NTB: Bug Fixes and New Features

Allocate and use a DMA engine channel to transmit and receive data over NTB. If none is allocated, fall back to using the CPU to transfer data.

Cc Dan Williams <[email protected]>

Cc Vinod Koul <[email protected]>

Cc Dave Jiang <[email protected]>

Signed-off-by Jon Mason <[email protected]>

---
 drivers/ntb/ntb_hw.c        |   17 +++
 drivers/ntb/ntb_hw.h        |    1 +
 drivers/ntb/ntb_transport.c |  285 ++++++++++++++++++++++++++++++++++++-------
 3 files changed, 258 insertions(+), 45 deletions(-)

drivers/ntb/ntb_hw.c +17 -0

index 1d8e551..014222c 100644
@@ -350,6 +350,23 @@ int ntb_read_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 *val)
 }
 
 /**
+ * ntb_get_mw_base() - get addr for the NTB memory window
+ * @ndev: pointer to ntb_device instance
+ * @mw: memory window number
+ *
+ * This function provides the base address of the memory window specified.
+ *
+ * RETURNS: address, or NULL on error.
+ */
+resource_size_t ntb_get_mw_base(struct ntb_device *ndev, unsigned int mw)
+{
+       if (mw >= ntb_max_mw(ndev))
+               return 0;
+
+       return pci_resource_start(ndev->pdev, MW_TO_BAR(mw));
+}
+
+/**
  * ntb_get_mw_vbase() - get virtual addr for the NTB memory window
  * @ndev: pointer to ntb_device instance
  * @mw: memory window number

drivers/ntb/ntb_hw.h +1 -0

index b03de80..ab5f768 100644
@@ -240,6 +240,7 @@ int ntb_write_local_spad(struct ntb_device *ndev, unsigned int idx, u32 val);
 int ntb_read_local_spad(struct ntb_device *ndev, unsigned int idx, u32 *val);
 int ntb_write_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 val);
 int ntb_read_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 *val);
+resource_size_t ntb_get_mw_base(struct ntb_device *ndev, unsigned int mw);
 void __iomem *ntb_get_mw_vbase(struct ntb_device *ndev, unsigned int mw);
 u64 ntb_get_mw_size(struct ntb_device *ndev, unsigned int mw);
 void ntb_ring_sdb(struct ntb_device *ndev, unsigned int idx);

drivers/ntb/ntb_transport.c +35 -4

index f7380e9..73a35e4 100644
@@ -47,6 +47,7 @@
  */
 #include <linux/debugfs.h>
 #include <linux/delay.h>
+#include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
 #include <linux/errno.h>
 #include <linux/export.h>

@@ -68,6 +69,10 @@ static unsigned char max_num_clients;
 module_param(max_num_clients, byte, 0644);
 MODULE_PARM_DESC(max_num_clients, "Maximum number of NTB transport clients");
 
+static unsigned int copy_bytes = 1024;
+module_param(copy_bytes, uint, 0644);
+MODULE_PARM_DESC(copy_bytes, "Threshold under which NTB will use the CPU to copy instead of DMA");
+
 struct ntb_queue_entry {
        /* ntb_queue list reference */
        struct list_head entry;

@@ -76,6 +81,13 @@ struct ntb_queue_entry {
        void *buf;
        unsigned int len;
        unsigned int flags;
+
+       struct ntb_transport_qp *qp;
+       union {
+               struct ntb_payload_header __iomem *tx_hdr;
+               struct ntb_payload_header *rx_hdr;
+       };
+       unsigned int index;
 };
 
 struct ntb_rx_info {

@@ -86,6 +98,7 @@ struct ntb_transport_qp {
        struct ntb_transport *transport;
        struct ntb_device *ndev;
        void *cb_data;
+       struct dma_chan *dma_chan;
 
        bool client_ready;
        bool qp_link;

@@ -99,6 +112,7 @@ struct ntb_transport_qp {
        struct list_head tx_free_q;
        spinlock_t ntb_tx_free_q_lock;
        void __iomem *tx_mw;
+       dma_addr_t tx_mw_raw;
        unsigned int tx_index;
        unsigned int tx_max_entry;
        unsigned int tx_max_frame;

@@ -114,6 +128,7 @@ struct ntb_transport_qp {
        unsigned int rx_index;
        unsigned int rx_max_entry;
        unsigned int rx_max_frame;
+       dma_cookie_t last_cookie;
 
        void (*event_handler) (void *data, int status);
        struct delayed_work link_work;

@@ -129,9 +144,14 @@ struct ntb_transport_qp {
        u64 rx_err_no_buf;
        u64 rx_err_oflow;
        u64 rx_err_ver;
+       u64 rx_memcpy;
+       u64 rx_async;
        u64 tx_bytes;
        u64 tx_pkts;
        u64 tx_ring_full;
+       u64 tx_err_no_buf;
+       u64 tx_memcpy;
+       u64 tx_async;
 };
 
 struct ntb_transport_mw {

@@ -381,7 +401,7 @@ static ssize_t debugfs_read(struct file *filp, char __user *ubuf, size_t count,
        char *buf;
        ssize_t ret, out_offset, out_count;
 
-       out_count = 600;
+       out_count = 1000;
 
        buf = kmalloc(out_count, GFP_KERNEL);
        if (!buf)

@@ -396,6 +416,10 @@ static ssize_t debugfs_read(struct file *filp, char __user *ubuf, size_t count,
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "rx_pkts - \t%llu\n", qp->rx_pkts);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "rx_memcpy - \t%llu\n", qp->rx_memcpy);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "rx_async - \t%llu\n", qp->rx_async);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "rx_ring_empty - %llu\n", qp->rx_ring_empty);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "rx_err_no_buf - %llu\n", qp->rx_err_no_buf);

@@ -415,8 +439,14 @@ static ssize_t debugfs_read(struct file *filp, char __user *ubuf, size_t count,
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_pkts - \t%llu\n", qp->tx_pkts);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "tx_memcpy - \t%llu\n", qp->tx_memcpy);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "tx_async - \t%llu\n", qp->tx_async);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_ring_full - \t%llu\n", qp->tx_ring_full);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "tx_err_no_buf - %llu\n", qp->tx_err_no_buf);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_mw - \t%p\n", qp->tx_mw);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_index - \t%u\n", qp->tx_index);

@@ -488,11 +518,11 @@ static void ntb_transport_setup_qp_mw(struct ntb_transport *nt,
                num_qps_mw = nt->max_qps / mw_max;
 
        rx_size = (unsigned int) nt->mw[mw_num].size / num_qps_mw;
-       qp->remote_rx_info = nt->mw[mw_num].virt_addr +
-                            (qp_num / mw_max * rx_size);
+       qp->rx_buff = nt->mw[mw_num].virt_addr + qp_num / mw_max * rx_size;
        rx_size -= sizeof(struct ntb_rx_info);
 
-       qp->rx_buff = qp->remote_rx_info + 1;
+       qp->remote_rx_info = qp->rx_buff + rx_size;
+
        /* Due to housekeeping, there must be atleast 2 buffs */
        qp->rx_max_frame = min(transport_mtu, rx_size / 2);
        qp->rx_max_entry = rx_size / qp->rx_max_frame;

@@ -802,6 +832,7 @@ static void ntb_transport_init_queue(struct ntb_transport *nt,
        struct ntb_transport_qp *qp;
        unsigned int num_qps_mw, tx_size;
        u8 mw_num, mw_max;
+       u64 qp_offset;
 
        mw_max = ntb_max_mw(nt->ndev);
        mw_num = QP_TO_MW(nt->ndev, qp_num);

drivers/ntb/ntb_transport.c +1 -2

        tx_size -= sizeof(struct ntb_rx_info);
+       qp->rx_info = qp->tx_mw + tx_size;
 
-       qp->tx_mw = qp->rx_info + 1;
        /* Due to housekeeping, there must be atleast 2 buffs */
        qp->tx_max_frame = min(transport_mtu, tx_size / 2);
        qp->tx_max_entry = tx_size / qp->tx_max_frame;

@@ -956,13 +988,22 @@ void ntb_transport_free(void *transport)
        kfree(nt);
 }
 
-static void ntb_rx_copy_task(struct ntb_transport_qp *qp,

drivers/ntb/ntb_transport.c +22 -0

+       iowrite32(entry->index, &qp->rx_info->entry);
 
        ntb_list_add(&qp->ntb_rx_free_q_lock, &entry->entry, &qp->rx_free_q);
 

@@ -970,6 +1011,80 @@ static void ntb_rx_copy_task(struct ntb_transport_qp *qp,
                qp->rx_handler(qp, qp->cb_data, cb_data, len);
 }
 
+static void ntb_memcpy_rx(struct ntb_queue_entry *entry, void *offset)
+{
+       void *buf = entry->buf;
+       size_t len = entry->len;
+
+       memcpy(buf, offset, len);
+
+       ntb_rx_copy_callback(entry);
+}
+
+static void ntb_async_rx(struct ntb_queue_entry *entry, void *offset,
+                        size_t len)
+{
+       struct dma_async_tx_descriptor *txd;
+       struct ntb_transport_qp *qp = entry->qp;
+       struct dma_chan *chan = qp->dma_chan;
+       struct dma_device *device;
+       dma_addr_t src, dest;
+       dma_cookie_t cookie;
+       void *buf = entry->buf;
+       unsigned long flags;

drivers/ntb/ntb_transport.c +8 -0

+               goto err1;
+
+       dest = dma_map_single(device->dev, buf, len, DMA_FROM_DEVICE);
+       if (dma_mapping_error(device->dev, dest))
+               goto err1;
+
+       src = dma_map_single(device->dev, offset, len, DMA_TO_DEVICE);
+       if (dma_mapping_error(device->dev, src))

drivers/ntb/ntb_transport.c +63 -29

+       qp->rx_async++;
+
+       return;
+
+err3:
+       dma_unmap_single(device->dev, src, len, DMA_TO_DEVICE);
+err2:
+       dma_unmap_single(device->dev, dest, len, DMA_FROM_DEVICE);
+err1:
+       dma_sync_wait(chan, qp->last_cookie);
+err:
+       ntb_memcpy_rx(entry, offset);
+       qp->rx_memcpy++;
+}
+
 static int ntb_process_rxc(struct ntb_transport_qp *qp)
 {
        struct ntb_payload_header *hdr;

