Message ID | 20230529092840.40413-2-linyunsheng@huawei.com (mailing list archive) |
---|---|
State | Changes Requested |
Delegated to: | Netdev Maintainers |
Headers | show |
Series | support non-frag page for page_pool_alloc_frag() | expand |
On Mon, 2023-05-29 at 17:28 +0800, Yunsheng Lin wrote: > Currently page_pool_dev_alloc_pages() can not be called > when PP_FLAG_PAGE_FRAG is set, because it does not use > the frag reference counting. > > As we are already doing a optimization by not updating > page->pp_frag_count in page_pool_defrag_page() for the > last frag user, and non-frag page only have one user, > so we utilize that to unify frag page and non-frag page > handling, so that page_pool_dev_alloc_pages() can also > be called with PP_FLAG_PAGE_FRAG set. > > Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com> > CC: Lorenzo Bianconi <lorenzo@kernel.org> > CC: Alexander Duyck <alexander.duyck@gmail.com> I"m not really a huge fan of the approach. Basically it looks like you are trying to turn every page pool page into a fragmented page. Why not just stick to keeping the fragemented pages and have a special case that just generates a mono-frag page for your allocator instead. The problem is there are some architectures where we just cannot support having pp_frag_count due to the DMA size. So it makes sense to leave those with just basic page pool instead of trying to fake that it is a fragmented page. > --- > include/net/page_pool.h | 38 +++++++++++++++++++++++++++++++------- > net/core/page_pool.c | 1 + > 2 files changed, 32 insertions(+), 7 deletions(-) > > diff --git a/include/net/page_pool.h b/include/net/page_pool.h > index c8ec2f34722b..ea7a0c0592a5 100644 > --- a/include/net/page_pool.h > +++ b/include/net/page_pool.h > @@ -50,6 +50,9 @@ > PP_FLAG_DMA_SYNC_DEV |\ > PP_FLAG_PAGE_FRAG) > > +#define PAGE_POOL_DMA_USE_PP_FRAG_COUNT \ > + (sizeof(dma_addr_t) > sizeof(unsigned long)) > + > /* > * Fast allocation side cache array/stack > * > @@ -295,13 +298,20 @@ void page_pool_put_defragged_page(struct page_pool *pool, struct page *page, > */ > static inline void page_pool_fragment_page(struct page *page, long nr) > { > - atomic_long_set(&page->pp_frag_count, nr); > + if (!PAGE_POOL_DMA_USE_PP_FRAG_COUNT) > + atomic_long_set(&page->pp_frag_count, nr); > } > > +/* We need to reset frag_count back to 1 for the last user to allow > + * only one user in case the page is recycled and allocated as non-frag > + * page. > + */ > static inline long page_pool_defrag_page(struct page *page, long nr) > { > long ret; > > + BUILD_BUG_ON(__builtin_constant_p(nr) && nr != 1); > + What is the point of this line? It doesn't make much sense to me. Are you just trying to force an optiimization? You would be better off just taking the BUILD_BUG_ON contents and feeding them into an if statement below since the statement will compile out anyway. It seems like what you would want here is: BUG_ON(!PAGE_POOL_DMA_USE_PP_FRAG_COUNT); Otherwise you are potentially writing to a variable that shouldn't exist. > /* If nr == pp_frag_count then we have cleared all remaining > * references to the page. No need to actually overwrite it, instead > * we can leave this to be overwritten by the calling function. > @@ -311,19 +321,36 @@ static inline long page_pool_defrag_page(struct page *page, long nr) > * especially when dealing with a page that may be partitioned > * into only 2 or 3 pieces. > */ > - if (atomic_long_read(&page->pp_frag_count) == nr) > + if (atomic_long_read(&page->pp_frag_count) == nr) { > + /* As we have ensured nr is always one for constant case > + * using the BUILD_BUG_ON() as above, only need to handle > + * the non-constant case here for frag count draining. > + */ > + if (!__builtin_constant_p(nr)) > + atomic_long_set(&page->pp_frag_count, 1); > + > return 0; > + } > > ret = atomic_long_sub_return(nr, &page->pp_frag_count); > WARN_ON(ret < 0); > + > + /* Reset frag count back to 1, this should be the rare case when > + * two users call page_pool_defrag_page() currently. > + */ > + if (!ret) > + atomic_long_set(&page->pp_frag_count, 1); > + > return ret; > } > > static inline bool page_pool_is_last_frag(struct page_pool *pool, > struct page *page) > { > - /* If fragments aren't enabled or count is 0 we were the last user */ > - return !(pool->p.flags & PP_FLAG_PAGE_FRAG) || > + /* When dma_addr_upper is overlapped with pp_frag_count > + * or we were the last page frag user. > + */ > + return PAGE_POOL_DMA_USE_PP_FRAG_COUNT || > (page_pool_defrag_page(page, 1) == 0); > } > > @@ -357,9 +384,6 @@ static inline void page_pool_recycle_direct(struct page_pool *pool, > page_pool_put_full_page(pool, page, true); > } > > -#define PAGE_POOL_DMA_USE_PP_FRAG_COUNT \ > - (sizeof(dma_addr_t) > sizeof(unsigned long)) > - > static inline dma_addr_t page_pool_get_dma_addr(struct page *page) > { > dma_addr_t ret = page->dma_addr; > diff --git a/net/core/page_pool.c b/net/core/page_pool.c > index e212e9d7edcb..0868aa8f6323 100644 > --- a/net/core/page_pool.c > +++ b/net/core/page_pool.c > @@ -334,6 +334,7 @@ static void page_pool_set_pp_info(struct page_pool *pool, > { > page->pp = pool; > page->pp_magic |= PP_SIGNATURE; > + page_pool_fragment_page(page, 1); > if (pool->p.init_callback) > pool->p.init_callback(page, pool->p.init_arg); > } Again, you are adding statements here that now have to be stripped out under specific circumstances. In my opinion it would be better to not modify base page pool pages and instead just have your allocator provide a 1 frag page pool page via a special case allocator rather then messing with all the other drivers.
On 2023/5/30 23:07, Alexander H Duyck wrote: > On Mon, 2023-05-29 at 17:28 +0800, Yunsheng Lin wrote: >> Currently page_pool_dev_alloc_pages() can not be called >> when PP_FLAG_PAGE_FRAG is set, because it does not use >> the frag reference counting. >> >> As we are already doing a optimization by not updating >> page->pp_frag_count in page_pool_defrag_page() for the >> last frag user, and non-frag page only have one user, >> so we utilize that to unify frag page and non-frag page >> handling, so that page_pool_dev_alloc_pages() can also >> be called with PP_FLAG_PAGE_FRAG set. >> >> Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com> >> CC: Lorenzo Bianconi <lorenzo@kernel.org> >> CC: Alexander Duyck <alexander.duyck@gmail.com> > > I"m not really a huge fan of the approach. Basically it looks like you > are trying to turn every page pool page into a fragmented page. Why not > just stick to keeping the fragemented pages and have a special case > that just generates a mono-frag page for your allocator instead. Let me try to describe what does this patch try to do and how it do that in more detailed in order to have more common understanding. Before this patch: As we use PP_FLAG_PAGE_FRAG to decide whether to check the frag count in page_pool_is_last_frag() when page is returned back to page pool, so: 1. PAGE_POOL_DMA_USE_PP_FRAG_COUNT being true case: page_pool_create() fails when it is called with PP_FLAG_PAGE_FRAG set, if user calls page_pool_alloc_frag(), then we warn about it and return NULL. 2. PAGE_POOL_DMA_USE_PP_FRAG_COUNT being false case: (1). page_pool_create() is called with PP_FLAG_PAGE_FRAG set: page_pool_alloc_pages() is not allowed to be called as the page->pp_frag_count is not setup correctly for the case. (2). page_pool_create() is called without PP_FLAG_PAGE_FRAG set: page_pool_alloc_frag() is not allowed to be called as the page->pp_frag_count is not checked in page_pool_is_last_frag(). and mlx5 using a mix of the about: page_pool_create() is called with with PP_FLAG_PAGE_FRAG, page_pool_dev_alloc_pages() is called to allocate a page and page_pool_fragment_page() is called to setup the page->pp_frag_count correctly so the page_pool_is_last_frag() can see the correct page->pp_frag_count, mlx5 driver handling the frag count is in the below, it is complicated and I am not sure if there are any added benefit that can justify the complication yet: https://www.spinics.net/lists/netdev/msg892893.html There are usecases for veth and virtio_net to use frag support in page pool to reduce memory usage, and it may request different frag size depending on the head/tail room space for xdp_frame/shinfo and mtu/packet size. When the requested frag size is large enough that a single page can not be split into more than one frag, using frag support only have performance penalty because of the extra frag count handling for frag support. So to avoid driver handling the page->pp_frag_count directly and driver calling different page pool API according to memory size, we need to way to unify the page_pool_is_last_frag() for frag and non-frag page. 1. https://patchwork.kernel.org/project/netdevbpf/patch/d3ae6bd3537fbce379382ac6a42f67e22f27ece2.1683896626.git.lorenzo@kernel.org/ 2. https://patchwork.kernel.org/project/netdevbpf/patch/20230526054621.18371-3-liangchen.linux@gmail.com/ After this patch: This patch ensure pp_frag_count of page from pool->alloc/pool->ring or newly allocated from page allocator is one, which means we assume that all pages have one frag user initially so that we can have a unified handling for frag and non-frag page in page_pool_is_last_frag(). So the key point in this patch is about unified handling in page_pool_is_last_frag(), which is the free/put side of page pool, not the alloc/generate side. Utilizing the page->pp_frag_count being one initially for every page is the least costly way to do that as best as I can think of. As it only add the cost of page_pool_fragment_page() for non-frag page case as you have mentioned below, which I think it is negligible as we are already dirtying the same cache line in page_pool_set_pp_info(). And