page_pool: recycle buffers
Subsystems (13)
v6 cover page_pool: recycle buffers
From: Matteo Croce <[email protected]>
This is a respin of [1]
This patchset shows the plans for allowing page_pool to handle and maintain DMA map/unmap of the pages it serves to the driver. For this to work a return hook in the network core is introduced.
The overall purpose is to simplify drivers, by providing a page allocation API that does recycling, such that each driver doesn't have to reinvent its own recycling scheme. Using page_pool in a driver does not require implementing XDP support, but it makes it trivially easy to do so. Instead of allocating buffers specifically for SKBs we now allocate a generic buffer and either wrap it on an SKB (via build_skb) or create an XDP frame.
The recycling code leverages the XDP recycle APIs.
The Marvell mvpp2 and mvneta drivers are used in this patchset to demonstrate how to use the API, and tested on a MacchiatoBIN and EspressoBIN boards respectively.
Please let this going in on a future -rc1 so to allow enough time to have wider tests.
Note that this series depends on the change "mm: fix struct page layout on 32-bit systems"[2] which is not yet in master.
v5 -> v6
- preserve pfmemalloc bit when setting signature
- fix typo in mvneta
- rebase on next-next with the new cache
- don't clear the skb->pp_recycle in pskb_expand_head()
v4 -> v5:
- move the signature so it doesn't alias with page->mapping
- use an invalid pointer as magic
- incorporate Matthew Wilcox's changes for pfmemalloc pages
- move the __skb_frag_unref() changes to a preliminary patch
- refactor some cpp directives
- only attempt recycling if skb->head_frag
- clear skb->pp_recycle in pskb_expand_head()
v3 -> v4:
- store a pointer to page_pool instead of xdp_mem_info
- drop a patch which reduces xdp_mem_info size
- do the recycling in the page_pool code instead of xdp_return
- remove some unused headers include
- remove some useless forward declaration
v2 -> v3:
- added missing SOBs
- CCed the MM people
v1 -> v2:
- fix a commit message
- avoid setting pp_recycle multiple times on mvneta
- squash two patches to avoid breaking bisect
[1] https://lore.kernel.org/netdev/154413868810.21735.572808840657728172.stgit@firesoul/
[2] https://lore.kernel.org/linux-mm/[email protected]/
Ilias Apalodimas (1):
page_pool: Allow drivers to hint on SKB recycling
Matteo Croce (4):
mm: add a signature in struct page
skbuff: add a parameter to __skb_frag_unref
mvpp2: recycle buffers
mvneta: recycle buffers
drivers/net/ethernet/marvell/mvneta.c | 11 +++--- .../net/ethernet/marvell/mvpp2/mvpp2_main.c | 2 +- drivers/net/ethernet/marvell/sky2.c | 2 +- drivers/net/ethernet/mellanox/mlx4/en_rx.c | 2 +- include/linux/mm.h | 12 ++++--- include/linux/mm_types.h | 12 ++++++- include/linux/poison.h | 3 ++ include/linux/skbuff.h | 34 ++++++++++++++++--- include/net/page_pool.h | 9 +++++ net/core/page_pool.c | 29 ++++++++++++++++ net/core/skbuff.c | 24 ++++++++++--- net/tls/tls_device.c | 2 +- 12 files changed, 119 insertions(+), 23 deletions(-)
Signature · 1 line
2.31.1
PATCH v6 1/5 mm: add a signature in struct page
From: Matteo Croce <[email protected]>
This is needed by the page_pool to avoid recycling a page not allocated via page_pool.
The page->signature field is aliased to page->lru.next and page->compound_head, but it can't be set by mistake because the signature value is a bad pointer, and can't trigger a false positive in PageTail() because the last bit is 0.
--- include/linux/mm.h | 12 +++++++----- include/linux/mm_types.h | 12 +++++++++++- include/linux/poison.h | 3 +++ net/core/page_pool.c | 6 ++++++ 4 files changed, 27 insertions(+), 6 deletions(-)
include/linux/mm.h +7 -5
@@ -1668,10 +1668,12 @@ struct address_space *page_mapping(struct page *page);static inline bool page_is_pfmemalloc (const struct page *page) {/* -* Page index cannot be this large so this must be -* a pfmemalloc page. +* This is not a tail page; compound_head of a head page is unused +* at return from the page allocator, and will be overwritten +* by callers who do not care whether the page came from the +* reserves. */ -return page->index ==-1UL ; +return page->compound_head &BIT (1 ); }/*
@@ -1680,12 +1682,12 @@ static inline bool page_is_pfmemalloc(const struct page *page)*/ static inline void set_page_pfmemalloc (struct page *page) { - page->index =-1UL ; + page->compound_head =BIT (1 ); }static inline void clear_page_pfmemalloc (struct page *page) { - page->index =0 ; + page->compound_head =0 ; }/*
include/linux/mm_types.h +11 -1
@@ -96,6 +96,13 @@ struct page {unsigned long private; };struct {/* page_pool used by netstack */ +/** +* @pp_magic: magic value to avoid recycling non +* page_pool allocated pages. +*/ +unsigned long pp_magic; +struct page_pool *pp; +unsigned long _pp_mapping_pad;/** * @dma_addr: might require a 64-bit value on * 32-bit architectures.