@@ -1008,41 +1123,44 @@ static int ntb_process_rxc(struct ntb_transport_qp *qp)
        if (hdr->flags & LINK_DOWN_FLAG) {
                ntb_qp_link_down(qp);
 
-               ntb_list_add(&qp->ntb_rx_pend_q_lock, &entry->entry,
-                            &qp->rx_pend_q);
-               goto out;
+               goto err;
        }
 
        dev_dbg(&ntb_query_pdev(qp->ndev)->dev,
                "rx offset %u, ver %u - %d payload received, buf size %d\n",
                qp->rx_index, hdr->ver, hdr->len, entry->len);
 
-       if (hdr->len <= entry->len) {
-               entry->len = hdr->len;
-               ntb_rx_copy_task(qp, entry, offset);
-       } else {
-               ntb_list_add(&qp->ntb_rx_pend_q_lock, &entry->entry,
-                            &qp->rx_pend_q);
+       qp->rx_bytes += hdr->len;
+       qp->rx_pkts++;
 
+       if (hdr->len > entry->len) {
                qp->rx_err_oflow++;
                dev_dbg(&ntb_query_pdev(qp->ndev)->dev,
                        "RX overflow! Wanted %d got %d\n",
                        hdr->len, entry->len);
+
+               goto err;
        }
 
-       qp->rx_bytes += hdr->len;
-       qp->rx_pkts++;
+       entry->index = qp->rx_index;
+       entry->rx_hdr = hdr;
 
-out:
-       /* Ensure that the data is fully copied out before clearing the flag */
-       wmb();
-       hdr->flags = 0;
-       iowrite32(qp->rx_index, &qp->rx_info->entry);
+       ntb_async_rx(entry, offset, hdr->len);
 
+out:
        qp->rx_index++;
        qp->rx_index %= qp->rx_max_entry;
 
        return 0;
+
+err:
+       ntb_list_add(&qp->ntb_rx_pend_q_lock, &entry->entry,
+                    &qp->rx_pend_q);
+       wmb();
+       hdr->flags = 0;
+       iowrite32(qp->rx_index, &qp->rx_info->entry);
+
+       goto out;
 }
 
 static void ntb_transport_rx(unsigned long data)

@@ -1070,19 +1188,12 @@ static void ntb_transport_rxc_db(void *data, int db_num)
        tasklet_schedule(&qp->rx_work);
 }
 
-static void ntb_tx_copy_task(struct ntb_transport_qp *qp,
-                            struct ntb_queue_entry *entry,
-                            void __iomem *offset)
+static void ntb_tx_copy_callback(void *data)
 {
-       struct ntb_payload_header __iomem *hdr;
-
-       memcpy_toio(offset, entry->buf, entry->len);
-
-       hdr = offset + qp->tx_max_frame - sizeof(struct ntb_payload_header);
-       iowrite32(entry->len, &hdr->len);
-       iowrite32((u32) qp->tx_pkts, &hdr->ver);
+       struct ntb_queue_entry *entry = data;
+       struct ntb_transport_qp *qp = entry->qp;
+       struct ntb_payload_header __iomem *hdr = entry->tx_hdr;
 
-       /* Ensure that the data is fully copied out before setting the flag */
        wmb();
        iowrite32(entry->flags | DESC_DONE_FLAG, &hdr->flags);
 

@@ -1103,15 +1214,79 @@ static void ntb_tx_copy_task(struct ntb_transport_qp *qp,
        ntb_list_add(&qp->ntb_tx_free_q_lock, &entry->entry, &qp->tx_free_q);
 }
 
-static int ntb_process_tx(struct ntb_transport_qp *qp,
-                         struct ntb_queue_entry *entry)
+static void ntb_memcpy_tx(struct ntb_queue_entry *entry, void __iomem *offset)
+{
+       memcpy_toio(offset, entry->buf, entry->len);
+
+       ntb_tx_copy_callback(entry);
+}
+
+static void ntb_async_tx(struct ntb_transport_qp *qp,
+                        struct ntb_queue_entry *entry)
 {
+       struct dma_async_tx_descriptor *txd;
+       struct dma_chan *chan = qp->dma_chan;
+       struct dma_device *device;
+       dma_addr_t src, dest;
+       dma_cookie_t cookie;
+       struct ntb_payload_header __iomem *hdr;
        void __iomem *offset;
+       size_t len = entry->len;
+       void *buf = entry->buf;
+       unsigned long flags;
 
        offset = qp->tx_mw + qp->tx_max_frame * qp->tx_index;
+       hdr = offset + qp->tx_max_frame - sizeof(struct ntb_payload_header);
+       entry->tx_hdr = hdr;
+
+       iowrite32(entry->len, &hdr->len);
+       iowrite32((u32) qp->tx_pkts, &hdr->ver);
+
+       if (!chan)

drivers/ntb/ntb_transport.c +24 -1

+       txd->callback = ntb_tx_copy_callback;
+       txd->callback_param = entry;
+
+       cookie = dmaengine_submit(txd);
+       if (dma_submit_error(cookie))
+               goto err1;
+
+       dma_async_issue_pending(chan);
+       qp->tx_async++;
+
+       return;
+err1:
+       dma_unmap_single(device->dev, src, len, DMA_TO_DEVICE);
+err:
+       ntb_memcpy_tx(entry, offset);
+       qp->tx_memcpy++;
+}
+
+static int ntb_process_tx(struct ntb_transport_qp *qp,
+                         struct ntb_queue_entry *entry)
+{
+       dev_dbg(&ntb_query_pdev(qp->ndev)->dev, "%lld - tx %u, entry len %d flags %x buff %p\n",
+               qp->tx_pkts, qp->tx_index, entry->len, entry->flags,
                entry->buf);
        if (qp->tx_index == qp->remote_rx_info->entry) {
                qp->tx_ring_full++;

@@ -1127,7 +1302,7 @@ static int ntb_process_tx(struct ntb_transport_qp *qp,
                return 0;
        }
 
-       ntb_tx_copy_task(qp, entry, offset);
+       ntb_async_tx(qp, entry);
 
        qp->tx_index++;
        qp->tx_index %= qp->tx_max_entry;

drivers/ntb/ntb_transport.c +3 -0

+
        for (i = 0; i < NTB_QP_DEF_NUM_ENTRIES; i++) {
                entry = kzalloc(sizeof(struct ntb_queue_entry), GFP_ATOMIC);
                if (!entry)
                        goto err1;
 
+               entry->qp = qp;
                ntb_list_add(&qp->ntb_rx_free_q_lock, &entry->entry,
                             &qp->rx_free_q);
        }

@@ -1227,6 +1407,7 @@ ntb_transport_create_queue(void *data, struct pci_dev *pdev,
                if (!entry)
                        goto err2;
 
+               entry->qp = qp;
                ntb_list_add(&qp->ntb_tx_free_q_lock, &entry->entry,
                             &qp->tx_free_q);
        }

drivers/ntb/ntb_transport.c +6 -1

+
        while ((entry = ntb_list_rm(&qp->ntb_rx_free_q_lock, &qp->rx_free_q)))
                kfree(entry);
 

@@ -1382,8 +1566,10 @@ int ntb_transport_tx_enqueue(struct ntb_transport_qp *qp, void *cb, void *data,
                return -EINVAL;
 
        entry = ntb_list_rm(&qp->ntb_tx_free_q_lock, &qp->tx_free_q);
-       if (!entry)
+       if (!entry) {
+               qp->tx_err_no_buf++;
                return -ENOMEM;
+       }
 
        entry->cb_data = cb;
        entry->buf = data;

@@ -1499,9 +1685,18 @@ EXPORT_SYMBOL_GPL(ntb_transport_qp_num);
  */
 unsigned int ntb_transport_max_size(struct ntb_transport_qp *qp)
 {
+       unsigned int max;
+

Allocate and use a DMA engine channel to transmit and receive data over NTB. If none is allocated, fall back to using the CPU to transfer data.

Cc Dan Williams <[email protected]>

Cc Vinod Koul <[email protected]>

Cc Dave Jiang <[email protected]>

Signed-off-by Jon Mason <[email protected]>

---
 drivers/ntb/ntb_hw.c        |   17 +++
 drivers/ntb/ntb_hw.h        |    1 +
 drivers/ntb/ntb_transport.c |  285 ++++++++++++++++++++++++++++++++++++-------
 3 files changed, 258 insertions(+), 45 deletions(-)

drivers/ntb/ntb_hw.c +17 -0

index 1d8e551..014222c 100644
@@ -350,6 +350,23 @@ int ntb_read_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 *val)
 }
 
 /**
+ * ntb_get_mw_base() - get addr for the NTB memory window
+ * @ndev: pointer to ntb_device instance
+ * @mw: memory window number
+ *
+ * This function provides the base address of the memory window specified.
+ *
+ * RETURNS: address, or NULL on error.
+ */
+resource_size_t ntb_get_mw_base(struct ntb_device *ndev, unsigned int mw)
+{
+       if (mw >= ntb_max_mw(ndev))
+               return 0;
+
+       return pci_resource_start(ndev->pdev, MW_TO_BAR(mw));
+}
+
+/**
  * ntb_get_mw_vbase() - get virtual addr for the NTB memory window
  * @ndev: pointer to ntb_device instance
  * @mw: memory window number

drivers/ntb/ntb_hw.h +1 -0

index b03de80..ab5f768 100644
@@ -240,6 +240,7 @@ int ntb_write_local_spad(struct ntb_device *ndev, unsigned int idx, u32 val);
 int ntb_read_local_spad(struct ntb_device *ndev, unsigned int idx, u32 *val);
 int ntb_write_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 val);
 int ntb_read_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 *val);
+resource_size_t ntb_get_mw_base(struct ntb_device *ndev, unsigned int mw);
 void __iomem *ntb_get_mw_vbase(struct ntb_device *ndev, unsigned int mw);
 u64 ntb_get_mw_size(struct ntb_device *ndev, unsigned int mw);
 void ntb_ring_sdb(struct ntb_device *ndev, unsigned int idx);

drivers/ntb/ntb_transport.c +35 -4

index f7380e9..73a35e4 100644
@@ -47,6 +47,7 @@
  */
 #include <linux/debugfs.h>
 #include <linux/delay.h>
+#include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
 #include <linux/errno.h>
 #include <linux/export.h>