for frag page, we avoid the reseting page->pp_frag_count to one by utilizing the optimization of not updating page->pp_frag_count in page_pool_defrag_page() for the last frag user. Please let me know if the above makes sense, or if misunderstood your concern here. > > The problem is there are some architectures where we just cannot > support having pp_frag_count due to the DMA size. So it makes sense to > leave those with just basic page pool instead of trying to fake that it > is a fragmented page. It kind of depend on how you veiw it, this patch view it as only supporting one frag when we can't support having pp_frag_count, so I would not call it faking. > >> --- >> include/net/page_pool.h | 38 +++++++++++++++++++++++++++++++------- >> net/core/page_pool.c | 1 + >> 2 files changed, 32 insertions(+), 7 deletions(-) >> >> diff --git a/include/net/page_pool.h b/include/net/page_pool.h >> index c8ec2f34722b..ea7a0c0592a5 100644 >> --- a/include/net/page_pool.h >> +++ b/include/net/page_pool.h >> @@ -50,6 +50,9 @@ >> PP_FLAG_DMA_SYNC_DEV |\ >> PP_FLAG_PAGE_FRAG) >> >> +#define PAGE_POOL_DMA_USE_PP_FRAG_COUNT \ >> + (sizeof(dma_addr_t) > sizeof(unsigned long)) >> + >> /* >> * Fast allocation side cache array/stack >> * >> @@ -295,13 +298,20 @@ void page_pool_put_defragged_page(struct page_pool *pool, struct page *page, >> */ >> static inline void page_pool_fragment_page(struct page *page, long nr) >> { >> - atomic_long_set(&page->pp_frag_count, nr); >> + if (!PAGE_POOL_DMA_USE_PP_FRAG_COUNT) >> + atomic_long_set(&page->pp_frag_count, nr); >> } >> >> +/* We need to reset frag_count back to 1 for the last user to allow >> + * only one user in case the page is recycled and allocated as non-frag >> + * page. >> + */ >> static inline long page_pool_defrag_page(struct page *page, long nr) >> { >> long ret; >> >> + BUILD_BUG_ON(__builtin_constant_p(nr) && nr != 1); >> + > > What is the point of this line? It doesn't make much sense to me. Are > you just trying to force an optiimization? You would be better off just > taking the BUILD_BUG_ON contents and feeding them into an if statement > below since the statement will compile out anyway. if the "if statement" you said refers to the below, then yes. >> + if (!__builtin_constant_p(nr)) >> + atomic_long_set(&page->pp_frag_count, 1); But it is a *BUILD*_BUG_ON(), isn't it compiled out anywhere we put it? Will move it down anyway to avoid confusion. > > It seems like what you would want here is: > BUG_ON(!PAGE_POOL_DMA_USE_PP_FRAG_COUNT); > > Otherwise you are potentially writing to a variable that shouldn't > exist. Not if the driver use the page_pool_alloc_frag() API instead of manipulating the page->pp_frag_count directly using the page_pool_defrag_page() like mlx5. The mlx5 call the page_pool_create() with with PP_FLAG_PAGE_FRAG set, and it does not seems to have a failback for PAGE_POOL_DMA_USE_PP_FRAG_COUNT case, and we may need to keep PP_FLAG_PAGE_FRAG for it. That's why we need to keep the driver from implementation detail(pp_frag_count handling specifically) of the frag support unless we have a very good reason. > >> /* If nr == pp_frag_count then we have cleared all remaining >> * references to the page. No need to actually overwrite it, instead >> * we can leave this to be overwritten by the calling function. >> @@ -311,19 +321,36 @@ static inline long page_pool_defrag_page(struct page *page, long nr) >> * especially when dealing with a page that may be partitioned >> * into only 2 or 3 pieces. >> */ >> - if (atomic_long_read(&page->pp_frag_count) == nr) >> + if (atomic_long_read(&page->pp_frag_count) == nr) { >> + /* As we have ensured nr is always one for constant case >> + * using the BUILD_BUG_ON() as above, only need to handle >> + * the non-constant case here for frag count draining. >> + */ >> + if (!__builtin_constant_p(nr)) >> + atomic_long_set(&page->pp_frag_count, 1); >> + >> return 0; >> + } >> >> ret = atomic_long_sub_return(nr, &page->pp_frag_count); >> WARN_ON(ret < 0); >> + >> + /* Reset frag count back to 1, this should be the rare case when >> + * two users call page_pool_defrag_page() currently. >> + */ >> + if (!ret) >> + atomic_long_set(&page->pp_frag_count, 1); >> + >> return ret; >> } >> >> static inline bool page_pool_is_last_frag(struct page_pool *pool, >> struct page *page) >> { >> - /* If fragments aren't enabled or count is 0 we were the last user */ >> - return !(pool->p.flags & PP_FLAG_PAGE_FRAG) || >> + /* When dma_addr_upper is overlapped with pp_frag_count >> + * or we were the last page frag user. >> + */ >> + return PAGE_POOL_DMA_USE_PP_FRAG_COUNT || >> (page_pool_defrag_page(page, 1) == 0); >> } >> >> @@ -357,9 +384,6 @@ static inline void page_pool_recycle_direct(struct page_pool *pool, >> page_pool_put_full_page(pool, page, true); >> } >> >> -#define PAGE_POOL_DMA_USE_PP_FRAG_COUNT \ >> - (sizeof(dma_addr_t) > sizeof(unsigned long)) >> - >> static inline dma_addr_t page_pool_get_dma_addr(struct page *page) >> { >> dma_addr_t ret = page->dma_addr; >> diff --git a/net/core/page_pool.c b/net/core/page_pool.c >> index e212e9d7edcb..0868aa8f6323 100644 >> --- a/net/core/page_pool.c >> +++ b/net/core/page_pool.c >> @@ -334,6 +334,7 @@ static void page_pool_set_pp_info(struct page_pool *pool, >> { >> page->pp = pool; >> page->pp_magic |= PP_SIGNATURE; >> + page_pool_fragment_page(page, 1); >> if (pool->p.init_callback) >> pool->p.init_callback(page, pool->p.init_arg); >> } > > Again, you are adding statements here that now have to be stripped out > under specific circumstances. In my opinion it would be better to not > modify base page pool pages and instead just have your allocator > provide a 1 frag page pool page via a special case allocator rather > then messing with all the other drivers. As above, it is about unifying handling for frag and non-frag page in page_pool_is_last_frag(). please let me know if there is any better way to do it without adding statements here. > . >
On Wed, 2023-05-31 at 19:55 +0800, Yunsheng Lin wrote: > On 2023/5/30 23:07, Alexander H Duyck wrote: > > On Mon, 2023-05-29 at 17:28 +0800, Yunsheng Lin wrote: > > > Currently page_pool_dev_alloc_pages() can not be called > > > when PP_FLAG_PAGE_FRAG is set, because it does not use > > > the frag reference counting. > > > > > > As we are already doing a optimization by not updating > > > page->pp_frag_count in page_pool_defrag_page() for the > > > last frag user, and non-frag page only have one user, > > > so we utilize that to unify frag page and non-frag page > > > handling, so that page_pool_dev_alloc_pages() can also > > > be called with PP_FLAG_PAGE_FRAG set. > > > > > > Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com> > > > CC: Lorenzo Bianconi <lorenzo@kernel.org> > > > CC: Alexander Duyck <alexander.duyck@gmail.com> > > > > I"m not really a huge fan of the approach. Basically it looks like you > > are trying to turn every page pool page into a fragmented page. Why not > > just stick to keeping the fragemented pages and have a special case > > that just generates a mono-frag page for your allocator instead. > > Let me try to describe what does this patch try to do and how it > do that in more detailed in order to have more common understanding. > > Before this patch: > > As we use PP_FLAG_PAGE_FRAG to decide whether to check the frag count > in page_pool_is_last_frag() when page is returned back to page pool, > so: > > 1. PAGE_POOL_DMA_USE_PP_FRAG_COUNT being true case: page_pool_create() > fails when it is called with PP_FLAG_PAGE_FRAG set, if user calls > page_pool_alloc_frag(), then we warn about it and return NULL. > > 2. PAGE_POOL_DMA_USE_PP_FRAG_COUNT being false case: > (1). page_pool_create() is called with PP_FLAG_PAGE_FRAG set: > page_pool_alloc_pages() is not allowed to be called as the > page->pp_frag_count is not setup correctly for the case. > (2). page_pool_create() is called without PP_FLAG_PAGE_FRAG set: > page_pool_alloc_frag() is not allowed to be called as the > page->pp_frag_count is not checked in page_pool_is_last_frag(). > > and mlx5 using a mix of the about: > page_pool_create() is called with with PP_FLAG_PAGE_FRAG, > page_pool_dev_alloc_pages() is called to allocate a page and > page_pool_fragment_page() is called to setup the page->pp_frag_count > correctly so the page_pool_is_last_frag() can see the correct > page->pp_frag_count, mlx5 driver handling the frag count is in the > below, it is complicated and I am not sure if there are any added > benefit that can justify the complication yet: > https://www.spinics.net/lists/netdev/msg892893.html > > There are usecases for veth and virtio_net to use frag support > in page pool to reduce memory usage, and it may request different > frag size depending on the head/tail room space for xdp_frame/shinfo > and mtu/packet size. When the requested frag size is large enough > that a single page can not be split into more than one frag, using > frag support only have performance penalty because of the extra frag > count handling for frag support. > So to avoid driver handling the page->pp_frag_count directly and driver > calling different page pool API according to memory size, we need to > way to unify the page_pool_is_last_frag() for frag and non-frag page. > > 1. https://patchwork.kernel.org/project/netdevbpf/patch/d3ae6bd3537fbce379382ac6a42f67e22f27ece2.1683896626.git.lorenzo@kernel.org/ > 2. https://patchwork.kernel.org/project/netdevbpf/patch/20230526054621.18371-3-liangchen.linux@gmail.com/ > > After this patch: > This patch ensure pp_frag_count of page from pool->alloc/pool->ring or > newly allocated from page allocator is one, which means we assume that > all pages have one frag user initially so that we can have a unified > handling for frag and non-frag page in page_pool_is_last_frag(). > > So the key point in this patch is about unified handling in > page_pool_is_last_frag(), which is the free/put side of page pool, > not the alloc/generate side. > > Utilizing the page->pp_frag_count being one initially for every page > is the least costly way to do that as best as I can think of. > As it only add the cost of page_pool_fragment_page() for non-frag page > case as you have mentioned below, which I think it is negligible as we > are already dirtying the same cache line in page_pool_set_pp_info(). > And for frag page, we avoid the reseting page->pp_frag_count to one by > utilizing the optimization of not updating page->pp_frag_count in > page_pool_defrag_page() for the last frag user. > > Please let me know if the above makes sense, or if misunderstood your > concern here. So my main concern is that what this is doing is masking things so that the veth and virtio_net drivers can essentially lie about the truesize of the memory they are using in order to improve their performance by misleading the socket layer about how much memory it is actually holding onto. We have historically had an issue with reporting the truesize of fragments, but generally the underestimation was kept to something less than 100% because pages were generally split by at least half. Where it would get messy is if a misbehaviing socket held onto packets for an exceedingly long time. What this patch set is doing is enabling explicit lying about the truesize, and it compounds that by allowing for mixing small allocations w/ large ones. > > > > The problem is there are some architectures where we just cannot > > support having pp_frag_count due to the DMA size. So it makes sense to > > leave those with just basic page pool instead of trying to fake that it > > is a fragmented page. > > It kind of depend on how you veiw it, this patch view it as only supporting > one frag when we can't support having pp_frag_count, so I would not call it > faking. So the big thing that make it "faking" is the truesize underestimation that will occur with these frames. > > > > > > --- > > > include/net/page_pool.h | 38 +++++++++++++++++++++++++++++++------- > > > net/core/page_pool.c | 1 + > > > 2 files changed, 32 insertions(+), 7 deletions(-) > > > > > > diff --git a/include/net/page_pool.h b/include/net/page_pool.h > > > index c8ec2f34722b..ea7a0c0592a5 100644 > > > --- a/include/net/page_pool.h > > > +++ b/include/net/page_pool.h > > > @@ -50,6 +50,9 @@ > > > PP_FLAG_DMA_SYNC_DEV |\ > > > PP_FLAG_PAGE_FRAG) > > > > > > +#define PAGE_POOL_DMA_USE_PP_FRAG_COUNT \ > > > + (sizeof(dma_addr_t) > sizeof(unsigned long)) > > > + > > > /* > > > * Fast allocation side cache array/stack > > > * > > > @@ -295,13 +298,20 @@ void page_pool_put_defragged_page(struct page_pool *pool, struct page *page, > > > */ > > > static inline void page_pool_fragment_page(struct page *page, long nr) > > > { > > > - atomic_long_set(&page->pp_frag_count, nr); > > > + if (!PAGE_POOL_DMA_USE_PP_FRAG_COUNT) > > > + atomic_long_set(&page->pp_frag_count, nr); > > > } > > > > > > +/* We need to reset frag_count back to 1 for the last user to allow > > > + * only one user in case the page is recycled and allocated as non-frag > > > + * page. > > > + */ > > > static inline long page_pool_defrag_page(struct page *page, long nr) > > > { > > > long ret; > > > > > > + BUILD_BUG_ON(__builtin_constant_p(nr) && nr != 1); > > > + > > > > What is the point of this line? It doesn't make much sense to me. Are > > you just trying to force an optiimization? You would be better off just > > taking the BUILD_BUG_ON contents and feeding them into an if statement > > below since the statement will compile out anyway. > > if the "if statement" you said refers to the below, then yes. > > > > + if (!__builtin_constant_p(nr)) > > > + atomic_long_set(&page->pp_frag_count, 1); > > But it is a *BUILD*_BUG_ON(), isn't it compiled out anywhere we put it? > > Will move it down anyway to avoid confusion. Actually now that I look at this more it is even more confusing. The whole point of this function was that we were supposed to be getting pp_frag_count to 0. However you are setting it to 1. This is seriously flawed. If we are going to treat non-fragmented pages as mono-frags then that is what we should do. We should be pulling this acounting into all of the page pool freeing paths, not trying to force the value up to 1 for the non-fragmented case. > > > > It seems like what you would want here is: > > BUG_ON(!PAGE_POOL_DMA_USE_PP_FRAG_COUNT); > > > > Otherwise you are potentially writing to a variable that shouldn't > > exist. > > Not if the driver use the page_pool_alloc_frag() API instead of manipulating > the page->pp_frag_count directly using the page_pool_defrag_page() like mlx5. > The mlx5 call the page_pool_create() with with PP_FLAG_PAGE_FRAG set, and > it does not seems to have a failback for PAGE_POOL_DMA_USE_PP_FRAG_COUNT > case, and we may need to keep PP_FLAG_PAGE_FRAG for it. That's why we need > to keep the driver from implementation detail(pp_frag_count handling specifically) > of the frag support unless we have a very good reason. > Getting the truesize is that "very good reason". The fact is the drivers were doing this long before page pool came around. Trying to pull that away from them is the wrong way to go in my opinion. > > > /* If nr == pp_frag_count then we have cleared all remaining > > > * references to the page. No need to actually overwrite it, instead > > > * we can leave this to be overwritten by the calling function. > > > @@ -311,19 +321,36 @@ static inline long page_pool_defrag_page(struct page *page, long nr) > > > * especially when dealing with a page that may be partitioned > > > * into only 2 or 3 pieces. > > > */ > > > - if (atomic_long_read(&page->pp_frag_count) == nr) > > > + if (atomic_long_read(&page->pp_frag_count) == nr) { > > > + /* As we have ensured nr is always one for constant case > > > + * using the BUILD_BUG_ON() as above, only need to handle > > > + * the non-constant case here for frag count draining. > > > + */ > > > + if (!__builtin_constant_p(nr)) > > > + atomic_long_set(&page->pp_frag_count, 1); > > > + > > > return 0; > > > + } > > > The optimization here was already the comparison since we didn't have to do anything if pp_frag_count == nr. The whole point of pp_frag_count going to 0 is that is considered non-fragmented in that case and ready to be freed. By resetting it to 1 you are implying that there is still one *other* user that is holding a fragment so the page cannot be freed. We weren't bothering with writing the value since the page is in the free path and this value is going to be unused until the page is reallocated anyway. > > > ret = atomic_long_sub_return(nr, &page->pp_frag_count); > > > WARN_ON(ret < 0); > > > + > > > + /* Reset frag count back to 1, this should be the rare case when > > > + * two users call page_pool_defrag_page() currently. > > > + */ > > > + if (!ret) > > > + atomic_long_set(&page->pp_frag_count, 1); > > > + > > > return ret; > > > } > > > > > > static inline bool page_pool_is_last_frag(struct page_pool *pool, > > > struct page *page) > > > { > > > - /* If fragments aren't enabled or count is 0 we were the last user */ > > > - return !(pool->p.flags & PP_FLAG_PAGE_FRAG) || > > > + /* When dma_addr_upper is overlapped with pp_frag_count > > > + * or we were the last page frag user. > > > + */ > > > + return PAGE_POOL_DMA_USE_PP_FRAG_COUNT || > > > (page_pool_defrag_page(page, 1) == 0); > > > } > > > > > > @@ -357,9 +384,6 @@ static inline void page_pool_recycle_direct(struct page_pool *pool, > > > page_pool_put_full_page(pool, page, true); > > > } > > > > > > -#define PAGE_POOL_DMA_USE_PP_FRAG_COUNT \ > > > - (sizeof(dma_addr_t) > sizeof(unsigned long)) > > > - > > > static inline dma_addr_t page_pool_get_dma_addr(struct page *page) > > > { > > > dma_addr_t ret = page->dma_addr; > > > diff --git a/net/core/page_pool.c b/net/core/page_pool.c > > > index e212e9d7edcb..0868aa8f6323 100644 > > > --- a/net/core/page_pool.c > > > +++ b/net/core/page_pool.c > > > @@ -334,6 +334,7 @@ static void page_pool_set_pp_info(struct page_pool *pool, > > > { > > > page->pp = pool; > > > page->pp_magic |= PP_SIGNATURE; > > > + page_pool_fragment_page(page, 1); > > > if (pool->p.init_callback) > > > pool->p.init_callback(page, pool->p.init_arg); > > > } > > > > Again, you are adding statements here that now have to be stripped out > > under specific circumstances. In my opinion it would be better to not > > modify base page pool pages and instead just have your allocator > > provide a 1 frag page pool page via a special case allocator rather > > then messing with all the other drivers. > > As above, it is about unifying handling for frag and non-frag page in > page_pool_is_last_frag(). please let me know if there is any better way > to do it without adding statements here. I get what you are trying to get at but I feel like the implementation is going to cause more problems than it helps. The problem is it is going to hurt base page pool performance and it just makes the fragmented pages that much more confusing to deal with. My advice as a first step would be to look at first solving how to enable the PP_FLAG_PAGE_FRAG mode when you have PAGE_POOL_DMA_USE_PP_FRAG_COUNT as true. That should be creating mono- frags as we are calling them, and we should have a way to get the truesize for those so we know when we are consuming significant amount of memory. Once that is solved then we can look at what it would take to apply mono-frags to the standard page pool case. Ideally we would need to find a way to do it with minimal overhead.
