffer offsets. */
+	uint16_t offset_align_log2:4; /**< Required offset alignment. */

4 bits are even insufficient to specify cache-line alignment. IMHO at least 8 bits are required.

Quoted 2 lines

4 bits seems to be quite sufficient. It is a log2, tells how many lsbs in offset should be zeroes. 2^15 is 32K, it covers all reasonable alignments for uint16_t type.

bitfields with uint16_t is not standard I think,
we could experience build issues.
If I remember well uint32_t is safer.

#118 lore raw
Quoted 31 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Friday, October 16, 2020 12:37
To: Andrew Rybchenko <[email protected]>; Slava Ovsiienko <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH v8 1/6] ethdev: introduce Rx buffer split

16/10/2020 11:15, Slava Ovsiienko:

Quoted 20 lines

From: Andrew Rybchenko <[email protected]>

Quoted 10 lines · Viacheslav Ovsiienko

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1445

Quoted 6 lines

/**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */

lib/librte_ethdev/rte_ethdev.h +3 -0
+__extension__
+struct rte_eth_rxseg_capa {
+	uint16_t max_seg; /**< Maximum amount of segments to split. */

May be 'max_segs' to avoid confusing vs maximum segment length.

OK, max_nseg would be more appropriate.

Quoted 6 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1448

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+	uint16_t multi_pools:1; /**< Supports receiving to multiple pools.*/
+	uint16_t offset_allowed:1; /**< Supports buffer offsets. */
+	uint16_t offset_align_log2:4; /**< Required offset alignment. */

4 bits are even insufficient to specify cache-line alignment. IMHO at least 8 bits are required.

Quoted 7 lines
Quoted 1 line

4 bits seems to be quite sufficient. It is a log2, tells how many lsbs in offset

should be zeroes.

Quoted 1 line

2^15 is 32K, it covers all reasonable alignments for uint16_t type.

bitfields with uint16_t is not standard I think, we could experience build issues. If I remember well uint32_t is safer.

Agree, it is being replaced with uint32_t and embraces max_nseg as well.

#119 lore raw
In reply to Viacheslav Ovsiienko
Quoted 100 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf.
This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

     seg0 - pool0, len0=14B, off0=2
     seg1 - pool1, len1=20B, off1=128B
     seg2 - pool2, len2=20B, off2=0B
     seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

<...>

Quoted 38 lines
lib/librte_ethdev/rte_ethdev.h +31 -0
+/**
   * A structure used to configure an RX ring of an Ethernet port.
   */
  struct rte_eth_rxconf {
@@ -977,6 +998,46 @@ struct rte_eth_rxconf {
  	uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
  	uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
  	uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
+	uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
+	/**
+	 * Points to the array of segment descriptions. Each array element
+	 * describes the properties for each segment in the receiving
+	 * buffer according to feature descripting structure.
+	 *
+	 * The supported capabilities of receiving segmentation is reported
+	 * in rte_eth_dev_info ->rx_seg_capa field.
+	 *
+	 * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
+	 * the PMD will split the received packets into multiple segments
+	 * according to the specification in the description array:
+	 *
+	 * - the first network buffer will be allocated from the memory pool,
+	 *   specified in the first array element, the second buffer, from the
+	 *   pool in the second element, and so on.
+	 *
+	 * - the offsets from the segment description elements specify
+	 *   the data offset from the buffer beginning except the first mbuf.
+	 *   For this one the offset is added with RTE_PKTMBUF_HEADROOM.
+	 *
+	 * - the lengths in the elements define the maximal data amount
+	 *   being received to each segment. The receiving starts with filling
+	 *   up the first mbuf data buffer up to specified length. If the
+	 *   there are data remaining (packet is longer than buffer in the first
+	 *   mbuf) the following data will be pushed to the next segment
+	 *   up to its own length, and so on.
+	 *
+	 * - If the length in the segment description element is zero
+	 *   the actual buffer size will be deduced from the appropriate

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1040

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+	 *   memory pool properties.
+	 *
+	 * - if there is not enough elements to describe the buffer for entire
+	 *   packet of maximal length the following parameters will be used
+	 *   for the all remaining segments:
+	 *     - pool from the last valid element
+	 *     - the buffer size from this pool
+	 *     - zero offset
+	 */
+	struct rte_eth_rxseg *rx_seg;

"struct rte_eth_rxconf" is very commonly used, I think all applications does the 'rte_eth_rx_queue_setup()', but "buffer split" is not a common usage,

I am against the "struct rte_eth_rxseg *rx_seg;" field creating this much noise in the "struct rte_eth_rxconf" documentation.
As mentioned before, can you please move the above detailed documentation to where "struct rte_eth_rxseg" defined, and in this struct put a single comment for "struct rte_eth_rxseg *rx_seg" ?

#120 lore raw
Quoted 166 lines · Ferruh Yigit
Quoted 100 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

     struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length, configures "split point" */
     uint16_t offset; /* data offset from beginning
                   of mbuf data buffer */
     uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
    rx_seg - pointer the array of segment descriptions, each element describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
    rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:
     seg0 - pool0, len0=14B, off0=2
     seg1 - pool1, len1=20B, off1=128B
     seg2 - pool2, len2=20B, off2=0B
     seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
     seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
     seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
     seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
     seg2 - 20B @ 0 in mbuf from pool2
     seg3 - 512B @ 0 in mbuf from pool3
     seg4 - 512B @ 0 in mbuf from pool3
     seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

<...>

Quoted 49 lines

+/**
   * A structure used to configure an RX ring of an Ethernet port. */
  struct rte_eth_rxconf {

(quoted diff) +0 -0
@@ -977,6 +998,46 @@ struct rte_eth_rxconf {

      uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
      uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
      uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
+    uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */ +    /**
+     * Points to the array of segment descriptions. Each array element +     * describes the properties for each segment in the receiving +     * buffer according to feature descripting structure.
+     *
+     * The supported capabilities of receiving segmentation is reported +     * in rte_eth_dev_info ->rx_seg_capa field.
+     *
+     * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field, +     * the PMD will split the received packets into multiple segments +     * according to the specification in the description array:
+     *
+     * - the first network buffer will be allocated from the memory pool,
+     *   specified in the first array element, the second buffer, from the
+     *   pool in the second element, and so on.
+     *
+     * - the offsets from the segment description elements specify +     *   the data offset from the buffer beginning except the first mbuf.
+     *   For this one the offset is added with RTE_PKTMBUF_HEADROOM. +     *
+     * - the lengths in the elements define the maximal data amount +     *   being received to each segment. The receiving starts with filling
+     *   up the first mbuf data buffer up to specified length. If the +     *   there are data remaining (packet is longer than buffer in the first
+     *   mbuf) the following data will be pushed to the next segment +     *   up to its own length, and so on.
+     *
+     * - If the length in the segment description element is zero +     *   the actual buffer size will be deduced from the appropriate +     *   memory pool properties.
+     *
+     * - if there is not enough elements to describe the buffer for

Quoted 8 lines

entire
+     *   packet of maximal length the following parameters will be used +     *   for the all remaining segments:
+     *     - pool from the last valid element
+     *     - the buffer size from this pool
+     *     - zero offset
+     */
+    struct rte_eth_rxseg *rx_seg;

"struct rte_eth_rxconf" is very commonly used, I think all applications does the 'rte_eth_rx_queue_setup()', but "buffer split" is not a common usage,

Quoted 5 lines

I am against the "struct rte_eth_rxseg *rx_seg;" field creating this much noise in the "struct rte_eth_rxconf" documentation. As mentioned before, can you please move the above detailed documentation to where "struct rte_eth_rxseg" defined, and in this struct put a single comment for "struct rte_eth_rxseg *rx_seg" ?

+1

#121 lore raw
In reply to Ferruh Yigit
Quoted 166 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, October 16, 2020 12:19
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH v8 1/6] ethdev: introduce Rx buffer split

Quoted 77 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */ };

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf.
This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

     seg0 - pool0, len0=14B, off0=2
     seg1 - pool1, len1=20B, off1=128B
     seg2 - pool2, len2=20B, off2=0B
     seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

Quoted 1 line

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and

configured

Quoted 19 lines

to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

<...>

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 999

Quoted 7 lines
lib/librte_ethdev/rte_ethdev.h +1 -0
+/**
   * A structure used to configure an RX ring of an Ethernet port.
   */
  struct rte_eth_rxconf {
@@ -977,6 +998,46 @@ struct rte_eth_rxconf {
  	uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
  	uint8_t rx_drop_en; /**< Drop packets if no descriptors are available.

*/

Quoted 32 lines
(quoted diff) +0 -0
  	uint8_t rx_deferred_start; /**< Do not start queue with

rte_eth_dev_start(). */

lib/librte_ethdev/rte_ethdev.h +30 -0
+	uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
+	/**
+	 * Points to the array of segment descriptions. Each array element
+	 * describes the properties for each segment in the receiving
+	 * buffer according to feature descripting structure.
+	 *
+	 * The supported capabilities of receiving segmentation is reported
+	 * in rte_eth_dev_info ->rx_seg_capa field.
+	 *
+	 * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
+	 * the PMD will split the received packets into multiple segments
+	 * according to the specification in the description array:
+	 *
+	 * - the first network buffer will be allocated from the memory pool,
+	 *   specified in the first array element, the second buffer, from the
+	 *   pool in the second element, and so on.
+	 *
+	 * - the offsets from the segment description elements specify
+	 *   the data offset from the buffer beginning except the first mbuf.
+	 *   For this one the offset is added with RTE_PKTMBUF_HEADROOM.
+	 *
+	 * - the lengths in the elements define the maximal data amount
+	 *   being received to each segment. The receiving starts with filling
+	 *   up the first mbuf data buffer up to specified length. If the
+	 *   there are data remaining (packet is longer than buffer in the first
+	 *   mbuf) the following data will be pushed to the next segment
+	 *   up to its own length, and so on.
+	 *
+	 * - If the length in the segment description element is zero
+	 *   the actual buffer size will be deduced from the appropriate

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1040

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+	 *   memory pool properties.
+	 *
+	 * - if there is not enough elements to describe the buffer for entire
+	 *   packet of maximal length the following parameters will be used
+	 *   for the all remaining segments:
+	 *     - pool from the last valid element
+	 *     - the buffer size from this pool
+	 *     - zero offset
+	 */
+	struct rte_eth_rxseg *rx_seg;

"struct rte_eth_rxconf" is very commonly used, I think all applications does the 'rte_eth_rx_queue_setup()', but "buffer split" is not a common usage,

Quoted 5 lines

I am against the "struct rte_eth_rxseg *rx_seg;" field creating this much noise in the "struct rte_eth_rxconf" documentation.
As mentioned before, can you please move the above detailed documentation to where "struct rte_eth_rxseg" defined, and in this struct put a single comment for "struct rte_eth_rxseg *rx_seg" ?

Sure, we had doubts about putting this wordy comment to the rxconf structure either. Now we can move the comment to the rte_eth_rxseg_split declaration.

With best regards, Slava

#122 lore raw

PATCH v8 2/6 app/testpmd: add multiple pools per core creation

In reply to Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

---
 app/test-pmd/bpf_cmd.c                |  4 +--
 app/test-pmd/cmdline.c                |  2 +-
 app/test-pmd/config.c                 |  6 ++--
 app/test-pmd/parameters.c             | 24 +++++++++----
 app/test-pmd/testpmd.c                | 63 +++++++++++++++++++----------------
 app/test-pmd/testpmd.h                | 24 ++++++++++---
 doc/guides/testpmd_app_ug/run_app.rst |  7 ++--
 7 files changed, 83 insertions(+), 47 deletions(-)
app/test-pmd/bpf_cmd.c +2 -2
index 16e3c3b..0a1a178 100644
@@ -69,7 +69,7 @@ struct cmd_bpf_ld_result {
 
 	*flags = RTE_BPF_ETH_F_NONE;
 	arg->type = RTE_BPF_ARG_PTR;
-	arg->size = mbuf_data_size;
+	arg->size = mbuf_data_size[0];
 
 	for (i = 0; str[i] != 0; i++) {
 		v = toupper(str[i]);
@@ -78,7 +78,7 @@ struct cmd_bpf_ld_result {
 		else if (v == 'M') {
 			arg->type = RTE_BPF_ARG_PTR_MBUF;
 			arg->size = sizeof(struct rte_mbuf);
-			arg->buf_size = mbuf_data_size;
+			arg->buf_size = mbuf_data_size[0];
 		} else if (v == '-')
 			continue;
 		else
app/test-pmd/cmdline.c +1 -1
index 273fb1a..a585cf0 100644
@@ -2907,7 +2907,7 @@ struct cmd_setup_rxtx_queue {
 		if (!numa_support || socket_id == NUMA_NO_CONFIG)
 			socket_id = port->socket_id;
 
-		mp = mbuf_pool_find(socket_id);
+		mp = mbuf_pool_find(socket_id, 0);
 		if (mp == NULL) {
 			printf("Failed to setup RX queue: "
 				"No mempool allocation"
app/test-pmd/config.c +3 -3
index d4be694..5f501f6 100644
@@ -690,7 +690,7 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 	printf("\nConnect to socket: %u", port->socket_id);
 
 	if (port_numa[port_id] != NUMA_NO_CONFIG) {
-		mp = mbuf_pool_find(port_numa[port_id]);
+		mp = mbuf_pool_find(port_numa[port_id], 0);
 		if (mp)
 			printf("\nmemory allocation on the socket: %d",
 							port_numa[port_id]);
@@ -3352,9 +3352,9 @@ struct igb_ring_desc_16_bytes {
 	 */
 	tx_pkt_len = 0;
 	for (i = 0; i < nb_segs; i++) {
-		if (seg_lengths[i] > (unsigned) mbuf_data_size) {
+		if (seg_lengths[i] > mbuf_data_size[0]) {
 			printf("length[%u]=%u > mbuf_data_size=%u - give up\n",
-			       i, seg_lengths[i], (unsigned) mbuf_data_size);
+			       i, seg_lengths[i], mbuf_data_size[0]);
 			return;
 		}
 		tx_pkt_len = (uint16_t)(tx_pkt_len + seg_lengths[i]);
app/test-pmd/parameters.c +18 -6
index 15ce8c1..4db4987 100644
@@ -106,7 +106,9 @@
 	       "(flag: 1 for RX; 2 for TX; 3 for RX and TX).\n");
 	printf("  --socket-num=N: set socket from which all memory is allocated "
 	       "in NUMA mode.\n");
-	printf("  --mbuf-size=N: set the data size of mbuf to N bytes.\n");
+	printf("  --mbuf-size=N,[N1[,..Nn]: set the data size of mbuf to "
+	       "N bytes. If multiple numbers are specified the extra pools "
+	       "will be created to receive with packet split features\n");
 	printf("  --total-num-mbufs=N: set the number of mbufs to be allocated "
 	       "in mbuf pools.\n");
 	printf("  --max-pkt-len=N: set the maximum size of packet to N bytes.\n");
@@ -892,12 +894,22 @@
 				}
 			}
 			if (!strcmp(lgopts[opt_idx].name, "mbuf-size")) {
-				n = atoi(optarg);
-				if (n > 0 && n <= 0xFFFF)
-					mbuf_data_size = (uint16_t) n;
-				else
+				unsigned int mb_sz[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs, i;
+
+				nb_segs = parse_item_list(optarg, "mbuf-size",
+					MAX_SEGS_BUFFER_SPLIT, mb_sz, 0);
+				if (nb_segs <= 0)
 					rte_exit(EXIT_FAILURE,
-						 "mbuf-size should be > 0 and < 65536\n");
+						 "bad mbuf-size\n");
+				for (i = 0; i < nb_segs; i++) {
+					if (mb_sz[i] <= 0 || mb_sz[i] > 0xFFFF)
+						rte_exit(EXIT_FAILURE,
+							 "mbuf-size should be "
+							 "> 0 and < 65536\n");
+					mbuf_data_size[i] = (uint16_t) mb_sz[i];
+				}
+				mbuf_data_size_n = nb_segs;
 			}
 			if (!strcmp(lgopts[opt_idx].name, "total-num-mbufs")) {
 				n = atoi(optarg);
app/test-pmd/testpmd.c +35 -28

145 lines not on this page · raw message

app/test-pmd/testpmd.h +19 -5
index 9a29d7a..b42d710 100644
@@ -42,6 +42,13 @@
  */
 #define RTE_MAX_SEGS_PER_PKT 255 /**< nb_segs is a 8-bit unsigned char. */
 
+/*
+ * The maximum number of segments per packet is used to configure
+ * buffer split feature, also specifies the maximum amount of
+ * optional Rx pools to allocate mbufs to split.
+ */
+#define MAX_SEGS_BUFFER_SPLIT 8 /**< nb_segs is a 8-bit unsigned char. */
+
 #define MAX_PKT_BURST 512
 #define DEF_PKT_BURST 32
 
@@ -393,7 +400,9 @@ struct queue_stats_mappings {
 extern uint8_t dcb_config;
 extern uint8_t dcb_test;
 
-extern uint16_t mbuf_data_size; /**< Mbuf data space size. */
+extern uint32_t mbuf_data_size_n;
+extern uint16_t mbuf_data_size[MAX_SEGS_BUFFER_SPLIT];
+/**< Mbuf data space size. */
 extern uint32_t param_total_num_mbufs;
 
 extern uint16_t stats_period;
@@ -605,17 +614,22 @@ struct mplsoudp_decap_conf {
 
 /* Mbuf Pools */
 static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+		    int name_size, unsigned int idx)
 {
-	snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	if (!idx)
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	else
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+			 sock_id, idx);
 }
 
 static inline struct rte_mempool *
-mbuf_pool_find(unsigned int sock_id)
+mbuf_pool_find(unsigned int sock_id, unsigned int idx)
 {
 	char pool_name[RTE_MEMPOOL_NAMESIZE];
 
-	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name));
+	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name), idx);
 	return rte_mempool_lookup((const char *)pool_name);
 }
 
doc/guides/testpmd_app_ug/run_app.rst +5 -2
index ec085c2..1eb0a10 100644
@@ -107,9 +107,12 @@ The command line options are:
     Set the socket from which all memory is allocated in NUMA mode,
     where 0 <= N < number of sockets on the board.
 
-*   ``--mbuf-size=N``
+*   ``--mbuf-size=N[,N1[,...Nn]``
 
-    Set the data size of the mbufs used to N bytes, where N < 65536. The default value is 2048.
+    Set the data size of the mbufs used to N bytes, where N < 65536.
+    The default value is 2048. If multiple mbuf-size values are specified the
+    extra memory pools will be created for allocating mbufs to receive packets
+    with buffer splittling features.
 
 *   ``--total-num-mbufs=N``
 
Signature · 1 line
1.8.3.1
#123 lore raw

PATCH v8 3/6 app/testpmd: add buffer split offload configuration

In reply to Viacheslav Ovsiienko

This patch add support for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT providing per queue configuration for this offload.