@@ -68,6 +69,10 @@ static unsigned char max_num_clients;
 module_param(max_num_clients, byte, 0644);
 MODULE_PARM_DESC(max_num_clients, "Maximum number of NTB transport clients");
 
+static unsigned int copy_bytes = 1024;
+module_param(copy_bytes, uint, 0644);
+MODULE_PARM_DESC(copy_bytes, "Threshold under which NTB will use the CPU to copy instead of DMA");
+
 struct ntb_queue_entry {
        /* ntb_queue list reference */
        struct list_head entry;

@@ -76,6 +81,13 @@ struct ntb_queue_entry {
        void *buf;
        unsigned int len;
        unsigned int flags;
+
+       struct ntb_transport_qp *qp;
+       union {
+               struct ntb_payload_header __iomem *tx_hdr;
+               struct ntb_payload_header *rx_hdr;
+       };
+       unsigned int index;
 };
 
 struct ntb_rx_info {

@@ -86,6 +98,7 @@ struct ntb_transport_qp {
        struct ntb_transport *transport;
        struct ntb_device *ndev;
        void *cb_data;
+       struct dma_chan *dma_chan;
 
        bool client_ready;
        bool qp_link;

@@ -99,6 +112,7 @@ struct ntb_transport_qp {
        struct list_head tx_free_q;
        spinlock_t ntb_tx_free_q_lock;
        void __iomem *tx_mw;
+       dma_addr_t tx_mw_raw;
        unsigned int tx_index;
        unsigned int tx_max_entry;
        unsigned int tx_max_frame;

@@ -114,6 +128,7 @@ struct ntb_transport_qp {
        unsigned int rx_index;
        unsigned int rx_max_entry;
        unsigned int rx_max_frame;
+       dma_cookie_t last_cookie;
 
        void (*event_handler) (void *data, int status);
        struct delayed_work link_work;

@@ -129,9 +144,14 @@ struct ntb_transport_qp {
        u64 rx_err_no_buf;
        u64 rx_err_oflow;
        u64 rx_err_ver;
+       u64 rx_memcpy;
+       u64 rx_async;
        u64 tx_bytes;
        u64 tx_pkts;
        u64 tx_ring_full;
+       u64 tx_err_no_buf;
+       u64 tx_memcpy;
+       u64 tx_async;
 };
 
 struct ntb_transport_mw {

@@ -381,7 +401,7 @@ static ssize_t debugfs_read(struct file *filp, char __user *ubuf, size_t count,
        char *buf;
        ssize_t ret, out_offset, out_count;
 
-       out_count = 600;
+       out_count = 1000;
 
        buf = kmalloc(out_count, GFP_KERNEL);
        if (!buf)

@@ -396,6 +416,10 @@ static ssize_t debugfs_read(struct file *filp, char __user *ubuf, size_t count,
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "rx_pkts - \t%llu\n", qp->rx_pkts);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "rx_memcpy - \t%llu\n", qp->rx_memcpy);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "rx_async - \t%llu\n", qp->rx_async);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "rx_ring_empty - %llu\n", qp->rx_ring_empty);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "rx_err_no_buf - %llu\n", qp->rx_err_no_buf);

@@ -415,8 +439,14 @@ static ssize_t debugfs_read(struct file *filp, char __user *ubuf, size_t count,
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_pkts - \t%llu\n", qp->tx_pkts);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "tx_memcpy - \t%llu\n", qp->tx_memcpy);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "tx_async - \t%llu\n", qp->tx_async);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_ring_full - \t%llu\n", qp->tx_ring_full);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "tx_err_no_buf - %llu\n", qp->tx_err_no_buf);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_mw - \t%p\n", qp->tx_mw);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_index - \t%u\n", qp->tx_index);

@@ -488,11 +518,11 @@ static void ntb_transport_setup_qp_mw(struct ntb_transport *nt,
                num_qps_mw = nt->max_qps / mw_max;
 
        rx_size = (unsigned int) nt->mw[mw_num].size / num_qps_mw;
-       qp->remote_rx_info = nt->mw[mw_num].virt_addr +
-                            (qp_num / mw_max * rx_size);
+       qp->rx_buff = nt->mw[mw_num].virt_addr + qp_num / mw_max * rx_size;
        rx_size -= sizeof(struct ntb_rx_info);
 
-       qp->rx_buff = qp->remote_rx_info + 1;
+       qp->remote_rx_info = qp->rx_buff + rx_size;
+
        /* Due to housekeeping, there must be atleast 2 buffs */
        qp->rx_max_frame = min(transport_mtu, rx_size / 2);
        qp->rx_max_entry = rx_size / qp->rx_max_frame;

@@ -802,6 +832,7 @@ static void ntb_transport_init_queue(struct ntb_transport *nt,
        struct ntb_transport_qp *qp;
        unsigned int num_qps_mw, tx_size;
        u8 mw_num, mw_max;
+       u64 qp_offset;
 
        mw_max = ntb_max_mw(nt->ndev);
        mw_num = QP_TO_MW(nt->ndev, qp_num);

drivers/ntb/ntb_transport.c +1 -2

        tx_size -= sizeof(struct ntb_rx_info);
+       qp->rx_info = qp->tx_mw + tx_size;
 
-       qp->tx_mw = qp->rx_info + 1;
        /* Due to housekeeping, there must be atleast 2 buffs */
        qp->tx_max_frame = min(transport_mtu, tx_size / 2);
        qp->tx_max_entry = tx_size / qp->tx_max_frame;

@@ -956,13 +988,22 @@ void ntb_transport_free(void *transport)
        kfree(nt);
 }
 
-static void ntb_rx_copy_task(struct ntb_transport_qp *qp,

drivers/ntb/ntb_transport.c +22 -0

+       iowrite32(entry->index, &qp->rx_info->entry);
 
        ntb_list_add(&qp->ntb_rx_free_q_lock, &entry->entry, &qp->rx_free_q);
 

@@ -970,6 +1011,80 @@ static void ntb_rx_copy_task(struct ntb_transport_qp *qp,
                qp->rx_handler(qp, qp->cb_data, cb_data, len);
 }
 
+static void ntb_memcpy_rx(struct ntb_queue_entry *entry, void *offset)
+{
+       void *buf = entry->buf;
+       size_t len = entry->len;
+
+       memcpy(buf, offset, len);
+
+       ntb_rx_copy_callback(entry);
+}
+
+static void ntb_async_rx(struct ntb_queue_entry *entry, void *offset,
+                        size_t len)
+{
+       struct dma_async_tx_descriptor *txd;
+       struct ntb_transport_qp *qp = entry->qp;
+       struct dma_chan *chan = qp->dma_chan;
+       struct dma_device *device;
+       dma_addr_t src, dest;
+       dma_cookie_t cookie;
+       void *buf = entry->buf;
+       unsigned long flags;

drivers/ntb/ntb_transport.c +8 -0

+               goto err1;
+
+       dest = dma_map_single(device->dev, buf, len, DMA_FROM_DEVICE);
+       if (dma_mapping_error(device->dev, dest))
+               goto err1;
+
+       src = dma_map_single(device->dev, offset, len, DMA_TO_DEVICE);
+       if (dma_mapping_error(device->dev, src))

drivers/ntb/ntb_transport.c +63 -29

+       qp->rx_async++;
+
+       return;
+
+err3:
+       dma_unmap_single(device->dev, src, len, DMA_TO_DEVICE);
+err2:
+       dma_unmap_single(device->dev, dest, len, DMA_FROM_DEVICE);
+err1:
+       dma_sync_wait(chan, qp->last_cookie);
+err:
+       ntb_memcpy_rx(entry, offset);
+       qp->rx_memcpy++;
+}
+
 static int ntb_process_rxc(struct ntb_transport_qp *qp)
 {
        struct ntb_payload_header *hdr;

@@ -1008,41 +1123,44 @@ static int ntb_process_rxc(struct ntb_transport_qp *qp)
        if (hdr->flags & LINK_DOWN_FLAG) {
                ntb_qp_link_down(qp);
 
-               ntb_list_add(&qp->ntb_rx_pend_q_lock, &entry->entry,
-                            &qp->rx_pend_q);
-               goto out;
+               goto err;
        }
 
        dev_dbg(&ntb_query_pdev(qp->ndev)->dev,
                "rx offset %u, ver %u - %d payload received, buf size %d\n",
                qp->rx_index, hdr->ver, hdr->len, entry->len);
 
-       if (hdr->len <= entry->len) {
-               entry->len = hdr->len;
-               ntb_rx_copy_task(qp, entry, offset);
-       } else {
-               ntb_list_add(&qp->ntb_rx_pend_q_lock, &entry->entry,
-                            &qp->rx_pend_q);
+       qp->rx_bytes += hdr->len;
+       qp->rx_pkts++;
 
+       if (hdr->len > entry->len) {
                qp->rx_err_oflow++;
                dev_dbg(&ntb_query_pdev(qp->ndev)->dev,
                        "RX overflow! Wanted %d got %d\n",
                        hdr->len, entry->len);
+
+               goto err;
        }
 
-       qp->rx_bytes += hdr->len;
-       qp->rx_pkts++;
+       entry->index = qp->rx_index;
+       entry->rx_hdr = hdr;
 
-out:
-       /* Ensure that the data is fully copied out before clearing the flag */
-       wmb();
-       hdr->flags = 0;
-       iowrite32(qp->rx_index, &qp->rx_info->entry);
+       ntb_async_rx(entry, offset, hdr->len);
 
+out:
        qp->rx_index++;
        qp->rx_index %= qp->rx_max_entry;
 
        return 0;
+
+err:
+       ntb_list_add(&qp->ntb_rx_pend_q_lock, &entry->entry,
+                    &qp->rx_pend_q);
+       wmb();
+       hdr->flags = 0;
+       iowrite32(qp->rx_index, &qp->rx_info->entry);
+
+       goto out;
 }
 
 static void ntb_transport_rx(unsigned long data)