On 2023/6/3 0:37, Alexander H Duyck wrote: ... >> >> Please let me know if the above makes sense, or if misunderstood your >> concern here. > > So my main concern is that what this is doing is masking things so that > the veth and virtio_net drivers can essentially lie about the truesize > of the memory they are using in order to improve their performance by > misleading the socket layer about how much memory it is actually > holding onto. > > We have historically had an issue with reporting the truesize of > fragments, but generally the underestimation was kept to something less > than 100% because pages were generally split by at least half. Where it > would get messy is if a misbehaviing socket held onto packets for an > exceedingly long time. > > What this patch set is doing is enabling explicit lying about the > truesize, and it compounds that by allowing for mixing small > allocations w/ large ones. > >>> >>> The problem is there are some architectures where we just cannot >>> support having pp_frag_count due to the DMA size. So it makes sense to >>> leave those with just basic page pool instead of trying to fake that it >>> is a fragmented page. >> >> It kind of depend on how you veiw it, this patch view it as only supporting >> one frag when we can't support having pp_frag_count, so I would not call it >> faking. > > So the big thing that make it "faking" is the truesize underestimation > that will occur with these frames. Let's discuss truesize issue in patch 2 instead of here. Personally, I still believe that if the driver can compute the truesize correctly by manipulating the page->pp_frag_count and frag offset directly, the page pool can do that too. > >> >>> >>>> --- ... >>> >>> What is the point of this line? It doesn't make much sense to me. Are >>> you just trying to force an optiimization? You would be better off just >>> taking the BUILD_BUG_ON contents and feeding them into an if statement >>> below since the statement will compile out anyway. >> >> if the "if statement" you said refers to the below, then yes. >> >>>> + if (!__builtin_constant_p(nr)) >>>> + atomic_long_set(&page->pp_frag_count, 1); >> >> But it is a *BUILD*_BUG_ON(), isn't it compiled out anywhere we put it? >> >> Will move it down anyway to avoid confusion. > > Actually now that I look at this more it is even more confusing. The > whole point of this function was that we were supposed to be getting > pp_frag_count to 0. However you are setting it to 1. > > This is seriously flawed. If we are going to treat non-fragmented pages > as mono-frags then that is what we should do. We should be pulling this > acounting into all of the page pool freeing paths, not trying to force > the value up to 1 for the non-fragmented case. I am not sure I understand what do you mean by 'non-fragmented ', 'mono-frags', 'page pool freeing paths' and 'non-fragmented case' here. maybe describe it more detailed with something like the pseudocode? > >>> >>> It seems like what you would want here is: >>> BUG_ON(!PAGE_POOL_DMA_USE_PP_FRAG_COUNT); >>> >>> Otherwise you are potentially writing to a variable that shouldn't >>> exist. >> >> Not if the driver use the page_pool_alloc_frag() API instead of manipulating >> the page->pp_frag_count directly using the page_pool_defrag_page() like mlx5. >> The mlx5 call the page_pool_create() with with PP_FLAG_PAGE_FRAG set, and >> it does not seems to have a failback for PAGE_POOL_DMA_USE_PP_FRAG_COUNT >> case, and we may need to keep PP_FLAG_PAGE_FRAG for it. That's why we need >> to keep the driver from implementation detail(pp_frag_count handling specifically) >> of the frag support unless we have a very good reason. >> > > Getting the truesize is that "very good reason". The fact is the > drivers were doing this long before page pool came around. Trying to > pull that away from them is the wrong way to go in my opinion. If the truesize is really the concern here, I think it make more sense to enforce it in the page pool instead of each driver doing their trick, so I also think we can do better here to handle pp_frag_count in the page pool instead of driver handling it, so let's continue the truesize disscussion in patch 2 to see if we can come up with something better there. > >>>> /* If nr == pp_frag_count then we have cleared all remaining >>>> * references to the page. No need to actually overwrite it, instead >>>> * we can leave this to be overwritten by the calling function. >>>> @@ -311,19 +321,36 @@ static inline long page_pool_defrag_page(struct page *page, long nr) >>>> * especially when dealing with a page that may be partitioned >>>> * into only 2 or 3 pieces. >>>> */ >>>> - if (atomic_long_read(&page->pp_frag_count) == nr) >>>> + if (atomic_long_read(&page->pp_frag_count) == nr) { >>>> + /* As we have ensured nr is always one for constant case >>>> + * using the BUILD_BUG_ON() as above, only need to handle >>>> + * the non-constant case here for frag count draining. >>>> + */ >>>> + if (!__builtin_constant_p(nr)) >>>> + atomic_long_set(&page->pp_frag_count, 1); >>>> + >>>> return 0; >>>> + } >>>> > > The optimization here was already the comparison since we didn't have > to do anything if pp_frag_count == nr. The whole point of pp_frag_count > going to 0 is that is considered non-fragmented in that case and ready > to be freed. By resetting it to 1 you are implying that there is still > one *other* user that is holding a fragment so the page cannot be > freed. > > We weren't bothering with writing the value since the page is in the > free path and this value is going to be unused until the page is > reallocated anyway. I am not sure what you meant above. But I will describe what is this patch trying to do again: When PP_FLAG_PAGE_FRAG is set and that flag is per page pool, not per page, so page_pool_alloc_pages() is not allowed to be called as the page->pp_frag_count is not setup correctly for the case. So in order to allow calling page_pool_alloc_pages(), as best as I can think of, either we need a per page flag/bit to decide whether to do something like dec_and_test for page->pp_frag_count in page_pool_is_last_frag(), or we unify the page->pp_frag_count handling in page_pool_is_last_frag() so that we don't need a per page flag/bit. This patch utilizes the optimization you mentioned above to unify the page->pp_frag_count handling. > >>>> ret = atomic_long_sub_return(nr, &page->pp_frag_count); >>>> WARN_ON(ret < 0); >>>> + >>>> + /* Reset frag count back to 1, this should be the rare case when >>>> + * two users call page_pool_defrag_page() currently. >>>> + */ >>>> + if (!ret) >>>> + atomic_long_set(&page->pp_frag_count, 1); >>>> + >>>> return ret; >>>> } >>>> ... >> As above, it is about unifying handling for frag and non-frag page in >> page_pool_is_last_frag(). please let me know if there is any better way >> to do it without adding statements here. > > I get what you are trying to get at but I feel like the implementation > is going to cause more problems than it helps. The problem is it is > going to hurt base page pool performance and it just makes the > fragmented pages that much more confusing to deal with. For base page pool performance, as I mentioned before: It remove PP_FLAG_PAGE_FRAG checking and only add the cost of page_pool_fragment_page() in page_pool_set_pp_info(), which I think it is negligible as we are already dirtying the same cache line in page_pool_set_pp_info(). For the confusing, sometimes it is about personal taste, so I am not going to argue with it:) But it would be good to provide a non-confusing way to do that with minimal overhead. I feel like you have provided it in the begin, but I am not able to understand it yet. > > My advice as a first step would be to look at first solving how to > enable the PP_FLAG_PAGE_FRAG mode when you have > PAGE_POOL_DMA_USE_PP_FRAG_COUNT as true. That should be creating mono- > frags as we are calling them, and we should have a way to get the > truesize for those so we know when we are consuming significant amount > of memory. Does the way to get the truesize in the below RFC make sense to you? https://patchwork.kernel.org/project/netdevbpf/patch/20230516124801.2465-4-linyunsheng@huawei.com/ > > Once that is solved then we can look at what it would take to apply > mono-frags to the standard page pool case. Ideally we would need to > find a way to do it with minimal overhead. > >
On Sat, Jun 3, 2023 at 5:59 AM Yunsheng Lin <yunshenglin0825@gmail.com> wrote: > > On 2023/6/3 0:37, Alexander H Duyck wrote: > ... > > >> > >> Please let me know if the above makes sense, or if misunderstood your > >> concern here. > > > > So my main concern is that what this is doing is masking things so that > > the veth and virtio_net drivers can essentially lie about the truesize > > of the memory they are using in order to improve their performance by > > misleading the socket layer about how much memory it is actually > > holding onto. > > > > We have historically had an issue with reporting the truesize of > > fragments, but generally the underestimation was kept to something less > > than 100% because pages were generally split by at least half. Where it > > would get messy is if a misbehaviing socket held onto packets for an > > exceedingly long time. > > > > What this patch set is doing is enabling explicit lying about the > > truesize, and it compounds that by allowing for mixing small > > allocations w/ large ones. > > > >>> > >>> The problem is there are some architectures where we just cannot > >>> support having pp_frag_count due to the DMA size. So it makes sense to > >>> leave those with just basic page pool instead of trying to fake that it > >>> is a fragmented page. > >> > >> It kind of depend on how you veiw it, this patch view it as only supporting > >> one frag when we can't support having pp_frag_count, so I would not call it > >> faking. > > > > So the big thing that make it "faking" is the truesize underestimation > > that will occur with these frames. > > Let's discuss truesize issue in patch 2 instead of here. > Personally, I still believe that if the driver can compute the > truesize correctly by manipulating the page->pp_frag_count and > frag offset directly, the page pool can do that too. > > > > >> > >>> > >>>> --- > > ... > > >>> > >>> What is the point of this line? It doesn't make much sense to me. Are > >>> you just trying to force an optiimization? You would be better off just > >>> taking the BUILD_BUG_ON contents and feeding them into an if statement > >>> below since the statement will compile out anyway. > >> > >> if the "if statement" you said refers to the below, then yes. > >> > >>>> + if (!