---
 app/test-pmd/cmdline.c | 21 +++++++++++----------
 app/test-pmd/config.c  |  9 +++++++++
 2 files changed, 20 insertions(+), 10 deletions(-)
app/test-pmd/cmdline.c +11 -10
index a585cf0..fa71039 100644
@@ -883,16 +883,16 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"port config <port_id> rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"     Enable or disable a per port Rx offloading"
 			" on all Rx queues of a port\n\n"
 
 			"port (port_id) rxq (queue_id) rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"    Enable or disable a per queue Rx offloading"
 			" only on a specific Rx queue\n\n"
 
@@ -18417,7 +18417,8 @@ struct cmd_config_per_port_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc#rss_hash");
+			   "scatter#buffer_split#timestamp#security#"
+			   "keep_crc#rss_hash");
 cmdline_parse_token_string_t cmd_config_per_port_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_port_rx_offload_result,
@@ -18497,8 +18498,8 @@ struct cmd_config_per_port_rx_offload_result {
 	.help_str = "port config <port_id> rx_offload vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc|rss_hash "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc|rss_hash on|off",
 	.tokens = {
 		(void *)&cmd_config_per_port_rx_offload_result_port,
 		(void *)&cmd_config_per_port_rx_offload_result_config,
@@ -18547,7 +18548,7 @@ struct cmd_config_per_queue_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc");
+			   "scatter#buffer_split#timestamp#security#keep_crc");
 cmdline_parse_token_string_t cmd_config_per_queue_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_queue_rx_offload_result,
@@ -18603,8 +18604,8 @@ struct cmd_config_per_queue_rx_offload_result {
 		    "vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc on|off",
 	.tokens = {
 		(void *)&cmd_config_per_queue_rx_offload_result_port,
 		(void *)&cmd_config_per_queue_rx_offload_result_port_id,
app/test-pmd/config.c +9 -0
index 5f501f6..7126d91 100644
@@ -1092,6 +1092,15 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 			printf("off\n");
 	}
 
+	if (dev_info.rx_offload_capa & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+		printf("RX offload buffer split:       ");
+		if (ports[port_id].dev_conf.rxmode.offloads &
+		    RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT)
+			printf("on\n");
+		else
+			printf("off\n");
+	}
+
 	if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_VLAN_INSERT) {
 		printf("VLAN insert:                   ");
 		if (ports[port_id].dev_conf.txmode.offloads &
Signature · 1 line
1.8.3.1
#124 lore raw

PATCH v8 4/6 app/testpmd: add rxpkts commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxpkts=X[,Y]

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command:

testpmd> set rxpkts (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only). Optionally the multiple memory pools can be specified with --mbuf-size command line parameter and the mbufs to receive will be allocated sequentially from these extra memory pools.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 48 ++++++++++++++++++++++-
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  7 ++++
 app/test-pmd/testpmd.h                      | 11 +++++-
 doc/guides/testpmd_app_ug/run_app.rst       |  9 +++++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 21 +++++++++-
 7 files changed, 165 insertions(+), 7 deletions(-)
app/test-pmd/cmdline.c +58 -3
index fa71039..d8dba54 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxpkts (x[,y]*)\n"
+			"    Set the length of each segment to scatter"
+			" packets on receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set txpkts (x[,y]*)\n"
 			"    Set the length of each segment of TXONLY"
 			" and optionally CSUM packets.\n\n"
@@ -3889,6 +3895,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxpkts_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxpkts;
+	cmdline_fixed_string_t seg_lengths;
+};
+
+static void
+cmd_set_rxpkts_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxpkts_result *res;
+	unsigned int seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_lengths, "segment lengths",
+				  MAX_SEGS_BUFFER_SPLIT, seg_lengths, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_segments(seg_lengths, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxpkts_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxpkts_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 rxpkts, "rxpkts");
+cmdline_parse_token_string_t cmd_set_rxpkts_lengths =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 seg_lengths, NULL);
+
+cmdline_parse_inst_t cmd_set_rxpkts = {
+	.f = cmd_set_rxpkts_parsed,
+	.data = NULL,
+	.help_str = "set rxpkts <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxpkts_keyword,
+		(void *)&cmd_set_rxpkts_name,
+		(void *)&cmd_set_rxpkts_lengths,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF TXONLY PACKETS *** */
 
 struct cmd_set_txpkts_result {
@@ -7517,6 +7569,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxpkts"))
+		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
 		show_tx_pkt_segments();
 	else if (!strcmp(res->what, "txtimes"))
@@ -7529,12 +7583,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19807,6 +19861,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
 	(cmdline_parse_inst_t *)&cmd_set_txtimes,
app/test-pmd/config.c +46 -2

62 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index 4db4987..e4e3635 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
 	printf("  --txonly-multi-flow: generate multiple flows in txonly mode\n");
@@ -662,6 +663,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
 		{ "disable-link-check",		0, 0, 0 },
@@ -1272,6 +1274,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
+				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs;
+
+				nb_segs = parse_item_list
+						(optarg, "rxpkt segments",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_len, 0);
+				if (nb_segs > 0)
+					set_rx_pkt_segments(seg_len, nb_segs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxpkts\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "txpkts")) {
 				unsigned seg_lengths[RTE_MAX_SEGS_PER_PKT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +7 -0
index 7e6ef80..f88c1e2 100644
@@ -210,6 +210,13 @@ struct fwd_engine * fwd_engines[] = {
 uint8_t f_quit;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 uint16_t tx_pkt_length = TXONLY_DEF_PACKET_LEN; /**< TXONLY packet length. */
app/test-pmd/testpmd.h +10 -1
index b42d710..8e5ba6a 100644
@@ -420,6 +420,13 @@ struct queue_stats_mappings {
 extern struct rte_fdir_conf fdir_conf;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 #define TXONLY_DEF_PACKET_LEN 64
@@ -816,7 +823,9 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_record_core_cycles(uint8_t on_off);
 void set_record_burst_stats(uint8_t on_off);
 void set_verbose_level(uint16_t vb_level);
-void set_tx_pkt_segments(unsigned *seg_lengths, unsigned nb_segs);
+void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
+void show_rx_pkt_segments(void);
+void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
 void show_tx_pkt_times(void);
doc/guides/testpmd_app_ug/run_app.rst +9 -0
index 1eb0a10..463b76c 100644
@@ -361,6 +361,15 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxpkts=X[,Y]``
+
+    Set the length of segments to scatter packets on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+    Optionally the multiple memory pools can be specified with --mbuf-size
+    command line parameter and the mbufs to receive will be allocated
+    sequentially from these extra memory pools.
+
 *   ``--txpkts=X[,Y]``
 
     Set TX segment sizes or total packet length. Valid for ``tx-only``
doc/guides/testpmd_app_ug/testpmd_funcs.rst +20 -1
index 795c739..ff88762 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxpkts``: Packets to RX split configuration.
+
 * ``txpkts``: Packets to TX configuration.
 
 * ``txtimes``: Burst time pattern for Tx only mode.
@@ -774,6 +776,23 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxpkts
+~~~~~~~~~~
+
+Set the length of segments to scatter packets on receiving if split
+feature is engaged. Affects only the queues configured with split offloads
+(currently BUFFER_SPLIT is supported only). Optionally the multiple memory
+pools can be specified with --mbuf-size command line parameter and the mbufs
+to receive will be allocated sequentially from these extra memory pools (the
+mbuf for the first segment is allocated from the first pool, the second one
+from the second pool, and so on, if segment number is greater then pool's the
+mbuf for remaining segments will be allocated from the last valid pool).
+
+   testpmd> set rxpkts (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. Zero value
+means to use the corresponding memory pool data buffer size.
+
 set txpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#125 lore raw

PATCH v8 5/6 app/testpmd: add rxoffs commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxoffs=X[,Y]

Sets the offsets of packet segments from the beginning of the receiving buffer if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command, providing the same:

testpmd> set rxoffs (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 44 +++++++++++++++++++++
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  2 +
 app/test-pmd/testpmd.h                      |  4 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  6 +++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 17 +++++++-
 7 files changed, 145 insertions(+), 4 deletions(-)
app/test-pmd/cmdline.c +58 -3
index d8dba54..7182bba 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxoffs (x[,y]*)\n"
+			"    Set the offset of each packet segment on"
+			" receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set rxpkts (x[,y]*)\n"
 			"    Set the length of each segment to scatter"
 			" packets on receiving if split feature is engaged."
@@ -3895,6 +3901,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT OFFSETS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxoffs_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxoffs;
+	cmdline_fixed_string_t seg_offsets;
+};
+
+static void
+cmd_set_rxoffs_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxoffs_result *res;
+	unsigned int seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_offsets, "segment offsets",
+				  MAX_SEGS_BUFFER_SPLIT, seg_offsets, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_offsets(seg_offsets, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxoffs_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxoffs_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 rxoffs, "rxoffs");
+cmdline_parse_token_string_t cmd_set_rxoffs_offsets =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 seg_offsets, NULL);
+
+cmdline_parse_inst_t cmd_set_rxoffs = {
+	.f = cmd_set_rxoffs_parsed,
+	.data = NULL,
+	.help_str = "set rxoffs <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxoffs_keyword,
+		(void *)&cmd_set_rxoffs_name,
+		(void *)&cmd_set_rxoffs_offsets,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
 
 struct cmd_set_rxpkts_result {
@@ -7569,6 +7621,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxoffs"))
+		show_rx_pkt_offsets();
 	else if (!strcmp(res->what, "rxpkts"))
 		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
@@ -7583,12 +7637,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxoffs#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19861,6 +19915,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxoffs,
 	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
app/test-pmd/config.c +44 -0

50 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index e4e3635..2298ba5 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxoffs=X[,Y]*: set RX segment offsets for split.\n");
 	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
@@ -663,6 +664,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxoffs",			1, 0, 0 },
 		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
@@ -1274,6 +1276,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxoffs")) {
+				unsigned int seg_off[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_offs;
+
+				nb_offs = parse_item_list
+						(optarg, "rxpkt offsets",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_off, 0);
+				if (nb_offs > 0)
+					set_rx_pkt_offsets(seg_off, nb_offs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxoffs\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
 				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +2 -0
index f88c1e2..580178d 100644
@@ -215,6 +215,8 @@ struct fwd_engine * fwd_engines[] = {
  */
 uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
app/test-pmd/testpmd.h +4 -0
index 8e5ba6a..fc56b60 100644
@@ -425,6 +425,8 @@ struct queue_stats_mappings {
  */
 extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+extern uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
@@ -825,6 +827,8 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_verbose_level(uint16_t vb_level);
 void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_rx_pkt_segments(void);
+void set_rx_pkt_offsets(unsigned int *seg_offsets, unsigned int nb_offs);
+void show_rx_pkt_offsets(void);
 void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
doc/guides/testpmd_app_ug/run_app.rst +6 -0
index 463b76c..9b0a84a 100644
@@ -361,6 +361,12 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxoffs=X[,Y]``
+
+    Set the offsets of packet segments on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+
 *   ``--rxpkts=X[,Y]``
 
     Set the length of segments to scatter packets on receiving if split
doc/guides/testpmd_app_ug/testpmd_funcs.rst +16 -1
index ff88762..c99d887 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxoffs``: Packet offsets for RX split.
+
 * ``rxpkts``: Packets to RX split configuration.
 
 * ``txpkts``: Packets to TX configuration.
@@ -776,6 +778,19 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxoffs
+~~~~~~~~~~
+
+Set the offsets of segments relating to the data buffer beginning on receiving
+if split feature is engaged. Affects only the queues configured with split
+offloads (currently BUFFER_SPLIT is supported only).
+
+   testpmd> set rxoffs (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. If the list
+of offsets is shorter than the list of segments the zero offsets will be used
+for the remaining segments.
+
 set rxpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#126 lore raw

PATCH v8 6/6 app/testpmd: add extended Rx queue setup

In reply to Viacheslav Ovsiienko

If Rx queue is configured with split feature the extended setup with specified segment sizes and pool will be performed.

---
 app/test-pmd/cmdline.c | 12 ++++++------
 app/test-pmd/testpmd.c | 48 ++++++++++++++++++++++++++++++++++++++++++++++--
 app/test-pmd/testpmd.h |  5 +++++
 3 files changed, 57 insertions(+), 8 deletions(-)
app/test-pmd/cmdline.c +6 -6
index 7182bba..204221f 100644
@@ -2927,12 +2927,12 @@ struct cmd_setup_rxtx_queue {
 				rxring_numa[res->portid]);
 			return;
 		}
-		ret = rte_eth_rx_queue_setup(res->portid,
-					     res->qid,
-					     port->nb_rx_desc[res->qid],
-					     socket_id,
-					     &port->rx_conf[res->qid],
-					     mp);
+		ret = rx_queue_setup(res->portid,
+				     res->qid,
+				     port->nb_rx_desc[res->qid],
+				     socket_id,
+				     &port->rx_conf[res->qid],
+				     mp);
 		if (ret)
 			printf("Failed to setup RX queue\n");
 	} else {
app/test-pmd/testpmd.c +46 -2

66 lines not on this page · raw message

app/test-pmd/testpmd.h +5 -0
index fc56b60..af654ea 100644
@@ -876,6 +876,11 @@ void port_rss_reta_info(portid_t port_id,
 
 void set_vf_traffic(portid_t port_id, uint8_t is_rx, uint16_t vf, uint8_t on);
 
+int
+rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
+	       uint16_t nb_rx_desc, unsigned int socket_id,
+	       struct rte_eth_rxconf *rx_conf, struct rte_mempool *mp);
+
 int set_queue_rate_limit(portid_t port_id, uint16_t queue_idx, uint16_t rate);
 int set_vf_rate_limit(portid_t port_id, uint16_t vf, uint16_t rate,
 				uint64_t q_msk);
Signature · 1 line
1.8.3.1
#127 lore raw

v9 cover ethdev: introduce Rx buffer split

In reply to Slava Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---

v1: http://patches.dpdk.org/patch/79594/
v2: http://patches.dpdk.org/patch/79893/
- add feature support to mlx5 PMD

v3: http://patches.dpdk.org/patch/80389/
- rte_eth_rx_queue_setup_ex is renamed to rte_eth_rxseg_queue_setup
- DEV_RX_OFFLOAD_BUFFER_SPLIT is renamed to RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT
- commit message update
- documentaion provided
- release notes update
- minor bug fixes in testpmd related part

v4: http://patches.dpdk.org/patch/80401/
- common part of rx_queue_setup/rxseg_queue_setup

v5: http://patches.dpdk.org/patch/80609/
- refactored to approach of providing split configuration in the rte_eth_rxconf structure instead of introducing the new API routine
- added support for rxoffs command to testpmd to provide segment offsets for complete testing of split configurations
- patchset is split into two parts - PMD part will be presented as separate series

v6: http://patches.dpdk.org/patch/80785/
- wordy comments rephrased
- typos fixed
- rte_eth_rx_queue_setup configuration check isolated for two main options
- the rest of comments addressed

v7: http://patches.dpdk.org/patch/80978/
- description as union of feature dedicated segment split description
- the split limitations reported in the rte_eth_dev_info->rx_seg_capa
- comments made less wordy

v8: http://patches.dpdk.org/patch/81038/
- bitfield instead of uint8;
- named union to avoid warinings;
- typos fixed

v9: - simplified rte_eth_rxseg union
- moved out wordy comment from rte_eth_rxconf
- rte_eth_rxseg_capa uint32 bitfield and naming
- added experimental tags

---
Viacheslav Ovsiienko (6):
ethdev: introduce Rx buffer split app/testpmd: add multiple pools per core creation app/testpmd: add buffer split offload configuration app/testpmd: add rxpkts commands and parameters app/testpmd: add rxoffs commands and parameters app/testpmd: add extended Rx queue setup

 app/test-pmd/bpf_cmd.c                      |   4 +-
 app/test-pmd/cmdline.c                      | 151 +++++++++++++++++++++++----
 app/test-pmd/config.c                       | 107 ++++++++++++++++++-
 app/test-pmd/parameters.c                   |  54 ++++++++--
 app/test-pmd/testpmd.c                      | 121 ++++++++++++++++------
 app/test-pmd/testpmd.h                      |  44 ++++++--
 doc/guides/nics/features.rst                |  15 +++
 doc/guides/rel_notes/deprecation.rst        |   5 -
 doc/guides/rel_notes/release_20_11.rst      |   9 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  22 +++-
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  36 ++++++-
 lib/librte_ethdev/rte_ethdev.c              | 153 +++++++++++++++++++++++-----
 lib/librte_ethdev/rte_ethdev.h              |  89 +++++++++++++++-
 13 files changed, 707 insertions(+), 103 deletions(-)
Signature · 1 line
1.8.3.1
#128 lore raw

PATCH v9 1/6 ethdev: introduce Rx buffer split

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---
 doc/guides/nics/features.rst           |  15 ++++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 ++
 lib/librte_ethdev/rte_ethdev.c         | 153 +++++++++++++++++++++++++++------
 lib/librte_ethdev/rte_ethdev.h         |  89 ++++++++++++++++++-
 5 files changed, 240 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +15 -0
index dd8c955..832ea3b 100644
@@ -185,6 +185,21 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[provides]   eth_dev_ops**: ``rxq_info_get:buffer_split``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index d1f5ed3..dde464b 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index f8686a5..e12cf96 100644
@@ -89,6 +89,12 @@ New Features
     as flow action.
   * Added new flow APIs to create/update/destroy/query shared action.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -321,6 +327,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +128 -25

187 lines not on this page · raw message

lib/librte_ethdev/rte_ethdev.h +88 -1

133 lines not on this page · raw message

Signature · 1 line
1.8.3.1
#129 lore raw
Quoted 101 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf.
This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

     seg0 - pool0, len0=14B, off0=2
     seg1 - pool1, len1=20B, off1=128B
     seg2 - pool2, len2=20B, off2=0B
     seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

<...>

doc/guides/nics/features.rst Supports receiving segmented mbufs. · line 199

Quoted 12 lines

+.. _nic_features_buffer_split:

doc/guides/nics/features.rst +11 -0
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[provides]   eth_dev_ops**: ``rxq_info_get:buffer_split``.

Previously you mentioned this is because 'rxq_info_get()' can provide buffer_split information, but with current implementation it doesn't and there is no filed in the struct to report such.

I suggest either add it now, while you can :) [with a techboard approval], or remove above documentation of it.

<...>

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info { · line 1450

Quoted 11 lines

/**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */

lib/librte_ethdev/rte_ethdev.h +8 -0
+__rte_experimental
+struct rte_eth_rxseg_capa {
+	__extension__
+	uint32_t max_nseg:16; /**< Maximum amount of segments to split. */
+	uint32_t multi_pools:1; /**< Supports receiving to multiple pools.*/
+	uint32_t offset_allowed:1; /**< Supports buffer offsets. */
+	uint32_t offset_align_log2:4; /**< Required offset alignment. */
+};

Now we are fiddling details, but,

I am not fun of the bitfields [1], but I assumed Thomas' request was to enable expanding capabilities later without breaking the ABI, which makes senses and suits to this kind of capability struct, if this is correct why made the 'max_nseg' a bitfield too?

Why not,
uint16_t max_nseg;
uint16_t multi_pools:1
uint16_t offset_allowed:1;
uint16_t offset_align_log2:4;
< This still leaves 10 bits to expand without ABI break>

[1]
unles very space critical use case, otherwise they just add more code to extract the same value, and not as simple as a simple variable :)

#130 lore raw
Quoted 17 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, October 16, 2020 14:21
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH v9 1/6] ethdev: introduce Rx buffer split

Quoted 6 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

[snip]

Quoted 9 lines

Review

Quoted 2 lines
(quoted diff) +2 -0
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**:

``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.

doc/guides/nics/features.rst Supports receiving segmented mbufs.

Quoted 3 lines
doc/guides/nics/features.rst +3 -0
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**:

``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.

Quoted 1 line

+* **[provides] eth_dev_ops**: ``rxq_info_get:buffer_split``.

Quoted 9 lines

Previously you mentioned this is because 'rxq_info_get()' can provide buffer_split information, but with current implementation it doesn't and there is no filed in the struct to report such.

I suggest either add it now, while you can :) [with a techboard approval], or remove above documentation of it.

<...>

Mmm, I messed up with rx_burst_mode_get(). Will fix, thanks.

Quoted 25 lines

lib/librte_ethdev/rte_ethdev.h struct rte_eth_switch_info {

Quoted 10 lines

/**
+ * Ethernet device Rx buffer segmentation capabilities.
+ */

lib/librte_ethdev/rte_ethdev.h +7 -0
+__rte_experimental
+struct rte_eth_rxseg_capa {
+	__extension__
+	uint32_t max_nseg:16; /**< Maximum amount of segments to split. */
+	uint32_t multi_pools:1; /**< Supports receiving to multiple pools.*/
+	uint32_t offset_allowed:1; /**< Supports buffer offsets. */
+	uint32_t offset_align_log2:4; /**< Required offset alignment. */ };

Now we are fiddling details, but,

I am not fun of the bitfields [1], but I assumed Thomas' request was to enable expanding capabilities later without breaking the ABI, which makes senses and suits to this kind of capability struct, if this is correct why made the 'max_nseg' a bitfield too?

Why not,
uint16_t max_nseg;
uint16_t multi_pools:1
uint16_t offset_allowed:1;
uint16_t offset_align_log2:4;
< This still leaves 10 bits to expand without ABI break>

Quoted 3 lines

[1]
unles very space critical use case, otherwise they just add more code to extract the same value, and not as simple as a simple variable :)

It seems not to be the case, there is the listing of the rte_eth_rx_queue_check_split():

8963 4b67 440FB784 movzwl 188(%rsp),%r8d ; [SO] max_nseg is fetched as regular uint16_t
8963 24BC0000
8963 00
8964 4b70 664539C1 cmpw %r8w,%r9w
8965 4b74 0F87A402 ja .L1749
8965 0000

I would prefer to keep uint32_t - it is more generic, IMO.

With best regards, Slava

#131 lore raw

PATCH v9 2/6 app/testpmd: add multiple pools per core creation

In reply to Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

---
 app/test-pmd/bpf_cmd.c                |  4 +--
 app/test-pmd/cmdline.c                |  2 +-
 app/test-pmd/config.c                 |  6 ++--
 app/test-pmd/parameters.c             | 24 +++++++++----
 app/test-pmd/testpmd.c                | 63 +++++++++++++++++++----------------
 app/test-pmd/testpmd.h                | 24 ++++++++++---
 doc/guides/testpmd_app_ug/run_app.rst |  7 ++--
 7 files changed, 83 insertions(+), 47 deletions(-)
app/test-pmd/bpf_cmd.c +2 -2
index 16e3c3b..0a1a178 100644
@@ -69,7 +69,7 @@ struct cmd_bpf_ld_result {
 
 	*flags = RTE_BPF_ETH_F_NONE;
 	arg->type = RTE_BPF_ARG_PTR;
-	arg->size = mbuf_data_size;
+	arg->size = mbuf_data_size[0];
 
 	for (i = 0; str[i] != 0; i++) {
 		v = toupper(str[i]);
@@ -78,7 +78,7 @@ struct cmd_bpf_ld_result {
 		else if (v == 'M') {
 			arg->type = RTE_BPF_ARG_PTR_MBUF;
 			arg->size = sizeof(struct rte_mbuf);
-			arg->buf_size = mbuf_data_size;
+			arg->buf_size = mbuf_data_size[0];
 		} else if (v == '-')
 			continue;
 		else
app/test-pmd/cmdline.c +1 -1
index 078f063..19b1896 100644
@@ -2925,7 +2925,7 @@ struct cmd_setup_rxtx_queue {
 		if (!numa_support || socket_id == NUMA_NO_CONFIG)
 			socket_id = port->socket_id;
 