@@ -130,7 +137,10 @@ struct page { }; };struct {/* Tail pages of compound page */ -unsigned long compound_head;/* Bit zero is set */ +/* Bit zero is set +* Bit one if pfmemalloc page +*/ +unsigned long compound_head;/* First tail page only */ unsigned char compound_dtor;
include/linux/poison.h +3 -0
@@ -78,4 +78,7 @@/********** security/ **********/ #define KEY_DESTROY 0xbd +/********** net/core/page_pool.c **********/ +#define PP_SIGNATURE (0x40 + POISON_POINTER_DELTA) +#endif
net/core/page_pool.c +6 -0
@@ -17,6 +17,7 @@#include <linux/dma-mapping.h> #include <linux/page-flags.h> #include <linux/mm.h> /* for __put_page() */ +#include <linux/poison.h> #include <trace/events/page_pool.h>
@@ -221,6 +222,8 @@ static struct page *__page_pool_alloc_page_order(struct page_pool *pool,return NULL ; } + page->pp_magic |=PP_SIGNATURE ; +/* Track how many pages are held 'in-flight' */ pool->pages_state_hold_cnt++;trace_page_pool_state_hold (pool, page, pool->pages_state_hold_cnt);
@@ -263,6 +266,7 @@ static struct page *__page_pool_alloc_pages_slow(struct page_pool *pool,put_page (page);continue ; } + page->pp_magic |=PP_SIGNATURE ; pool->alloc.cache[pool->alloc.count++] = page;/* Track how many pages are held 'in-flight' */ pool->pages_state_hold_cnt++;
@@ -341,6 +345,8 @@ void page_pool_release_page(struct page_pool *pool, struct page *page) DMA_ATTR_SKIP_CPU_SYNC);page_pool_set_dma_addr (page,0 ); skip_dma_unmap: + page->pp_magic =0 ; +/* This may be the last page returned, releasing the pool, so * it is not safe to reference pool afterwards. */
Signature · 1 line
2.31.1
PATCH v6 2/5 skbuff: add a parameter to __skb_frag_unref
In reply to Matteo CroceFrom: Matteo Croce <[email protected]>
This is a prerequisite patch, the next one is enabling recycling of skbs and fragments. Add an extra argument on __skb_frag_unref() to handle recycling, and update the current users of the function with that.
--- drivers/net/ethernet/marvell/sky2.c | 2 +- drivers/net/ethernet/mellanox/mlx4/en_rx.c | 2 +- include/linux/skbuff.h | 8 +++++--- net/core/skbuff.c | 4 ++-- net/tls/tls_device.c | 2 +- 5 files changed, 10 insertions(+), 8 deletions(-)
drivers/net/ethernet/marvell/sky2.c +1 -1
@@ -2503,7 +2503,7 @@ static void skb_put_frags(struct sk_buff *skb, unsigned int hdr_space,if (length ==0 ) {/* don't need this page */ -__skb_frag_unref (frag); +__skb_frag_unref (frag, false); --skb_shinfo (skb)->nr_frags; }else { size =min (length, (unsigned )PAGE_SIZE );
drivers/net/ethernet/mellanox/mlx4/en_rx.c +1 -1
@@ -526,7 +526,7 @@ static int mlx4_en_complete_rx_desc(struct mlx4_en_priv *priv, fail:while (nr >0 ) { nr--; -__skb_frag_unref (skb_shinfo (skb)->frags + nr); +__skb_frag_unref (skb_shinfo (skb)->frags + nr, false); }return 0 ; }
include/linux/skbuff.h +5 -3
@@ -3081,10 +3081,12 @@ static inline void skb_frag_ref(struct sk_buff *skb, int f)/** * __skb_frag_unref - release a reference on a paged fragment. * @frag: the paged fragment +* @recycle: recycle the page if allocated via page_pool * -* Releases a reference on the paged fragment @frag. +* Releases a reference on the paged fragment @frag +* or recycles the page via the page_pool API. */ -static inline void __skb_frag_unref (skb_frag_t *frag) +static inline void __skb_frag_unref (skb_frag_t *frag,bool recycle) {put_page (skb_frag_page (frag)); }
@@ -3098,7 +3100,7 @@ static inline void __skb_frag_unref(skb_frag_t *frag)*/ static inline void skb_frag_unref (struct sk_buff *skb,int f) { -__skb_frag_unref (&skb_shinfo (skb)->frags[f]); +__skb_frag_unref (&skb_shinfo (skb)->frags[f], false); }/**
net/core/skbuff.c +2 -2
@@ -664,7 +664,7 @@ static void skb_release_data(struct sk_buff *skb)skb_zcopy_clear (skb, true);for (i =0 ; i < shinfo->nr_frags; i++) -__skb_frag_unref (&shinfo->frags[i]); +__skb_frag_unref (&shinfo->frags[i], false);if (shinfo->frag_list)kfree_skb_list (shinfo->frag_list);
@@ -3495,7 +3495,7 @@ int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen) fragto = &skb_shinfo (tgt)->frags[merge];skb_frag_size_add (fragto,skb_frag_size (fragfrom)); -__skb_frag_unref (fragfrom); +__skb_frag_unref (fragfrom, false); }/* Reposition in the original skb */
net/tls/tls_device.c +1 -1
@@ -127,7 +127,7 @@ static void destroy_record(struct tls_record_info *record)int i;for (i =0 ; i < record->num_frags; i++) -__skb_frag_unref (&record->frags[i]); +__skb_frag_unref (&record->frags[i], false);kfree (record); }
Signature · 1 line
2.31.1
PATCH v6 3/5 page_pool: Allow drivers to hint on SKB recycling
In reply to Matteo CroceFrom: Ilias Apalodimas <[email protected]>
Up to now several high speed NICs have custom mechanisms of recycling the allocated memory they use for their payloads.
Our page_pool API already has recycling capabilities that are always used when we are running in 'XDP mode'. So let's tweak the API and the kernel network stack slightly and allow the recycling to happen even during the standard operation.
The API doesn't take into account 'split page' policies used by those drivers currently, but can be extended once we have users for that.
The idea is to be able to intercept the packet on skb_release_data(). If it's a buffer coming from our page_pool API recycle it back to the pool for further usage or just release the packet entirely.