__builtin_constant_p(nr)) > >>>> + atomic_long_set(&page->pp_frag_count, 1); > >> > >> But it is a *BUILD*_BUG_ON(), isn't it compiled out anywhere we put it? > >> > >> Will move it down anyway to avoid confusion. > > > > Actually now that I look at this more it is even more confusing. The > > whole point of this function was that we were supposed to be getting > > pp_frag_count to 0. However you are setting it to 1. > > > > This is seriously flawed. If we are going to treat non-fragmented pages > > as mono-frags then that is what we should do. We should be pulling this > > acounting into all of the page pool freeing paths, not trying to force > > the value up to 1 for the non-fragmented case. > > I am not sure I understand what do you mean by 'non-fragmented ', > 'mono-frags', 'page pool freeing paths' and 'non-fragmented case' > here. maybe describe it more detailed with something like the > pseudocode? What you are attempting to generate are "mono-frags" where a page pool page has a frag count of 1. I refer to "non-fragmented pages" as the legacy page pool page without pp_frags set. The "page-pool freeing paths" are the ones outside of the fragmented bits here. Basically __page_pool_put_page and the like. What you should be doing is pushing the reference counting code down deeper into the page pool logic. Currently it is more of a surface setup. The whole point I am getting at with this is that we should see the number of layers reduced for the fragmented pages, and by converting the non-fragmented pages to mono-frags we should see that maintain its current performance and total number of layers instead of having more layers added to it. > > > >>> > >>> It seems like what you would want here is: > >>> BUG_ON(!PAGE_POOL_DMA_USE_PP_FRAG_COUNT); > >>> > >>> Otherwise you are potentially writing to a variable that shouldn't > >>> exist. > >> > >> Not if the driver use the page_pool_alloc_frag() API instead of manipulating > >> the page->pp_frag_count directly using the page_pool_defrag_page() like mlx5. > >> The mlx5 call the page_pool_create() with with PP_FLAG_PAGE_FRAG set, and > >> it does not seems to have a failback for PAGE_POOL_DMA_USE_PP_FRAG_COUNT > >> case, and we may need to keep PP_FLAG_PAGE_FRAG for it. That's why we need > >> to keep the driver from implementation detail(pp_frag_count handling specifically) > >> of the frag support unless we have a very good reason. > >> > > > > Getting the truesize is that "very good reason". The fact is the > > drivers were doing this long before page pool came around. Trying to > > pull that away from them is the wrong way to go in my opinion. > > If the truesize is really the concern here, I think it make more > sense to enforce it in the page pool instead of each driver doing > their trick, so I also think we can do better here to handle > pp_frag_count in the page pool instead of driver handling it, so > let's continue the truesize disscussion in patch 2 to see if we > can come up with something better there. The problem is we don't free the page until the next allocation so the truesize will be false as the remainder of the page should be added to the truesize. The drivers tend to know what they are doing with the page and when they are freeing it. We don't have that sort of knowledge when we are doing the allocation. > > > >>>> /* If nr == pp_frag_count then we have cleared all remaining > >>>> * references to the page. No need to actually overwrite it, instead > >>>> * we can leave this to be overwritten by the calling function. > >>>> @@ -311,19 +321,36 @@ static inline long page_pool_defrag_page(struct page *page, long nr) > >>>> * especially when dealing with a page that may be partitioned > >>>> * into only 2 or 3 pieces. > >>>> */ > >>>> - if (atomic_long_read(&page->pp_frag_count) == nr) > >>>> + if (atomic_long_read(&page->pp_frag_count) == nr) { > >>>> + /* As we have ensured nr is always one for constant case > >>>> + * using the BUILD_BUG_ON() as above, only need to handle > >>>> + * the non-constant case here for frag count draining. > >>>> + */ > >>>> + if (!__builtin_constant_p(nr)) > >>>> + atomic_long_set(&page->pp_frag_count, 1); > >>>> + > >>>> return 0; > >>>> + } > >>>> > > > > The optimization here was already the comparison since we didn't have > > to do anything if pp_frag_count == nr. The whole point of pp_frag_count > > going to 0 is that is considered non-fragmented in that case and ready > > to be freed. By resetting it to 1 you are implying that there is still > > one *other* user that is holding a fragment so the page cannot be > > freed. > > > > We weren't bothering with writing the value since the page is in the > > free path and this value is going to be unused until the page is > > reallocated anyway. > > I am not sure what you meant above. > But I will describe what is this patch trying to do again: > When PP_FLAG_PAGE_FRAG is set and that flag is per page pool, not per > page, so page_pool_alloc_pages() is not allowed to be called as the > page->pp_frag_count is not setup correctly for the case. > > So in order to allow calling page_pool_alloc_pages(), as best as I > can think of, either we need a per page flag/bit to decide whether > to do something like dec_and_test for page->pp_frag_count in > page_pool_is_last_frag(), or we unify the page->pp_frag_count handling > in page_pool_is_last_frag() so that we don't need a per page flag/bit. > > This patch utilizes the optimization you mentioned above to unify the > page->pp_frag_count handling. Basically what should be happening if all page-pool pages are to be considered "fragmented" is that we should be folding this into the freeing logic. What we now have a 2 stage setup where we are dropping the count to 0, then rebounding it and setting it back to 1. If we are going to have all page pool pages fragmented then the freeing path for page pool pages should just be handling frag count directly instead of hacking on it here and ignoring it in the actual freeing paths. > > > >>>> ret = atomic_long_sub_return(nr, &page->pp_frag_count); > >>>> WARN_ON(ret < 0); > >>>> + > >>>> + /* Reset frag count back to 1, this should be the rare case when > >>>> + * two users call page_pool_defrag_page() currently. > >>>> + */ > >>>> + if (!ret) > >>>> + atomic_long_set(&page->pp_frag_count, 1); > >>>> + > >>>> return ret; > >>>> } > >>>> > > ... > > >> As above, it is about unifying handling for frag and non-frag page in > >> page_pool_is_last_frag(). please let me know if there is any better way > >> to do it without adding statements here. > > > > I get what you are trying to get at but I feel like the implementation > > is going to cause more problems than it helps. The problem is it is > > going to hurt base page pool performance and it just makes the > > fragmented pages that much more confusing to deal with. > > For base page pool performance, as I mentioned before: > It remove PP_FLAG_PAGE_FRAG checking and only add the cost of > page_pool_fragment_page() in page_pool_set_pp_info(), which I > think it is negligible as we are already dirtying the same cache > line in page_pool_set_pp_info(). I have no problem with getting rid of the flag. > For the confusing, sometimes it is about personal taste, so I am > not going to argue with it:) But it would be good to provide a > non-confusing way to do that with minimal overhead. I feel like > you have provided it in the begin, but I am not able to understand > it yet. The problem here is that instead of treating all page pool pages as fragmented, what the patch set has done is added a shim layer so that you are layering fragmentation on top of page pool pages which was already the case. That is why I have suggested make page pool pages a "mono-frag" as your first patch. Basically it is going to force you to have to set the pp_frag value for these pages, and verify it is 1 when you are freeing it. Then you are going to have to modify the fragmented cases to make use of lower level calls because now instead of us defragging a fragmented page, and then freeing it the two operations essentially have to be combined into one operation. > > > > My advice as a first step would be to look at first solving how to > > enable the PP_FLAG_PAGE_FRAG mode when you have > > PAGE_POOL_DMA_USE_PP_FRAG_COUNT as true. That should be creating mono- > > frags as we are calling them, and we should have a way to get the > > truesize for those so we know when we are consuming significant amount > > of memory. > > Does the way to get the truesize in the below RFC make sense to you? > https://patchwork.kernel.org/project/netdevbpf/patch/20230516124801.2465-4-linyunsheng@huawei.com/ It doesn't add any value. All you are doing is passing the "size" value as "truesize". The whole point of the "truesize" would be to report the actual size. So a step in that direction would be to bump truesize to include the remainder that isn't used when you decide it is time to allocate a new page. The problem is that it requires some fore-knowledge of what the next requested size is going to be. That is why it is better to just have the drivers manage this since they know what size they typically request and when they are going to close pages. Like I said, if you are wanting to go down this path you are better off starting with page pool and making all regular page pool pages into mono-frags. Then from there we can start building things out. With that you could then let drivers like the Mellanox one handle its own fragmenting knowing it has to return things to a mono-frag in order for it to be working correctly.