-		mp = mbuf_pool_find(socket_id);
+		mp = mbuf_pool_find(socket_id, 0);
 		if (mp == NULL) {
 			printf("Failed to setup RX queue: "
 				"No mempool allocation"
app/test-pmd/config.c +3 -3
index 2c00b55..c554dd0 100644
@@ -690,7 +690,7 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 	printf("\nConnect to socket: %u", port->socket_id);
 
 	if (port_numa[port_id] != NUMA_NO_CONFIG) {
-		mp = mbuf_pool_find(port_numa[port_id]);
+		mp = mbuf_pool_find(port_numa[port_id], 0);
 		if (mp)
 			printf("\nmemory allocation on the socket: %d",
 							port_numa[port_id]);
@@ -3580,9 +3580,9 @@ struct igb_ring_desc_16_bytes {
 	 */
 	tx_pkt_len = 0;
 	for (i = 0; i < nb_segs; i++) {
-		if (seg_lengths[i] > (unsigned) mbuf_data_size) {
+		if (seg_lengths[i] > mbuf_data_size[0]) {
 			printf("length[%u]=%u > mbuf_data_size=%u - give up\n",
-			       i, seg_lengths[i], (unsigned) mbuf_data_size);
+			       i, seg_lengths[i], mbuf_data_size[0]);
 			return;
 		}
 		tx_pkt_len = (uint16_t)(tx_pkt_len + seg_lengths[i]);
app/test-pmd/parameters.c +18 -6
index 15ce8c1..4db4987 100644
@@ -106,7 +106,9 @@
 	       "(flag: 1 for RX; 2 for TX; 3 for RX and TX).\n");
 	printf("  --socket-num=N: set socket from which all memory is allocated "
 	       "in NUMA mode.\n");
-	printf("  --mbuf-size=N: set the data size of mbuf to N bytes.\n");
+	printf("  --mbuf-size=N,[N1[,..Nn]: set the data size of mbuf to "
+	       "N bytes. If multiple numbers are specified the extra pools "
+	       "will be created to receive with packet split features\n");
 	printf("  --total-num-mbufs=N: set the number of mbufs to be allocated "
 	       "in mbuf pools.\n");
 	printf("  --max-pkt-len=N: set the maximum size of packet to N bytes.\n");
@@ -892,12 +894,22 @@
 				}
 			}
 			if (!strcmp(lgopts[opt_idx].name, "mbuf-size")) {
-				n = atoi(optarg);
-				if (n > 0 && n <= 0xFFFF)
-					mbuf_data_size = (uint16_t) n;
-				else
+				unsigned int mb_sz[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs, i;
+
+				nb_segs = parse_item_list(optarg, "mbuf-size",
+					MAX_SEGS_BUFFER_SPLIT, mb_sz, 0);
+				if (nb_segs <= 0)
 					rte_exit(EXIT_FAILURE,
-						 "mbuf-size should be > 0 and < 65536\n");
+						 "bad mbuf-size\n");
+				for (i = 0; i < nb_segs; i++) {
+					if (mb_sz[i] <= 0 || mb_sz[i] > 0xFFFF)
+						rte_exit(EXIT_FAILURE,
+							 "mbuf-size should be "
+							 "> 0 and < 65536\n");
+					mbuf_data_size[i] = (uint16_t) mb_sz[i];
+				}
+				mbuf_data_size_n = nb_segs;
 			}
 			if (!strcmp(lgopts[opt_idx].name, "total-num-mbufs")) {
 				n = atoi(optarg);
app/test-pmd/testpmd.c +35 -28

145 lines not on this page · raw message

app/test-pmd/testpmd.h +19 -5
index a4dfd4f..0975305 100644
@@ -42,6 +42,13 @@
  */
 #define RTE_MAX_SEGS_PER_PKT 255 /**< nb_segs is a 8-bit unsigned char. */
 
+/*
+ * The maximum number of segments per packet is used to configure
+ * buffer split feature, also specifies the maximum amount of
+ * optional Rx pools to allocate mbufs to split.
+ */
+#define MAX_SEGS_BUFFER_SPLIT 8 /**< nb_segs is a 8-bit unsigned char. */
+
 #define MAX_PKT_BURST 512
 #define DEF_PKT_BURST 32
 
@@ -403,7 +410,9 @@ struct queue_stats_mappings {
 extern uint8_t dcb_config;
 extern uint8_t dcb_test;
 
-extern uint16_t mbuf_data_size; /**< Mbuf data space size. */
+extern uint32_t mbuf_data_size_n;
+extern uint16_t mbuf_data_size[MAX_SEGS_BUFFER_SPLIT];
+/**< Mbuf data space size. */
 extern uint32_t param_total_num_mbufs;
 
 extern uint16_t stats_period;
@@ -615,17 +624,22 @@ struct mplsoudp_decap_conf {
 
 /* Mbuf Pools */
 static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+		    int name_size, unsigned int idx)
 {
-	snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	if (!idx)
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	else
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+			 sock_id, idx);
 }
 
 static inline struct rte_mempool *
-mbuf_pool_find(unsigned int sock_id)
+mbuf_pool_find(unsigned int sock_id, unsigned int idx)
 {
 	char pool_name[RTE_MEMPOOL_NAMESIZE];
 
-	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name));
+	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name), idx);
 	return rte_mempool_lookup((const char *)pool_name);
 }
 
doc/guides/testpmd_app_ug/run_app.rst +5 -2
index ec085c2..1eb0a10 100644
@@ -107,9 +107,12 @@ The command line options are:
     Set the socket from which all memory is allocated in NUMA mode,
     where 0 <= N < number of sockets on the board.
 
-*   ``--mbuf-size=N``
+*   ``--mbuf-size=N[,N1[,...Nn]``
 
-    Set the data size of the mbufs used to N bytes, where N < 65536. The default value is 2048.
+    Set the data size of the mbufs used to N bytes, where N < 65536.
+    The default value is 2048. If multiple mbuf-size values are specified the
+    extra memory pools will be created for allocating mbufs to receive packets
+    with buffer splittling features.
 
 *   ``--total-num-mbufs=N``
 
Signature · 1 line
1.8.3.1
#132 lore raw

PATCH v9 3/6 app/testpmd: add buffer split offload configuration

In reply to Viacheslav Ovsiienko

This patch add support for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT providing per queue configuration for this offload.

---
 app/test-pmd/cmdline.c | 21 +++++++++++----------
 app/test-pmd/config.c  |  9 +++++++++
 2 files changed, 20 insertions(+), 10 deletions(-)
app/test-pmd/cmdline.c +11 -10
index 19b1896..0ef87c9 100644
@@ -883,16 +883,16 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"port config <port_id> rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"     Enable or disable a per port Rx offloading"
 			" on all Rx queues of a port\n\n"
 
 			"port (port_id) rxq (queue_id) rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"    Enable or disable a per queue Rx offloading"
 			" only on a specific Rx queue\n\n"
 
@@ -18435,7 +18435,8 @@ struct cmd_config_per_port_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc#rss_hash");
+			   "scatter#buffer_split#timestamp#security#"
+			   "keep_crc#rss_hash");
 cmdline_parse_token_string_t cmd_config_per_port_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_port_rx_offload_result,
@@ -18515,8 +18516,8 @@ struct cmd_config_per_port_rx_offload_result {
 	.help_str = "port config <port_id> rx_offload vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc|rss_hash "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc|rss_hash on|off",
 	.tokens = {
 		(void *)&cmd_config_per_port_rx_offload_result_port,
 		(void *)&cmd_config_per_port_rx_offload_result_config,
@@ -18565,7 +18566,7 @@ struct cmd_config_per_queue_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc");
+			   "scatter#buffer_split#timestamp#security#keep_crc");
 cmdline_parse_token_string_t cmd_config_per_queue_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_queue_rx_offload_result,
@@ -18621,8 +18622,8 @@ struct cmd_config_per_queue_rx_offload_result {
 		    "vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc on|off",
 	.tokens = {
 		(void *)&cmd_config_per_queue_rx_offload_result_port,
 		(void *)&cmd_config_per_queue_rx_offload_result_port_id,
app/test-pmd/config.c +9 -0
index c554dd0..0f041e1 100644
@@ -1092,6 +1092,15 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 			printf("off\n");
 	}
 
+	if (dev_info.rx_offload_capa & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+		printf("RX offload buffer split:       ");
+		if (ports[port_id].dev_conf.rxmode.offloads &
+		    RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT)
+			printf("on\n");
+		else
+			printf("off\n");
+	}
+
 	if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_VLAN_INSERT) {
 		printf("VLAN insert:                   ");
 		if (ports[port_id].dev_conf.txmode.offloads &
Signature · 1 line
1.8.3.1
#133 lore raw

PATCH v9 4/6 app/testpmd: add rxpkts commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxpkts=X[,Y]

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command:

testpmd> set rxpkts (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only). Optionally the multiple memory pools can be specified with --mbuf-size command line parameter and the mbufs to receive will be allocated sequentially from these extra memory pools.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 48 ++++++++++++++++++++++-
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  7 ++++
 app/test-pmd/testpmd.h                      | 11 +++++-
 doc/guides/testpmd_app_ug/run_app.rst       |  9 +++++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 21 +++++++++-
 7 files changed, 165 insertions(+), 7 deletions(-)
app/test-pmd/cmdline.c +58 -3
index 0ef87c9..a195e37 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxpkts (x[,y]*)\n"
+			"    Set the length of each segment to scatter"
+			" packets on receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set txpkts (x[,y]*)\n"
 			"    Set the length of each segment of TXONLY"
 			" and optionally CSUM packets.\n\n"
@@ -3907,6 +3913,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxpkts_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxpkts;
+	cmdline_fixed_string_t seg_lengths;
+};
+
+static void
+cmd_set_rxpkts_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxpkts_result *res;
+	unsigned int seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_lengths, "segment lengths",
+				  MAX_SEGS_BUFFER_SPLIT, seg_lengths, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_segments(seg_lengths, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxpkts_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxpkts_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 rxpkts, "rxpkts");
+cmdline_parse_token_string_t cmd_set_rxpkts_lengths =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 seg_lengths, NULL);
+
+cmdline_parse_inst_t cmd_set_rxpkts = {
+	.f = cmd_set_rxpkts_parsed,
+	.data = NULL,
+	.help_str = "set rxpkts <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxpkts_keyword,
+		(void *)&cmd_set_rxpkts_name,
+		(void *)&cmd_set_rxpkts_lengths,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF TXONLY PACKETS *** */
 
 struct cmd_set_txpkts_result {
@@ -7535,6 +7587,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxpkts"))
+		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
 		show_tx_pkt_segments();
 	else if (!strcmp(res->what, "txtimes"))
@@ -7547,12 +7601,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19825,6 +19879,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
 	(cmdline_parse_inst_t *)&cmd_set_txtimes,
app/test-pmd/config.c +46 -2

62 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index 4db4987..e4e3635 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
 	printf("  --txonly-multi-flow: generate multiple flows in txonly mode\n");
@@ -662,6 +663,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
 		{ "disable-link-check",		0, 0, 0 },
@@ -1272,6 +1274,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
+				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs;
+
+				nb_segs = parse_item_list
+						(optarg, "rxpkt segments",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_len, 0);
+				if (nb_segs > 0)
+					set_rx_pkt_segments(seg_len, nb_segs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxpkts\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "txpkts")) {
 				unsigned seg_lengths[RTE_MAX_SEGS_PER_PKT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +7 -0
index 7e6ef80..f88c1e2 100644
@@ -210,6 +210,13 @@ struct fwd_engine * fwd_engines[] = {
 uint8_t f_quit;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 uint16_t tx_pkt_length = TXONLY_DEF_PACKET_LEN; /**< TXONLY packet length. */
app/test-pmd/testpmd.h +10 -1
index 0975305..c5c942a 100644
@@ -430,6 +430,13 @@ struct queue_stats_mappings {
 extern struct rte_fdir_conf fdir_conf;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 #define TXONLY_DEF_PACKET_LEN 64
@@ -836,7 +843,9 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_record_core_cycles(uint8_t on_off);
 void set_record_burst_stats(uint8_t on_off);
 void set_verbose_level(uint16_t vb_level);
-void set_tx_pkt_segments(unsigned *seg_lengths, unsigned nb_segs);
+void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
+void show_rx_pkt_segments(void);
+void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
 void show_tx_pkt_times(void);
doc/guides/testpmd_app_ug/run_app.rst +9 -0
index 1eb0a10..463b76c 100644
@@ -361,6 +361,15 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxpkts=X[,Y]``
+
+    Set the length of segments to scatter packets on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+    Optionally the multiple memory pools can be specified with --mbuf-size
+    command line parameter and the mbufs to receive will be allocated
+    sequentially from these extra memory pools.
+
 *   ``--txpkts=X[,Y]``
 
     Set TX segment sizes or total packet length. Valid for ``tx-only``
doc/guides/testpmd_app_ug/testpmd_funcs.rst +20 -1
index 43c0ea0..b9fad3a 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxpkts``: Packets to RX split configuration.
+
 * ``txpkts``: Packets to TX configuration.
 
 * ``txtimes``: Burst time pattern for Tx only mode.
@@ -774,6 +776,23 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxpkts
+~~~~~~~~~~
+
+Set the length of segments to scatter packets on receiving if split
+feature is engaged. Affects only the queues configured with split offloads
+(currently BUFFER_SPLIT is supported only). Optionally the multiple memory
+pools can be specified with --mbuf-size command line parameter and the mbufs
+to receive will be allocated sequentially from these extra memory pools (the
+mbuf for the first segment is allocated from the first pool, the second one
+from the second pool, and so on, if segment number is greater then pool's the
+mbuf for remaining segments will be allocated from the last valid pool).
+
+   testpmd> set rxpkts (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. Zero value
+means to use the corresponding memory pool data buffer size.
+
 set txpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#134 lore raw

PATCH v9 5/6 app/testpmd: add rxoffs commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxoffs=X[,Y]

Sets the offsets of packet segments from the beginning of the receiving buffer if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command, providing the same:

testpmd> set rxoffs (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 44 +++++++++++++++++++++
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  2 +
 app/test-pmd/testpmd.h                      |  4 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  6 +++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 17 +++++++-
 7 files changed, 145 insertions(+), 4 deletions(-)
app/test-pmd/cmdline.c +58 -3
index a195e37..e09df36 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxoffs (x[,y]*)\n"
+			"    Set the offset of each packet segment on"
+			" receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set rxpkts (x[,y]*)\n"
 			"    Set the length of each segment to scatter"
 			" packets on receiving if split feature is engaged."
@@ -3913,6 +3919,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT OFFSETS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxoffs_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxoffs;
+	cmdline_fixed_string_t seg_offsets;
+};
+
+static void
+cmd_set_rxoffs_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxoffs_result *res;
+	unsigned int seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_offsets, "segment offsets",
+				  MAX_SEGS_BUFFER_SPLIT, seg_offsets, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_offsets(seg_offsets, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxoffs_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxoffs_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 rxoffs, "rxoffs");
+cmdline_parse_token_string_t cmd_set_rxoffs_offsets =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 seg_offsets, NULL);
+
+cmdline_parse_inst_t cmd_set_rxoffs = {
+	.f = cmd_set_rxoffs_parsed,
+	.data = NULL,
+	.help_str = "set rxoffs <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxoffs_keyword,
+		(void *)&cmd_set_rxoffs_name,
+		(void *)&cmd_set_rxoffs_offsets,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
 
 struct cmd_set_rxpkts_result {
@@ -7587,6 +7639,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxoffs"))
+		show_rx_pkt_offsets();
 	else if (!strcmp(res->what, "rxpkts"))
 		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
@@ -7601,12 +7655,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxoffs#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19879,6 +19933,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxoffs,
 	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
app/test-pmd/config.c +44 -0

50 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index e4e3635..2298ba5 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxoffs=X[,Y]*: set RX segment offsets for split.\n");
 	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
@@ -663,6 +664,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxoffs",			1, 0, 0 },
 		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
@@ -1274,6 +1276,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxoffs")) {
+				unsigned int seg_off[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_offs;
+
+				nb_offs = parse_item_list
+						(optarg, "rxpkt offsets",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_off, 0);
+				if (nb_offs > 0)
+					set_rx_pkt_offsets(seg_off, nb_offs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxoffs\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
 				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +2 -0
index f88c1e2..580178d 100644
@@ -215,6 +215,8 @@ struct fwd_engine * fwd_engines[] = {
  */
 uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
app/test-pmd/testpmd.h +4 -0
index c5c942a..dc37148 100644
@@ -435,6 +435,8 @@ struct queue_stats_mappings {
  */
 extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+extern uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
@@ -845,6 +847,8 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_verbose_level(uint16_t vb_level);
 void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_rx_pkt_segments(void);
+void set_rx_pkt_offsets(unsigned int *seg_offsets, unsigned int nb_offs);
+void show_rx_pkt_offsets(void);
 void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
doc/guides/testpmd_app_ug/run_app.rst +6 -0
index 463b76c..9b0a84a 100644
@@ -361,6 +361,12 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxoffs=X[,Y]``
+
+    Set the offsets of packet segments on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+
 *   ``--rxpkts=X[,Y]``
 
     Set the length of segments to scatter packets on receiving if split
doc/guides/testpmd_app_ug/testpmd_funcs.rst +16 -1
index b9fad3a..e73530e 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxoffs``: Packet offsets for RX split.
+
 * ``rxpkts``: Packets to RX split configuration.
 
 * ``txpkts``: Packets to TX configuration.
@@ -776,6 +778,19 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxoffs
+~~~~~~~~~~
+
+Set the offsets of segments relating to the data buffer beginning on receiving
+if split feature is engaged. Affects only the queues configured with split
+offloads (currently BUFFER_SPLIT is supported only).
+
+   testpmd> set rxoffs (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. If the list
+of offsets is shorter than the list of segments the zero offsets will be used
+for the remaining segments.
+
 set rxpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#135 lore raw

PATCH v9 6/6 app/testpmd: add extended Rx queue setup

In reply to Viacheslav Ovsiienko

If Rx queue is configured with split feature the extended setup with specified segment sizes and pool will be performed.

---
 app/test-pmd/cmdline.c | 12 ++++++------
 app/test-pmd/testpmd.c | 49 +++++++++++++++++++++++++++++++++++++++++++++++--
 app/test-pmd/testpmd.h |  5 +++++
 3 files changed, 58 insertions(+), 8 deletions(-)
app/test-pmd/cmdline.c +6 -6
index e09df36..07ee4e4 100644
@@ -2945,12 +2945,12 @@ struct cmd_setup_rxtx_queue {
 				rxring_numa[res->portid]);
 			return;
 		}
-		ret = rte_eth_rx_queue_setup(res->portid,
-					     res->qid,
-					     port->nb_rx_desc[res->qid],
-					     socket_id,
-					     &port->rx_conf[res->qid],
-					     mp);
+		ret = rx_queue_setup(res->portid,
+				     res->qid,
+				     port->nb_rx_desc[res->qid],
+				     socket_id,
+				     &port->rx_conf[res->qid],
+				     mp);
 		if (ret)
 			printf("Failed to setup RX queue\n");
 	} else {
app/test-pmd/testpmd.c +47 -2

67 lines not on this page · raw message

app/test-pmd/testpmd.h +5 -0
index dc37148..c7b2cb4 100644
@@ -896,6 +896,11 @@ void port_rss_reta_info(portid_t port_id,
 
 void set_vf_traffic(portid_t port_id, uint8_t is_rx, uint16_t vf, uint8_t on);
 
+int
+rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
+	       uint16_t nb_rx_desc, unsigned int socket_id,
+	       struct rte_eth_rxconf *rx_conf, struct rte_mempool *mp);
+
 int set_queue_rate_limit(portid_t port_id, uint16_t queue_idx, uint16_t rate);
 int set_vf_rate_limit(portid_t port_id, uint16_t vf, uint16_t rate,
 				uint64_t q_msk);
Signature · 1 line
1.8.3.1
#136 lore raw

v10 cover ethdev: introduce Rx buffer split

In reply to Slava Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---

v1: http://patches.dpdk.org/patch/79594/
v2: http://patches.dpdk.org/patch/79893/
- add feature support to mlx5 PMD

v3: http://patches.dpdk.org/patch/80389/
- rte_eth_rx_queue_setup_ex is renamed to rte_eth_rxseg_queue_setup
- DEV_RX_OFFLOAD_BUFFER_SPLIT is renamed to RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT
- commit message update
- documentaion provided
- release notes update
- minor bug fixes in testpmd related part

v4: http://patches.dpdk.org/patch/80401/
- common part of rx_queue_setup/rxseg_queue_setup

v5: http://patches.dpdk.org/patch/80609/
- refactored to approach of providing split configuration in the rte_eth_rxconf structure instead of introducing the new API routine
- added support for rxoffs command to testpmd to provide segment offsets for complete testing of split configurations
- patchset is split into two parts - PMD part will be presented as separate series

v6: http://patches.dpdk.org/patch/80785/
- wordy comments rephrased
- typos fixed
- rte_eth_rx_queue_setup configuration check isolated for two main options
- the rest of comments addressed

v7: http://patches.dpdk.org/patch/80978/
- description as union of feature dedicated segment split description
- the split limitations reported in the rte_eth_dev_info->rx_seg_capa
- comments made less wordy

v8: - bitfield instead of uint8;
- named union to avoid warinings;
- typos fixed

v9: http://patches.dpdk.org/patch/81080/
- simplified rte_eth_rxseg union
- moved out wordy comment from rte_eth_rxconf
- rte_eth_rxseg_capa uint32 bitfield and naming
- added experimental tags

v10: - documentation clenup (removed rxq_info_get mention)
---

Viacheslav Ovsiienko (6):
ethdev: introduce Rx buffer split app/testpmd: add multiple pools per core creation app/testpmd: add buffer split offload configuration app/testpmd: add rxpkts commands and parameters app/testpmd: add rxoffs commands and parameters app/testpmd: add extended Rx queue setup

 app/test-pmd/bpf_cmd.c                      |   4 +-
 app/test-pmd/cmdline.c                      | 151 +++++++++++++++++++++++----
 app/test-pmd/config.c                       | 107 ++++++++++++++++++-
 app/test-pmd/parameters.c                   |  54 ++++++++--
 app/test-pmd/testpmd.c                      | 121 ++++++++++++++++------
 app/test-pmd/testpmd.h                      |  44 ++++++--
 doc/guides/nics/features.rst                |  14 +++
 doc/guides/rel_notes/deprecation.rst        |   5 -
 doc/guides/rel_notes/release_20_11.rst      |   9 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  22 +++-
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  36 ++++++-
 lib/librte_ethdev/rte_ethdev.c              | 153 +++++++++++++++++++++++-----
 lib/librte_ethdev/rte_ethdev.h              |  89 +++++++++++++++-
 13 files changed, 706 insertions(+), 103 deletions(-)
Signature · 1 line
1.8.3.1
#137 lore raw

PATCH v10 1/6 ethdev: introduce Rx buffer split

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---
 doc/guides/nics/features.rst           |  14 +++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 ++
 lib/librte_ethdev/rte_ethdev.c         | 153 +++++++++++++++++++++++++++------
 lib/librte_ethdev/rte_ethdev.h         |  89 ++++++++++++++++++-
 5 files changed, 239 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +14 -0
index dd8c955..16e00b8 100644
@@ -185,6 +185,20 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index d1f5ed3..dde464b 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index f8686a5..e12cf96 100644
@@ -89,6 +89,12 @@ New Features
     as flow action.
   * Added new flow APIs to create/update/destroy/query shared action.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -321,6 +327,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +128 -25