To achieve that we introduce a bit in struct sk_buff (pp_recycle:1) and a field in struct page (page->pp) to store the page_pool pointer. Storing the information in page->pp allows us to recycle both SKBs and their fragments.
We could have skipped the skb bit entirely, since identical information can bederived from struct page. However, in an effort to affect the free path as less as possible, reading a single bit in the skb which is already in cache, is better that trying to derive identical information for the page stored data.
The driver or page_pool has to take care of the sync operations on it's own during the buffer recycling since the buffer is, after opting-in to the recycling, never unmapped.
Since the gain on the drivers depends on the architecture, we are not enabling recycling by default if the page_pool API is used on a driver. In order to enable recycling the driver must call skb_mark_for_recycle() to store the information we need for recycling in page->pp and enabling the recycling bit, or page_pool_store_mem_info() for a fragment.
--- include/linux/skbuff.h | 28 +++++++++++++++++++++++++--- include/net/page_pool.h | 9 +++++++++ net/core/page_pool.c | 23 +++++++++++++++++++++++ net/core/skbuff.c | 24 ++++++++++++++++++++---- 4 files changed, 77 insertions(+), 7 deletions(-)
include/linux/skbuff.h +25 -3
@@ -40,6 +40,9 @@#if IS_ENABLED (CONFIG_NF_CONNTRACK )#include <linux/netfilter/nf_conntrack_common.h> #endif +#ifdef CONFIG_PAGE_POOL +#include <net/page_pool.h> +#endif /* The interface for checksum offload between the stack and networking drivers * is as follows...
@@ -667,6 +670,8 @@ typedef unsigned char *sk_buff_data_t;* @head_frag: skb was allocated from page fragments, * not allocated by kmalloc() or vmalloc(). * @pfmemalloc: skbuff was allocated from PFMEMALLOC reserves +* @pp_recycle: mark the packet for recycling instead of freeing (implies +* page_pool support on driver) * @active_extensions: active extensions (skb_ext_id types) * @ndisc_nodetype: router type (from link layer) * @ooo_okay: allow the mapping of a socket to a queue to be changed
@@ -791,10 +796,12 @@ struct sk_buff { fclone:2, peeked:1, head_frag:1, - pfmemalloc:1; + pfmemalloc:1, + pp_recycle:1; /* page_pool recycle indicator */#ifdef CONFIG_SKB_EXTENSIONS __u8 active_extensions;#endif +/* fields enclosed in headers_start/headers_end are copied * using a single memcpy() in __copy_skb_header() */
@@ -3088,7 +3095,13 @@ static inline void skb_frag_ref(struct sk_buff *skb, int f)*/ static inline void __skb_frag_unref (skb_frag_t *frag,bool recycle) { -put_page (skb_frag_page (frag)); +struct page *page =skb_frag_page (frag); + +#ifdef CONFIG_PAGE_POOL +if (recycle &&page_pool_return_skb_page (page_address (page))) +return ; +#endif +put_page (page); }/**
@@ -3100,7 +3113,7 @@ static inline void __skb_frag_unref(skb_frag_t *frag, bool recycle)*/ static inline void skb_frag_unref (struct sk_buff *skb,int f) { -__skb_frag_unref (&skb_shinfo (skb)->frags[f], false); +__skb_frag_unref (&skb_shinfo (skb)->frags[f], skb->pp_recycle); }/**
@@ -4699,5 +4712,14 @@ static inline u64 skb_get_kcov_handle(struct sk_buff *skb)#endif } +#ifdef CONFIG_PAGE_POOL +static inline void skb_mark_for_recycle (struct sk_buff *skb,struct page *page, +struct page_pool *pp) +{ + skb->pp_recycle =1 ; +page_pool_store_mem_info (page, pp); +} +#endif +#endif /* __KERNEL__ */ #endif /* _LINUX_SKBUFF_H */
include/net/page_pool.h +9 -0
@@ -146,6 +146,8 @@ inline enum dma_data_direction page_pool_get_dma_dir(struct page_pool *pool)return pool->p.dma_dir; } +bool page_pool_return_skb_page (void *data); +struct page_pool *page_pool_create (const struct page_pool_params *params);#ifdef CONFIG_PAGE_POOL
@@ -251,4 +253,11 @@ static inline void page_pool_ring_unlock(struct page_pool *pool) spin_unlock_bh(&pool->ring.producer_lock); } +/* Store mem_info on struct page and use it while recycling skb frags */ +static inline +void page_pool_store_mem_info (struct page *page,struct page_pool *pp) +{ + page->pp = pp; +} +#endif /* _NET_PAGE_POOL_H */
net/core/page_pool.c +23 -0
@@ -628,3 +628,26 @@ void page_pool_update_nid(struct page_pool *pool, int new_nid) } }EXPORT_SYMBOL (page_pool_update_nid); + +bool page_pool_return_skb_page (void *data) +{ +struct page_pool *pp; +struct page *page; + + page =virt_to_head_page (data); +if (unlikely (page->pp_magic !=PP_SIGNATURE )) +return false; + + pp = (struct page_pool *)page->pp; + +/* Driver set this to memory recycling info. Reset it on recycle. +* This will *not* work for NIC using a split-page memory model. +* The page will be returned to the pool here regardless of the +* 'flipped' fragment being in use or not. +*/ + page->pp =NULL ; +page_pool_put_full_page (pp,virt_to_head_page (data), false); + +return true; +} +EXPORT_SYMBOL (page_pool_return_skb_page);