@@ -1070,19 +1188,12 @@ static void ntb_transport_rxc_db(void *data, int db_num)
        tasklet_schedule(&qp->rx_work);
 }
 
-static void ntb_tx_copy_task(struct ntb_transport_qp *qp,
-                            struct ntb_queue_entry *entry,
-                            void __iomem *offset)
+static void ntb_tx_copy_callback(void *data)
 {
-       struct ntb_payload_header __iomem *hdr;
-
-       memcpy_toio(offset, entry->buf, entry->len);
-
-       hdr = offset + qp->tx_max_frame - sizeof(struct ntb_payload_header);
-       iowrite32(entry->len, &hdr->len);
-       iowrite32((u32) qp->tx_pkts, &hdr->ver);
+       struct ntb_queue_entry *entry = data;
+       struct ntb_transport_qp *qp = entry->qp;
+       struct ntb_payload_header __iomem *hdr = entry->tx_hdr;
 
-       /* Ensure that the data is fully copied out before setting the flag */
        wmb();
        iowrite32(entry->flags | DESC_DONE_FLAG, &hdr->flags);
 

@@ -1103,15 +1214,79 @@ static void ntb_tx_copy_task(struct ntb_transport_qp *qp,
        ntb_list_add(&qp->ntb_tx_free_q_lock, &entry->entry, &qp->tx_free_q);
 }
 
-static int ntb_process_tx(struct ntb_transport_qp *qp,
-                         struct ntb_queue_entry *entry)
+static void ntb_memcpy_tx(struct ntb_queue_entry *entry, void __iomem *offset)
+{
+       memcpy_toio(offset, entry->buf, entry->len);
+
+       ntb_tx_copy_callback(entry);
+}
+
+static void ntb_async_tx(struct ntb_transport_qp *qp,
+                        struct ntb_queue_entry *entry)
 {
+       struct dma_async_tx_descriptor *txd;
+       struct dma_chan *chan = qp->dma_chan;
+       struct dma_device *device;
+       dma_addr_t src, dest;
+       dma_cookie_t cookie;
+       struct ntb_payload_header __iomem *hdr;
        void __iomem *offset;
+       size_t len = entry->len;
+       void *buf = entry->buf;
+       unsigned long flags;
 
        offset = qp->tx_mw + qp->tx_max_frame * qp->tx_index;
+       hdr = offset + qp->tx_max_frame - sizeof(struct ntb_payload_header);
+       entry->tx_hdr = hdr;
+
+       iowrite32(entry->len, &hdr->len);
+       iowrite32((u32) qp->tx_pkts, &hdr->ver);
+
+       if (!chan)

drivers/ntb/ntb_transport.c +24 -1

+       txd->callback = ntb_tx_copy_callback;
+       txd->callback_param = entry;
+
+       cookie = dmaengine_submit(txd);
+       if (dma_submit_error(cookie))
+               goto err1;
+
+       dma_async_issue_pending(chan);
+       qp->tx_async++;
+
+       return;
+err1:
+       dma_unmap_single(device->dev, src, len, DMA_TO_DEVICE);
+err:
+       ntb_memcpy_tx(entry, offset);
+       qp->tx_memcpy++;
+}
+
+static int ntb_process_tx(struct ntb_transport_qp *qp,
+                         struct ntb_queue_entry *entry)
+{
+       dev_dbg(&ntb_query_pdev(qp->ndev)->dev, "%lld - tx %u, entry len %d flags %x buff %p\n",
+               qp->tx_pkts, qp->tx_index, entry->len, entry->flags,
                entry->buf);
        if (qp->tx_index == qp->remote_rx_info->entry) {
                qp->tx_ring_full++;

@@ -1127,7 +1302,7 @@ static int ntb_process_tx(struct ntb_transport_qp *qp,
                return 0;
        }
 
-       ntb_tx_copy_task(qp, entry, offset);
+       ntb_async_tx(qp, entry);
 
        qp->tx_index++;
        qp->tx_index %= qp->tx_max_entry;

drivers/ntb/ntb_transport.c +3 -0

+
        for (i = 0; i < NTB_QP_DEF_NUM_ENTRIES; i++) {
                entry = kzalloc(sizeof(struct ntb_queue_entry), GFP_ATOMIC);
                if (!entry)
                        goto err1;
 
+               entry->qp = qp;
                ntb_list_add(&qp->ntb_rx_free_q_lock, &entry->entry,
                             &qp->rx_free_q);
        }

@@ -1227,6 +1407,7 @@ ntb_transport_create_queue(void *data, struct pci_dev *pdev,
                if (!entry)
                        goto err2;
 
+               entry->qp = qp;
                ntb_list_add(&qp->ntb_tx_free_q_lock, &entry->entry,
                             &qp->tx_free_q);
        }

drivers/ntb/ntb_transport.c +7 -1

        ntb_unregister_db_callback(qp->ndev, qp->qp_num);
        tasklet_disable(&qp->rx_work);
 
+       cancel_delayed_work_sync(&qp->link_work);
+
        while ((entry = ntb_list_rm(&qp->ntb_rx_free_q_lock, &qp->rx_free_q)))
                kfree(entry);
 

@@ -1382,8 +1566,10 @@ int ntb_transport_tx_enqueue(struct ntb_transport_qp *qp, void *cb, void *data,
                return -EINVAL;
 
        entry = ntb_list_rm(&qp->ntb_tx_free_q_lock, &qp->tx_free_q);
-       if (!entry)
+       if (!entry) {
+               qp->tx_err_no_buf++;
                return -ENOMEM;
+       }
 
        entry->cb_data = cb;
        entry->buf = data;

@@ -1499,9 +1685,18 @@ EXPORT_SYMBOL_GPL(ntb_transport_qp_num);
  */
 unsigned int ntb_transport_max_size(struct ntb_transport_qp *qp)
 {
+       unsigned int max;
+

Quoted 190 lines · Jon Mason

Allocate and use a DMA engine channel to transmit and receive data over NTB. If none is allocated, fall back to using the CPU to transfer data.

Cc Dan Williams <[email protected]>

Cc Vinod Koul <[email protected]>

Cc Dave Jiang <[email protected]>

Signed-off-by Jon Mason <[email protected]>

---
 drivers/ntb/ntb_hw.c        |   17 +++
 drivers/ntb/ntb_hw.h        |    1 +
 drivers/ntb/ntb_transport.c |  285 ++++++++++++++++++++++++++++++++++++-------
 3 files changed, 258 insertions(+), 45 deletions(-)

drivers/ntb/ntb_hw.c +17 -0

index 1d8e551..014222c 100644
@@ -350,6 +350,23 @@ int ntb_read_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 *val)
 }
 
 /**
+ * ntb_get_mw_base() - get addr for the NTB memory window
+ * @ndev: pointer to ntb_device instance
+ * @mw: memory window number
+ *
+ * This function provides the base address of the memory window specified.
+ *
+ * RETURNS: address, or NULL on error.
+ */
+resource_size_t ntb_get_mw_base(struct ntb_device *ndev, unsigned int mw)
+{
+       if (mw >= ntb_max_mw(ndev))
+               return 0;
+
+       return pci_resource_start(ndev->pdev, MW_TO_BAR(mw));
+}
+
+/**
  * ntb_get_mw_vbase() - get virtual addr for the NTB memory window
  * @ndev: pointer to ntb_device instance
  * @mw: memory window number

drivers/ntb/ntb_hw.h +1 -0

index b03de80..ab5f768 100644
@@ -240,6 +240,7 @@ int ntb_write_local_spad(struct ntb_device *ndev, unsigned int idx, u32 val);
 int ntb_read_local_spad(struct ntb_device *ndev, unsigned int idx, u32 *val);
 int ntb_write_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 val);
 int ntb_read_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 *val);
+resource_size_t ntb_get_mw_base(struct ntb_device *ndev, unsigned int mw);
 void __iomem *ntb_get_mw_vbase(struct ntb_device *ndev, unsigned int mw);
 u64 ntb_get_mw_size(struct ntb_device *ndev, unsigned int mw);
 void ntb_ring_sdb(struct ntb_device *ndev, unsigned int idx);

drivers/ntb/ntb_transport.c +35 -4

index f7380e9..73a35e4 100644
@@ -47,6 +47,7 @@
  */
 #include <linux/debugfs.h>
 #include <linux/delay.h>
+#include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
 #include <linux/errno.h>
 #include <linux/export.h>

@@ -68,6 +69,10 @@ static unsigned char max_num_clients;
 module_param(max_num_clients, byte, 0644);
 MODULE_PARM_DESC(max_num_clients, "Maximum number of NTB transport clients");
 
+static unsigned int copy_bytes = 1024;
+module_param(copy_bytes, uint, 0644);
+MODULE_PARM_DESC(copy_bytes, "Threshold under which NTB will use the CPU to copy instead of DMA");
+
 struct ntb_queue_entry {
        /* ntb_queue list reference */
        struct list_head entry;

@@ -76,6 +81,13 @@ struct ntb_queue_entry {
        void *buf;
        unsigned int len;
        unsigned int flags;
+
+       struct ntb_transport_qp *qp;
+       union {
+               struct ntb_payload_header __iomem *tx_hdr;
+               struct ntb_payload_header *rx_hdr;
+       };
+       unsigned int index;
 };
 
 struct ntb_rx_info {

@@ -86,6 +98,7 @@ struct ntb_transport_qp {
        struct ntb_transport *transport;
        struct ntb_device *ndev;
        void *cb_data;
+       struct dma_chan *dma_chan;
 
        bool client_ready;
        bool qp_link;

@@ -99,6 +112,7 @@ struct ntb_transport_qp {
        struct list_head tx_free_q;
        spinlock_t ntb_tx_free_q_lock;
        void __iomem *tx_mw;
+       dma_addr_t tx_mw_raw;
        unsigned int tx_index;
        unsigned int tx_max_entry;
        unsigned int tx_max_frame;

@@ -114,6 +128,7 @@ struct ntb_transport_qp {
        unsigned int rx_index;
        unsigned int rx_max_entry;
        unsigned int rx_max_frame;
+       dma_cookie_t last_cookie;
 
        void (*event_handler) (void *data, int status);
        struct delayed_work link_work;