187 lines not on this page · raw message

lib/librte_ethdev/rte_ethdev.h +88 -1

133 lines not on this page · raw message

Signature · 1 line
1.8.3.1
#138 lore raw

PATCH v10 2/6 app/testpmd: add multiple pools per core creation

In reply to Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

---
 app/test-pmd/bpf_cmd.c                |  4 +--
 app/test-pmd/cmdline.c                |  2 +-
 app/test-pmd/config.c                 |  6 ++--
 app/test-pmd/parameters.c             | 24 +++++++++----
 app/test-pmd/testpmd.c                | 63 +++++++++++++++++++----------------
 app/test-pmd/testpmd.h                | 24 ++++++++++---
 doc/guides/testpmd_app_ug/run_app.rst |  7 ++--
 7 files changed, 83 insertions(+), 47 deletions(-)
app/test-pmd/bpf_cmd.c +2 -2
index 16e3c3b..0a1a178 100644
@@ -69,7 +69,7 @@ struct cmd_bpf_ld_result {
 
 	*flags = RTE_BPF_ETH_F_NONE;
 	arg->type = RTE_BPF_ARG_PTR;
-	arg->size = mbuf_data_size;
+	arg->size = mbuf_data_size[0];
 
 	for (i = 0; str[i] != 0; i++) {
 		v = toupper(str[i]);
@@ -78,7 +78,7 @@ struct cmd_bpf_ld_result {
 		else if (v == 'M') {
 			arg->type = RTE_BPF_ARG_PTR_MBUF;
 			arg->size = sizeof(struct rte_mbuf);
-			arg->buf_size = mbuf_data_size;
+			arg->buf_size = mbuf_data_size[0];
 		} else if (v == '-')
 			continue;
 		else
app/test-pmd/cmdline.c +1 -1
index 078f063..19b1896 100644
@@ -2925,7 +2925,7 @@ struct cmd_setup_rxtx_queue {
 		if (!numa_support || socket_id == NUMA_NO_CONFIG)
 			socket_id = port->socket_id;
 
-		mp = mbuf_pool_find(socket_id);
+		mp = mbuf_pool_find(socket_id, 0);
 		if (mp == NULL) {
 			printf("Failed to setup RX queue: "
 				"No mempool allocation"
app/test-pmd/config.c +3 -3
index 2c00b55..c554dd0 100644
@@ -690,7 +690,7 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 	printf("\nConnect to socket: %u", port->socket_id);
 
 	if (port_numa[port_id] != NUMA_NO_CONFIG) {
-		mp = mbuf_pool_find(port_numa[port_id]);
+		mp = mbuf_pool_find(port_numa[port_id], 0);
 		if (mp)
 			printf("\nmemory allocation on the socket: %d",
 							port_numa[port_id]);
@@ -3580,9 +3580,9 @@ struct igb_ring_desc_16_bytes {
 	 */
 	tx_pkt_len = 0;
 	for (i = 0; i < nb_segs; i++) {
-		if (seg_lengths[i] > (unsigned) mbuf_data_size) {
+		if (seg_lengths[i] > mbuf_data_size[0]) {
 			printf("length[%u]=%u > mbuf_data_size=%u - give up\n",
-			       i, seg_lengths[i], (unsigned) mbuf_data_size);
+			       i, seg_lengths[i], mbuf_data_size[0]);
 			return;
 		}
 		tx_pkt_len = (uint16_t)(tx_pkt_len + seg_lengths[i]);
app/test-pmd/parameters.c +18 -6
index 15ce8c1..4db4987 100644
@@ -106,7 +106,9 @@
 	       "(flag: 1 for RX; 2 for TX; 3 for RX and TX).\n");
 	printf("  --socket-num=N: set socket from which all memory is allocated "
 	       "in NUMA mode.\n");
-	printf("  --mbuf-size=N: set the data size of mbuf to N bytes.\n");
+	printf("  --mbuf-size=N,[N1[,..Nn]: set the data size of mbuf to "
+	       "N bytes. If multiple numbers are specified the extra pools "
+	       "will be created to receive with packet split features\n");
 	printf("  --total-num-mbufs=N: set the number of mbufs to be allocated "
 	       "in mbuf pools.\n");
 	printf("  --max-pkt-len=N: set the maximum size of packet to N bytes.\n");
@@ -892,12 +894,22 @@
 				}
 			}
 			if (!strcmp(lgopts[opt_idx].name, "mbuf-size")) {
-				n = atoi(optarg);
-				if (n > 0 && n <= 0xFFFF)
-					mbuf_data_size = (uint16_t) n;
-				else
+				unsigned int mb_sz[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs, i;
+
+				nb_segs = parse_item_list(optarg, "mbuf-size",
+					MAX_SEGS_BUFFER_SPLIT, mb_sz, 0);
+				if (nb_segs <= 0)
 					rte_exit(EXIT_FAILURE,
-						 "mbuf-size should be > 0 and < 65536\n");
+						 "bad mbuf-size\n");
+				for (i = 0; i < nb_segs; i++) {
+					if (mb_sz[i] <= 0 || mb_sz[i] > 0xFFFF)
+						rte_exit(EXIT_FAILURE,
+							 "mbuf-size should be "
+							 "> 0 and < 65536\n");
+					mbuf_data_size[i] = (uint16_t) mb_sz[i];
+				}
+				mbuf_data_size_n = nb_segs;
 			}
 			if (!strcmp(lgopts[opt_idx].name, "total-num-mbufs")) {
 				n = atoi(optarg);
app/test-pmd/testpmd.c +35 -28

145 lines not on this page · raw message

app/test-pmd/testpmd.h +19 -5
index a4dfd4f..0975305 100644
@@ -42,6 +42,13 @@
  */
 #define RTE_MAX_SEGS_PER_PKT 255 /**< nb_segs is a 8-bit unsigned char. */
 
+/*
+ * The maximum number of segments per packet is used to configure
+ * buffer split feature, also specifies the maximum amount of
+ * optional Rx pools to allocate mbufs to split.
+ */
+#define MAX_SEGS_BUFFER_SPLIT 8 /**< nb_segs is a 8-bit unsigned char. */
+
 #define MAX_PKT_BURST 512
 #define DEF_PKT_BURST 32
 
@@ -403,7 +410,9 @@ struct queue_stats_mappings {
 extern uint8_t dcb_config;
 extern uint8_t dcb_test;
 
-extern uint16_t mbuf_data_size; /**< Mbuf data space size. */
+extern uint32_t mbuf_data_size_n;
+extern uint16_t mbuf_data_size[MAX_SEGS_BUFFER_SPLIT];
+/**< Mbuf data space size. */
 extern uint32_t param_total_num_mbufs;
 
 extern uint16_t stats_period;
@@ -615,17 +624,22 @@ struct mplsoudp_decap_conf {
 
 /* Mbuf Pools */
 static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+		    int name_size, unsigned int idx)
 {
-	snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	if (!idx)
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	else
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+			 sock_id, idx);
 }
 
 static inline struct rte_mempool *
-mbuf_pool_find(unsigned int sock_id)
+mbuf_pool_find(unsigned int sock_id, unsigned int idx)
 {
 	char pool_name[RTE_MEMPOOL_NAMESIZE];
 
-	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name));
+	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name), idx);
 	return rte_mempool_lookup((const char *)pool_name);
 }
 
doc/guides/testpmd_app_ug/run_app.rst +5 -2
index ec085c2..1eb0a10 100644
@@ -107,9 +107,12 @@ The command line options are:
     Set the socket from which all memory is allocated in NUMA mode,
     where 0 <= N < number of sockets on the board.
 
-*   ``--mbuf-size=N``
+*   ``--mbuf-size=N[,N1[,...Nn]``
 
-    Set the data size of the mbufs used to N bytes, where N < 65536. The default value is 2048.
+    Set the data size of the mbufs used to N bytes, where N < 65536.
+    The default value is 2048. If multiple mbuf-size values are specified the
+    extra memory pools will be created for allocating mbufs to receive packets
+    with buffer splittling features.
 
 *   ``--total-num-mbufs=N``
 
Signature · 1 line
1.8.3.1
#139 lore raw

PATCH v10 3/6 app/testpmd: add buffer split offload configuration

In reply to Viacheslav Ovsiienko

This patch add support for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT providing per queue configuration for this offload.

---
 app/test-pmd/cmdline.c | 21 +++++++++++----------
 app/test-pmd/config.c  |  9 +++++++++
 2 files changed, 20 insertions(+), 10 deletions(-)
app/test-pmd/cmdline.c +11 -10
index 19b1896..0ef87c9 100644
@@ -883,16 +883,16 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"port config <port_id> rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"     Enable or disable a per port Rx offloading"
 			" on all Rx queues of a port\n\n"
 
 			"port (port_id) rxq (queue_id) rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"    Enable or disable a per queue Rx offloading"
 			" only on a specific Rx queue\n\n"
 
@@ -18435,7 +18435,8 @@ struct cmd_config_per_port_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc#rss_hash");
+			   "scatter#buffer_split#timestamp#security#"
+			   "keep_crc#rss_hash");
 cmdline_parse_token_string_t cmd_config_per_port_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_port_rx_offload_result,
@@ -18515,8 +18516,8 @@ struct cmd_config_per_port_rx_offload_result {
 	.help_str = "port config <port_id> rx_offload vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc|rss_hash "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc|rss_hash on|off",
 	.tokens = {
 		(void *)&cmd_config_per_port_rx_offload_result_port,
 		(void *)&cmd_config_per_port_rx_offload_result_config,
@@ -18565,7 +18566,7 @@ struct cmd_config_per_queue_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc");
+			   "scatter#buffer_split#timestamp#security#keep_crc");
 cmdline_parse_token_string_t cmd_config_per_queue_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_queue_rx_offload_result,
@@ -18621,8 +18622,8 @@ struct cmd_config_per_queue_rx_offload_result {
 		    "vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc on|off",
 	.tokens = {
 		(void *)&cmd_config_per_queue_rx_offload_result_port,
 		(void *)&cmd_config_per_queue_rx_offload_result_port_id,
app/test-pmd/config.c +9 -0
index c554dd0..0f041e1 100644
@@ -1092,6 +1092,15 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 			printf("off\n");
 	}
 
+	if (dev_info.rx_offload_capa & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+		printf("RX offload buffer split:       ");
+		if (ports[port_id].dev_conf.rxmode.offloads &
+		    RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT)
+			printf("on\n");
+		else
+			printf("off\n");
+	}
+
 	if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_VLAN_INSERT) {
 		printf("VLAN insert:                   ");
 		if (ports[port_id].dev_conf.txmode.offloads &
Signature · 1 line
1.8.3.1
#140 lore raw

PATCH v10 4/6 app/testpmd: add rxpkts commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxpkts=X[,Y]

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command:

testpmd> set rxpkts (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only). Optionally the multiple memory pools can be specified with --mbuf-size command line parameter and the mbufs to receive will be allocated sequentially from these extra memory pools.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 48 ++++++++++++++++++++++-
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  7 ++++
 app/test-pmd/testpmd.h                      | 11 +++++-
 doc/guides/testpmd_app_ug/run_app.rst       |  9 +++++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 21 +++++++++-
 7 files changed, 165 insertions(+), 7 deletions(-)
app/test-pmd/cmdline.c +58 -3
index 0ef87c9..a195e37 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxpkts (x[,y]*)\n"
+			"    Set the length of each segment to scatter"
+			" packets on receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set txpkts (x[,y]*)\n"
 			"    Set the length of each segment of TXONLY"
 			" and optionally CSUM packets.\n\n"
@@ -3907,6 +3913,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxpkts_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxpkts;
+	cmdline_fixed_string_t seg_lengths;
+};
+
+static void
+cmd_set_rxpkts_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxpkts_result *res;
+	unsigned int seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_lengths, "segment lengths",
+				  MAX_SEGS_BUFFER_SPLIT, seg_lengths, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_segments(seg_lengths, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxpkts_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxpkts_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 rxpkts, "rxpkts");
+cmdline_parse_token_string_t cmd_set_rxpkts_lengths =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 seg_lengths, NULL);
+
+cmdline_parse_inst_t cmd_set_rxpkts = {
+	.f = cmd_set_rxpkts_parsed,
+	.data = NULL,
+	.help_str = "set rxpkts <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxpkts_keyword,
+		(void *)&cmd_set_rxpkts_name,
+		(void *)&cmd_set_rxpkts_lengths,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF TXONLY PACKETS *** */
 
 struct cmd_set_txpkts_result {
@@ -7535,6 +7587,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxpkts"))
+		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
 		show_tx_pkt_segments();
 	else if (!strcmp(res->what, "txtimes"))
@@ -7547,12 +7601,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19825,6 +19879,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
 	(cmdline_parse_inst_t *)&cmd_set_txtimes,
app/test-pmd/config.c +46 -2

62 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index 4db4987..e4e3635 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
 	printf("  --txonly-multi-flow: generate multiple flows in txonly mode\n");
@@ -662,6 +663,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
 		{ "disable-link-check",		0, 0, 0 },
@@ -1272,6 +1274,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
+				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs;
+
+				nb_segs = parse_item_list
+						(optarg, "rxpkt segments",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_len, 0);
+				if (nb_segs > 0)
+					set_rx_pkt_segments(seg_len, nb_segs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxpkts\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "txpkts")) {
 				unsigned seg_lengths[RTE_MAX_SEGS_PER_PKT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +7 -0
index 7e6ef80..f88c1e2 100644
@@ -210,6 +210,13 @@ struct fwd_engine * fwd_engines[] = {
 uint8_t f_quit;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 uint16_t tx_pkt_length = TXONLY_DEF_PACKET_LEN; /**< TXONLY packet length. */
app/test-pmd/testpmd.h +10 -1
index 0975305..c5c942a 100644
@@ -430,6 +430,13 @@ struct queue_stats_mappings {
 extern struct rte_fdir_conf fdir_conf;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 #define TXONLY_DEF_PACKET_LEN 64
@@ -836,7 +843,9 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_record_core_cycles(uint8_t on_off);
 void set_record_burst_stats(uint8_t on_off);
 void set_verbose_level(uint16_t vb_level);
-void set_tx_pkt_segments(unsigned *seg_lengths, unsigned nb_segs);
+void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
+void show_rx_pkt_segments(void);
+void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
 void show_tx_pkt_times(void);
doc/guides/testpmd_app_ug/run_app.rst +9 -0
index 1eb0a10..463b76c 100644
@@ -361,6 +361,15 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxpkts=X[,Y]``
+
+    Set the length of segments to scatter packets on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+    Optionally the multiple memory pools can be specified with --mbuf-size
+    command line parameter and the mbufs to receive will be allocated
+    sequentially from these extra memory pools.
+
 *   ``--txpkts=X[,Y]``
 
     Set TX segment sizes or total packet length. Valid for ``tx-only``
doc/guides/testpmd_app_ug/testpmd_funcs.rst +20 -1
index 43c0ea0..b9fad3a 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxpkts``: Packets to RX split configuration.
+
 * ``txpkts``: Packets to TX configuration.
 
 * ``txtimes``: Burst time pattern for Tx only mode.
@@ -774,6 +776,23 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxpkts
+~~~~~~~~~~
+
+Set the length of segments to scatter packets on receiving if split
+feature is engaged. Affects only the queues configured with split offloads
+(currently BUFFER_SPLIT is supported only). Optionally the multiple memory
+pools can be specified with --mbuf-size command line parameter and the mbufs
+to receive will be allocated sequentially from these extra memory pools (the
+mbuf for the first segment is allocated from the first pool, the second one
+from the second pool, and so on, if segment number is greater then pool's the
+mbuf for remaining segments will be allocated from the last valid pool).
+
+   testpmd> set rxpkts (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. Zero value
+means to use the corresponding memory pool data buffer size.
+
 set txpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#141 lore raw

PATCH v10 5/6 app/testpmd: add rxoffs commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxoffs=X[,Y]

Sets the offsets of packet segments from the beginning of the receiving buffer if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command, providing the same:

testpmd> set rxoffs (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 44 +++++++++++++++++++++
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  2 +
 app/test-pmd/testpmd.h                      |  4 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  6 +++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 17 +++++++-
 7 files changed, 145 insertions(+), 4 deletions(-)
app/test-pmd/cmdline.c +58 -3
index a195e37..e09df36 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxoffs (x[,y]*)\n"
+			"    Set the offset of each packet segment on"
+			" receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set rxpkts (x[,y]*)\n"
 			"    Set the length of each segment to scatter"
 			" packets on receiving if split feature is engaged."
@@ -3913,6 +3919,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT OFFSETS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxoffs_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxoffs;
+	cmdline_fixed_string_t seg_offsets;
+};
+
+static void
+cmd_set_rxoffs_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxoffs_result *res;
+	unsigned int seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_offsets, "segment offsets",
+				  MAX_SEGS_BUFFER_SPLIT, seg_offsets, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_offsets(seg_offsets, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxoffs_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxoffs_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 rxoffs, "rxoffs");
+cmdline_parse_token_string_t cmd_set_rxoffs_offsets =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 seg_offsets, NULL);
+
+cmdline_parse_inst_t cmd_set_rxoffs = {
+	.f = cmd_set_rxoffs_parsed,
+	.data = NULL,
+	.help_str = "set rxoffs <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxoffs_keyword,
+		(void *)&cmd_set_rxoffs_name,
+		(void *)&cmd_set_rxoffs_offsets,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
 
 struct cmd_set_rxpkts_result {
@@ -7587,6 +7639,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxoffs"))
+		show_rx_pkt_offsets();
 	else if (!strcmp(res->what, "rxpkts"))
 		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
@@ -7601,12 +7655,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxoffs#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19879,6 +19933,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxoffs,
 	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
app/test-pmd/config.c +44 -0

50 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index e4e3635..2298ba5 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxoffs=X[,Y]*: set RX segment offsets for split.\n");
 	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
@@ -663,6 +664,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxoffs",			1, 0, 0 },
 		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
@@ -1274,6 +1276,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxoffs")) {
+				unsigned int seg_off[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_offs;
+
+				nb_offs = parse_item_list
+						(optarg, "rxpkt offsets",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_off, 0);
+				if (nb_offs > 0)
+					set_rx_pkt_offsets(seg_off, nb_offs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxoffs\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
 				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +2 -0
index f88c1e2..580178d 100644
@@ -215,6 +215,8 @@ struct fwd_engine * fwd_engines[] = {
  */
 uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
app/test-pmd/testpmd.h +4 -0
index c5c942a..dc37148 100644
@@ -435,6 +435,8 @@ struct queue_stats_mappings {
  */
 extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+extern uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
@@ -845,6 +847,8 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_verbose_level(uint16_t vb_level);
 void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_rx_pkt_segments(void);
+void set_rx_pkt_offsets(unsigned int *seg_offsets, unsigned int nb_offs);
+void show_rx_pkt_offsets(void);
 void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
doc/guides/testpmd_app_ug/run_app.rst +6 -0
index 463b76c..9b0a84a 100644
@@ -361,6 +361,12 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxoffs=X[,Y]``
+
+    Set the offsets of packet segments on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+
 *   ``--rxpkts=X[,Y]``
 
     Set the length of segments to scatter packets on receiving if split
doc/guides/testpmd_app_ug/testpmd_funcs.rst +16 -1
index b9fad3a..e73530e 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxoffs``: Packet offsets for RX split.
+
 * ``rxpkts``: Packets to RX split configuration.
 
 * ``txpkts``: Packets to TX configuration.
@@ -776,6 +778,19 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxoffs
+~~~~~~~~~~
+
+Set the offsets of segments relating to the data buffer beginning on receiving
+if split feature is engaged. Affects only the queues configured with split
+offloads (currently BUFFER_SPLIT is supported only).
+
+   testpmd> set rxoffs (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. If the list
+of offsets is shorter than the list of segments the zero offsets will be used
+for the remaining segments.
+
 set rxpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#142 lore raw

PATCH v10 6/6 app/testpmd: add extended Rx queue setup

In reply to Viacheslav Ovsiienko

If Rx queue is configured with split feature the extended setup with specified segment sizes and pool will be performed.