net/core/skbuff.c +20 -4
@@ -70,6 +70,9 @@#include <net/xfrm.h> #include <net/mpls.h> #include <net/mptcp.h> +#ifdef CONFIG_PAGE_POOL +#include <net/page_pool.h> +#endif #include <linux/uaccess.h> #include <trace/events/skb.h>
@@ -645,10 +648,15 @@ static void skb_free_head(struct sk_buff *skb) {unsigned char *head = skb->head; -if (skb->head_frag) +if (skb->head_frag) { +#ifdef CONFIG_PAGE_POOL +if (skb->pp_recycle &&page_pool_return_skb_page (head)) +return ; +#endif skb_free_frag (head); -else + }else {kfree (head); + } }static void skb_release_data (struct sk_buff *skb)
@@ -664,7 +672,7 @@ static void skb_release_data(struct sk_buff *skb)skb_zcopy_clear (skb, true);for (i =0 ; i < shinfo->nr_frags; i++) -__skb_frag_unref (&shinfo->frags[i], false); +__skb_frag_unref (&shinfo->frags[i], skb->pp_recycle);if (shinfo->frag_list)kfree_skb_list (shinfo->frag_list);
@@ -1046,6 +1054,7 @@ static struct sk_buff *__skb_clone(struct sk_buff *n, struct sk_buff *skb) n->nohdr =0 ; n->peeked =0 ;C (pfmemalloc); +C (pp_recycle); n->destructor =NULL ;C (tail);C (end);
@@ -3495,7 +3504,7 @@ int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen) fragto = &skb_shinfo (tgt)->frags[merge];skb_frag_size_add (fragto,skb_frag_size (fragfrom)); -__skb_frag_unref (fragfrom, false); +__skb_frag_unref (fragfrom, skb->pp_recycle); }/* Reposition in the original skb */
@@ -5285,6 +5294,13 @@ bool skb_try_coalesce(struct sk_buff *to, struct sk_buff *from,if (skb_cloned (to))return false; +/* The page pool signature of struct page will eventually figure out +* which pages can be recycled or not but for now let's prohibit slab +* allocated and page_pool allocated SKBs from being coalesced. +*/ +if (to->pp_recycle != from->pp_recycle) +return false; +if (len <=skb_tailroom (to)) {if (len)BUG_ON (skb_copy_bits (from,0 ,skb_put (to, len), len));
Signature · 1 line
2.31.1
PATCH v6 4/5 mvpp2: recycle buffers
In reply to Matteo CroceFrom: Matteo Croce <[email protected]>
Use the new recycling API for page_pool.
In a drop rate test, the packet rate is almost doubled, from 1110 Kpps to 2128 Kpps.
perf top on a stock system shows:
Overhead Shared Object Symbol 34.88% [kernel] [k] page_pool_release_page 8.06% [kernel] [k] free_unref_page 6.42% [mvpp2] [k] mvpp2_rx 6.07% [kernel] [k] eth_type_trans 5.18% [kernel] [k] __netif_receive_skb_core 4.95% [kernel] [k] build_skb 4.88% [kernel] [k] kmem_cache_free 3.97% [kernel] [k] kmem_cache_alloc 3.45% [kernel] [k] dev_gro_receive 2.73% [kernel] [k] page_frag_free 2.07% [kernel] [k] __alloc_pages_bulk 1.99% [kernel] [k] arch_local_irq_save 1.84% [kernel] [k] skb_release_data 1.20% [kernel] [k] netif_receive_skb_list_internal
With packet rate stable at 1100 Kpps:
tx: 0 bps 0 pps rx: 532.7 Mbps 1110 Kpps
tx: 0 bps 0 pps rx: 532.6 Mbps 1110 Kpps
tx: 0 bps 0 pps rx: 532.4 Mbps 1109 Kpps
tx: 0 bps 0 pps rx: 532.1 Mbps 1109 Kpps
tx: 0 bps 0 pps rx: 531.9 Mbps 1108 Kpps
tx: 0 bps 0 pps rx: 531.9 Mbps 1108 Kpps
And this is the same output with recycling enabled:
Overhead Shared Object Symbol 12.91% [kernel] [k] eth_type_trans 12.54% [mvpp2] [k] mvpp2_rx 9.67% [kernel] [k] build_skb 9.63% [kernel] [k] __netif_receive_skb_core 8.44% [kernel] [k] page_pool_put_page 8.07% [kernel] [k] kmem_cache_free 7.79% [kernel] [k] kmem_cache_alloc 6.86% [kernel] [k] dev_gro_receive 3.19% [kernel] [k] skb_release_data 2.41% [kernel] [k] netif_receive_skb_list_internal 2.18% [kernel] [k] page_pool_refill_alloc_cache 1.76% [kernel] [k] napi_gro_receive 1.61% [kernel] [k] kfree_skb 1.20% [kernel] [k] dma_sync_single_for_device 1.16% [mvpp2] [k] mvpp2_poll 1.12% [mvpp2] [k] mvpp2_read
With packet rate above 2100 Kpps:
tx: 0 bps 0 pps rx: 1021 Mbps 2128 Kpps
tx: 0 bps 0 pps rx: 1021 Mbps 2127 Kpps
tx: 0 bps 0 pps rx: 1021 Mbps 2128 Kpps
tx: 0 bps 0 pps rx: 1021 Mbps 2128 Kpps
tx: 0 bps 0 pps rx: 1022 Mbps 2128 Kpps
tx: 0 bps 0 pps rx: 1022 Mbps 2129 Kpps
The major performance increase is explained by the fact that the most CPU consuming functions (page_pool_release_page, page_frag_free and free_unref_page) are no longer called on a per packet basis.
The test was done by sending to the macchiatobin 64 byte ethernet frames with an invalid ethertype, so the packets are dropped early in the RX path.
--- drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c +1 -1
@@ -3964,7 +3964,7 @@ static int mvpp2_rx(struct mvpp2_port *port, struct napi_struct *napi, }if (pp) -page_pool_release_page (pp,virt_to_page (data)); +skb_mark_for_recycle (skb,virt_to_page (data), pp);else dma_unmap_single_attrs(dev->dev.parent, dma_addr, bm_pool->buf_size, DMA_FROM_DEVICE,
Signature · 1 line
2.31.1
PATCH v6 5/5 mvneta: recycle buffers
In reply to Matteo CroceFrom: Matteo Croce <[email protected]>
Use the new recycling API for page_pool.
In a drop rate test, the packet rate increased by 10%, from 269 Kpps to 296 Kpps.
perf top on a stock system shows:
Overhead Shared Object Symbol 21.78% [kernel] [k] __pi___inval_dcache_area 21.66% [mvneta] [k] mvneta_rx_swbm 7.00% [kernel] [k] kmem_cache_alloc 6.05% [kernel] [k] eth_type_trans 4.44% [kernel] [k] kmem_cache_free.part.0 3.80% [kernel] [k] __netif_receive_skb_core 3.68% [kernel] [k] dev_gro_receive 3.65% [kernel] [k] get_page_from_freelist 3.43% [kernel] [k] page_pool_release_page 3.35% [kernel] [k] free_unref_page
And this is the same output with recycling enabled:
Overhead Shared Object Symbol 24.10% [kernel] [k] __pi___inval_dcache_area 23.02% [mvneta] [k] mvneta_rx_swbm 7.19% [kernel] [k] kmem_cache_alloc 6.50% [kernel] [k] eth_type_trans 4.93% [kernel] [k] __netif_receive_skb_core 4.77% [kernel] [k] kmem_cache_free.part.0 3.93% [kernel] [k] dev_gro_receive 3.03% [kernel] [k] build_skb 2.91% [kernel] [k] page_pool_put_page 2.85% [kernel] [k] __xdp_return
The test was done with mausezahn on the TX side with 64 byte raw ethernet frames.
--- drivers/net/ethernet/marvell/mvneta.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-)
drivers/net/ethernet/marvell/mvneta.c +7 -4
@@ -2320,7 +2320,7 @@ mvneta_swbm_add_rx_fragment(struct mvneta_port *pp, }static struct sk_buff * -mvneta_swbm_build_skb (struct mvneta_port *pp,struct mvneta_rx_queue *rxq, +mvneta_swbm_build_skb (struct mvneta_port *pp,struct page_pool *pool,struct xdp_buff *xdp,u32 desc_status) {struct skb_shared_info *sinfo =xdp_get_shared_info_from_buff (xdp);
@@ -2331,7 +2331,7 @@ mvneta_swbm_build_skb(struct mvneta_port *pp, struct mvneta_rx_queue *rxq,if (!skb)return ERR_PTR (-ENOMEM ); -page_pool_release_page (rxq->page_pool,virt_to_page (xdp->data)); +skb_mark_for_recycle (skb,virt_to_page (xdp->data), pool);skb_reserve (skb, xdp->data - xdp->data_hard_start);skb_put (skb, xdp->data_end - xdp->data);
@@ -2343,7 +2343,10 @@ mvneta_swbm_build_skb(struct mvneta_port *pp, struct mvneta_rx_queue *rxq,skb_add_rx_frag (skb,skb_shinfo (skb)->nr_frags,skb_frag_page (frag),skb_frag_off (frag),skb_frag_size (frag),PAGE_SIZE ); -page_pool_release_page (rxq->page_pool,skb_frag_page (frag)); +/* We don't need to reset pp_recycle here. It's already set, so +* just mark fragments for recycling. +*/ +page_pool_store_mem_info (skb_frag_page (frag), pool); }return skb;
@@ -2425,7 +2428,7 @@ static int mvneta_rx_swbm(struct napi_struct *napi, mvneta_run_xdp(pp, rxq, xdp_prog, &xdp_buf, frame_sz, &ps)) goto next; - skb =mvneta_swbm_build_skb (pp, rxq, &xdp_buf, desc_status); + skb =mvneta_swbm_build_skb (pp, rxq->page_pool, &xdp_buf, desc_status);if (IS_ERR (skb)) {struct mvneta_pcpu_stats *stats =this_cpu_ptr (pp->stats);
Signature · 1 line
2.31.1
Quoted 3 lines · Matteo Croce
Note that this series depends on the change "mm: fix struct page layout on 32-bit systems"[2] which is not yet in master.
I see that it just entered net-next:
commit 9ddb3c14afba8bc5950ed297f02d4ae05ff35cd1
Author: Matthew Wilcox (Oracle) <[email protected]> Date: Fri May 14 17:27:24 2021 -0700
mm: fix struct page layout on 32-bit systems
Regards,
Signature · 1 line
per aspera ad upstream
Re: [PATCH net-next v6 3/5] page_pool: Allow drivers to hint on SKB recycling
In reply to Matteo CroceQuoted 8 lines · Matteo Croce
net/core/page_pool.c +8
-0+bool page_pool_return_skb_page (void *data) +{ +struct page_pool *pp; +struct page *page; + + page =virt_to_head_page (data); +if (unlikely (page->pp_magic !=PP_SIGNATURE )) +return false;
net/core/page_pool.c void page_pool_update_nid(struct page_pool *pool, int new_nid) · line 649
Quoted 10 lines
net/core/page_pool.c +10
-0+ + pp = (struct page_pool *)page->pp; + +/* Driver set this to memory recycling info. Reset it on recycle. +* This will *not* work for NIC using a split-page memory model. +* The page will be returned to the pool here regardless of the +* 'flipped' fragment being in use or not. +*/ + page->pp =NULL ; +page_pool_put_full_page (pp,virt_to_head_page (data), false);
Here I could just use the cached "page" instead of calling virt_to_head_page() once again.