On 2023/6/5 22:58, Alexander Duyck wrote: ... >> >> I am not sure I understand what do you mean by 'non-fragmented ', >> 'mono-frags', 'page pool freeing paths' and 'non-fragmented case' >> here. maybe describe it more detailed with something like the >> pseudocode? > > What you are attempting to generate are "mono-frags" where a page pool > page has a frag count of 1. I refer to "non-fragmented pages" as the > legacy page pool page without pp_frags set. > > The "page-pool freeing paths" are the ones outside of the fragmented > bits here. Basically __page_pool_put_page and the like. What you > should be doing is pushing the reference counting code down deeper > into the page pool logic. Currently it is more of a surface setup. > > The whole point I am getting at with this is that we should see the > number of layers reduced for the fragmented pages, and by converting > the non-fragmented pages to mono-frags we should see that maintain its > current performance and total number of layers instead of having more > layers added to it. Do you mean reducing the number of layers for the fragmented pages by moving the page->pp_frag_count handling from page_pool_defrag_page() to __page_pool_put_page() where page->_refcount is checked? Or merge page->pp_frag_count into page->_refcount so that we don't need page->pp_frag_count anymore? As my understanding, when a page from page pool is passed to the stack to be processed, the stack may hold onto that page by taking page->_refcount too, which means page pool has no control over who will hold onto and when that taken will be released, that is why page pool do the "page_ref_count(page) == 1" checking in __page_pool_put_page(), if it is not true, the page pool can't recycle the page, so pp_frag_count and _refcount have different meaning and serve different purpose, merging them doesn't work, and moving them to one place doesn't make much sense too? Or is there other obvious consideration that I missed? >>> >> I am not sure what you meant above. >> But I will describe what is this patch trying to do again: >> When PP_FLAG_PAGE_FRAG is set and that flag is per page pool, not per >> page, so page_pool_alloc_pages() is not allowed to be called as the >> page->pp_frag_count is not setup correctly for the case. >> >> So in order to allow calling page_pool_alloc_pages(), as best as I >> can think of, either we need a per page flag/bit to decide whether >> to do something like dec_and_test for page->pp_frag_count in >> page_pool_is_last_frag(), or we unify the page->pp_frag_count handling >> in page_pool_is_last_frag() so that we don't need a per page flag/bit. >> >> This patch utilizes the optimization you mentioned above to unify the >> page->pp_frag_count handling. > > Basically what should be happening if all page-pool pages are to be > considered "fragmented" is that we should be folding this into the > freeing logic. What we now have a 2 stage setup where we are dropping > the count to 0, then rebounding it and setting it back to 1. If we are > going to have all page pool pages fragmented then the freeing path for > page pool pages should just be handling frag count directly instead of > hacking on it here and ignoring it in the actual freeing paths. Do you mean doing something like below? isn't it dirtying the cache line of 'struct page' whenever a page is recycled, which means we may not be able to the maintain current performance for non-fragmented or mono-frag case? --- a/net/core/page_pool.c +++ b/net/core/page_pool.c @@ -583,6 +583,10 @@ static __always_inline struct page * __page_pool_put_page(struct page_pool *pool, struct page *page, unsigned int dma_sync_size, bool allow_direct) { + + if (!page_pool_defrag_page(page, 1)) + return NULL; + /* This allocator is optimized for the XDP mode that uses * one-frame-per-page, but have fallbacks that act like the * regular page allocator APIs. @@ -594,6 +598,7 @@ __page_pool_put_page(struct page_pool *pool, struct page *page, */ if (likely(page_ref_count(page) == 1 && !page_is_pfmemalloc(page))) { /* Read barrier done in page_ref_count / READ_ONCE */ + page_pool_fragment_page(page, 1); if (pool->p.flags & PP_FLAG_DMA_SYNC_DEV) page_pool_dma_sync_for_device(pool, page, > >>> >>>>>> ret = atomic_long_sub_return(nr, &page->pp_frag_count); >>>>>> WARN_ON(ret < 0); >>>>>> + >>>>>> + /* Reset frag count back to 1, this should be the rare case when >>>>>> + * two users call page_pool_defrag_page() currently. >>>>>> + */ >>>>>> + if (!ret) >>>>>> + atomic_long_set(&page->pp_frag_count, 1); >>>>>> + >>>>>> return ret; >>>>>> } >>>>>> >> >> ... >> >>>> As above, it is about unifying handling for frag and non-frag page in >>>> page_pool_is_last_frag(). please let me know if there is any better way >>>> to do it without adding statements here. >>> >>> I get what you are trying to get at but I feel like the implementation >>> is going to cause more problems than it helps. The problem is it is >>> going to hurt base page pool performance and it just makes the >>> fragmented pages that much more confusing to deal with. >> >> For base page pool performance, as I mentioned before: >> It remove PP_FLAG_PAGE_FRAG checking and only add the cost of >> page_pool_fragment_page() in page_pool_set_pp_info(), which I >> think it is negligible as we are already dirtying the same cache >> line in page_pool_set_pp_info(). > > I have no problem with getting rid of the flag. > >> For the confusing, sometimes it is about personal taste, so I am >> not going to argue with it:) But it would be good to provide a >> non-confusing way to do that with minimal overhead. I feel like >> you have provided it in the begin, but I am not able to understand >> it yet. > > The problem here is that instead of treating all page pool pages as > fragmented, what the patch set has done is added a shim layer so that > you are layering fragmentation on top of page pool pages which was > already the case. > > That is why I have suggested make page pool pages a "mono-frag" as > your first patch. Basically it is going to force you to have to set > the pp_frag value for these pages, and verify it is 1 when you are > freeing it. It seems it is bascially what this patch do with minimal overhead to the previous users. Let me try again with what this patch mainly do: Currently when page_pool_create() is called with PP_FLAG_PAGE_FRAG flag, page_pool_alloc_pages() is only allowed to be called under the below constraints: 1. page_pool_fragment_page() need to be called to setup page->pp_frag_count immediately. 2. page_pool_defrag_page() often need to be called to drain the page->pp_frag_count when there is no more user will be holding on to that page. Those constraints exist in order to support a page to be splitted into multi frags. And those constraints have some overhead because of the cache line dirtying/bouncing and atomic update. Those constraints are unavoidable for case when we need a page to be splitted into more than one frag, but there is also case that we want to avoid the above constraints and their overhead when a page can't be splitted as it can only hold a big frag as requested by user, depending on different use cases: use case 1: allocate page without page splitting. use case 2: allocate page with page splitting. use case 3: allocate page with or without page splitting depending on the frag size. Currently page pool only provide page_pool_alloc_pages() and page_pool_alloc_frag() API to enable the above 1 & 2 separately, so we can not use a combination of 1 & 2 to enable 3, it is not possible yet because of the per page_pool flag PP_FLAG_PAGE_FRAG. So in order to allow allocating unsplitted page without the overhead of splitted page while still allow allocating splitted page, we need to remove the per page_pool flag in page_pool_is_last_frag(), as best as I can think of, it seems there are two methods as below: 1. Add per page flag/bit to indicate a page is splitted or not, which means we might need to update that flag/bit everytime the page is recycled, dirtying the cache line of 'struct page' for use case 1. 2. Unify the page->pp_frag_count handling for both splitted and unsplitted page by assuming all pages in the page pool is splitted into a big frag initially. Because we want to support the above use case 3 with minimal overhead, especially not adding any noticable overhead for use case 1, and we are already doing an optimization by not updating pp_frag_count in page_pool_defrag_page() for the last frag user, this patch chooses to unify the pp_frag_count handling to support the above use case 3. Let me know if it is making any sense here. > > Then you are going to have to modify the fragmented cases to make use > of lower level calls because now instead of us defragging a fragmented > page, and then freeing it the two operations essentially have to be > combined into one operation. Does 'defragging a fragmented page' mean doing decrementing pp_frag_count? "freeing it" mean calling put_page()? What does 'combined' really means here? > >>> >>> My advice as a first step would be to look at first solving how to >>> enable the PP_FLAG_PAGE_FRAG mode when you have >>> PAGE_POOL_DMA_USE_PP_FRAG_COUNT as true. That should be creating mono- >>> frags as we are calling them, and we should have a way to get the >>> truesize for those so we know when we are consuming significant amount >>> of memory. >> >> Does the way to get the truesize in the below RFC make sense to you? >> https://patchwork.kernel.org/project/netdevbpf/patch/20230516124801.2465-4-linyunsheng@huawei.com/ > > It doesn't add any value. All you are doing is passing the "size" > value as "truesize". The whole point of the "truesize" would be to > report the actual size. So a step in that direction would be to bump > truesize to include the remainder that isn't used when you decide it > is time to allocate a new page. The problem is that it requires some > fore-knowledge of what the next requested size is going to be. That is > why it is better to just have the drivers manage this since they know > what size they typically request and when they are going to close > pages. > > Like I said, if you are wanting to go down this path you are better > off starting with page pool and making all regular page pool pages > into mono-frags. Then from there we can start building things out. 'mono-frag' means page with pp_frag_count being one. If yes, then I feel like we have the same goal here, but may have different opinion on how to implement it. > > With that you could then let drivers like the Mellanox one handle its > own fragmenting knowing it has to return things to a mono-frag in > order for it to be working correctly. I still really don't how it will be better for mlx5 to handle its own fragmenting yet? +cc Dragos & Saeed to share some info here, so that we can see if page pool learn from it. > > . >
On Tue, Jun 6, 2023 at 5:41 AM Yunsheng Lin <linyunsheng@huawei.com> wrote: > > On 2023/6/5 22:58, Alexander Duyck wrote: > > ... > > >> > >> I am not sure I understand what do you mean by 'non-fragmented ', > >> 'mono-frags', 'page pool freeing paths' and 'non-fragmented case' > >> here. maybe describe it more detailed with something like the > >> pseudocode? > > > > What you are attempting to generate are "mono-frags" where a page pool > > page has a frag count of 1. I refer to "non-fragmented pages" as the > > legacy page pool page without pp_frags set. > > > > The "page-pool freeing paths" are the ones outside of the fragmented > > bits here. Basically __page_pool_put_page and the like. What you > > should be doing is pushing the reference counting code down deeper > > into the page pool logic. Currently it is more of a surface setup. > > > > The whole point I am getting at with this is that we should see the > > number of layers reduced for the fragmented pages, and by converting > > the non-fragmented pages to mono-frags we should see that maintain its > > current performance and total number of layers instead of having more > > layers added to it. > > Do you mean reducing the number of layers for the fragmented pages by > moving the page->pp_frag_count handling from page_pool_defrag_page() > to __page_pool_put_page() where page->_refcount is checked? I was thinking you could move pp_frag_count down into __page_pool_put_page(). Basically it is doing the static check against 1, and if it isn't 1 we will have to subtract 1 from it. > Or merge page->pp_frag_count into page->_refcount so that we don't > need page->pp_frag_count anymore? > > As my understanding, when a page from page pool is passed to the stack > to be processed, the stack may hold onto that page by taking > page->_refcount too, which means page pool has no control over who will > hold onto and when that taken will be released, that is why page pool > do the "page_ref_count(page) == 