@@ -129,9 +144,14 @@ struct ntb_transport_qp {
        u64 rx_err_no_buf;
        u64 rx_err_oflow;
        u64 rx_err_ver;
+       u64 rx_memcpy;
+       u64 rx_async;
        u64 tx_bytes;
        u64 tx_pkts;
        u64 tx_ring_full;
+       u64 tx_err_no_buf;
+       u64 tx_memcpy;
+       u64 tx_async;
 };
 
 struct ntb_transport_mw {

@@ -381,7 +401,7 @@ static ssize_t debugfs_read(struct file *filp, char __user *ubuf, size_t count,
        char *buf;
        ssize_t ret, out_offset, out_count;
 
-       out_count = 600;
+       out_count = 1000;
 
        buf = kmalloc(out_count, GFP_KERNEL);
        if (!buf)

@@ -396,6 +416,10 @@ static ssize_t debugfs_read(struct file *filp, char __user *ubuf, size_t count,
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "rx_pkts - \t%llu\n", qp->rx_pkts);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "rx_memcpy - \t%llu\n", qp->rx_memcpy);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "rx_async - \t%llu\n", qp->rx_async);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "rx_ring_empty - %llu\n", qp->rx_ring_empty);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "rx_err_no_buf - %llu\n", qp->rx_err_no_buf);

@@ -415,8 +439,14 @@ static ssize_t debugfs_read(struct file *filp, char __user *ubuf, size_t count,
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_pkts - \t%llu\n", qp->tx_pkts);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "tx_memcpy - \t%llu\n", qp->tx_memcpy);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "tx_async - \t%llu\n", qp->tx_async);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_ring_full - \t%llu\n", qp->tx_ring_full);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
+                              "tx_err_no_buf - %llu\n", qp->tx_err_no_buf);
+       out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_mw - \t%p\n", qp->tx_mw);
        out_offset += snprintf(buf + out_offset, out_count - out_offset,
                               "tx_index - \t%u\n", qp->tx_index);

@@ -488,11 +518,11 @@ static void ntb_transport_setup_qp_mw(struct ntb_transport *nt,
                num_qps_mw = nt->max_qps / mw_max;
 
        rx_size = (unsigned int) nt->mw[mw_num].size / num_qps_mw;
-       qp->remote_rx_info = nt->mw[mw_num].virt_addr +
-                            (qp_num / mw_max * rx_size);
+       qp->rx_buff = nt->mw[mw_num].virt_addr + qp_num / mw_max * rx_size;
        rx_size -= sizeof(struct ntb_rx_info);
 
-       qp->rx_buff = qp->remote_rx_info + 1;
+       qp->remote_rx_info = qp->rx_buff + rx_size;
+
        /* Due to housekeeping, there must be atleast 2 buffs */
        qp->rx_max_frame = min(transport_mtu, rx_size / 2);
        qp->rx_max_entry = rx_size / qp->rx_max_frame;

@@ -802,6 +832,7 @@ static void ntb_transport_init_queue(struct ntb_transport *nt,
        struct ntb_transport_qp *qp;
        unsigned int num_qps_mw, tx_size;
        u8 mw_num, mw_max;
+       u64 qp_offset;
 
        mw_max = ntb_max_mw(nt->ndev);
        mw_num = QP_TO_MW(nt->ndev, qp_num);

@@ -820,11 +851,12 @@ static void ntb_transport_init_queue(struct ntb_transport *nt,
                num_qps_mw = nt->max_qps / mw_max;
 

drivers/ntb/ntb_transport.c static void ntb_transport_init_queue(struct ntb_transport *nt, · line 856

Quoted 6 lines

drivers/ntb/ntb_transport.c +3 -2

        tx_size = (unsigned int) ntb_get_mw_size(qp->ndev, mw_num) / num_qps_mw;
-       qp->rx_info = ntb_get_mw_vbase(nt->ndev, mw_num) +
-                     (qp_num / mw_max * tx_size);
+       qp_offset = qp_num / mw_max * tx_size;
+       qp->tx_mw = ntb_get_mw_vbase(nt->ndev, mw_num) + qp_offset;
+       qp->tx_mw_raw = ntb_get_mw_base(qp->ndev, mw_num) + qp_offset;

Just a quibble with the name, why "raw" instead of "phys"?

Quoted 12 lines

drivers/ntb/ntb_transport.c +1 -1

        tx_size -= sizeof(struct ntb_rx_info);
+       qp->rx_info = qp->tx_mw + tx_size;
 
-       qp->tx_mw = qp->rx_info + 1;
        /* Due to housekeeping, there must be atleast 2 buffs */
        qp->tx_max_frame = min(transport_mtu, tx_size / 2);
        qp->tx_max_entry = tx_size / qp->tx_max_frame;

@@ -956,13 +988,22 @@ void ntb_transport_free(void *transport)
        kfree(nt);
 }
 

drivers/ntb/ntb_transport.c void ntb_transport_free(void *transport) · line 999

Quoted 11 lines

drivers/ntb/ntb_transport.c +6 -2

-static void ntb_rx_copy_task(struct ntb_transport_qp *qp,
-                            struct ntb_queue_entry *entry, void *offset)
+static void ntb_rx_copy_callback(void *data)
 {
+       struct ntb_queue_entry *entry = data;
+       struct ntb_transport_qp *qp = entry->qp;
        void *cb_data = entry->cb_data;
        unsigned int len = entry->len;
+       struct ntb_payload_header *hdr = entry->rx_hdr;
+
+       wmb();

What is this barrier paired with?

drivers/ntb/ntb_transport.c void ntb_transport_free(void *transport) · line 1005

Quoted 7 lines

drivers/ntb/ntb_transport.c +5 -1

+       hdr->flags = 0;
 
-       memcpy(entry->buf, offset, entry->len);
+       /* If the callbacks come out of order, the writing of the index to the
+        * last completed will be out of order.  This may result in the the
+        * receive stalling forever.
+        */

Is this for the case where we are bouncing back and forth between sync/async? Otherwise I do not see how transactions could get out of order given you allocate a channel once per queue. Is this comment saying that the iowrite32 is somehow a fix, or is this comment a FIXME?

Quoted 34 lines

drivers/ntb/ntb_transport.c +27 -0

+       iowrite32(entry->index, &qp->rx_info->entry);
 
        ntb_list_add(&qp->ntb_rx_free_q_lock, &entry->entry, &qp->rx_free_q);
 

@@ -970,6 +1011,80 @@ static void ntb_rx_copy_task(struct ntb_transport_qp *qp,
                qp->rx_handler(qp, qp->cb_data, cb_data, len);
 }
 
+static void ntb_memcpy_rx(struct ntb_queue_entry *entry, void *offset)
+{
+       void *buf = entry->buf;
+       size_t len = entry->len;
+
+       memcpy(buf, offset, len);
+
+       ntb_rx_copy_callback(entry);
+}
+
+static void ntb_async_rx(struct ntb_queue_entry *entry, void *offset,
+                        size_t len)
+{
+       struct dma_async_tx_descriptor *txd;
+       struct ntb_transport_qp *qp = entry->qp;
+       struct dma_chan *chan = qp->dma_chan;
+       struct dma_device *device;
+       dma_addr_t src, dest;
+       dma_cookie_t cookie;
+       void *buf = entry->buf;
+       unsigned long flags;
+
+       entry->len = len;
+
+       if (!chan)
+               goto err;

drivers/ntb/ntb_transport.c static void ntb_rx_copy_task(struct ntb_transport_qp *qp, · line 1044

Quoted 5 lines

drivers/ntb/ntb_transport.c +5 -0

+
+       device = chan->device;
+
+       if (len < copy_bytes ||
+           !is_dma_copy_aligned(device, __pa(offset), __pa(buf), len))

__pa()'s necessary here? I don't think the alignment requirements will ever cross PAGE_OFFSET.

Quoted 12 lines

drivers/ntb/ntb_transport.c +12 -0

+               goto err1;
+
+       dest = dma_map_single(device->dev, buf, len, DMA_FROM_DEVICE);
+       if (dma_mapping_error(device->dev, dest))
+               goto err1;
+
+       src = dma_map_single(device->dev, offset, len, DMA_TO_DEVICE);
+       if (dma_mapping_error(device->dev, src))
+               goto err2;
+
+       flags = DMA_COMPL_DEST_UNMAP_SINGLE | DMA_COMPL_SRC_UNMAP_SINGLE |
+               DMA_PREP_INTERRUPT;

drivers/ntb/ntb_transport.c static void ntb_rx_copy_task(struct ntb_transport_qp *qp, · line 1062

Quoted 6 lines

drivers/ntb/ntb_transport.c +6 -0

+       txd = device->device_prep_dma_memcpy(chan, dest, src, len, flags);
+       if (!txd)
+               goto err3;
+
+       txd_clear_parent(txd);
+       txd_clear_next(txd);

Neither of these should be necessary.

Quoted 5 lines

drivers/ntb/ntb_transport.c +5 -0

+       txd->callback = ntb_rx_copy_callback;
+       txd->callback_param = entry;
+
+       cookie = dmaengine_submit(txd);
+       if (dma_submit_error(cookie))

drivers/ntb/ntb_transport.c static void ntb_rx_copy_task(struct ntb_transport_qp *qp, · line 1072

Quoted 5 lines

drivers/ntb/ntb_transport.c +5 -0

+               goto err3;
+
+       qp->last_cookie = cookie;
+
+       dma_async_issue_pending(chan);

hmm... can this go in ntb_process_rx() so that the submission is batched? Cuts down on mmio.