Signature · 1 line
per aspera ad upstream
Re: [PATCH net-next v6 3/5] page_pool: Allow drivers to hint on SKB recycling
In reply to Matteo CroceQuoted 29 lines · Matteo Croce
From: Ilias Apalodimas <[email protected]>
Up to now several high speed NICs have custom mechanisms of recycling the allocated memory they use for their payloads.
Ourpage_poolAPI already has recycling capabilities that are always used when we are running in 'XDP mode'. So let's tweak the API and the kernel network stack slightly and allow the recycling to happen even during the standard operation.
The API doesn't take into account 'split page' policies used by those drivers currently, but can be extended once we have users for that.The idea is to be able to intercept the packet on
skb_release_data(). If it's a buffer coming from ourpage_poolAPI recycle it back to the pool for further usage or just release the packet entirely.To achieve that we introduce a bit in struct
sk_buff(pp_recycle:1) and a field in struct page (page->pp) to store thepage_poolpointer. Storing the information inpage->ppallows us to recycle both SKBs and their fragments.
We could have skipped the skb bit entirely, since identical information can bederived from struct page. However, in an effort to affect the free path as less as possible, reading a single bit in the skb which is already in cache, is better that trying to derive identical information for the page stored data.The driver or
page_poolhas to take care of the sync operations on it's own during the buffer recycling since the buffer is, after opting-in to the recycling, never unmapped.
Quoted 5 lines
Since the gain on the drivers depends on the architecture, we are not enabling recycling by default if the
page_poolAPI is used on a driver. In order to enable recycling the driver must callskb_mark_for_recycle()to store the information we need for recycling inpage->ppand enabling the recycling bit, orpage_pool_store_mem_info()for a fragment.
The state of this patch in patchwork is "Not Applicable", so you may need to respin it again.
Some minor comment below:
Quoted 37 lines
Co-developed-by Jesper Dangaard Brouer <[email protected]>Signed-off-by Jesper Dangaard Brouer <[email protected]>Co-developed-by Matteo Croce <[email protected]>Signed-off-by Matteo Croce <[email protected]>Signed-off-by Ilias Apalodimas <[email protected]>--- include/linux/skbuff.h | 28 +++++++++++++++++++++++++--- include/net/page_pool.h | 9 +++++++++ net/core/page_pool.c | 23 +++++++++++++++++++++++ net/core/skbuff.c | 24 ++++++++++++++++++++---- 4 files changed, 77 insertions(+), 7 deletions(-)include/linux/skbuff.h +5
-0@@ -40,6 +40,9 @@#if IS_ENABLED (CONFIG_NF_CONNTRACK )#include <linux/netfilter/nf_conntrack_common.h> #endif +#ifdef CONFIG_PAGE_POOL +#include <net/page_pool.h> +#endif /* The interface for checksum offload between the stack and networking drivers * is as follows... @@ -667,6 +670,8 @@ typedef unsigned char *sk_buff_data_t;* @head_frag: skb was allocated from page fragments, * not allocated by kmalloc() or vmalloc(). * @pfmemalloc: skbuff was allocated from PFMEMALLOC reserves +* @pp_recycle: mark the packet for recycling instead of freeing (implies +* page_pool support on driver) * @active_extensions: active extensions (skb_ext_id types) * @ndisc_nodetype: router type (from link layer) * @ooo_okay: allow the mapping of a socket to a queue to be changed
include/linux/skbuff.h struct sk_buff { · line 800
Quoted 6 lines
include/linux/skbuff.h +2
-1@@ -791,10 +796,12 @@ struct sk_buff { fclone:2, peeked:1, head_frag:1, - pfmemalloc:1; + pfmemalloc:1, + pp_recycle:1; /* page_pool recycle indicator */
The about comment seems unnecessary, for there is comment added above in this patch to explain that.
include/linux/skbuff.h struct sk_buff { · line 804
Quoted 4 lines
include/linux/skbuff.h +1
-0#ifdef CONFIG_SKB_EXTENSIONS __u8 active_extensions;#endif +
Unnecessary change?
Quoted 56 lines
include/linux/skbuff.h +17
-2/* fields enclosed in headers_start/headers_end are copied * using a single memcpy() in __copy_skb_header() */ @@ -3088,7 +3095,13 @@ static inline void skb_frag_ref(struct sk_buff *skb, int f)*/ static inline void __skb_frag_unref (skb_frag_t *frag,bool recycle) { -put_page (skb_frag_page (frag)); +struct page *page =skb_frag_page (frag); + +#ifdef CONFIG_PAGE_POOL +if (recycle &&page_pool_return_skb_page (page_address (page))) +return ; +#endif +put_page (page); }/** @@ -3100,7 +3113,7 @@ static inline void __skb_frag_unref(skb_frag_t *frag, bool recycle)*/ static inline void skb_frag_unref (struct sk_buff *skb,int f) { -__skb_frag_unref (&skb_shinfo (skb)->frags[f], false); +__skb_frag_unref (&skb_shinfo (skb)->frags[f], skb->pp_recycle); }/** @@ -4699,5 +4712,14 @@ static inline u64 skb_get_kcov_handle(struct sk_buff *skb)#endif } +#ifdef CONFIG_PAGE_POOL +static inline void skb_mark_for_recycle (struct sk_buff *skb,struct page *page, +struct page_pool *pp) +{ + skb->pp_recycle =1 ; +page_pool_store_mem_info (page, pp); +} +#endif +#endif /* __KERNEL__ */ #endif /* _LINUX_SKBUFF_H */ include/net/page_pool.h +2
-0@@ -146,6 +146,8 @@ inline enum dma_data_direction page_pool_get_dma_dir(struct page_pool *pool)return pool->p.dma_dir; } +bool page_pool_return_skb_page (void *data); +struct page_pool *page_pool_create (const struct page_pool_params *params);#ifdef CONFIG_PAGE_POOL
include/net/page_pool.h static inline void page_pool_ring_unlock(struct page_pool *pool) · line 258
Quoted 6 lines
include/net/page_pool.h +3
-0@@ -251,4 +253,11 @@ static inline void page_pool_ring_unlock(struct page_pool *pool)spin_unlock_bh (&pool->ring.producer_lock); } +/* Store mem_info on struct page and use it while recycling skb frags */ +static inline +void page_pool_store_mem_info (struct page *page,struct page_pool *pp)
The conventional practice seems to put "struct page_pool" before other parameter in page_pool.h.