---
 app/test-pmd/cmdline.c | 12 ++++++------
 app/test-pmd/testpmd.c | 49 +++++++++++++++++++++++++++++++++++++++++++++++--
 app/test-pmd/testpmd.h |  5 +++++
 3 files changed, 58 insertions(+), 8 deletions(-)
app/test-pmd/cmdline.c +6 -6
index e09df36..07ee4e4 100644
@@ -2945,12 +2945,12 @@ struct cmd_setup_rxtx_queue {
 				rxring_numa[res->portid]);
 			return;
 		}
-		ret = rte_eth_rx_queue_setup(res->portid,
-					     res->qid,
-					     port->nb_rx_desc[res->qid],
-					     socket_id,
-					     &port->rx_conf[res->qid],
-					     mp);
+		ret = rx_queue_setup(res->portid,
+				     res->qid,
+				     port->nb_rx_desc[res->qid],
+				     socket_id,
+				     &port->rx_conf[res->qid],
+				     mp);
 		if (ret)
 			printf("Failed to setup RX queue\n");
 	} else {
app/test-pmd/testpmd.c +47 -2

67 lines not on this page · raw message

app/test-pmd/testpmd.h +5 -0
index dc37148..c7b2cb4 100644
@@ -896,6 +896,11 @@ void port_rss_reta_info(portid_t port_id,
 
 void set_vf_traffic(portid_t port_id, uint8_t is_rx, uint16_t vf, uint8_t on);
 
+int
+rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
+	       uint16_t nb_rx_desc, unsigned int socket_id,
+	       struct rte_eth_rxconf *rx_conf, struct rte_mempool *mp);
+
 int set_queue_rate_limit(portid_t port_id, uint16_t queue_idx, uint16_t rate);
 int set_vf_rate_limit(portid_t port_id, uint16_t vf, uint16_t rate,
 				uint64_t q_msk);
Signature · 1 line
1.8.3.1
#143 lore raw

v11 cover ethdev: introduce Rx buffer split

In reply to Slava Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---

v1: http://patches.dpdk.org/patch/79594/
v2: http://patches.dpdk.org/patch/79893/
- add feature support to mlx5 PMD

v3: http://patches.dpdk.org/patch/80389/
- rte_eth_rx_queue_setup_ex is renamed to rte_eth_rxseg_queue_setup
- DEV_RX_OFFLOAD_BUFFER_SPLIT is renamed to RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT
- commit message update
- documentaion provided
- release notes update
- minor bug fixes in testpmd related part

v4: http://patches.dpdk.org/patch/80401/
- common part of rx_queue_setup/rxseg_queue_setup

v5: http://patches.dpdk.org/patch/80609/
- refactored to approach of providing split configuration in the rte_eth_rxconf structure instead of introducing the new API routine
- added support for rxoffs command to testpmd to provide segment offsets for complete testing of split configurations
- patchset is split into two parts - PMD part will be presented as separate series

v6: http://patches.dpdk.org/patch/80785/
- wordy comments rephrased
- typos fixed
- rte_eth_rx_queue_setup configuration check isolated for two main options
- the rest of comments addressed

v7: http://patches.dpdk.org/patch/80978/
- description as union of feature dedicated segment split description
- the split limitations reported in the rte_eth_dev_info->rx_seg_capa
- comments made less wordy

v8: - bitfield instead of uint8;
- named union to avoid warinings;
- typos fixed

v9: http://patches.dpdk.org/patch/81080/
- simplified rte_eth_rxseg union
- moved out wordy comment from rte_eth_rxconf
- rte_eth_rxseg_capa uint32 bitfield and naming
- added experimental tags

v10: - missed in mailing list, see v11 for changes

v11: - documentation cleanup (removed rxq_info_get mention)
- max_nseg moved from the recored and converted to uint16_t;
- reserved uint16_t is added to rte_eth_rxseg_capa;

---

Viacheslav Ovsiienko (6):
ethdev: introduce Rx buffer split app/testpmd: add multiple pools per core creation app/testpmd: add buffer split offload configuration app/testpmd: add rxpkts commands and parameters app/testpmd: add rxoffs commands and parameters app/testpmd: add extended Rx queue setup

 app/test-pmd/bpf_cmd.c                      |   4 +-
 app/test-pmd/cmdline.c                      | 151 +++++++++++++++++++++++----
 app/test-pmd/config.c                       | 107 ++++++++++++++++++-
 app/test-pmd/parameters.c                   |  54 ++++++++--
 app/test-pmd/testpmd.c                      | 121 ++++++++++++++++------
 app/test-pmd/testpmd.h                      |  44 ++++++--
 doc/guides/nics/features.rst                |  14 +++
 doc/guides/rel_notes/deprecation.rst        |   5 -
 doc/guides/rel_notes/release_20_11.rst      |   9 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  22 +++-
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  36 ++++++-
 lib/librte_ethdev/rte_ethdev.c              | 153 +++++++++++++++++++++++-----
 lib/librte_ethdev/rte_ethdev.h              |  90 +++++++++++++++-
 13 files changed, 707 insertions(+), 103 deletions(-)
Signature · 1 line
1.8.3.1
#144 lore raw

PATCH v11 1/6 ethdev: introduce Rx buffer split

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---
 doc/guides/nics/features.rst           |  14 +++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 ++
 lib/librte_ethdev/rte_ethdev.c         | 153 +++++++++++++++++++++++++++------
 lib/librte_ethdev/rte_ethdev.h         |  90 ++++++++++++++++++-
 5 files changed, 240 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +14 -0
index dd8c955..16e00b8 100644
@@ -185,6 +185,20 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index d1f5ed3..dde464b 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index f8686a5..e12cf96 100644
@@ -89,6 +89,12 @@ New Features
     as flow action.
   * Added new flow APIs to create/update/destroy/query shared action.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -321,6 +327,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +128 -25

187 lines not on this page · raw message

lib/librte_ethdev/rte_ethdev.h +89 -1

134 lines not on this page · raw message

Signature · 1 line
1.8.3.1
Acked-by on v11 1/6
#145 lore raw

16/10/2020 15:39, Viacheslav Ovsiienko:

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1862

Quoted 7 lines
lib/librte_ethdev/rte_ethdev.c +7 -0
+		if (offset != 0) {
+			if (seg_capa->offset_allowed == 0) {
+				RTE_ETHDEV_LOG(ERR, "Rx segmentation with offset is not supported\n");
+				return -ENOTSUP;
+			}
+			if (offset & offset_mask) {
+				RTE_ETHDEV_LOG(ERR, "Rx segmentat invalid offset alignment %u, %u\n",

small typo: segmentat -> segmentation

[...]

Quoted 7 lines

/**
+ * A structure used to configure an Rx packet segment to split.
+ *
+ * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
+ * the PMD will split the received packets into multiple segments
+ * according to the specification in the description array:
+ *

Minor detail: below bullets should start with a capital letter.

Quoted 7 lines

+ * - the first network buffer will be allocated from the memory pool,
+ * specified in the first array element, the second buffer, from the
+ * pool in the second element, and so on.
+ *
+ * - the offsets from the segment description elements specify
+ * the data offset from the buffer beginning except the first mbuf.
+ * For this one the offset is added with RTE_PKTMBUF_HEADROOM.

"this one" -> "the first segment, "

Quoted 15 lines

+ *
+ * - the lengths in the elements define the maximal data amount

+ *   being received to each segment. The receiving starts with filling
+ *   up the first mbuf data buffer up to specified length. If the
+ *   there are data remaining (packet is longer than buffer in the first
+ *   mbuf) the following data will be pushed to the next segment
+ *   up to its own length, and so on.

+ *
+ * - If the length in the segment description element is zero
+ * the actual buffer size will be deduced from the appropriate
+ * memory pool properties.
+ *
+ * - if there is not enough elements to describe the buffer for entire
+ * packet of maximal length the following parameters will be used
+ * for the all remaining segments:

Quoted 4 lines
+ *     - pool from the last valid element
+ *     - the buffer size from this pool
+ *     - zero offset

+ */

I think the description is precise enough, thanks.

Quoted 7 lines
lib/librte_ethdev/rte_ethdev.h +7 -0
+__rte_experimental
+struct rte_eth_rxseg_split {
+	struct rte_mempool *mp; /**< Memory pool to allocate segment from. */
+	uint16_t length; /**< Segment data length, configures split point. */
+	uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
+	uint32_t reserved; /**< Reserved field. */
+};

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 1021

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+
+/**
+ * A common structure used to describe Rx packet segment properties.
+ */
+__rte_experimental
+union rte_eth_rxseg {
+	/* The settings for buffer split offload. */
+	struct rte_eth_rxseg_split split;
+	/* The other features settings should be added here. */
+};

OK for having an union.

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+
+/**
  * A structure used to configure an RX ring of an Ethernet port.
  */
 struct rte_eth_rxconf {
@@ -977,12 +1028,22 @@ struct rte_eth_rxconf {
 	uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
 	uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
 	uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
+	uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1041

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +4 -0
 	/**
 	 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
 	 * Only offloads set on rx_queue_offload_capa or rx_offload_capa
 	 * fields on rte_eth_dev_info structure are allowed to be set.
 	 */
 	uint64_t offloads;
+	/**
+	 * Points to the array of segment descriptions. Each array element
+	 * describes the properties for each segment in the receiving
+	 * buffer according to feature descripting structure.

Alternative (shorter) wording:
Points to the array of segment descriptions for an entire packet. Array elements are properties for consecutive Rx segments.

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1044

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+	 *
+	 * The supported capabilities of receiving segmentation is reported
+	 * in rte_eth_dev_info ->rx_seg_capa field.

Rx segmentation capabilities are reported in rte_eth_dev_info.rx_seg_capa.

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+	 */
+	union rte_eth_rxseg *rx_seg;

[...]

Quoted 11 lines
+ *   The configuration structure also contains the pointer to the array
+ *   of the receiving buffer segment descriptions, see rx_seg and rx_nseg
+ *   fields, this extended configuration might be used by split offloads like
+ *   RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mp_pool is not NULL,
+ *   the extended configuration fields must be set to NULL and zero.
  * @param mb_pool
  *   The pointer to the memory pool from which to allocate *rte_mbuf* network
- *   memory buffers to populate each descriptor of the receive ring.
+ *   memory buffers to populate each descriptor of the receive ring. There are
+ *   two options to provide Rx buffer configuration:
+ *   - single pool:
Quoted 4 lines

+ * mb_pool is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is 0.
+ * - multiple segments description:
+ * mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0.
+ * Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads.

As said previously, I feel rx_conf.rx_seg does not need to be part of the requirements, as long as rx_conf.rx_nseg is defined.

All my comments are optional, this version is already very good.

#146 lore raw
Quoted 141 lines

-----Original Message-----
From: Thomas Monjalon <[email protected]>
Sent: Friday, October 16, 2020 18:14
To: Slava Ovsiienko <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [dpdk-dev] [PATCH v11 1/6] ethdev: introduce Rx buffer split

16/10/2020 15:39, Viacheslav Ovsiienko:

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev *

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.c +3 -0
+		if (offset != 0) {
+			if (seg_capa->offset_allowed == 0) {
+				RTE_ETHDEV_LOG(ERR, "Rx segmentation

with offset is not supported\n");

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev *

Quoted 4 lines
lib/librte_ethdev/rte_ethdev.c +4 -0
+				return -ENOTSUP;
+			}
+			if (offset & offset_mask) {
+				RTE_ETHDEV_LOG(ERR, "Rx segmentat invalid

offset alignment %u,

Quoted 1 line

+%u\n",

small typo: segmentat -> segmentation

[...]

Quoted 7 lines

/**
+ * A structure used to configure an Rx packet segment to split.
+ *
+ * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
+ * the PMD will split the received packets into multiple segments
+ * according to the specification in the description array:
+ *

Minor detail: below bullets should start with a capital letter.

Quoted 7 lines

+ * - the first network buffer will be allocated from the memory pool,
+ * specified in the first array element, the second buffer, from the
+ * pool in the second element, and so on.
+ *
+ * - the offsets from the segment description elements specify
+ * the data offset from the buffer beginning except the first mbuf.
+ * For this one the offset is added with RTE_PKTMBUF_HEADROOM.

"this one" -> "the first segment, "

Quoted 15 lines

+ *
+ * - the lengths in the elements define the maximal data amount

+ *   being received to each segment. The receiving starts with filling
+ *   up the first mbuf data buffer up to specified length. If the
+ *   there are data remaining (packet is longer than buffer in the first
+ *   mbuf) the following data will be pushed to the next segment
+ *   up to its own length, and so on.

+ *
+ * - If the length in the segment description element is zero
+ * the actual buffer size will be deduced from the appropriate
+ * memory pool properties.
+ *
+ * - if there is not enough elements to describe the buffer for entire
+ * packet of maximal length the following parameters will be used
+ * for the all remaining segments:

Quoted 4 lines
+ *     - pool from the last valid element
+ *     - the buffer size from this pool
+ *     - zero offset

+ */

I think the description is precise enough, thanks.

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode {

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+__rte_experimental
+struct rte_eth_rxseg_split {
+	struct rte_mempool *mp; /**< Memory pool to allocate segment

from. */

Review

Quoted 2 lines
(quoted diff) +2 -0
+	uint16_t length; /**< Segment data length, configures split point. */
+	uint16_t offset; /**< Data offset from beginning of mbuf data buffer.

*/

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode {

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+	uint32_t reserved; /**< Reserved field. */ };
+
+/**
+ * A common structure used to describe Rx packet segment properties.
+ */
+__rte_experimental
+union rte_eth_rxseg {
+	/* The settings for buffer split offload. */
+	struct rte_eth_rxseg_split split;
+	/* The other features settings should be added here. */ };

OK for having an union.

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1029

Quoted 8 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+
+/**
  * A structure used to configure an RX ring of an Ethernet port.
  */
 struct rte_eth_rxconf {
@@ -977,12 +1028,22 @@ struct rte_eth_rxconf {
 	uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
 	uint8_t rx_drop_en; /**< Drop packets if no descriptors are available.

*/

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1041

Quoted 13 lines
(quoted diff) +0 -0
 	uint8_t rx_deferred_start; /**< Do not start queue with

rte_eth_dev_start(). */

lib/librte_ethdev/rte_ethdev.h +5 -0
+	uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
 	/**
 	 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
 	 * Only offloads set on rx_queue_offload_capa or rx_offload_capa
 	 * fields on rte_eth_dev_info structure are allowed to be set.
 	 */
 	uint64_t offloads;
+	/**
+	 * Points to the array of segment descriptions. Each array element
+	 * describes the properties for each segment in the receiving
+	 * buffer according to feature descripting structure.

Alternative (shorter) wording:
Points to the array of segment descriptions for an entire packet. Array elements are properties for consecutive Rx segments.

lib/librte_ethdev/rte_ethdev.h struct rte_eth_rxconf { · line 1044

Quoted 3 lines
lib/librte_ethdev/rte_ethdev.h +3 -0
+	 *
+	 * The supported capabilities of receiving segmentation is reported
+	 * in rte_eth_dev_info ->rx_seg_capa field.

Rx segmentation capabilities are reported in rte_eth_dev_info.rx_seg_capa.

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.h +2 -0
+	 */
+	union rte_eth_rxseg *rx_seg;

[...]

Quoted 7 lines
+ *   The configuration structure also contains the pointer to the array
+ *   of the receiving buffer segment descriptions, see rx_seg and rx_nseg
+ *   fields, this extended configuration might be used by split offloads like
+ *   RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mp_pool is not NULL,
+ *   the extended configuration fields must be set to NULL and zero.

* @param mb_pool
* The pointer to the memory pool from which to allocate *rte_mbuf*

network

Quoted 2 lines

- * memory buffers to populate each descriptor of the receive ring.
+ * memory buffers to populate each descriptor of the receive ring. There

are

Quoted 6 lines

+ * two options to provide Rx buffer configuration:
+ * - single pool:
+ * mb_pool is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is 0.
+ * - multiple segments description:
+ * mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0.
+ * Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in

Quoted 7 lines

offloads.

As said previously, I feel rx_conf.rx_seg does not need to be part of the requirements, as long as rx_conf.rx_nseg is defined.

All my comments are optional, this version is already very good.

All comments are adopted (in coming v12), thank you.

With best regards, Slava

#147 lore raw
In reply to Viacheslav Ovsiienko
Quoted 101 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf.
This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

     seg0 - pool0, len0=14B, off0=2
     seg1 - pool1, len1=20B, off1=128B
     seg2 - pool2, len2=20B, off2=0B
     seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

<...>

Quoted 7 lines
lib/librte_ethdev/rte_ethdev.h +7 -0
+__rte_experimental
+struct rte_eth_rxseg_split {
+	struct rte_mempool *mp; /**< Memory pool to allocate segment from. */
+	uint16_t length; /**< Segment data length, configures split point. */
+	uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
+	uint32_t reserved; /**< Reserved field. */
+};

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 1021

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+
+/**
+ * A common structure used to describe Rx packet segment properties.
+ */
+__rte_experimental
+union rte_eth_rxseg {
+	/* The settings for buffer split offload. */
+	struct rte_eth_rxseg_split split;
+	/* The other features settings should be added here. */
+};

Is the '__rte_experimental' on the data structs do anything other than documentation?

icc is complaining about them [1], if they are only for documenation, would it be OK to move it to the comment and remove the '__rte_experimental' tag?

[1]
../lib/librte_ethdev/rte_ethdev.h(1005): error #2651: attribute does not apply to any entity
__rte_experimental
^

#148 lore raw

16/10/2020 17:47, Ferruh Yigit:

Quoted 22 lines · Viacheslav Ovsiienko
Quoted 7 lines
lib/librte_ethdev/rte_ethdev.h +7 -0
+__rte_experimental
+struct rte_eth_rxseg_split {
+	struct rte_mempool *mp; /**< Memory pool to allocate segment from. */
+	uint16_t length; /**< Segment data length, configures split point. */
+	uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
+	uint32_t reserved; /**< Reserved field. */
+};

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 1021

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+
+/**
+ * A common structure used to describe Rx packet segment properties.
+ */
+__rte_experimental
+union rte_eth_rxseg {
+	/* The settings for buffer split offload. */
+	struct rte_eth_rxseg_split split;
+	/* The other features settings should be added here. */
+};

Is the '__rte_experimental' on the data structs do anything other than documentation?

Quoted 8 lines

icc is complaining about them [1], if they are only for documenation, would it be OK to move it to the comment and remove the '__rte_experimental' tag?

[1]
../lib/librte_ethdev/rte_ethdev.h(1005): error #2651: attribute does not apply to any entity
__rte_experimental
^

Yes should be a comment.

#149 lore raw
Quoted 26 lines · Thomas Monjalon

16/10/2020 17:47, Ferruh Yigit:

Quoted 25 lines · Viacheslav Ovsiienko
Quoted 7 lines
lib/librte_ethdev/rte_ethdev.h +7 -0
+__rte_experimental
+struct rte_eth_rxseg_split {
+	struct rte_mempool *mp; /**< Memory pool to allocate segment from. */
+	uint16_t length; /**< Segment data length, configures split point. */
+	uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
+	uint32_t reserved; /**< Reserved field. */
+};

lib/librte_ethdev/rte_ethdev.h struct rte_eth_txmode { · line 1021

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.h +10 -0
+
+/**
+ * A common structure used to describe Rx packet segment properties.
+ */
+__rte_experimental
+union rte_eth_rxseg {
+	/* The settings for buffer split offload. */
+	struct rte_eth_rxseg_split split;
+	/* The other features settings should be added here. */
+};

Is the '__rte_experimental' on the data structs do anything other than documentation?

icc is complaining about them [1], if they are only for documenation, would it be OK to move it to the comment and remove the '__rte_experimental' tag?

Quoted 8 lines
Quoted 5 lines

[1]
../lib/librte_ethdev/rte_ethdev.h(1005): error #2651: attribute does not apply to any entity
__rte_experimental
^

Yes should be a comment.

Following is a sample from rte_flow.h as comment:
https://git.dpdk.org/dpdk/tree/lib/librte_ethdev/rte_flow.h?h=v20.08#n1392

#150 lore raw

PATCH v11 2/6 app/testpmd: add multiple pools per core creation

In reply to Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

---
 app/test-pmd/bpf_cmd.c                |  4 +--
 app/test-pmd/cmdline.c                |  2 +-
 app/test-pmd/config.c                 |  6 ++--
 app/test-pmd/parameters.c             | 24 +++++++++----
 app/test-pmd/testpmd.c                | 63 +++++++++++++++++++----------------
 app/test-pmd/testpmd.h                | 24 ++++++++++---
 doc/guides/testpmd_app_ug/run_app.rst |  7 ++--
 7 files changed, 83 insertions(+), 47 deletions(-)
app/test-pmd/bpf_cmd.c +2 -2
index 16e3c3b..0a1a178 100644
@@ -69,7 +69,7 @@ struct cmd_bpf_ld_result {
 
 	*flags = RTE_BPF_ETH_F_NONE;
 	arg->type = RTE_BPF_ARG_PTR;
-	arg->size = mbuf_data_size;
+	arg->size = mbuf_data_size[0];
 
 	for (i = 0; str[i] != 0; i++) {
 		v = toupper(str[i]);
@@ -78,7 +78,7 @@ struct cmd_bpf_ld_result {
 		else if (v == 'M') {
 			arg->type = RTE_BPF_ARG_PTR_MBUF;
 			arg->size = sizeof(struct rte_mbuf);
-			arg->buf_size = mbuf_data_size;
+			arg->buf_size = mbuf_data_size[0];
 		} else if (v == '-')
 			continue;
 		else
app/test-pmd/cmdline.c +1 -1
index 078f063..19b1896 100644
@@ -2925,7 +2925,7 @@ struct cmd_setup_rxtx_queue {
 		if (!numa_support || socket_id == NUMA_NO_CONFIG)
 			socket_id = port->socket_id;
 
-		mp = mbuf_pool_find(socket_id);
+		mp = mbuf_pool_find(socket_id, 0);
 		if (mp == NULL) {
 			printf("Failed to setup RX queue: "
 				"No mempool allocation"
app/test-pmd/config.c +3 -3
index 2c00b55..c554dd0 100644
@@ -690,7 +690,7 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 	printf("\nConnect to socket: %u", port->socket_id);
 
 	if (port_numa[port_id] != NUMA_NO_CONFIG) {
-		mp = mbuf_pool_find(port_numa[port_id]);
+		mp = mbuf_pool_find(port_numa[port_id], 0);
 		if (mp)
 			printf("\nmemory allocation on the socket: %d",
 							port_numa[port_id]);
@@ -3580,9 +3580,9 @@ struct igb_ring_desc_16_bytes {
 	 */
 	tx_pkt_len = 0;
 	for (i = 0; i < nb_segs; i++) {
-		if (seg_lengths[i] > (unsigned) mbuf_data_size) {
+		if (seg_lengths[i] > mbuf_data_size[0]) {
 			printf("length[%u]=%u > mbuf_data_size=%u - give up\n",
-			       i, seg_lengths[i], (unsigned) mbuf_data_size);
+			       i, seg_lengths[i], mbuf_data_size[0]);
 			return;
 		}
 		tx_pkt_len = (uint16_t)(tx_pkt_len + seg_lengths[i]);
app/test-pmd/parameters.c +18 -6
index 15ce8c1..4db4987 100644
@@ -106,7 +106,9 @@
 	       "(flag: 1 for RX; 2 for TX; 3 for RX and TX).\n");
 	printf("  --socket-num=N: set socket from which all memory is allocated "
 	       "in NUMA mode.\n");
-	printf("  --mbuf-size=N: set the data size of mbuf to N bytes.\n");
+	printf("  --mbuf-size=N,[N1[,..Nn]: set the data size of mbuf to "
+	       "N bytes. If multiple numbers are specified the extra pools "
+	       "will be created to receive with packet split features\n");
 	printf("  --total-num-mbufs=N: set the number of mbufs to be allocated "
 	       "in mbuf pools.\n");
 	printf("  --max-pkt-len=N: set the maximum size of packet to N bytes.\n");
@@ -892,12 +894,22 @@
 				}
 			}
 			if (!strcmp(lgopts[opt_idx].name, "mbuf-size")) {
-				n = atoi(optarg);
-				if (n > 0 && n <= 0xFFFF)
-					mbuf_data_size = (uint16_t) n;
-				else
+				unsigned int mb_sz[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs, i;
+
+				nb_segs = parse_item_list(optarg, "mbuf-size",
+					MAX_SEGS_BUFFER_SPLIT, mb_sz, 0);
+				if (nb_segs <= 0)
 					rte_exit(EXIT_FAILURE,
-						 "mbuf-size should be > 0 and < 65536\n");
+						 "bad mbuf-size\n");
+				for (i = 0; i < nb_segs; i++) {
+					if (mb_sz[i] <= 0 || mb_sz[i] > 0xFFFF)
+						rte_exit(EXIT_FAILURE,
+							 "mbuf-size should be "
+							 "> 0 and < 65536\n");
+					mbuf_data_size[i] = (uint16_t) mb_sz[i];
+				}
+				mbuf_data_size_n = nb_segs;
 			}
 			if (!strcmp(lgopts[opt_idx].name, "total-num-mbufs")) {
 				n = atoi(optarg);
app/test-pmd/testpmd.c +35 -28

145 lines not on this page · raw message

app/test-pmd/testpmd.h +19 -5
index a4dfd4f..0975305 100644
@@ -42,6 +42,13 @@
  */
 #define RTE_MAX_SEGS_PER_PKT 255 /**< nb_segs is a 8-bit unsigned char. */
 
+/*
+ * The maximum number of segments per packet is used to configure
+ * buffer split feature, also specifies the maximum amount of
+ * optional Rx pools to allocate mbufs to split.
+ */
+#define MAX_SEGS_BUFFER_SPLIT 8 /**< nb_segs is a 8-bit unsigned char. */
+
 #define MAX_PKT_BURST 512
 #define DEF_PKT_BURST 32
 
@@ -403,7 +410,9 @@ struct queue_stats_mappings {
 extern uint8_t dcb_config;
 extern uint8_t dcb_test;
 
-extern uint16_t mbuf_data_size; /**< Mbuf data space size. */
+extern uint32_t mbuf_data_size_n;
+extern uint16_t mbuf_data_size[MAX_SEGS_BUFFER_SPLIT];
+/**< Mbuf data space size. */
 extern uint32_t param_total_num_mbufs;
 
 extern uint16_t stats_period;
@@ -615,17 +624,22 @@ struct mplsoudp_decap_conf {
 
 /* Mbuf Pools */
 static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+		    int name_size, unsigned int idx)
 {
-	snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	if (!idx)
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	else
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+			 sock_id, idx);
 }
 
 static inline struct rte_mempool *
-mbuf_pool_find(unsigned int sock_id)
+mbuf_pool_find(unsigned int sock_id, unsigned int idx)
 {
 	char pool_name[RTE_MEMPOOL_NAMESIZE];
 
-	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name));
+	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name), idx);
 	return rte_mempool_lookup((const char *)pool_name);
 }
 
doc/guides/testpmd_app_ug/run_app.rst +5 -2
index ec085c2..1eb0a10 100644
@@ -107,9 +107,12 @@ The command line options are:
     Set the socket from which all memory is allocated in NUMA mode,
     where 0 <= N < number of sockets on the board.
 