1" checking in __page_pool_put_page(), > if it is not true, the page pool can't recycle the page, so pp_frag_count > and _refcount have different meaning and serve different purpose, merging > them doesn't work, and moving them to one place doesn't make much sense > too? > > Or is there other obvious consideration that I missed? You have the right understanding. Basically we cannot recycle it until the fragcount is one or reaches 0 after subtraction and the refcount is 1. If we attempt to free it and fragcount hits 0 and refcount is != 1 we have to free it. > >>> > >> I am not sure what you meant above. > >> But I will describe what is this patch trying to do again: > >> When PP_FLAG_PAGE_FRAG is set and that flag is per page pool, not per > >> page, so page_pool_alloc_pages() is not allowed to be called as the > >> page->pp_frag_count is not setup correctly for the case. > >> > >> So in order to allow calling page_pool_alloc_pages(), as best as I > >> can think of, either we need a per page flag/bit to decide whether > >> to do something like dec_and_test for page->pp_frag_count in > >> page_pool_is_last_frag(), or we unify the page->pp_frag_count handling > >> in page_pool_is_last_frag() so that we don't need a per page flag/bit. > >> > >> This patch utilizes the optimization you mentioned above to unify the > >> page->pp_frag_count handling. > > > > Basically what should be happening if all page-pool pages are to be > > considered "fragmented" is that we should be folding this into the > > freeing logic. What we now have a 2 stage setup where we are dropping > > the count to 0, then rebounding it and setting it back to 1. If we are > > going to have all page pool pages fragmented then the freeing path for > > page pool pages should just be handling frag count directly instead of > > hacking on it here and ignoring it in the actual freeing paths. > > Do you mean doing something like below? isn't it dirtying the cache line > of 'struct page' whenever a page is recycled, which means we may not be > able to the maintain current performance for non-fragmented or mono-frag > case? > > --- a/net/core/page_pool.c > +++ b/net/core/page_pool.c > @@ -583,6 +583,10 @@ static __always_inline struct page * > __page_pool_put_page(struct page_pool *pool, struct page *page, > unsigned int dma_sync_size, bool allow_direct) > { > + > + if (!page_pool_defrag_page(page, 1)) > + return NULL; > + Yes, that is pretty much it. This would be your standard case page pool put path. Basically it allows us to start getting rid of a bunch of noise in the fragmented path. > /* This allocator is optimized for the XDP mode that uses > * one-frame-per-page, but have fallbacks that act like the > * regular page allocator APIs. > @@ -594,6 +598,7 @@ __page_pool_put_page(struct page_pool *pool, struct page *page, > */ > if (likely(page_ref_count(page) == 1 && !page_is_pfmemalloc(page))) { > /* Read barrier done in page_ref_count / READ_ONCE */ > + page_pool_fragment_page(page, 1); I wouldn't bother resetting this to 1 until after you have recycled it and pulled it back out again as an allocation. Basically when the pages are sitting in the pool the frag_count should be 0. That way it makes it easier to track and is similar to how the memory allocator actually deals with the page reference count. Basically if the page is sitting in the pool the frag_count is 0, once it comes out it should be 1 or more indicating it is in use. > > if (pool->p.flags & PP_FLAG_DMA_SYNC_DEV) > page_pool_dma_sync_for_device(pool, page, > > > > > > >>> > >>>>>> ret = atomic_long_sub_return(nr, &page->pp_frag_count); > >>>>>> WARN_ON(ret < 0); > >>>>>> + > >>>>>> + /* Reset frag count back to 1, this should be the rare case when > >>>>>> + * two users call page_pool_defrag_page() currently. > >>>>>> + */ > >>>>>> + if (!ret) > >>>>>> + atomic_long_set(&page->pp_frag_count, 1); > >>>>>> + > >>>>>> return ret; > >>>>>> } > >>>>>> > >> > >> ... > >> > >>>> As above, it is about unifying handling for frag and non-frag page in > >>>> page_pool_is_last_frag(). please let me know if there is any better way > >>>> to do it without adding statements here. > >>> > >>> I get what you are trying to get at but I feel like the implementation > >>> is going to cause more problems than it helps. The problem is it is > >>> going to hurt base page pool performance and it just makes the > >>> fragmented pages that much more confusing to deal with. > >> > >> For base page pool performance, as I mentioned before: > >> It remove PP_FLAG_PAGE_FRAG checking and only add the cost of > >> page_pool_fragment_page() in page_pool_set_pp_info(), which I > >> think it is negligible as we are already dirtying the same cache > >> line in page_pool_set_pp_info(). > > > > I have no problem with getting rid of the flag. > > > >> For the confusing, sometimes it is about personal taste, so I am > >> not going to argue with it:) But it would be good to provide a > >> non-confusing way to do that with minimal overhead. I feel like > >> you have provided it in the begin, but I am not able to understand > >> it yet. > > > > The problem here is that instead of treating all page pool pages as > > fragmented, what the patch set has done is added a shim layer so that > > you are layering fragmentation on top of page pool pages which was > > already the case. > > > > That is why I have suggested make page pool pages a "mono-frag" as > > your first patch. Basically it is going to force you to have to set > > the pp_frag value for these pages, and verify it is 1 when you are > > freeing it. > > It seems it is bascially what this patch do with minimal > overhead to the previous users. > > Let me try again with what this patch mainly do: > > Currently when page_pool_create() is called with > PP_FLAG_PAGE_FRAG flag, page_pool_alloc_pages() is only > allowed to be called under the below constraints: > 1. page_pool_fragment_page() need to be called to setup > page->pp_frag_count immediately. > 2. page_pool_defrag_page() often need to be called to > drain the page->pp_frag_count when there is no more > user will be holding on to that page. Right. Basically you will need to assign the value much larger in page_pool_fragment_page assuming you have a frag count of 1 after allocation. On free you should be able to do either an atomic sub and verify non-zero when you have to defrag at the end of fragmenting a page. > Those constraints exist in order to support a page to > be splitted into multi frags. Right. However it isn't much different then how we were dealing with the page reference count in drivers such as ixgbe. You can take a look at ixgbe_alloc_mapped_page() for an example of that. > And those constraints have some overhead because of the > cache line dirtying/bouncing and atomic update. We already have dirtied it on allocation and we were already dirtying it on freeing as well. > Those constraints are unavoidable for case when we need > a page to be splitted into more than one frag, but there > is also case that we want to avoid the above constraints > and their overhead when a page can't be splitted as it > can only hold a big frag as requested by user, depending > on different use cases: > use case 1: allocate page without page splitting. > use case 2: allocate page with page splitting. > use case 3: allocate page with or without page splitting > depending on the frag size. > > Currently page pool only provide page_pool_alloc_pages() > and page_pool_alloc_frag() API to enable the above 1 & 2 > separately, so we can not use a combination of 1 & 2 to > enable 3, it is not possible yet because of the per > page_pool flag PP_FLAG_PAGE_FRAG. I get that we need to get rid of the flag. The general idea here is one step at a time. If we want to get rid of the flag then we have to make the page pool set the frag_count in all cases where it can. In the cases where we run into the DMA issue the functions that frag/defrag pages have to succeed w/ only support for 1 frag. Most likely we will need to wrap the inline helpers for that. What it means is that the "DMA_USES" case will make the frag allocator synonymous with the non-fragmented allocator so both will be providing full pages. > So in order to allow allocating unsplitted page without > the overhead of splitted page while still allow allocating > splitted page, we need to remove the per page_pool flag > in page_pool_is_last_frag(), as best as I can think of, it > seems there are two methods as below: > 1. Add per page flag/bit to indicate a page is splitted or > not, which means we might need to update that flag/bit > everytime the page is recycled, dirtying the cache line > of 'struct page' for use case 1. > 2. Unify the page->pp_frag_count handling for both splitted > and unsplitted page by assuming all pages in the page > pool is splitted into a big frag initially. I am in support of 2. It is the simplest approach here. Basically if we cannot support it due to the DMA variable definition then we should have auto-magically succeeding versions of the defrag and fragment functions that only support allowing 1 fragment per page. > Because we want to support the above use case 3 with minimal > overhead, especially not adding any noticable overhead for > use case 1, and we are already doing an optimization by not > updating pp_frag_count in page_pool_defrag_page() for the > last frag user, this patch chooses to unify the pp_frag_count > handling to support the above use case 3. > > Let me know if it is making any sense here. Yes, that is pretty much it. > > > > Then you are going to have to modify the fragmented cases to make use > > of lower level calls because now instead of us defragging a fragmented > > page, and then freeing it the two operations essentially have to be > > combined into one operation. > > Does 'defragging a fragmented page' mean doing decrementing pp_frag_count? > "freeing it" mean calling put_page()? What does 'combined' really means > here? The change is that the code would do the subtraction and if it hit 0 it was freeing the page. That is the one piece that gets more complicated because we really should be hitting 1. So we may be adding a few more operations to that case. > > > >>> > >>> My advice as a first step would be to look at first solving how to > >>> enable the PP_FLAG_PAGE_FRAG mode when you have > >>> PAGE_POOL_DMA_USE_PP_FRAG_COUNT as true. That should be creating mono- > >>> frags as we are calling them, and we should have a way to get the > >>> truesize for those so we know when we are consuming significant amount > >>> of memory. > >> > >> Does the way to get the truesize in the below RFC make sense to you? > >> https://patchwork.kernel.org/project/netdevbpf/patch/20230516124801.2465-4-linyunsheng@huawei.com/ > > > > It doesn't add any value. All you are doing is passing the "size" > > value as "truesize". The whole point of the "truesize" would be to > > report the actual size. So a step in that direction would be to bump > > truesize to include the remainder that isn't used when you decide it > > is time to allocate a new page. The problem is that it requires some > > fore-knowledge of what the next requested size is going to be. That is > > why it is better to just have the drivers manage this since they know > > what size they typically request and when they are going to close > > pages. > > > > Like I said, if you are wanting to go down this path you are better > > off starting with page pool and making all regular page pool pages > > into mono-frags. Then from there we can start building things out. > > 'mono-frag' means page with pp_frag_count being one. If yes, then I > feel like we have the same goal here, but may have different opinion > on how to implement it. Yeah, I think it is mostly implementation differences. I thought back when we did this I had advocated for just frag counting all the pages right from the start. > > > > With that you could then let drivers like the Mellanox one handle its > > own fragmenting knowing it has to return things to a mono-frag in > > order for it to be working correctly. > > I still really don't how it will be better for mlx5 to handle its > own fragmenting yet? > > +cc Dragos & Saeed to share some info here, so that we can see > if page pool learn from it. It has more to do with the fact that the driver knows what it is going to do beforehand. In many cases it can look at the page and know that it isn't going to reuse it again so it can just report the truesize being the length from the current pointer to the end of the page. You can think of it as the performance advantage of a purpose built ASIC versus a general purpose CPU. The fact is we are able to cut out much of the unnecessary overhead if we know exactly how we are going to use the memory in the driver versus having to guess at it in the page pool API.