Quoted 20 lines
include/net/page_pool.h +4
-0+{ + page->pp = pp; +} +#endif /* _NET_PAGE_POOL_H */ net/core/page_pool.c +7
-0@@ -628,3 +628,26 @@ void page_pool_update_nid(struct page_pool *pool, int new_nid) } }EXPORT_SYMBOL (page_pool_update_nid); + +bool page_pool_return_skb_page (void *data) +{ +struct page_pool *pp; +struct page *page; + + page =virt_to_head_page (data);
net/core/page_pool.c void page_pool_update_nid(struct page_pool *pool, int new_nid) · line 647
Quoted 10 lines
net/core/page_pool.c +10
-0+if (unlikely (page->pp_magic !=PP_SIGNATURE )) +return false; + + pp = (struct page_pool *)page->pp; + +/* Driver set this to memory recycling info. Reset it on recycle. +* This will *not* work for NIC using a split-page memory model. +* The page will be returned to the pool here regardless of the +* 'flipped' fragment being in use or not. +*/
I am not sure I understand how does the last part of comment related to the code below, as there is no driver using split-page memory model will reach here because those driver will not call skb_mark_for_recycle(), right?
Quoted 6 lines
net/core/page_pool.c +6
-0+ page->pp =NULL ; +page_pool_put_full_page (pp,virt_to_head_page (data), false); + +return true; +} +EXPORT_SYMBOL (page_pool_return_skb_page);
Ilias Apalodimas
maintainerRe: [PATCH net-next v6 3/5] page_pool: Allow drivers to hint on SKB recycling
[...]
Quoted 11 lines
net/core/page_pool.c void page_pool_update_nid(struct page_pool *pool, int new_nid) · line 647
Quoted 5 lines
net/core/page_pool.c +5
-0+/* Driver set this to memory recycling info. Reset it on recycle. +* This will *not* work for NIC using a split-page memory model. +* The page will be returned to the pool here regardless of the +* 'flipped' fragment being in use or not. +*/ I am not sure I understand how does the last part of comment related to the code below, as there is no driver using split-page memory model will reach here because those driver will not call
skb_mark_for_recycle(), right?
Yes the comment is there to prohibit people (mlx5 only actually) to add the recycling bit on their driver. Because if they do it will *probably* work but they might get random corrupted packets which will be hard to debug.
Quoted 7 lines
Quoted 6 lines
net/core/page_pool.c +6
-0+ page->pp =NULL ; +page_pool_put_full_page (pp,virt_to_head_page (data), false); + +return true; +} +EXPORT_SYMBOL (page_pool_return_skb_page);
Re: [PATCH net-next v6 3/5] page_pool: Allow drivers to hint on SKB recycling
Quoted 13 lines · Ilias Apalodimas
[...]
Quoted 11 lines
net/core/page_pool.c void page_pool_update_nid(struct page_pool *pool, int new_nid) · line 647
Quoted 5 lines
net/core/page_pool.c +5
-0+/* Driver set this to memory recycling info. Reset it on recycle. +* This will *not* work for NIC using a split-page memory model. +* The page will be returned to the pool here regardless of the +* 'flipped' fragment being in use or not. +*/ I am not sure I understand how does the last part of comment related to the code below, as there is no driver using split-page memory model will reach here because those driver will not call
skb_mark_for_recycle(), right?
Quoted 4 lines
Yes the comment is there to prohibit people (mlx5 only actually) to add the recycling bit on their driver. Because if they do it will *probably* work but they might get random corrupted packets which will be hard to debug.
What's the complexity for getting it to work with split page model? Since 1500 is the default MTU, requiring a page per packet means a lot of wasted memory.
Re: [PATCH net-next v6 3/5] page_pool: Allow drivers to hint on SKB recycling
Quoted 15 lines · David Ahern
Quoted 13 lines · Ilias Apalodimas
[...]
Quoted 11 lines
net/core/page_pool.c void page_pool_update_nid(struct page_pool *pool, int new_nid) · line 647
Quoted 5 lines
net/core/page_pool.c +5
-0+/* Driver set this to memory recycling info. Reset it on recycle. +* This will *not* work for NIC using a split-page memory model. +* The page will be returned to the pool here regardless of the +* 'flipped' fragment being in use or not. +*/ I am not sure I understand how does the last part of comment related to the code below, as there is no driver using split-page memory model will reach here because those driver will not call
skb_mark_for_recycle(), right?
Quoted 8 lines
Quoted 4 lines
Yes the comment is there to prohibit people (mlx5 only actually) to add the recycling bit on their driver. Because if they do it will *probably* work but they might get random corrupted packets which will be hard to debug.
What's the complexity for getting it to work with split page model? Since 1500 is the default MTU, requiring a page per packet means a lot of wasted memory.
We could create a new memory model, e.g. MEM_TYPE_PAGE_SPLIT, and restore the behavior present in the previous versions of this serie, which is, save xdp_mem_info in struct page.
As this could slightly impact the performances, this can be added in a future change when the drivers which are doing it want to use this recycling api.
Signature · 1 line
per aspera ad upstream
Re: [PATCH net-next v6 3/5] page_pool: Allow drivers to hint on SKB recycling
Quoted 25 lines · Matteo Croce
Quoted 15 lines · David Ahern
Quoted 13 lines · Ilias Apalodimas
[...]