Quoted 138 lines

drivers/ntb/ntb_transport.c +63 -29

+       qp->rx_async++;
+
+       return;
+
+err3:
+       dma_unmap_single(device->dev, src, len, DMA_TO_DEVICE);
+err2:
+       dma_unmap_single(device->dev, dest, len, DMA_FROM_DEVICE);
+err1:
+       dma_sync_wait(chan, qp->last_cookie);
+err:
+       ntb_memcpy_rx(entry, offset);
+       qp->rx_memcpy++;
+}
+
 static int ntb_process_rxc(struct ntb_transport_qp *qp)
 {
        struct ntb_payload_header *hdr;

@@ -1008,41 +1123,44 @@ static int ntb_process_rxc(struct ntb_transport_qp *qp)
        if (hdr->flags & LINK_DOWN_FLAG) {
                ntb_qp_link_down(qp);
 
-               ntb_list_add(&qp->ntb_rx_pend_q_lock, &entry->entry,
-                            &qp->rx_pend_q);
-               goto out;
+               goto err;
        }
 
        dev_dbg(&ntb_query_pdev(qp->ndev)->dev,
                "rx offset %u, ver %u - %d payload received, buf size %d\n",
                qp->rx_index, hdr->ver, hdr->len, entry->len);
 
-       if (hdr->len <= entry->len) {
-               entry->len = hdr->len;
-               ntb_rx_copy_task(qp, entry, offset);
-       } else {
-               ntb_list_add(&qp->ntb_rx_pend_q_lock, &entry->entry,
-                            &qp->rx_pend_q);
+       qp->rx_bytes += hdr->len;
+       qp->rx_pkts++;
 
+       if (hdr->len > entry->len) {
                qp->rx_err_oflow++;
                dev_dbg(&ntb_query_pdev(qp->ndev)->dev,
                        "RX overflow! Wanted %d got %d\n",
                        hdr->len, entry->len);
+
+               goto err;
        }
 
-       qp->rx_bytes += hdr->len;
-       qp->rx_pkts++;
+       entry->index = qp->rx_index;
+       entry->rx_hdr = hdr;
 
-out:
-       /* Ensure that the data is fully copied out before clearing the flag */
-       wmb();
-       hdr->flags = 0;
-       iowrite32(qp->rx_index, &qp->rx_info->entry);
+       ntb_async_rx(entry, offset, hdr->len);
 
+out:
        qp->rx_index++;
        qp->rx_index %= qp->rx_max_entry;
 
        return 0;
+
+err:
+       ntb_list_add(&qp->ntb_rx_pend_q_lock, &entry->entry,
+                    &qp->rx_pend_q);
+       wmb();
+       hdr->flags = 0;
+       iowrite32(qp->rx_index, &qp->rx_info->entry);
+
+       goto out;
 }
 
 static void ntb_transport_rx(unsigned long data)

@@ -1070,19 +1188,12 @@ static void ntb_transport_rxc_db(void *data, int db_num)
        tasklet_schedule(&qp->rx_work);
 }
 
-static void ntb_tx_copy_task(struct ntb_transport_qp *qp,
-                            struct ntb_queue_entry *entry,
-                            void __iomem *offset)
+static void ntb_tx_copy_callback(void *data)
 {
-       struct ntb_payload_header __iomem *hdr;
-
-       memcpy_toio(offset, entry->buf, entry->len);
-
-       hdr = offset + qp->tx_max_frame - sizeof(struct ntb_payload_header);
-       iowrite32(entry->len, &hdr->len);
-       iowrite32((u32) qp->tx_pkts, &hdr->ver);
+       struct ntb_queue_entry *entry = data;
+       struct ntb_transport_qp *qp = entry->qp;
+       struct ntb_payload_header __iomem *hdr = entry->tx_hdr;
 
-       /* Ensure that the data is fully copied out before setting the flag */
        wmb();
        iowrite32(entry->flags | DESC_DONE_FLAG, &hdr->flags);
 

@@ -1103,15 +1214,79 @@ static void ntb_tx_copy_task(struct ntb_transport_qp *qp,
        ntb_list_add(&qp->ntb_tx_free_q_lock, &entry->entry, &qp->tx_free_q);
 }
 
-static int ntb_process_tx(struct ntb_transport_qp *qp,
-                         struct ntb_queue_entry *entry)
+static void ntb_memcpy_tx(struct ntb_queue_entry *entry, void __iomem *offset)
+{
+       memcpy_toio(offset, entry->buf, entry->len);
+
+       ntb_tx_copy_callback(entry);
+}
+
+static void ntb_async_tx(struct ntb_transport_qp *qp,
+                        struct ntb_queue_entry *entry)
 {
+       struct dma_async_tx_descriptor *txd;
+       struct dma_chan *chan = qp->dma_chan;
+       struct dma_device *device;
+       dma_addr_t src, dest;
+       dma_cookie_t cookie;
+       struct ntb_payload_header __iomem *hdr;
        void __iomem *offset;
+       size_t len = entry->len;
+       void *buf = entry->buf;
+       unsigned long flags;
 
        offset = qp->tx_mw + qp->tx_max_frame * qp->tx_index;
+       hdr = offset + qp->tx_max_frame - sizeof(struct ntb_payload_header);
+       entry->tx_hdr = hdr;
+
+       iowrite32(entry->len, &hdr->len);
+       iowrite32((u32) qp->tx_pkts, &hdr->ver);
+
+       if (!chan)

drivers/ntb/ntb_transport.c static void ntb_tx_copy_task(struct ntb_transport_qp *qp, · line 1253

Quoted 10 lines

drivers/ntb/ntb_transport.c +7 -2

+               goto err;
 
-       dev_dbg(&ntb_query_pdev(qp->ndev)->dev, "%lld - offset %p, tx %u, entry len %d flags %x buff %p\n",
-               qp->tx_pkts, offset, qp->tx_index, entry->len, entry->flags,
+       device = chan->device;
+
+       dest = qp->tx_mw_raw + qp->tx_max_frame * qp->tx_index;
+
+       if (len < copy_bytes ||
+           !is_dma_copy_aligned(device, __pa(buf), dest, len))

ditto on the use of __pa here.

drivers/ntb/ntb_transport.c static void ntb_tx_copy_task(struct ntb_transport_qp *qp, · line 1266

Quoted 13 lines

drivers/ntb/ntb_transport.c +13 -0

+               goto err;
+
+       src = dma_map_single(device->dev, buf, len, DMA_TO_DEVICE);
+       if (dma_mapping_error(device->dev, src))
+               goto err;
+
+       flags = DMA_COMPL_SRC_UNMAP_SINGLE | DMA_PREP_INTERRUPT;
+       txd = device->device_prep_dma_memcpy(chan, dest, src, len, flags);
+       if (!txd)
+               goto err1;
+
+       txd_clear_parent(txd);
+       txd_clear_next(txd);

...again not needed.

Quoted 39 lines

drivers/ntb/ntb_transport.c +24 -1

+       txd->callback = ntb_tx_copy_callback;
+       txd->callback_param = entry;
+
+       cookie = dmaengine_submit(txd);
+       if (dma_submit_error(cookie))
+               goto err1;
+
+       dma_async_issue_pending(chan);
+       qp->tx_async++;
+
+       return;
+err1:
+       dma_unmap_single(device->dev, src, len, DMA_TO_DEVICE);
+err:
+       ntb_memcpy_tx(entry, offset);
+       qp->tx_memcpy++;
+}
+
+static int ntb_process_tx(struct ntb_transport_qp *qp,
+                         struct ntb_queue_entry *entry)
+{
+       dev_dbg(&ntb_query_pdev(qp->ndev)->dev, "%lld - tx %u, entry len %d flags %x buff %p\n",
+               qp->tx_pkts, qp->tx_index, entry->len, entry->flags,
                entry->buf);
        if (qp->tx_index == qp->remote_rx_info->entry) {
                qp->tx_ring_full++;

@@ -1127,7 +1302,7 @@ static int ntb_process_tx(struct ntb_transport_qp *qp,
                return 0;
        }
 
-       ntb_tx_copy_task(qp, entry, offset);
+       ntb_async_tx(qp, entry);
 
        qp->tx_index++;
        qp->tx_index %= qp->tx_max_entry;

@@ -1213,11 +1388,16 @@ ntb_transport_create_queue(void *data, struct pci_dev *pdev,
        qp->tx_handler = handlers->tx_handler;
        qp->event_handler = handlers->event_handler;
 

drivers/ntb/ntb_transport.c ntb_transport_create_queue(void *data, struct pci_dev *pdev, · line 1393

Quoted 3 lines

drivers/ntb/ntb_transport.c +3 -0

+       qp->dma_chan = dma_find_channel(DMA_MEMCPY);
+       if (!qp->dma_chan)
+               dev_info(&pdev->dev, "Unable to allocate private DMA channel, using CPU instead\n");

You can drop "private" since this is a public allocation.

Quoted 20 lines

drivers/ntb/ntb_transport.c +3 -0

+
        for (i = 0; i < NTB_QP_DEF_NUM_ENTRIES; i++) {
                entry = kzalloc(sizeof(struct ntb_queue_entry), GFP_ATOMIC);
                if (!entry)
                        goto err1;
 
+               entry->qp = qp;
                ntb_list_add(&qp->ntb_rx_free_q_lock, &entry->entry,
                             &qp->rx_free_q);
        }

@@ -1227,6 +1407,7 @@ ntb_transport_create_queue(void *data, struct pci_dev *pdev,
                if (!entry)
                        goto err2;
 
+               entry->qp = qp;
                ntb_list_add(&qp->ntb_tx_free_q_lock, &entry->entry,
                             &qp->tx_free_q);
        }

@@ -1272,11 +1453,14 @@ void ntb_transport_free_queue(struct ntb_transport_qp *qp)
 

drivers/ntb/ntb_transport.c void ntb_transport_free_queue(struct ntb_transport_qp *qp) · line 1457

Quoted 5 lines

drivers/ntb/ntb_transport.c +2 -1

        pdev = ntb_query_pdev(qp->ndev);
 
-       cancel_delayed_work_sync(&qp->link_work);
+       if (qp->dma_chan)
+               dmaengine_terminate_all(qp->dma_chan);

Do you need this or can you just wait for all outstanding tx / rx to quiesce?