-*   ``--mbuf-size=N``
+*   ``--mbuf-size=N[,N1[,...Nn]``
 
-    Set the data size of the mbufs used to N bytes, where N < 65536. The default value is 2048.
+    Set the data size of the mbufs used to N bytes, where N < 65536.
+    The default value is 2048. If multiple mbuf-size values are specified the
+    extra memory pools will be created for allocating mbufs to receive packets
+    with buffer splittling features.
 
 *   ``--total-num-mbufs=N``
 
Signature · 1 line
1.8.3.1
#151 lore raw

Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 16 lines · Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

<...>

app/test-pmd/testpmd.h struct mplsoudp_decap_conf { · line 634

Quoted 12 lines

/* Mbuf Pools */
static inline void

app/test-pmd/testpmd.h +7 -2
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+		    int name_size, unsigned int idx)
  {
-	snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	if (!idx)
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	else
+		snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+			 sock_id, idx);

'mp_name' can theoretically overflow and gives a compiler warning, although not sure if this truncation is a problem in practice.

../app/test-pmd/testpmd.c: In function ‘rx_queue_setup’:
../app/test-pmd/testpmd.h:666:53: error: ‘%u’ directive output may be truncated writing between 1 and 10 bytes into a region of size between 0 and 7 [-Werror=format-truncation=]
666 | snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
| ^~
../app/test-pmd/testpmd.h:666:32: note: directive argument in the range [1, 4294967295]
666 | snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",

      |                                ^~~~~~~~~~~~~~~~~~~~~~~~
../app/test-pmd/testpmd.h:666:3: note: ‘snprintf’ output between 21 and 39 bytes into a destination of size 26
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
      |   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  667 |     sock_id, idx);
      |     ~~~~~~~~~~~~~

Any suggestion for fix? Can we shorten the string above, is it used somewhere else? Or casting variables to a smaller size may work too..

#152 lore raw

Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 48 lines · Ferruh Yigit
Quoted 16 lines · Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

<...>

Quoted 5 lines

  /* Mbuf Pools */
  static inline void

app/test-pmd/testpmd.h +2 -1
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+            int name_size, unsigned int idx)
Quoted 7 lines

  {
-    snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id); +    if (!idx)
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id); +    else
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", +             sock_id, idx);

'mp_name' can theoretically overflow and gives a compiler warning, although not sure if this truncation is a problem in practice.

../app/test-pmd/testpmd.c: In function ‘rx_queue_setup’:
../app/test-pmd/testpmd.h:666:53: error: ‘%u’ directive output may be truncated writing between 1 and 10 bytes into a region of size between 0 and 7 [-Werror=format-truncation=]
 666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
     |                                                     ^~
../app/test-pmd/testpmd.h:666:32: note: directive argument in the range [1, 4294967295]
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
      |                                ^~~~~~~~~~~~~~~~~~~~~~~~
../app/test-pmd/testpmd.h:666:3: note: ‘snprintf’ output between 21 and 39 bytes into a destination of size 26
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
      |   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  667 |     sock_id, idx);
      |     ~~~~~~~~~~~~~

Quoted 1 line

Any suggestion for fix? Can we shorten the string above, is it used somewhere else? Or casting variables to a smaller size may work too..

What do you think to following:

diff --git a/app/test-pmd/testpmd.h b/app/test-pmd/testpmd.h index 4cd0f967f0..4ac1c1f86e 100644
--- a/app/test-pmd/testpmd.h
+++ b/app/test-pmd/testpmd.h
@@ -661,10 +661,11 @@ mbuf_poolname_build(unsigned int sock_id, char *mp_name,
int name_size, unsigned int idx)

  {
         if (!idx)

- snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+ snprintf(mp_name, name_size, "mbuf_pool_s%u",
+ (uint16_t)sock_id);
else
- snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
- sock_id, idx);
+ snprintf(mp_name, name_size, "mbuf_pool_s%u_%u",
+ (uint16_t)sock_id, (uint16_t)idx);
}

static inline struct rte_mempool *

#153 lore raw

Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 91 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, October 16, 2020 18:39
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 55 lines · Ferruh Yigit
Quoted 16 lines · Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

<...>

Quoted 6 lines

  /* Mbuf Pools */
  static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name, +            int name_size, unsigned int idx)

Quoted 8 lines

  {
-    snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id); +    if (!idx)
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u", +sock_id);
+    else
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", +             sock_id, idx);

'mp_name' can theoretically overflow and gives a compiler warning, although not sure if this truncation is a problem in practice.

../app/test-pmd/testpmd.c: In function ‘rx_queue_setup’:
../app/test-pmd/testpmd.h:666:53: error: ‘%u’ directive output may be truncated writing between 1 and 10 bytes into a region of size between 0 and 7 [-Werror=format-truncation=]
 666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |                                                     ^~
../app/test-pmd/testpmd.h:666:32: note: directive argument in the range [1, 4294967295]
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |                                ^~~~~~~~~~~~~~~~~~~~~~~~
../app/test-pmd/testpmd.h:666:3: note: ‘snprintf’ output between 21 and 39 bytes into a destination of size 26
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 667 |     sock_id, idx);
      |     ~~~~~~~~~~~~~

Quoted 2 lines

Any suggestion for fix? Can we shorten the string above, is it used somewhere else? Or casting variables to a smaller size may work too..

What do you think to following:

diff --git a/app/test-pmd/testpmd.h b/app/test-pmd/testpmd.h index 4cd0f967f0..4ac1c1f86e 100644
--- a/app/test-pmd/testpmd.h
+++ b/app/test-pmd/testpmd.h
@@ -661,10 +661,11 @@ mbuf_poolname_build(unsigned int sock_id, char
*mp_name,
int name_size, unsigned int idx)

   {
          if (!idx)

- snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+ snprintf(mp_name, name_size, "mbuf_pool_s%u",
+ (uint16_t)sock_id);
else
- snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
- sock_id, idx);
+ snprintf(mp_name, name_size, "mbuf_pool_s%u_%u",
+ (uint16_t)sock_id, (uint16_t)idx);
}

Quoted 1 line

static inline struct rte_mempool *

Testpmd build mbuf pool names with mbuf_poolname_build() routine only (invoked from mbuf_pool_find/mbuf_pool_create). I suppose it is safe to replace "mbuf_pool_socket_" prefix with shorter one.

Say something like this: #define TESTPMD_MBUF_POOL_PFX "mbuf"

Truncating idx to uint16_t is OK, let me update the types of routines. sock_id seems to be OK either. If you agree on both - please, let me update the patch.

With best regards, Slava

#154 lore raw

Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 91 lines · Slava Ovsiienko
Quoted 91 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, October 16, 2020 18:39
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 55 lines · Ferruh Yigit
Quoted 16 lines · Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

<...>

Quoted 4 lines

  /* Mbuf Pools */
  static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)

app/test-pmd/testpmd.h struct mplsoudp_decap_conf { · line 634

Quoted 10 lines
app/test-pmd/testpmd.h +8 -1
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+            int name_size, unsigned int idx)
   {
-    snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+    if (!idx)
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u",
+sock_id);
+    else
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+             sock_id, idx);

'mp_name' can theoretically overflow and gives a compiler warning, although not sure if this truncation is a problem in practice.

../app/test-pmd/testpmd.c: In function ‘rx_queue_setup’:
../app/test-pmd/testpmd.h:666:53: error: ‘%u’ directive output may be truncated writing between 1 and 10 bytes into a region of size between 0 and 7 [-Werror=format-truncation=]
 666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |                                                     ^~
../app/test-pmd/testpmd.h:666:32: note: directive argument in the range [1, 4294967295]
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |                                ^~~~~~~~~~~~~~~~~~~~~~~~
../app/test-pmd/testpmd.h:666:3: note: ‘snprintf’ output between 21 and 39 bytes into a destination of size 26
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 667 |     sock_id, idx);
      |     ~~~~~~~~~~~~~

Quoted 2 lines

Any suggestion for fix? Can we shorten the string above, is it used somewhere else? Or casting variables to a smaller size may work too..

What do you think to following:

   diff --git a/app/test-pmd/testpmd.h b/app/test-pmd/testpmd.h
   index 4cd0f967f0..4ac1c1f86e 100644
   --- a/app/test-pmd/testpmd.h
   +++ b/app/test-pmd/testpmd.h
   @@ -661,10 +661,11 @@ mbuf_poolname_build(unsigned int sock_id, char
*mp_name,
                       int name_size, unsigned int idx)
    {
           if (!idx)
   -               snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
   +               snprintf(mp_name, name_size, "mbuf_pool_s%u",
   +                       (uint16_t)sock_id);
           else
   -               snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
   -                        sock_id, idx);
   +               snprintf(mp_name, name_size, "mbuf_pool_s%u_%u",
   +                       (uint16_t)sock_id, (uint16_t)idx);
    }
Quoted 8 lines
Quoted 1 line

static inline struct rte_mempool *

Testpmd build mbuf pool names with mbuf_poolname_build() routine only (invoked from mbuf_pool_find/mbuf_pool_create). I suppose it is safe to replace "mbuf_pool_socket_" prefix with shorter one.

Say something like this: #define TESTPMD_MBUF_POOL_PFX "mbuf"

Just 'mbuf' may not give enough context in the logs, what about "mbuf_pool_%u_%u"?

Quoted 3 lines

Truncating idx to uint16_t is OK, let me update the types of routines. sock_id seems to be OK either. If you agree on both - please, let me update the patch.

If variables are not truncated, it leaves very small place for prefix, 3 chars if I calculate correctly, so it seems we need to truncate it anyway.

#155 lore raw

Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 111 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, October 16, 2020 18:52
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 73 lines · Slava Ovsiienko
Quoted 73 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, October 16, 2020 18:39
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 56 lines · Ferruh Yigit
Quoted 16 lines · Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

<...>

Quoted 4 lines

  /* Mbuf Pools */
  static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)

app/test-pmd/testpmd.h struct mplsoudp_decap_conf { · line 634

Quoted 10 lines
app/test-pmd/testpmd.h +8 -1
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+            int name_size, unsigned int idx)
   {
-    snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+    if (!idx)
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u",
+sock_id);
+    else
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+             sock_id, idx);

'mp_name' can theoretically overflow and gives a compiler warning, although not sure if this truncation is a problem in practice.

../app/test-pmd/testpmd.c: In function ‘rx_queue_setup’:
../app/test-pmd/testpmd.h:666:53: error: ‘%u’ directive output may be truncated writing between 1 and 10 bytes into a region of size between
0 and 7 [-Werror=format-truncation=]
 666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |                                                     ^~
../app/test-pmd/testpmd.h:666:32: note: directive argument in the range [1, 4294967295]
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |                                ^~~~~~~~~~~~~~~~~~~~~~~~
../app/test-pmd/testpmd.h:666:3: note: ‘snprintf’ output between 21 and 39 bytes into a destination of size 26
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 667 |     sock_id, idx);
      |     ~~~~~~~~~~~~~

Quoted 2 lines

Any suggestion for fix? Can we shorten the string above, is it used somewhere else? Or casting variables to a smaller size may work too..

What do you think to following:

Quoted 10 lines
Quoted 10 lines
   diff --git a/app/test-pmd/testpmd.h b/app/test-pmd/testpmd.h
   index 4cd0f967f0..4ac1c1f86e 100644
   --- a/app/test-pmd/testpmd.h
   +++ b/app/test-pmd/testpmd.h
   @@ -661,10 +661,11 @@ mbuf_poolname_build(unsigned int sock_id,
char *mp_name,
                       int name_size, unsigned int idx)
    {
           if (!idx)

- snprintf(mp_name, name_size, "mbuf_pool_socket_%u",

sock_id);

Quoted 15 lines
Quoted 10 lines

+ snprintf(mp_name, name_size, "mbuf_pool_s%u",
+ (uint16_t)sock_id);
else
- snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
- sock_id, idx);
+ snprintf(mp_name, name_size, "mbuf_pool_s%u_%u",
+ (uint16_t)sock_id, (uint16_t)idx);
}

static inline struct rte_mempool *

Testpmd build mbuf pool names with mbuf_poolname_build() routine only (invoked from mbuf_pool_find/mbuf_pool_create). I suppose it is safe to replace "mbuf_pool_socket_" prefix with shorter one.

Quoted 5 lines
Quoted 2 lines

Say something like this: #define TESTPMD_MBUF_POOL_PFX "mbuf"

Just 'mbuf' may not give enough context in the logs, what about "mbuf_pool_%u_%u"?

It is place in the pool structure, so it is an object name and tightly coupled with the pool. OK, let's try "mb_pool" ?

#156 lore raw

Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 111 lines · Slava Ovsiienko
Quoted 111 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, October 16, 2020 18:52
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 73 lines · Slava Ovsiienko
Quoted 73 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Friday, October 16, 2020 18:39
To: Slava Ovsiienko <[email protected]>; [email protected]
Cc: NBU-Contact-Thomas Monjalon <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [dpdk-dev] [PATCH v11 2/6] app/testpmd: add multiple pools per core creation

Quoted 53 lines · Ferruh Yigit
Quoted 16 lines · Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

<...>

Quoted 4 lines

  /* Mbuf Pools */
  static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)

app/test-pmd/testpmd.h struct mplsoudp_decap_conf { · line 634

Quoted 10 lines
app/test-pmd/testpmd.h +8 -1
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+            int name_size, unsigned int idx)
    {
-    snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+    if (!idx)
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u",
+sock_id);
+    else
+        snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
+             sock_id, idx);

'mp_name' can theoretically overflow and gives a compiler warning, although not sure if this truncation is a problem in practice.

../app/test-pmd/testpmd.c: In function ‘rx_queue_setup’:
../app/test-pmd/testpmd.h:666:53: error: ‘%u’ directive output may be truncated writing between 1 and 10 bytes into a region of size between
0 and 7 [-Werror=format-truncation=]
 666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |                                                     ^~
../app/test-pmd/testpmd.h:666:32: note: directive argument in the range [1, 4294967295]
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |                                ^~~~~~~~~~~~~~~~~~~~~~~~
../app/test-pmd/testpmd.h:666:3: note: ‘snprintf’ output between 21 and 39 bytes into a destination of size 26
  666 |   snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u", |   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Quoted 5 lines
     667 |     sock_id, idx);
         |     ~~~~~~~~~~~~~

Any suggestion for fix? Can we shorten the string above, is it used somewhere else? Or casting variables to a smaller size may work too..

What do you think to following:

Quoted 10 lines
Quoted 10 lines
    diff --git a/app/test-pmd/testpmd.h b/app/test-pmd/testpmd.h
    index 4cd0f967f0..4ac1c1f86e 100644
    --- a/app/test-pmd/testpmd.h
    +++ b/app/test-pmd/testpmd.h
    @@ -661,10 +661,11 @@ mbuf_poolname_build(unsigned int sock_id,

char *mp_name,
int name_size, unsigned int idx)
{
if (!idx)
- snprintf(mp_name, name_size, "mbuf_pool_socket_%u",

sock_id);

Quoted 15 lines
Quoted 10 lines

+ snprintf(mp_name, name_size, "mbuf_pool_s%u",
+ (uint16_t)sock_id);
else
- snprintf(mp_name, name_size, "mbuf_pool_socket_%u_%u",
- sock_id, idx);
+ snprintf(mp_name, name_size, "mbuf_pool_s%u_%u",
+ (uint16_t)sock_id, (uint16_t)idx);
}

static inline struct rte_mempool *

Testpmd build mbuf pool names with mbuf_poolname_build() routine only (invoked from mbuf_pool_find/mbuf_pool_create). I suppose it is safe to replace "mbuf_pool_socket_" prefix with shorter one.

Quoted 8 lines
Quoted 5 lines
Quoted 2 lines

Say something like this: #define TESTPMD_MBUF_POOL_PFX "mbuf"

Just 'mbuf' may not give enough context in the logs, what about "mbuf_pool_%u_%u"?

It is place in the pool structure, so it is an object name and tightly coupled with the pool. OK, let's try "mb_pool" ?

sounds good

#157 lore raw

PATCH v11 3/6 app/testpmd: add buffer split offload configuration

In reply to Viacheslav Ovsiienko

This patch add support for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT providing per queue configuration for this offload.