Quoted 11 lines
net/core/page_pool.c void page_pool_update_nid(struct page_pool *pool, int new_nid) · line 647
Quoted 5 lines
net/core/page_pool.c +5
-0+/* Driver set this to memory recycling info. Reset it on recycle. +* This will *not* work for NIC using a split-page memory model. +* The page will be returned to the pool here regardless of the +* 'flipped' fragment being in use or not. +*/ I am not sure I understand how does the last part of comment related to the code below, as there is no driver using split-page memory model will reach here because those driver will not call
skb_mark_for_recycle(), right?Quoted 8 lines
Quoted 4 lines
Yes the comment is there to prohibit people (mlx5 only actually) to add the recycling bit on their driver. Because if they do it will *probably* work but they might get random corrupted packets which will be hard to debug.
What's the complexity for getting it to work with split page model? Since 1500 is the default MTU, requiring a page per packet means a lot of wasted memory.
Quoted 7 lines
We could create a new memory model, e.g.
MEM_TYPE_PAGE_SPLIT, and restore the behavior present in the previous versions of this serie, which is, savexdp_mem_infoin struct page.
As this could slightly impact the performances, this can be added in a future change when the drivers which are doing it want to use this recycling api.
page-split model doesn't only help reduce memory waste, but increase cache-locality, especially for aggregated GRO SKBs.
I'm looking forward to integrating the page-pool SKB recycling API into mlx5e datapath. For this we need it to support the page-split model.
Let's see what's missing and how we can help making this happen.
Regards,
Tariq
Ilias Apalodimas
maintainerRe: [PATCH net-next v6 3/5] page_pool: Allow drivers to hint on SKB recycling
In reply to David AhernHi David,
Quoted 14 lines · David Ahern
Quoted 13 lines · Ilias Apalodimas
[...]
Quoted 11 lines
net/core/page_pool.c void page_pool_update_nid(struct page_pool *pool, int new_nid) · line 647
Quoted 5 lines
net/core/page_pool.c +5
-0+/* Driver set this to memory recycling info. Reset it on recycle. +* This will *not* work for NIC using a split-page memory model. +* The page will be returned to the pool here regardless of the +* 'flipped' fragment being in use or not. +*/ I am not sure I understand how does the last part of comment related to the code below, as there is no driver using split-page memory model will reach here because those driver will not call
skb_mark_for_recycle(), right?
Quoted 8 lines
Quoted 4 lines
Yes the comment is there to prohibit people (mlx5 only actually) to add the recycling bit on their driver. Because if they do it will *probably* work but they might get random corrupted packets which will be hard to debug.
What's the complexity for getting it to work with split page model? Since 1500 is the default MTU, requiring a page per packet means a lot of wasted memory.
It boils down to 'can we re-use the page or is someone using it'. Yunsheng sent a patch in earlier series that implements this with ref counters. As Matteo mentions we can also add another page pool type.
In theory none of those sound too hard, but we'll have to code it and see.
/Ilias
Ilias Apalodimas
maintainerRe: [PATCH net-next v6 3/5] page_pool: Allow drivers to hint on SKB recycling
In reply to Tariq ToukanHi Tariq,
Quoted 18 lines
Quoted 10 lines
Quoted 9 lines
Quoted 5 lines
Yes the comment is there to prohibit people (mlx5 only actually) to add the recycling bit on their driver. Because if they do it will *probably* work but they might get random corrupted packets which will be hard to debug.
What's the complexity for getting it to work with split page model? Since 1500 is the default MTU, requiring a page per packet means a lot of wasted memory.
Quoted 7 lines
We could create a new memory model, e.g.
MEM_TYPE_PAGE_SPLIT, and restore the behavior present in the previous versions of this serie, which is, savexdp_mem_infoin struct page.
As this could slightly impact the performances, this can be added in a future change when the drivers which are doing it want to use this recycling api.
Quoted 7 lines
page-split model doesn't only help reduce memory waste, but increase cache-locality, especially for aggregated GRO SKBs.
I'm looking forward to integrating the page-pool SKB recycling API into mlx5e datapath. For this we need it to support the page-split model.
Let's see what's missing and how we can help making this happen.
Yes that's the final goal. As I said I don't think adding the page split model will fundamentally change the current patchset. So imho we should get this in first, make sure that everything is fine, and then add code for the mlx cards.
Regards
/Ilias
Re: [PATCH net-next v6 3/5] page_pool: Allow drivers to hint on SKB recycling
Quoted 26 lines · Ilias Apalodimas
Hi Tariq,
Quoted 24 lines
Quoted 10 lines
Quoted 9 lines
Quoted 5 lines
Yes the comment is there to prohibit people (mlx5 only actually) to add the recycling bit on their driver. Because if they do it will *probably* work but they might get random corrupted packets which will be hard to debug.
What's the complexity for getting it to work with split page model? Since 1500 is the default MTU, requiring a page per packet means a lot of wasted memory.
Quoted 7 lines
We could create a new memory model, e.g.
MEM_TYPE_PAGE_SPLIT, and restore the behavior present in the previous versions of this serie, which is, savexdp_mem_infoin struct page.
As this could slightly impact the performances, this can be added in a future change when the drivers which are doing it want to use this recycling api.page-split model doesn't only help reduce memory waste, but increase cache-locality, especially for aggregated GRO SKBs.
I'm looking forward to integrating the page-pool SKB recycling API into mlx5e datapath. For this we need it to support the page-split model.
Quoted 7 lines
Quoted 1 line
Let's see what's missing and how we can help making this happen.
Yes that's the final goal. As I said I don't think adding the page split model will fundamentally change the current patchset. So imho we should get this in first, make sure that everything is fine, and then add code for the mlx cards.
Sounds good