Quoted 29 lines

drivers/ntb/ntb_transport.c +7 -1

        ntb_unregister_db_callback(qp->ndev, qp->qp_num);
        tasklet_disable(&qp->rx_work);
 
+       cancel_delayed_work_sync(&qp->link_work);
+
        while ((entry = ntb_list_rm(&qp->ntb_rx_free_q_lock, &qp->rx_free_q)))
                kfree(entry);
 

@@ -1382,8 +1566,10 @@ int ntb_transport_tx_enqueue(struct ntb_transport_qp *qp, void *cb, void *data,
                return -EINVAL;
 
        entry = ntb_list_rm(&qp->ntb_tx_free_q_lock, &qp->tx_free_q);
-       if (!entry)
+       if (!entry) {
+               qp->tx_err_no_buf++;
                return -ENOMEM;
+       }
 
        entry->cb_data = cb;
        entry->buf = data;

@@ -1499,9 +1685,18 @@ EXPORT_SYMBOL_GPL(ntb_transport_qp_num);
  */
 unsigned int ntb_transport_max_size(struct ntb_transport_qp *qp)
 {
+       unsigned int max;
+
        if (!qp)
                return 0;
 

drivers/ntb/ntb_transport.c EXPORT_SYMBOL_GPL(ntb_transport_qp_num); · line 1698

Quoted 7 lines

drivers/ntb/ntb_transport.c +6 -1

-       return qp->tx_max_frame - sizeof(struct ntb_payload_header);
+       if (!qp->dma_chan)
+               return qp->tx_max_frame - sizeof(struct ntb_payload_header);
+
+       /* If DMA engine usage is possible, try to find the max size for that */
+       max = qp->tx_max_frame - sizeof(struct ntb_payload_header);
+       max -= max % (1 << qp->dma_chan->device->copy_align);

Unless I missed it you need a dmaengine_get() dmaengine_put() pair in your driver setup/teardown. dmaengine_get() notifies dmaengine of a dma_find_channel() user.

Quoted 36 lines · Dan Williams

Quoted 1 line · Jon Mason

Allocate and use a DMA engine channel to transmit and receive data

over

Quoted 1 line

NTB. If none is allocated, fall back to using the CPU to transfer

data.

Quoted 9 lines

Cc Dan Williams <[email protected]>

Cc Vinod Koul <[email protected]>

Cc Dave Jiang <[email protected]>

Signed-off-by Jon Mason <[email protected]>

---
 drivers/ntb/ntb_hw.c        |   17 +++
 drivers/ntb/ntb_hw.h        |    1 +
 drivers/ntb/ntb_transport.c |  285

++++++++++++++++++++++++++++++++++++-------

drivers/ntb/ntb_hw.c int ntb_read_remote_spad(struct ntb_device

Quoted 7 lines

 3 files changed, 258 insertions(+), 45 deletions(-)

drivers/ntb/ntb_hw.c +0 -0

index 1d8e551..014222c 100644
@@ -350,6 +350,23 @@ int ntb_read_remote_spad(struct ntb_device

*ndev, unsigned int idx, u32 *val)

drivers/ntb/ntb_hw.c int ntb_read_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 *val)

Quoted 8 lines

drivers/ntb/ntb_hw.c +5 -0

 }
 
 /**
+ * ntb_get_mw_base() - get addr for the NTB memory window
+ * @ndev: pointer to ntb_device instance
+ * @mw: memory window number
+ *
+ * This function provides the base address of the memory window

specified.

Quoted 4 lines

drivers/ntb/ntb_hw.c +4 -0

+ *
+ * RETURNS: address, or NULL on error.
+ */
+resource_size_t ntb_get_mw_base(struct ntb_device *ndev, unsigned

Quoted 7 lines

int mw)

drivers/ntb/ntb_hw.c int ntb_read_remote_spad(struct ntb_device *ndev, unsigned int idx, u32 *val) · line 367

Quoted 6 lines

drivers/ntb/ntb_hw.c +6 -0

+{
+       if (mw >= ntb_max_mw(ndev))
+               return 0;
+
+       return pci_resource_start(ndev->pdev, MW_TO_BAR(mw));
+}

Quoted 186 lines

drivers/ntb/ntb_hw.h int ntb_write_local_spad(struct ntb_device *ndev,

Quoted 10 lines

drivers/ntb/ntb_hw.c +2 -0

+
+/**
  * ntb_get_mw_vbase() - get virtual addr for the NTB memory window
  * @ndev: pointer to ntb_device instance
  * @mw: memory window number

drivers/ntb/ntb_hw.h +0 -0

index b03de80..ab5f768 100644
@@ -240,6 +240,7 @@ int ntb_write_local_spad(struct ntb_device *ndev,

unsigned int idx, u32 val);

drivers/ntb/ntb_hw.h int ntb_write_local_spad(struct ntb_device *ndev, · line 240

Quoted 1 line

drivers/ntb/ntb_hw.h +0 -0

 int ntb_read_local_spad(struct ntb_device *ndev, unsigned int idx,

u32 *val);

drivers/ntb/ntb_hw.h int ntb_write_local_spad(struct ntb_device *ndev, · line 241

Quoted 1 line

drivers/ntb/ntb_hw.h +0 -0

 int ntb_write_remote_spad(struct ntb_device *ndev, unsigned int idx,

u32 val);

drivers/ntb/ntb_hw.h int ntb_write_local_spad(struct ntb_device *ndev, · line 242

Quoted 1 line

drivers/ntb/ntb_hw.h +0 -0

 int ntb_read_remote_spad(struct ntb_device *ndev, unsigned int idx,

u32 *val);

drivers/ntb/ntb_hw.h int ntb_write_local_spad(struct ntb_device *ndev, · line 243

Quoted 1 line

drivers/ntb/ntb_hw.h +1 -0

+resource_size_t ntb_get_mw_base(struct ntb_device *ndev, unsigned

int mw);

drivers/ntb/ntb_hw.h int ntb_write_local_spad(struct ntb_device *ndev, · line 244

Quoted 1 line

drivers/ntb/ntb_hw.h +0 -0

 void __iomem *ntb_get_mw_vbase(struct ntb_device *ndev, unsigned int

mw);

Quoted 15 lines

drivers/ntb/ntb_hw.h +0 -0

 u64 ntb_get_mw_size(struct ntb_device *ndev, unsigned int mw);
 void ntb_ring_sdb(struct ntb_device *ndev, unsigned int idx);

drivers/ntb/ntb_transport.c +1 -0

index f7380e9..73a35e4 100644
@@ -47,6 +47,7 @@
  */
 #include <linux/debugfs.h>
 #include <linux/delay.h>
+#include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
 #include <linux/errno.h>
 #include <linux/export.h>

drivers/ntb/ntb_transport.c static unsigned char max_num_clients; · line 70

Quoted 2 lines

drivers/ntb/ntb_transport.c +0 -0

@@ -68,6 +69,10 @@ static unsigned char max_num_clients;
 module_param(max_num_clients, byte, 0644);
 MODULE_PARM_DESC(max_num_clients, "Maximum number of NTB transport

clients");

drivers/ntb/ntb_transport.c static unsigned char max_num_clients;

Quoted 4 lines

drivers/ntb/ntb_transport.c +3 -0

+static unsigned int copy_bytes = 1024;
+module_param(copy_bytes, uint, 0644);
+MODULE_PARM_DESC(copy_bytes, "Threshold under which NTB will use the

CPU to copy instead of DMA");

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp,

Quoted 58 lines

drivers/ntb/ntb_transport.c +16 -0

+
 struct ntb_queue_entry {
        /* ntb_queue list reference */
        struct list_head entry;

@@ -76,6 +81,13 @@ struct ntb_queue_entry {
        void *buf;
        unsigned int len;
        unsigned int flags;
+
+       struct ntb_transport_qp *qp;
+       union {
+               struct ntb_payload_header __iomem *tx_hdr;
+               struct ntb_payload_header *rx_hdr;
+       };
+       unsigned int index;
 };
 
 struct ntb_rx_info {

@@ -86,6 +98,7 @@ struct ntb_transport_qp {
        struct ntb_transport *transport;
        struct ntb_device *ndev;
        void *cb_data;
+       struct dma_chan *dma_chan;
 
        bool client_ready;
        bool qp_link;

@@ -99,6 +112,7 @@ struct ntb_transport_qp {
        struct list_head tx_free_q;
        spinlock_t ntb_tx_free_q_lock;
        void __iomem *tx_mw;
+       dma_addr_t tx_mw_raw;
        unsigned int tx_index;
        unsigned int tx_max_entry;
        unsigned int tx_max_frame;

@@ -114,6 +128,7 @@ struct ntb_transport_qp {
        unsigned int rx_index;
        unsigned int rx_max_entry;
        unsigned int rx_max_frame;
+       dma_cookie_t last_cookie;
 
        void (*event_handler) (void *data, int status);
        struct delayed_work link_work;

@@ -129,9 +144,14 @@ struct ntb_transport_qp {
        u64 rx_err_no_buf;
        u64 rx_err_oflow;
        u64 rx_err_ver;
+       u64 rx_memcpy;
+       u64 rx_async;
        u64 tx_bytes;
        u64 tx_pkts;
        u64 tx_ring_full;
+       u64 tx_err_no_buf;
+       u64 tx_memcpy;
+       u64 tx_async;
 };
 
 struct ntb_transport_mw {

@@ -381,7 +401,7 @@ static ssize_t debugfs_read(struct file *filp,

char __user *ubuf, size_t count,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp,

Quoted 9 lines

drivers/ntb/ntb_transport.c +1 -1

        char *buf;
        ssize_t ret, out_offset, out_count;
 
-       out_count = 600;
+       out_count = 1000;
 
        buf = kmalloc(out_count, GFP_KERNEL);
        if (!buf)

@@ -396,6 +416,10 @@ static ssize_t debugfs_read(struct file *filp,

char __user *ubuf, size_t count,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 416

Quoted 1 line

drivers/ntb/ntb_transport.c +0 -0

        out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 418

Quoted 2 lines

drivers/ntb/ntb_transport.c +0 -0

                               "rx_pkts - \t%llu\n", qp->rx_pkts);
        out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 420

Quoted 2 lines

drivers/ntb/ntb_transport.c +2 -0

+                              "rx_memcpy - \t%llu\n", qp->rx_memcpy);
+       out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 422

Quoted 2 lines

drivers/ntb/ntb_transport.c +2 -0

+                              "rx_async - \t%llu\n", qp->rx_async);
+       out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 423

Quoted 1 line

drivers/ntb/ntb_transport.c +0 -0

                               "rx_ring_empty - %llu\n",

qp->rx_ring_empty);