---
 app/test-pmd/cmdline.c | 21 +++++++++++----------
 app/test-pmd/config.c  |  9 +++++++++
 2 files changed, 20 insertions(+), 10 deletions(-)
app/test-pmd/cmdline.c +11 -10
index 19b1896..0ef87c9 100644
@@ -883,16 +883,16 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"port config <port_id> rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"     Enable or disable a per port Rx offloading"
 			" on all Rx queues of a port\n\n"
 
 			"port (port_id) rxq (queue_id) rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"    Enable or disable a per queue Rx offloading"
 			" only on a specific Rx queue\n\n"
 
@@ -18435,7 +18435,8 @@ struct cmd_config_per_port_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc#rss_hash");
+			   "scatter#buffer_split#timestamp#security#"
+			   "keep_crc#rss_hash");
 cmdline_parse_token_string_t cmd_config_per_port_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_port_rx_offload_result,
@@ -18515,8 +18516,8 @@ struct cmd_config_per_port_rx_offload_result {
 	.help_str = "port config <port_id> rx_offload vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc|rss_hash "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc|rss_hash on|off",
 	.tokens = {
 		(void *)&cmd_config_per_port_rx_offload_result_port,
 		(void *)&cmd_config_per_port_rx_offload_result_config,
@@ -18565,7 +18566,7 @@ struct cmd_config_per_queue_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc");
+			   "scatter#buffer_split#timestamp#security#keep_crc");
 cmdline_parse_token_string_t cmd_config_per_queue_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_queue_rx_offload_result,
@@ -18621,8 +18622,8 @@ struct cmd_config_per_queue_rx_offload_result {
 		    "vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc on|off",
 	.tokens = {
 		(void *)&cmd_config_per_queue_rx_offload_result_port,
 		(void *)&cmd_config_per_queue_rx_offload_result_port_id,
app/test-pmd/config.c +9 -0
index c554dd0..0f041e1 100644
@@ -1092,6 +1092,15 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 			printf("off\n");
 	}
 
+	if (dev_info.rx_offload_capa & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+		printf("RX offload buffer split:       ");
+		if (ports[port_id].dev_conf.rxmode.offloads &
+		    RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT)
+			printf("on\n");
+		else
+			printf("off\n");
+	}
+
 	if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_VLAN_INSERT) {
 		printf("VLAN insert:                   ");
 		if (ports[port_id].dev_conf.txmode.offloads &
Signature · 1 line
1.8.3.1
#158 lore raw

PATCH v11 4/6 app/testpmd: add rxpkts commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxpkts=X[,Y]

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command:

testpmd> set rxpkts (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

Sets the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only). Optionally the multiple memory pools can be specified with --mbuf-size command line parameter and the mbufs to receive will be allocated sequentially from these extra memory pools.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 48 ++++++++++++++++++++++-
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  7 ++++
 app/test-pmd/testpmd.h                      | 11 +++++-
 doc/guides/testpmd_app_ug/run_app.rst       |  9 +++++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 21 +++++++++-
 7 files changed, 165 insertions(+), 7 deletions(-)
app/test-pmd/cmdline.c +58 -3
index 0ef87c9..a195e37 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxpkts (x[,y]*)\n"
+			"    Set the length of each segment to scatter"
+			" packets on receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set txpkts (x[,y]*)\n"
 			"    Set the length of each segment of TXONLY"
 			" and optionally CSUM packets.\n\n"
@@ -3907,6 +3913,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxpkts_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxpkts;
+	cmdline_fixed_string_t seg_lengths;
+};
+
+static void
+cmd_set_rxpkts_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxpkts_result *res;
+	unsigned int seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_lengths, "segment lengths",
+				  MAX_SEGS_BUFFER_SPLIT, seg_lengths, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_segments(seg_lengths, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxpkts_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxpkts_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 rxpkts, "rxpkts");
+cmdline_parse_token_string_t cmd_set_rxpkts_lengths =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxpkts_result,
+				 seg_lengths, NULL);
+
+cmdline_parse_inst_t cmd_set_rxpkts = {
+	.f = cmd_set_rxpkts_parsed,
+	.data = NULL,
+	.help_str = "set rxpkts <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxpkts_keyword,
+		(void *)&cmd_set_rxpkts_name,
+		(void *)&cmd_set_rxpkts_lengths,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF TXONLY PACKETS *** */
 
 struct cmd_set_txpkts_result {
@@ -7535,6 +7587,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxpkts"))
+		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
 		show_tx_pkt_segments();
 	else if (!strcmp(res->what, "txtimes"))
@@ -7547,12 +7601,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19825,6 +19879,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
 	(cmdline_parse_inst_t *)&cmd_set_txtimes,
app/test-pmd/config.c +46 -2

62 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index 4db4987..e4e3635 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
 	printf("  --txonly-multi-flow: generate multiple flows in txonly mode\n");
@@ -662,6 +663,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
 		{ "disable-link-check",		0, 0, 0 },
@@ -1272,6 +1274,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
+				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs;
+
+				nb_segs = parse_item_list
+						(optarg, "rxpkt segments",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_len, 0);
+				if (nb_segs > 0)
+					set_rx_pkt_segments(seg_len, nb_segs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxpkts\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "txpkts")) {
 				unsigned seg_lengths[RTE_MAX_SEGS_PER_PKT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +7 -0
index 7e6ef80..f88c1e2 100644
@@ -210,6 +210,13 @@ struct fwd_engine * fwd_engines[] = {
 uint8_t f_quit;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 uint16_t tx_pkt_length = TXONLY_DEF_PACKET_LEN; /**< TXONLY packet length. */
app/test-pmd/testpmd.h +10 -1
index 0975305..c5c942a 100644
@@ -430,6 +430,13 @@ struct queue_stats_mappings {
 extern struct rte_fdir_conf fdir_conf;
 
 /*
+ * Configuration of packet segments used to scatter received packets
+ * if some of split features is configured.
+ */
+extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+
+/*
  * Configuration of packet segments used by the "txonly" processing engine.
  */
 #define TXONLY_DEF_PACKET_LEN 64
@@ -836,7 +843,9 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_record_core_cycles(uint8_t on_off);
 void set_record_burst_stats(uint8_t on_off);
 void set_verbose_level(uint16_t vb_level);
-void set_tx_pkt_segments(unsigned *seg_lengths, unsigned nb_segs);
+void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
+void show_rx_pkt_segments(void);
+void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
 void show_tx_pkt_times(void);
doc/guides/testpmd_app_ug/run_app.rst +9 -0
index 1eb0a10..463b76c 100644
@@ -361,6 +361,15 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxpkts=X[,Y]``
+
+    Set the length of segments to scatter packets on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+    Optionally the multiple memory pools can be specified with --mbuf-size
+    command line parameter and the mbufs to receive will be allocated
+    sequentially from these extra memory pools.
+
 *   ``--txpkts=X[,Y]``
 
     Set TX segment sizes or total packet length. Valid for ``tx-only``
doc/guides/testpmd_app_ug/testpmd_funcs.rst +20 -1
index 43c0ea0..b9fad3a 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxpkts``: Packets to RX split configuration.
+
 * ``txpkts``: Packets to TX configuration.
 
 * ``txtimes``: Burst time pattern for Tx only mode.
@@ -774,6 +776,23 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxpkts
+~~~~~~~~~~
+
+Set the length of segments to scatter packets on receiving if split
+feature is engaged. Affects only the queues configured with split offloads
+(currently BUFFER_SPLIT is supported only). Optionally the multiple memory
+pools can be specified with --mbuf-size command line parameter and the mbufs
+to receive will be allocated sequentially from these extra memory pools (the
+mbuf for the first segment is allocated from the first pool, the second one
+from the second pool, and so on, if segment number is greater then pool's the
+mbuf for remaining segments will be allocated from the last valid pool).
+
+   testpmd> set rxpkts (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. Zero value
+means to use the corresponding memory pool data buffer size.
+
 set txpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#159 lore raw

PATCH v11 5/6 app/testpmd: add rxoffs commands and parameters

In reply to Viacheslav Ovsiienko

Add command line parameter:

--rxoffs=X[,Y]

Sets the offsets of packet segments from the beginning of the receiving buffer if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).

Add interactive mode command, providing the same:

testpmd> set rxoffs (x[,y]*)

Where x[,y]* represents a CSV list of values, without white space.

---
 app/test-pmd/cmdline.c                      | 61 +++++++++++++++++++++++++++--
 app/test-pmd/config.c                       | 44 +++++++++++++++++++++
 app/test-pmd/parameters.c                   | 15 +++++++
 app/test-pmd/testpmd.c                      |  2 +
 app/test-pmd/testpmd.h                      |  4 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  6 +++
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 17 +++++++-
 7 files changed, 145 insertions(+), 4 deletions(-)
app/test-pmd/cmdline.c +58 -3
index a195e37..e09df36 100644
@@ -183,7 +183,7 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"show (rxq|txq) info (port_id) (queue_id)\n"
 			"    Display information for configured RX/TX queue.\n\n"
 
-			"show config (rxtx|cores|fwd|rxpkts|txpkts)\n"
+			"show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts)\n"
 			"    Display the given configuration.\n\n"
 
 			"read rxd (port_id) (queue_id) (rxd_id)\n"
@@ -294,6 +294,12 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"    Set the transmit delay time and number of retries,"
 			" effective when retry is enabled.\n\n"
 
+			"set rxoffs (x[,y]*)\n"
+			"    Set the offset of each packet segment on"
+			" receiving if split feature is engaged."
+			" Affects only the queues configured with split"
+			" offloads.\n\n"
+
 			"set rxpkts (x[,y]*)\n"
 			"    Set the length of each segment to scatter"
 			" packets on receiving if split feature is engaged."
@@ -3913,6 +3919,52 @@ struct cmd_set_log_result {
 	},
 };
 
+/* *** SET SEGMENT OFFSETS OF RX PACKETS SPLIT *** */
+
+struct cmd_set_rxoffs_result {
+	cmdline_fixed_string_t cmd_keyword;
+	cmdline_fixed_string_t rxoffs;
+	cmdline_fixed_string_t seg_offsets;
+};
+
+static void
+cmd_set_rxoffs_parsed(void *parsed_result,
+		      __rte_unused struct cmdline *cl,
+		      __rte_unused void *data)
+{
+	struct cmd_set_rxoffs_result *res;
+	unsigned int seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+	unsigned int nb_segs;
+
+	res = parsed_result;
+	nb_segs = parse_item_list(res->seg_offsets, "segment offsets",
+				  MAX_SEGS_BUFFER_SPLIT, seg_offsets, 0);
+	if (nb_segs > 0)
+		set_rx_pkt_offsets(seg_offsets, nb_segs);
+}
+
+cmdline_parse_token_string_t cmd_set_rxoffs_keyword =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_rxoffs_name =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 rxoffs, "rxoffs");
+cmdline_parse_token_string_t cmd_set_rxoffs_offsets =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxoffs_result,
+				 seg_offsets, NULL);
+
+cmdline_parse_inst_t cmd_set_rxoffs = {
+	.f = cmd_set_rxoffs_parsed,
+	.data = NULL,
+	.help_str = "set rxoffs <len0[,len1]*>",
+	.tokens = {
+		(void *)&cmd_set_rxoffs_keyword,
+		(void *)&cmd_set_rxoffs_name,
+		(void *)&cmd_set_rxoffs_offsets,
+		NULL,
+	},
+};
+
 /* *** SET SEGMENT LENGTHS OF RX PACKETS SPLIT *** */
 
 struct cmd_set_rxpkts_result {
@@ -7587,6 +7639,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
 		fwd_lcores_config_display();
 	else if (!strcmp(res->what, "fwd"))
 		pkt_fwd_config_display(&cur_fwd_config);
+	else if (!strcmp(res->what, "rxoffs"))
+		show_rx_pkt_offsets();
 	else if (!strcmp(res->what, "rxpkts"))
 		show_rx_pkt_segments();
 	else if (!strcmp(res->what, "txpkts"))
@@ -7601,12 +7655,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
 	TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-				 "rxtx#cores#fwd#rxpkts#txpkts#txtimes");
+				 "rxtx#cores#fwd#rxoffs#rxpkts#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
 	.f = cmd_showcfg_parsed,
 	.data = NULL,
-	.help_str = "show config rxtx|cores|fwd|rxpkts|txpkts|txtimes",
+	.help_str = "show config rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes",
 	.tokens = {
 		(void *)&cmd_showcfg_show,
 		(void *)&cmd_showcfg_port,
@@ -19879,6 +19933,7 @@ struct cmd_showport_macs_result {
 	(cmdline_parse_inst_t *)&cmd_reset,
 	(cmdline_parse_inst_t *)&cmd_set_numbers,
 	(cmdline_parse_inst_t *)&cmd_set_log,
+	(cmdline_parse_inst_t *)&cmd_set_rxoffs,
 	(cmdline_parse_inst_t *)&cmd_set_rxpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txpkts,
 	(cmdline_parse_inst_t *)&cmd_set_txsplit,
app/test-pmd/config.c +44 -0

50 lines not on this page · raw message

app/test-pmd/parameters.c +15 -0
index e4e3635..2298ba5 100644
@@ -184,6 +184,7 @@
 	       "(0 <= mapping <= %d).\n", RTE_ETHDEV_QUEUE_STAT_CNTRS - 1);
 	printf("  --no-flush-rx: Don't flush RX streams before forwarding."
 	       " Used mainly with PCAP drivers.\n");
+	printf("  --rxoffs=X[,Y]*: set RX segment offsets for split.\n");
 	printf("  --rxpkts=X[,Y]*: set RX segment sizes to split.\n");
 	printf("  --txpkts=X[,Y]*: set TX segment sizes"
 		" or total packet length.\n");
@@ -663,6 +664,7 @@
 		{ "rx-queue-stats-mapping",	1, 0, 0 },
 		{ "no-flush-rx",	0, 0, 0 },
 		{ "flow-isolate-all",	        0, 0, 0 },
+		{ "rxoffs",			1, 0, 0 },
 		{ "rxpkts",			1, 0, 0 },
 		{ "txpkts",			1, 0, 0 },
 		{ "txonly-multi-flow",		0, 0, 0 },
@@ -1274,6 +1276,19 @@
 						 "invalid RX queue statistics mapping config entered\n");
 				}
 			}
+			if (!strcmp(lgopts[opt_idx].name, "rxoffs")) {
+				unsigned int seg_off[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_offs;
+
+				nb_offs = parse_item_list
+						(optarg, "rxpkt offsets",
+						 MAX_SEGS_BUFFER_SPLIT,
+						 seg_off, 0);
+				if (nb_offs > 0)
+					set_rx_pkt_offsets(seg_off, nb_offs);
+				else
+					rte_exit(EXIT_FAILURE, "bad rxoffs\n");
+			}
 			if (!strcmp(lgopts[opt_idx].name, "rxpkts")) {
 				unsigned int seg_len[MAX_SEGS_BUFFER_SPLIT];
 				unsigned int nb_segs;
app/test-pmd/testpmd.c +2 -0
index f88c1e2..580178d 100644
@@ -215,6 +215,8 @@ struct fwd_engine * fwd_engines[] = {
  */
 uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
app/test-pmd/testpmd.h +4 -0
index c5c942a..dc37148 100644
@@ -435,6 +435,8 @@ struct queue_stats_mappings {
  */
 extern uint16_t rx_pkt_seg_lengths[MAX_SEGS_BUFFER_SPLIT];
 extern uint8_t  rx_pkt_nb_segs; /**< Number of segments to split */
+extern uint16_t rx_pkt_seg_offsets[MAX_SEGS_BUFFER_SPLIT];
+extern uint8_t  rx_pkt_nb_offs; /**< Number of specified offsets */
 
 /*
  * Configuration of packet segments used by the "txonly" processing engine.
@@ -845,6 +847,8 @@ void vlan_tpid_set(portid_t port_id, enum rte_vlan_type vlan_type,
 void set_verbose_level(uint16_t vb_level);
 void set_rx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_rx_pkt_segments(void);
+void set_rx_pkt_offsets(unsigned int *seg_offsets, unsigned int nb_offs);
+void show_rx_pkt_offsets(void);
 void set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs);
 void show_tx_pkt_segments(void);
 void set_tx_pkt_times(unsigned int *tx_times);
doc/guides/testpmd_app_ug/run_app.rst +6 -0
index 463b76c..9b0a84a 100644
@@ -361,6 +361,12 @@ The command line options are:
 
     Don't flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
 
+*   ``--rxoffs=X[,Y]``
+
+    Set the offsets of packet segments on receiving if split
+    feature is engaged. Affects only the queues configured
+    with split offloads (currently BUFFER_SPLIT is supported only).
+
 *   ``--rxpkts=X[,Y]``
 
     Set the length of segments to scatter packets on receiving if split
doc/guides/testpmd_app_ug/testpmd_funcs.rst +16 -1
index b9fad3a..e73530e 100644
@@ -273,7 +273,7 @@ show config
 Displays the configuration of the application.
 The configuration comes from the command-line, the runtime or the application defaults::
 
-   testpmd> show config (rxtx|cores|fwd|rxpkts|txpkts|txtimes)
+   testpmd> show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes)
 
 The available information categories are:
 
@@ -283,6 +283,8 @@ The available information categories are:
 
 * ``fwd``: Packet forwarding configuration.
 
+* ``rxoffs``: Packet offsets for RX split.
+
 * ``rxpkts``: Packets to RX split configuration.
 
 * ``txpkts``: Packets to TX configuration.
@@ -776,6 +778,19 @@ When retry is enabled, the transmit delay time and number of retries can also be
 
    testpmd> set burst tx delay (microseconds) retry (num)
 