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 424

Quoted 1 line

drivers/ntb/ntb_transport.c +0 -0

        out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 425

Quoted 1 line

drivers/ntb/ntb_transport.c +0 -0

                               "rx_err_no_buf - %llu\n",

qp->rx_err_no_buf);

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp,

Quoted 1 line

drivers/ntb/ntb_transport.c +0 -0

@@ -415,8 +439,14 @@ static ssize_t debugfs_read(struct file *filp,

char __user *ubuf, size_t count,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 439

Quoted 1 line

drivers/ntb/ntb_transport.c +0 -0

        out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 441

Quoted 2 lines

drivers/ntb/ntb_transport.c +0 -0

                               "tx_pkts - \t%llu\n", qp->tx_pkts);
        out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 443

Quoted 2 lines

drivers/ntb/ntb_transport.c +2 -0

+                              "tx_memcpy - \t%llu\n", qp->tx_memcpy);
+       out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 445

Quoted 2 lines

drivers/ntb/ntb_transport.c +2 -0

+                              "tx_async - \t%llu\n", qp->tx_async);
+       out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 446

Quoted 1 line

drivers/ntb/ntb_transport.c +0 -0

                               "tx_ring_full - \t%llu\n",

qp->tx_ring_full);

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 447

Quoted 1 line

drivers/ntb/ntb_transport.c +0 -0

        out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 448

Quoted 1 line

drivers/ntb/ntb_transport.c +1 -0

+                              "tx_err_no_buf - %llu\n",

qp->tx_err_no_buf);

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 449

Quoted 1 line

drivers/ntb/ntb_transport.c +1 -0

+       out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static ssize_t debugfs_read(struct file *filp, · line 451

Quoted 2 lines

drivers/ntb/ntb_transport.c +0 -0

                               "tx_mw - \t%p\n", qp->tx_mw);
        out_offset += snprintf(buf + out_offset, out_count -

out_offset,

drivers/ntb/ntb_transport.c static void ntb_transport_setup_qp_mw(struct

Quoted 2 lines

drivers/ntb/ntb_transport.c +0 -0

                               "tx_index - \t%u\n", qp->tx_index);

@@ -488,11 +518,11 @@ static void ntb_transport_setup_qp_mw(struct

ntb_transport *nt,

drivers/ntb/ntb_transport.c static void ntb_transport_setup_qp_mw(struct ntb_transport *nt,

Quoted 6 lines

drivers/ntb/ntb_transport.c +1 -2

                num_qps_mw = nt->max_qps / mw_max;
 
        rx_size = (unsigned int) nt->mw[mw_num].size / num_qps_mw;
-       qp->remote_rx_info = nt->mw[mw_num].virt_addr +
-                            (qp_num / mw_max * rx_size);
+       qp->rx_buff = nt->mw[mw_num].virt_addr + qp_num / mw_max *

rx_size;

drivers/ntb/ntb_transport.c static void ntb_transport_init_queue(struct

Quoted 9 lines

drivers/ntb/ntb_transport.c +2 -1

        rx_size -= sizeof(struct ntb_rx_info);
 
-       qp->rx_buff = qp->remote_rx_info + 1;
+       qp->remote_rx_info = qp->rx_buff + rx_size;
+
        /* Due to housekeeping, there must be atleast 2 buffs */
        qp->rx_max_frame = min(transport_mtu, rx_size / 2);
        qp->rx_max_entry = rx_size / qp->rx_max_frame;

@@ -802,6 +832,7 @@ static void ntb_transport_init_queue(struct

ntb_transport *nt,

drivers/ntb/ntb_transport.c static void ntb_transport_init_queue(struct

Quoted 8 lines

drivers/ntb/ntb_transport.c +1 -0

        struct ntb_transport_qp *qp;
        unsigned int num_qps_mw, tx_size;
        u8 mw_num, mw_max;
+       u64 qp_offset;
 
        mw_max = ntb_max_mw(nt->ndev);
        mw_num = QP_TO_MW(nt->ndev, qp_num);

@@ -820,11 +851,12 @@ static void ntb_transport_init_queue(struct

ntb_transport *nt,

drivers/ntb/ntb_transport.c static void ntb_transport_setup_qp_mw(struct ntb_transport *nt,

Quoted 3 lines

drivers/ntb/ntb_transport.c +0 -0

                num_qps_mw = nt->max_qps / mw_max;
 
        tx_size = (unsigned int) ntb_get_mw_size(qp->ndev, mw_num) /

Quoted 8 lines

num_qps_mw;

drivers/ntb/ntb_transport.c static void ntb_transport_init_queue(struct ntb_transport *nt, · line 856

Quoted 5 lines

drivers/ntb/ntb_transport.c +3 -2

-       qp->rx_info = ntb_get_mw_vbase(nt->ndev, mw_num) +
-                     (qp_num / mw_max * tx_size);
+       qp_offset = qp_num / mw_max * tx_size;
+       qp->tx_mw = ntb_get_mw_vbase(nt->ndev, mw_num) + qp_offset;
+       qp->tx_mw_raw = ntb_get_mw_base(qp->ndev, mw_num) + qp_offset;

Just a quibble with the name, why "raw" instead of "phys"?

What's in a name? ;-)

phys is fine.

Quoted 15 lines

Quoted 9 lines

drivers/ntb/ntb_transport.c +1 -1

        tx_size -= sizeof(struct ntb_rx_info);
+       qp->rx_info = qp->tx_mw + tx_size;
 
-       qp->tx_mw = qp->rx_info + 1;
        /* Due to housekeeping, there must be atleast 2 buffs */
        qp->tx_max_frame = min(transport_mtu, tx_size / 2);
        qp->tx_max_entry = tx_size / qp->tx_max_frame;

drivers/ntb/ntb_transport.c void ntb_transport_free(void *transport) · line 960

Quoted 5 lines

drivers/ntb/ntb_transport.c +0 -2

@@ -956,13 +988,22 @@ void ntb_transport_free(void *transport)
        kfree(nt);
 }
 
-static void ntb_rx_copy_task(struct ntb_transport_qp *qp,
-                            struct ntb_queue_entry *entry, void

*offset)

Quoted 11 lines

drivers/ntb/ntb_transport.c void ntb_transport_free(void *transport) · line 999

Quoted 9 lines

drivers/ntb/ntb_transport.c +6 -0

+static void ntb_rx_copy_callback(void *data)
 {
+       struct ntb_queue_entry *entry = data;
+       struct ntb_transport_qp *qp = entry->qp;
        void *cb_data = entry->cb_data;
        unsigned int len = entry->len;
+       struct ntb_payload_header *hdr = entry->rx_hdr;
+
+       wmb();

What is this barrier paired with?

You will see a wmb being removed from ntb_process_rxc below. It is VERY necessary to be here.

Quoted 8 lines

drivers/ntb/ntb_transport.c void ntb_transport_free(void *transport)

Quoted 5 lines

drivers/ntb/ntb_transport.c +2 -1

+       hdr->flags = 0;
 
-       memcpy(entry->buf, offset, entry->len);
+       /* If the callbacks come out of order, the writing of the

index to the

drivers/ntb/ntb_transport.c void ntb_transport_free(void *transport) · line 1003

Quoted 1 line

drivers/ntb/ntb_transport.c +1 -0

+        * last completed will be out of order.  This may result in

the the

Quoted 8 lines

drivers/ntb/ntb_transport.c void ntb_transport_free(void *transport) · line 1005

Quoted 2 lines

drivers/ntb/ntb_transport.c +2 -0

+        * receive stalling forever.
+        */

Is this for the case where we are bouncing back and forth between sync/async? Otherwise I do not see how transactions could get out of order given you allocate a channel once per queue. Is this comment saying that the iowrite32 is somehow a fix, or is this comment a FIXME?

There is a case for a mix, the "copy_bytes" variable above switches to CPU for small transfers (which greatly increases throughput on small transfers). The caveat to it is the need to flush the DMA engine to prevent out-of-order. This comment is mainly an reminder of this issue.

Quoted 45 lines

drivers/ntb/ntb_transport.c void ntb_transport_free(void *transport)

Quoted 4 lines

drivers/ntb/ntb_transport.c +1 -0

+       iowrite32(entry->index, &qp->rx_info->entry);
 
        ntb_list_add(&qp->ntb_rx_free_q_lock, &entry->entry,

&qp->rx_free_q);

drivers/ntb/ntb_transport.c static void ntb_rx_copy_task(struct

Quoted 2 lines

drivers/ntb/ntb_transport.c +0 -0

@@ -970,6 +1011,80 @@ static void ntb_rx_copy_task(struct

ntb_transport_qp *qp,

drivers/ntb/ntb_transport.c static void ntb_rx_copy_task(struct ntb_transport_qp *qp,

Quoted 4 lines

drivers/ntb/ntb_transport.c +1 -0

                qp->rx_handler(qp, qp->cb_data, cb_data, len);
 }
 
+static void ntb_memcpy_rx(struct ntb_queue_entry *entry, void

*offset)

Quoted 20 lines

drivers/ntb/ntb_transport.c +20 -0

+{
+       void *buf = entry->buf;
+       size_t len = entry->len;
+
+       memcpy(buf, offset, len);
+
+       ntb_rx_copy_callback(entry);
+}
+
+static void ntb_async_rx(struct ntb_queue_entry *entry, void *offset,
+                        size_t len)
+{
+       struct dma_async_tx_descriptor *txd;
+       struct ntb_transport_qp *qp = entry->qp;
+       struct dma_chan *chan = qp->dma_chan;
+       struct dma_device *device;
+       dma_addr_t src, dest;
+       dma_cookie_t cookie;
+       void *buf = entry->buf;
+       unsigned long flags;

drivers/ntb/ntb_transport.c static void ntb_rx_copy_task(struct ntb_transport_qp *qp,

Quoted 10 lines

drivers/ntb/ntb_transport.c +10 -0

+
+       entry->len = len;
+
+       if (!chan)
+               goto err;
+
+       device = chan->device;
+
+       if (len < copy_bytes ||
+           !is_dma_copy_aligned(device, __pa(offset), __pa(buf),

len))