+set rxoffs
+~~~~~~~~~~
+
+Set the offsets of segments relating to the data buffer beginning on receiving
+if split feature is engaged. Affects only the queues configured with split
+offloads (currently BUFFER_SPLIT is supported only).
+
+   testpmd> set rxoffs (x[,y]*)
+
+Where x[,y]* represents a CSV list of values, without white space. If the list
+of offsets is shorter than the list of segments the zero offsets will be used
+for the remaining segments.
+
 set rxpkts
 ~~~~~~~~~~
 
Signature · 1 line
1.8.3.1
#160 lore raw

PATCH v11 6/6 app/testpmd: add extended Rx queue setup

In reply to Viacheslav Ovsiienko

If Rx queue is configured with split feature the extended setup with specified segment sizes and pool will be performed.

---
 app/test-pmd/cmdline.c | 12 ++++++------
 app/test-pmd/testpmd.c | 49 +++++++++++++++++++++++++++++++++++++++++++++++--
 app/test-pmd/testpmd.h |  5 +++++
 3 files changed, 58 insertions(+), 8 deletions(-)
app/test-pmd/cmdline.c +6 -6
index e09df36..07ee4e4 100644
@@ -2945,12 +2945,12 @@ struct cmd_setup_rxtx_queue {
 				rxring_numa[res->portid]);
 			return;
 		}
-		ret = rte_eth_rx_queue_setup(res->portid,
-					     res->qid,
-					     port->nb_rx_desc[res->qid],
-					     socket_id,
-					     &port->rx_conf[res->qid],
-					     mp);
+		ret = rx_queue_setup(res->portid,
+				     res->qid,
+				     port->nb_rx_desc[res->qid],
+				     socket_id,
+				     &port->rx_conf[res->qid],
+				     mp);
 		if (ret)
 			printf("Failed to setup RX queue\n");
 	} else {
app/test-pmd/testpmd.c +47 -2

67 lines not on this page · raw message

app/test-pmd/testpmd.h +5 -0
index dc37148..c7b2cb4 100644
@@ -896,6 +896,11 @@ void port_rss_reta_info(portid_t port_id,
 
 void set_vf_traffic(portid_t port_id, uint8_t is_rx, uint16_t vf, uint8_t on);
 
+int
+rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
+	       uint16_t nb_rx_desc, unsigned int socket_id,
+	       struct rte_eth_rxconf *rx_conf, struct rte_mempool *mp);
+
 int set_queue_rate_limit(portid_t port_id, uint16_t queue_idx, uint16_t rate);
 int set_vf_rate_limit(portid_t port_id, uint16_t vf, uint16_t rate,
 				uint64_t q_msk);
Signature · 1 line
1.8.3.1
#161 lore raw

v12 cover ethdev: introduce Rx buffer split

In reply to Slava Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rx_conf.rx_seg is NULL, rx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---

v1: http://patches.dpdk.org/patch/79594/
v2: http://patches.dpdk.org/patch/79893/
- add feature support to mlx5 PMD

v3: http://patches.dpdk.org/patch/80389/
- rte_eth_rx_queue_setup_ex is renamed to rte_eth_rxseg_queue_setup
- DEV_RX_OFFLOAD_BUFFER_SPLIT is renamed to RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT
- commit message update
- documentaion provided
- release notes update
- minor bug fixes in testpmd related part

v4: http://patches.dpdk.org/patch/80401/
- common part of rx_queue_setup/rxseg_queue_setup

v5: http://patches.dpdk.org/patch/80609/
- refactored to approach of providing split configuration in the rte_eth_rxconf structure instead of introducing the new API routine
- added support for rxoffs command to testpmd to provide segment offsets for complete testing of split configurations
- patchset is split into two parts - PMD part will be presented as separate series

v6: http://patches.dpdk.org/patch/80785/
- wordy comments rephrased
- typos fixed
- rte_eth_rx_queue_setup configuration check isolated for two main options
- the rest of comments addressed

v7: http://patches.dpdk.org/patch/80978/
- description as union of feature dedicated segment split description
- the split limitations reported in the rte_eth_dev_info->rx_seg_capa
- comments made less wordy

v8: - bitfield instead of uint8;
- named union to avoid warinings;
- typos fixed

v9: http://patches.dpdk.org/patch/81080/
- simplified rte_eth_rxseg union
- moved out wordy comment from rte_eth_rxconf
- rte_eth_rxseg_capa uint32 bitfield and naming
- added experimental tags

v10: - missed in mailing list, see v11 for changes

v11: http://patches.dpdk.org/patch/81116/
- documentation cleanup (removed rxq_info_get mention)
- max_nseg moved from the recored and converted to uint16_t;
- reserved uint16_t is added to rte_eth_rxseg_capa;

v12: - __rte_experimental at structure declarations
converted to the tag in the comment
- testpmd pool names shortened, prefix provided
- comments reworded

---
Viacheslav Ovsiienko (6):
ethdev: introduce Rx buffer split app/testpmd: add multiple pools per core creation app/testpmd: add buffer split offload configuration app/testpmd: add rxpkts commands and parameters app/testpmd: add rxoffs commands and parameters app/testpmd: add extended Rx queue setup

 app/test-pmd/bpf_cmd.c                      |   4 +-
 app/test-pmd/cmdline.c                      | 151 +++++++++++++++++++++++----
 app/test-pmd/config.c                       | 107 ++++++++++++++++++-
 app/test-pmd/parameters.c                   |  54 ++++++++--
 app/test-pmd/testpmd.c                      | 121 ++++++++++++++++------
 app/test-pmd/testpmd.h                      |  48 +++++++--
 doc/guides/nics/features.rst                |  14 +++
 doc/guides/rel_notes/deprecation.rst        |   5 -
 doc/guides/rel_notes/release_20_11.rst      |   9 ++
 doc/guides/testpmd_app_ug/run_app.rst       |  22 +++-
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |  36 ++++++-
 lib/librte_ethdev/rte_ethdev.c              | 153 +++++++++++++++++++++++-----
 lib/librte_ethdev/rte_ethdev.h              |  95 ++++++++++++++++-
 13 files changed, 716 insertions(+), 103 deletions(-)
Signature · 1 line
1.8.3.1
#162 lore raw

PATCH v12 1/6 ethdev: introduce Rx buffer split

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rrx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides the single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

    seg0 - pool0, len0=14B, off0=2
    seg1 - pool1, len1=20B, off1=128B
    seg2 - pool2, len2=20B, off2=0B
    seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

---
 doc/guides/nics/features.rst           |  14 +++
 doc/guides/rel_notes/deprecation.rst   |   5 --
 doc/guides/rel_notes/release_20_11.rst |   9 ++
 lib/librte_ethdev/rte_ethdev.c         | 153 +++++++++++++++++++++++++++------
 lib/librte_ethdev/rte_ethdev.h         |  95 +++++++++++++++++++-
 5 files changed, 245 insertions(+), 31 deletions(-)
doc/guides/nics/features.rst +14 -0
index dd8c955..16e00b8 100644
@@ -185,6 +185,20 @@ Supports receiving segmented mbufs.
 * **[related]    eth_dev_ops**: ``rx_pkt_burst``.
 
 
+.. _nic_features_buffer_split:
+
+Buffer Split on Rx
+------------------
+
+Scatters the packets being received on specified boundaries to segmented mbufs.
+
+* **[uses]       rte_eth_rxconf,rte_eth_rxmode**: ``offloads:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[uses]       rte_eth_rxconf**: ``rx_conf.rx_seg, rx_conf.rx_nseg``.
+* **[implements] datapath**: ``Buffer Split functionality``.
+* **[provides]   rte_eth_dev_info**: ``rx_offload_capa:RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT``.
+* **[related] API**: ``rte_eth_rx_queue_setup()``.
+
+
 .. _nic_features_lro:
 
 LRO
doc/guides/rel_notes/deprecation.rst +0 -5
index d1f5ed3..dde464b 100644
@@ -138,11 +138,6 @@ Deprecation Notices
   In 19.11 PMDs will still update the field even when the offload is not
   enabled.
 
-* ethdev: Add new fields to ``rte_eth_rxconf`` to configure the receiving
-  queues to split ingress packets into multiple segments according to the
-  specified lengths into the buffers allocated from the specified
-  memory pools. The backward compatibility to existing API is preserved.
-
 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
   will be removed in 21.11.
   Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
doc/guides/rel_notes/release_20_11.rst +9 -0
index f8686a5..e12cf96 100644
@@ -89,6 +89,12 @@ New Features
     as flow action.
   * Added new flow APIs to create/update/destroy/query shared action.
 
+* **Introduced extended buffer description for receiving.**
+
+  Added the extended Rx buffer description for Rx queue setup routine
+  providing the individual settings for each Rx segment with maximal size,
+  buffer offset and memory pool to allocate data buffers from.
+
 * **Updated Broadcom bnxt driver.**
 
   Updated the Broadcom bnxt driver with new features and improvements, including:
@@ -321,6 +327,9 @@ API Changes
   As the data of ``uint8_t`` will be truncated when queue number under
   a TC is greater than 256.
 
+* ethdev: Added fields rx_seg and rx_nseg to rte_eth_rxconf structure
+  to provide extended description of the receiving buffer.
+
 * vhost: Moved vDPA APIs from experimental to stable.
 
 * rawdev: Added a structure size parameter to the functions
lib/librte_ethdev/rte_ethdev.c +128 -25

187 lines not on this page · raw message

lib/librte_ethdev/rte_ethdev.h +94 -1

139 lines not on this page · raw message

Signature · 1 line
1.8.3.1
#163 lore raw
Quoted 102 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length,
configures "split point" */
uint16_t offset; /* data offset from beginning
of mbuf data buffer */
uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
rx_seg - pointer the array of segment descriptions, each element
describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf.
This array allows to specify the different settings for each segment in individual fashion.
rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rrx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides the single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:

     seg0 - pool0, len0=14B, off0=2
     seg1 - pool1, len1=20B, off1=128B
     seg2 - pool2, len2=20B, off2=0B
     seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
seg2 - 20B @ 0 in mbuf from pool2
seg3 - 512B @ 0 in mbuf from pool3
seg4 - 512B @ 0 in mbuf from pool3
seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

<...>

lib/librte_ethdev/rte_ethdev.c struct rte_eth_dev * · line 1924

Quoted 8 lines
lib/librte_ethdev/rte_ethdev.c +7 -1
-	mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+	if (mp != NULL) {
+		/* Single pool configuration check. */
+		if (rx_conf->rx_nseg != 0) {
+			RTE_ETHDEV_LOG(ERR,
+				       "Ambiguous segment configuration\n");
+			return -EINVAL;
+		}

'rte_eth_rx_queue_setup()' accepts 'rx_conf' to be NULL, there is a NULL check for it somewhere below the function :)

Above will crash in that case, need to fix this, I am on it, any idea welcome.

#164 lore raw
Quoted 115 lines · Ferruh Yigit
Quoted 102 lines · Viacheslav Ovsiienko

The DPDK datapath in the transmit direction is very flexible. An application can build the multi-segment packet and manages almost all data aspects - the memory pools where segments are allocated from, the segment lengths, the memory attributes like external buffers, registered for DMA, etc.

In the receiving direction, the datapath is much less flexible, an application can only specify the memory pool to configure the receiving queue and nothing more. In order to extend receiving datapath capabilities it is proposed to add the way to provide extended information how to split the packets being received.

The new offload flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT in device capabilities is introduced to present the way for PMD to report to application about supporting Rx packet split to configurable segments. Prior invoking the rte_eth_rx_queue_setup() routine application should check RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag.

The following structure is introduced to specify the Rx packet segment for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT offload:

struct rte_eth_rxseg_split {

     struct rte_mempool *mp; /* memory pools to allocate segment from */ uint16_t length; /* segment maximal data length, configures "split point" */
     uint16_t offset; /* data offset from beginning
                   of mbuf data buffer */
     uint32_t reserved; /* reserved field */
};

The segment descriptions are added to the rte_eth_rxconf structure:
    rx_seg - pointer the array of segment descriptions, each element describes the memory pool, maximal data length, initial data offset from the beginning of data buffer in mbuf. This array allows to specify the different settings for each segment in individual fashion.
    rx_nseg - number of elements in the array

If the extended segment descriptions is provided with these new fields the mp parameter of the rte_eth_rx_queue_setup must be specified as NULL to avoid ambiguity.

There are two options to specify Rx buffer configuration:
- mp is not NULL, rrx_conf.rx_nseg is zero, it is compatible configuration, follows existing implementation, provides the single pool and no description for segment sizes and offsets.
- mp is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not zero, it provides the extended configuration, individually for each segment.

f the Rx queue is configured with new settings the packets being received will be split into multiple segments pushed to the mbufs with specified attributes. The PMD will split the received packets into multiple segments according to the specification in the description array.

For example, let's suppose we configured the Rx queue with the following segments:
     seg0 - pool0, len0=14B, off0=2
     seg1 - pool1, len1=20B, off1=128B
     seg2 - pool2, len2=20B, off2=0B
     seg3 - pool3, len3=512B, off3=0B

The packet 46 bytes long will look like the following:
     seg0 - 14B long @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B long @ 128 in mbuf from pool1
     seg2 - 12B long @ 0 in mbuf from pool2

The packet 1500 bytes long will look like the following:
     seg0 - 14B @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from pool0 seg1 - 20B @ 128 in mbuf from pool1
     seg2 - 20B @ 0 in mbuf from pool2
     seg3 - 512B @ 0 in mbuf from pool3
     seg4 - 512B @ 0 in mbuf from pool3
     seg5 - 422B @ 0 in mbuf from pool3

The offload RTE_ETH_RX_OFFLOAD_SCATTER must be present and configured to support new buffer split feature (if rx_nseg is greater than one).

The split limitations imposed by underlying PMD is reported in the new introduced rte_eth_dev_info->rx_seg_capa field.

The new approach would allow splitting the ingress packets into multiple parts pushed to the memory with different attributes. For example, the packet headers can be pushed to the embedded data buffers within mbufs and the application data into the external buffers attached to mbufs allocated from the different memory pools. The memory attributes for the split parts may differ either - for example the application data may be pushed into the external memory located on the dedicated physical device, say GPU or NVMe. This would improve the DPDK receiving datapath flexibility with preserving compatibility with existing API.

<...>

Quoted 8 lines

-    mbp_buf_size = rte_pktmbuf_data_room_size(mp); +    if (mp != NULL) {
+        /* Single pool configuration check. */
+        if (rx_conf->rx_nseg != 0) {
+            RTE_ETHDEV_LOG(ERR,
+                       "Ambiguous segment configuration\n"); +            return -EINVAL;
+        }

Quoted 3 lines

'rte_eth_rx_queue_setup()' accepts 'rx_conf' to be NULL, there is a NULL check for it somewhere below the function :)

Above will crash in that case, need to fix this, I am on it, any idea welcome.

adding additional checks as following in the next-net:

diff --git a/lib/librte_ethdev/rte_ethdev.c b/lib/librte_ethdev/rte_ethdev.c index 372a63bafa..ee2bc51671 100644
--- a/lib/librte_ethdev/rte_ethdev.c
+++ b/lib/librte_ethdev/rte_ethdev.c
@@ -1947,7 +1947,7 @@ rte_eth_rx_queue_setup(uint16_t port_id, uint16_t
rx_queue_id,

         if (mp != NULL) {
                 /* Single pool configuration check. */
 -               if (rx_conf->rx_nseg != 0) {
 +               if (rx_conf != NULL && rx_conf->rx_nseg != 0) {
                         RTE_ETHDEV_LOG(ERR,
                                        "Ambiguous segment configuration\n");
                         return -EINVAL;
 @@ -1983,7 +1983,7 @@ rte_eth_rx_queue_setup(uint16_t port_id, uint16_t 
rx_queue_id,
                 uint16_t n_seg = rx_conf->rx_nseg;

                 /* Extended multi-segment configuration check. */

- if (rx_conf->rx_seg == NULL || rx_conf->rx_nseg == 0) {
+ if (rx_conf == NULL || rx_conf->rx_seg == NULL ||
rx_conf->rx_nseg == 0) {
RTE_ETHDEV_LOG(ERR,
"Memory pool is null and no extended

configuration provided\n");
                         return -EINVAL;
#165 lore raw

PATCH v12 2/6 app/testpmd: add multiple pools per core creation

In reply to Viacheslav Ovsiienko

The command line parameter --mbuf-size is updated, it can handle the multiple values like the following:

--mbuf-size=2176,512,768,4096

specifying the creation the extra memory pools with the requested mbuf data buffer sizes. If some buffer split feature is engaged the extra memory pools can be used to configure the Rx queues with rte_the_dev_rx_queue_setup_ex().

The extra pools are created with requested sizes, and pool names are assigned with appended index: mbuf_pool_socket_%socket_%index. Index zero is used to specify the first mandatory pool to maintain compatibility with existing code.

---
 app/test-pmd/bpf_cmd.c                |  4 +--
 app/test-pmd/cmdline.c                |  2 +-
 app/test-pmd/config.c                 |  6 ++--
 app/test-pmd/parameters.c             | 24 +++++++++----
 app/test-pmd/testpmd.c                | 63 +++++++++++++++++++----------------
 app/test-pmd/testpmd.h                | 28 +++++++++++++---
 doc/guides/testpmd_app_ug/run_app.rst |  7 ++--
 7 files changed, 87 insertions(+), 47 deletions(-)
app/test-pmd/bpf_cmd.c +2 -2
index 16e3c3b..0a1a178 100644
@@ -69,7 +69,7 @@ struct cmd_bpf_ld_result {
 
 	*flags = RTE_BPF_ETH_F_NONE;
 	arg->type = RTE_BPF_ARG_PTR;
-	arg->size = mbuf_data_size;
+	arg->size = mbuf_data_size[0];
 
 	for (i = 0; str[i] != 0; i++) {
 		v = toupper(str[i]);
@@ -78,7 +78,7 @@ struct cmd_bpf_ld_result {
 		else if (v == 'M') {
 			arg->type = RTE_BPF_ARG_PTR_MBUF;
 			arg->size = sizeof(struct rte_mbuf);
-			arg->buf_size = mbuf_data_size;
+			arg->buf_size = mbuf_data_size[0];
 		} else if (v == '-')
 			continue;
 		else
app/test-pmd/cmdline.c +1 -1
index 078f063..19b1896 100644
@@ -2925,7 +2925,7 @@ struct cmd_setup_rxtx_queue {
 		if (!numa_support || socket_id == NUMA_NO_CONFIG)
 			socket_id = port->socket_id;
 
-		mp = mbuf_pool_find(socket_id);
+		mp = mbuf_pool_find(socket_id, 0);
 		if (mp == NULL) {
 			printf("Failed to setup RX queue: "
 				"No mempool allocation"
app/test-pmd/config.c +3 -3
index 2c00b55..c554dd0 100644
@@ -690,7 +690,7 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 	printf("\nConnect to socket: %u", port->socket_id);
 
 	if (port_numa[port_id] != NUMA_NO_CONFIG) {
-		mp = mbuf_pool_find(port_numa[port_id]);
+		mp = mbuf_pool_find(port_numa[port_id], 0);
 		if (mp)
 			printf("\nmemory allocation on the socket: %d",
 							port_numa[port_id]);
@@ -3580,9 +3580,9 @@ struct igb_ring_desc_16_bytes {
 	 */
 	tx_pkt_len = 0;
 	for (i = 0; i < nb_segs; i++) {
-		if (seg_lengths[i] > (unsigned) mbuf_data_size) {
+		if (seg_lengths[i] > mbuf_data_size[0]) {
 			printf("length[%u]=%u > mbuf_data_size=%u - give up\n",
-			       i, seg_lengths[i], (unsigned) mbuf_data_size);
+			       i, seg_lengths[i], mbuf_data_size[0]);
 			return;
 		}
 		tx_pkt_len = (uint16_t)(tx_pkt_len + seg_lengths[i]);
app/test-pmd/parameters.c +18 -6
index 15ce8c1..4db4987 100644
@@ -106,7 +106,9 @@
 	       "(flag: 1 for RX; 2 for TX; 3 for RX and TX).\n");
 	printf("  --socket-num=N: set socket from which all memory is allocated "
 	       "in NUMA mode.\n");
-	printf("  --mbuf-size=N: set the data size of mbuf to N bytes.\n");
+	printf("  --mbuf-size=N,[N1[,..Nn]: set the data size of mbuf to "
+	       "N bytes. If multiple numbers are specified the extra pools "
+	       "will be created to receive with packet split features\n");
 	printf("  --total-num-mbufs=N: set the number of mbufs to be allocated "
 	       "in mbuf pools.\n");
 	printf("  --max-pkt-len=N: set the maximum size of packet to N bytes.\n");
@@ -892,12 +894,22 @@
 				}
 			}
 			if (!strcmp(lgopts[opt_idx].name, "mbuf-size")) {
-				n = atoi(optarg);
-				if (n > 0 && n <= 0xFFFF)
-					mbuf_data_size = (uint16_t) n;
-				else
+				unsigned int mb_sz[MAX_SEGS_BUFFER_SPLIT];
+				unsigned int nb_segs, i;
+
+				nb_segs = parse_item_list(optarg, "mbuf-size",
+					MAX_SEGS_BUFFER_SPLIT, mb_sz, 0);
+				if (nb_segs <= 0)
 					rte_exit(EXIT_FAILURE,
-						 "mbuf-size should be > 0 and < 65536\n");
+						 "bad mbuf-size\n");
+				for (i = 0; i < nb_segs; i++) {
+					if (mb_sz[i] <= 0 || mb_sz[i] > 0xFFFF)
+						rte_exit(EXIT_FAILURE,
+							 "mbuf-size should be "
+							 "> 0 and < 65536\n");
+					mbuf_data_size[i] = (uint16_t) mb_sz[i];
+				}
+				mbuf_data_size_n = nb_segs;
 			}
 			if (!strcmp(lgopts[opt_idx].name, "total-num-mbufs")) {
 				n = atoi(optarg);
app/test-pmd/testpmd.c +35 -28

145 lines not on this page · raw message

app/test-pmd/testpmd.h +23 -5
index a4dfd4f..f889698 100644
@@ -42,6 +42,16 @@
  */
 #define RTE_MAX_SEGS_PER_PKT 255 /**< nb_segs is a 8-bit unsigned char. */
 
+/*
+ * The maximum number of segments per packet is used to configure
+ * buffer split feature, also specifies the maximum amount of
+ * optional Rx pools to allocate mbufs to split.
+ */
+#define MAX_SEGS_BUFFER_SPLIT 8 /**< nb_segs is a 8-bit unsigned char. */
+
+/* The prefix of the mbuf pool names created by the application. */
+#define MBUF_POOL_NAME_PFX "mb_pool"
+
 #define MAX_PKT_BURST 512
 #define DEF_PKT_BURST 32
 
@@ -403,7 +413,9 @@ struct queue_stats_mappings {
 extern uint8_t dcb_config;
 extern uint8_t dcb_test;
 
-extern uint16_t mbuf_data_size; /**< Mbuf data space size. */
+extern uint32_t mbuf_data_size_n;
+extern uint16_t mbuf_data_size[MAX_SEGS_BUFFER_SPLIT];
+/**< Mbuf data space size. */
 extern uint32_t param_total_num_mbufs;
 
 extern uint16_t stats_period;
@@ -615,17 +627,23 @@ struct mplsoudp_decap_conf {
 
 /* Mbuf Pools */
 static inline void
-mbuf_poolname_build(unsigned int sock_id, char* mp_name, int name_size)
+mbuf_poolname_build(unsigned int sock_id, char *mp_name,
+		    int name_size, uint16_t idx)
 {
-	snprintf(mp_name, name_size, "mbuf_pool_socket_%u", sock_id);
+	if (!idx)
+		snprintf(mp_name, name_size,
+			 MBUF_POOL_NAME_PFX "_%u", sock_id);
+	else
+		snprintf(mp_name, name_size,
+			 MBUF_POOL_NAME_PFX "_%hu_%hu", (uint16_t)sock_id, idx);
 }
 
 static inline struct rte_mempool *
-mbuf_pool_find(unsigned int sock_id)
+mbuf_pool_find(unsigned int sock_id, uint16_t idx)
 {
 	char pool_name[RTE_MEMPOOL_NAMESIZE];
 
-	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name));
+	mbuf_poolname_build(sock_id, pool_name, sizeof(pool_name), idx);
 	return rte_mempool_lookup((const char *)pool_name);
 }
 
doc/guides/testpmd_app_ug/run_app.rst +5 -2
index ec085c2..1eb0a10 100644
@@ -107,9 +107,12 @@ The command line options are:
     Set the socket from which all memory is allocated in NUMA mode,
     where 0 <= N < number of sockets on the board.
 
-*   ``--mbuf-size=N``
+*   ``--mbuf-size=N[,N1[,...Nn]``
 
-    Set the data size of the mbufs used to N bytes, where N < 65536. The default value is 2048.
+    Set the data size of the mbufs used to N bytes, where N < 65536.
+    The default value is 2048. If multiple mbuf-size values are specified the
+    extra memory pools will be created for allocating mbufs to receive packets
+    with buffer splittling features.
 
 *   ``--total-num-mbufs=N``
 
Signature · 1 line
1.8.3.1
#166 lore raw

PATCH v12 3/6 app/testpmd: add buffer split offload configuration

In reply to Viacheslav Ovsiienko

This patch add support for RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT providing per queue configuration for this offload.

---
 app/test-pmd/cmdline.c | 21 +++++++++++----------
 app/test-pmd/config.c  |  9 +++++++++
 2 files changed, 20 insertions(+), 10 deletions(-)
app/test-pmd/cmdline.c +11 -10
index 19b1896..0ef87c9 100644
@@ -883,16 +883,16 @@ static void cmd_help_long_parsed(void *parsed_result,
 			"port config <port_id> rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"     Enable or disable a per port Rx offloading"
 			" on all Rx queues of a port\n\n"
 
 			"port (port_id) rxq (queue_id) rx_offload vlan_strip|"
 			"ipv4_cksum|udp_cksum|tcp_cksum|tcp_lro|qinq_strip|"
 			"outer_ipv4_cksum|macsec_strip|header_split|"
-			"vlan_filter|vlan_extend|jumbo_frame|"
-			"scatter|timestamp|security|keep_crc on|off\n"
+			"vlan_filter|vlan_extend|jumbo_frame|scatter|"
+			"buffer_split|timestamp|security|keep_crc on|off\n"
 			"    Enable or disable a per queue Rx offloading"
 			" only on a specific Rx queue\n\n"
 
@@ -18435,7 +18435,8 @@ struct cmd_config_per_port_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc#rss_hash");
+			   "scatter#buffer_split#timestamp#security#"
+			   "keep_crc#rss_hash");
 cmdline_parse_token_string_t cmd_config_per_port_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_port_rx_offload_result,
@@ -18515,8 +18516,8 @@ struct cmd_config_per_port_rx_offload_result {
 	.help_str = "port config <port_id> rx_offload vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc|rss_hash "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc|rss_hash on|off",
 	.tokens = {
 		(void *)&cmd_config_per_port_rx_offload_result_port,
 		(void *)&cmd_config_per_port_rx_offload_result_config,
@@ -18565,7 +18566,7 @@ struct cmd_config_per_queue_rx_offload_result {
 		 offload, "vlan_strip#ipv4_cksum#udp_cksum#tcp_cksum#tcp_lro#"
 			   "qinq_strip#outer_ipv4_cksum#macsec_strip#"
 			   "header_split#vlan_filter#vlan_extend#jumbo_frame#"
-			   "scatter#timestamp#security#keep_crc");
+			   "scatter#buffer_split#timestamp#security#keep_crc");
 cmdline_parse_token_string_t cmd_config_per_queue_rx_offload_result_on_off =
 	TOKEN_STRING_INITIALIZER
 		(struct cmd_config_per_queue_rx_offload_result,
@@ -18621,8 +18622,8 @@ struct cmd_config_per_queue_rx_offload_result {
 		    "vlan_strip|ipv4_cksum|"
 		    "udp_cksum|tcp_cksum|tcp_lro|qinq_strip|outer_ipv4_cksum|"
 		    "macsec_strip|header_split|vlan_filter|vlan_extend|"
-		    "jumbo_frame|scatter|timestamp|security|keep_crc "
-		    "on|off",
+		    "jumbo_frame|scatter|buffer_split|timestamp|security|"
+		    "keep_crc on|off",
 	.tokens = {
 		(void *)&cmd_config_per_queue_rx_offload_result_port,
 		(void *)&cmd_config_per_queue_rx_offload_result_port_id,
app/test-pmd/config.c +9 -0
index c554dd0..0f041e1 100644
@@ -1092,6 +1092,15 @@ static int bus_match_all(const struct rte_bus *bus, const void *data)
 			printf("off\n");
 	}
 
+	if (dev_info.rx_offload_capa & RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
+		printf("RX offload buffer split:       ");
+		if (ports[port_id].dev_conf.rxmode.offloads &
+		    RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT)
+			printf("on\n");
+		else
+			printf("off\n");
+	}
+
 	if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_VLAN_INSERT) {
 		printf("VLAN insert:                   ");
 		if (ports[port_id].dev_conf.txmode.offloads &
Signature · 1 line
1.8.3.1