On 2023/6/6 23:33, Alexander Duyck wrote: >> Do you mean doing something like below? isn't it dirtying the cache line >> of 'struct page' whenever a page is recycled, which means we may not be >> able to the maintain current performance for non-fragmented or mono-frag >> case? >> >> --- a/net/core/page_pool.c >> +++ b/net/core/page_pool.c >> @@ -583,6 +583,10 @@ static __always_inline struct page * >> __page_pool_put_page(struct page_pool *pool, struct page *page, >> unsigned int dma_sync_size, bool allow_direct) >> { >> + >> + if (!page_pool_defrag_page(page, 1)) >> + return NULL; >> + > > Yes, that is pretty much it. This would be your standard case page > pool put path. Basically it allows us to start getting rid of a bunch > of noise in the fragmented path. > >> /* This allocator is optimized for the XDP mode that uses >> * one-frame-per-page, but have fallbacks that act like the >> * regular page allocator APIs. >> @@ -594,6 +598,7 @@ __page_pool_put_page(struct page_pool *pool, struct page *page, >> */ >> if (likely(page_ref_count(page) == 1 && !page_is_pfmemalloc(page))) { >> /* Read barrier done in page_ref_count / READ_ONCE */ >> + page_pool_fragment_page(page, 1); > > I wouldn't bother resetting this to 1 until after you have recycled it > and pulled it back out again as an allocation. Basically when the > pages are sitting in the pool the frag_count should be 0. That way it > makes it easier to track and is similar to how the memory allocator > actually deals with the page reference count. Basically if the page is > sitting in the pool the frag_count is 0, once it comes out it should > be 1 or more indicating it is in use. Let's be more specific about what we want to do here: For a specific page without splitting, the journey that it will go through is as below before this patch: 1. It is allocated from the page allocator. 2. It is initialized in page_pool_set_pp_info(). 3. It is passed to driver by page pool. 4. It is passed to stack by the driver. 5. It is passed back to the page pool by the stack, depending on the page_ref_count() checking: 5.1 page_ref_count() being one, the page is now owned by the page pool, and may be passed to the driver by going to step 3. 5.2 page_ref_count() not being one, the page is released by page pool doing resoure cleaning like dma mapping and put_page(). So a page may go through step 3 ~ 5.1 many times without dirtying the cache line of 'struct page' as my understanding. If I follow your suggestion here, It seems for a specific page without splitting, it may go through: 1. It is allocated from the page allocator. 2. It is initialized in page_pool_set_pp_info(). 3. It's pp_frag_count is set to one. 4. It is passed to driver by page pool. 5. It is passed to stack by the driver. 6. It is passed back to the page pool by the stack, depending on the pp_frag_count and page_ref_count() checking: 6.1 pp_frag_count and page_ref_count() being one, the page is now owned by the page pool, and may be passed to the driver by going to step 3. 6.2 otherwise the page is released by page pool doing resoure cleaning like dma mapping and put_page(). Aren't we dirtying the cache line of 'struct page' everytime the page is recycled? Or did I miss something obvious here? For my implementation, for a specific page without splitting, it may go through: 1. It is allocated from the page allocator. 2. It is initialized in page_pool_set_pp_info() and it's pp_frag_count set to one. 3. It is passed to driver by page pool. 4. It is passed to stack by the driver. 5. It is passed back to the page pool by the stack, depending on the page_ref_count() checking: 5.1 pp_frag_count and page_ref_count() being one, the page is now owned by the page pool, and as the optimization by not updating page->pp_frag_count in page_pool_defrag_page() for the last frag user, it can be passed to the driver by going to step 3 without resetting the pp_frag_count to 1, which may dirty the cache line of 'struct page'. 5.2 otherwise the page is released by page pool doing resoure cleaning like dma mapping and put_page(). Does it make any sense, or it doesn't really matter we are dirtying the cache line of 'struct page' whenever a page without splitted is recycled? >> >> Does 'defragging a fragmented page' mean doing decrementing pp_frag_count? >> "freeing it" mean calling put_page()? What does 'combined' really means >> here? > > The change is that the code would do the subtraction and if it hit 0 > it was freeing the page. That is the one piece that gets more > complicated because we really should be hitting 1. So we may be adding > a few more operations to that case. > >>> I am not sure I understand it. Does 'gets more complicated' means doing some optimization like page_pool_defrag_page() does? https://elixir.bootlin.com/linux/v6.4-rc5/source/include/net/page_pool.h#L314 > >>> >>> With that you could then let drivers like the Mellanox one handle its >>> own fragmenting knowing it has to return things to a mono-frag in >>> order for it to be working correctly. >> >> I still really don't how it will be better for mlx5 to handle its >> own fragmenting yet? >> >> +cc Dragos & Saeed to share some info here, so that we can see >> if page pool learn from it. > > It has more to do with the fact that the driver knows what it is going > to do beforehand. In many cases it can look at the page and know that > it isn't going to reuse it again so it can just report the truesize > being the length from the current pointer to the end of the page. > > You can think of it as the performance advantage of a purpose built > ASIC versus a general purpose CPU. The fact is we are able to cut out > much of the unnecessary overhead if we know exactly how we are going > to use the memory in the driver versus having to guess at it in the > page pool API. In general, I would agree with that. But for the specific case with mlx5, I am not sure about that, that's why I am curious about what is the exact reason about it doing the complicated frag_count handing in the driver instead of improving the page pool to support it's usecase, if it is about the last frag truesize problem here, we can do something like virtio_net do in the page pool too. > > . >
On Wed, Jun 7, 2023 at 5:46 AM Yunsheng Lin <linyunsheng@huawei.com> wrote: > > On 2023/6/6 23:33, Alexander Duyck wrote: > >> Do you mean doing something like below? isn't it dirtying the cache line > >> of 'struct page' whenever a page is recycled, which means we may not be > >> able to the maintain current performance for non-fragmented or mono-frag > >> case? > >> > >> --- a/net/core/page_pool.c > >> +++ b/net/core/page_pool.c > >> @@ -583,6 +583,10 @@ static __always_inline struct page * > >> __page_pool_put_page(struct page_pool *pool, struct page *page, > >> unsigned int dma_sync_size, bool allow_direct) > >> { > >> + > >> + if (!page_pool_defrag_page(page, 1)) > >> + return NULL; > >> + > > > > Yes, that is pretty much it. This would be your standard case page > > pool put path. Basically it allows us to start getting rid of a bunch > > of noise in the fragmented path. > > > >> /* This allocator is optimized for the XDP mode that uses > >> * one-frame-per-page, but have fallbacks that act like the > >> * regular page allocator APIs. > >> @@ -594,6 +598,7 @@ __page_pool_put_page(struct page_pool *pool, struct page *page, > >> */ > >> if (likely(page_ref_count(page) == 1 && !page_is_pfmemalloc(page))) { > >> /* Read barrier done in page_ref_count / READ_ONCE */ > >> + page_pool_fragment_page(page, 1); > > > > I wouldn't bother resetting this to 1 until after you have recycled it > > and pulled it back out again as an allocation. Basically when the > > pages are sitting in the pool the frag_count should be 0. That way it > > makes it easier to track and is similar to how the memory allocator > > actually deals with the page reference count. Basically if the page is > > sitting in the pool the frag_count is 0, once it comes out it should > > be 1 or more indicating it is in use. > > Let's be more specific about what we want to do here: > > For a specific page without splitting, the journey that it will go > through is as below before this patch: > 1. It is allocated from the page allocator. > 2. It is initialized in page_pool_set_pp_info(). > 3. It is passed to driver by page pool. > 4. It is passed to stack by the driver. > 5. It is passed back to the page pool by the stack, depending on the > page_ref_count() checking: > 5.1 page_ref_count() being one, the page is now owned by the page > pool, and may be passed to the driver by going to step 3. > 5.2 page_ref_count() not being one, the page is released by page > pool doing resoure cleaning like dma mapping and put_page(). > > So a page may go through step 3 ~ 5.1 many times without dirtying > the cache line of 'struct page' as my understanding. > > > If I follow your suggestion here, It seems for a specific page without > splitting, it may go through: > 1. It is allocated from the page allocator. > 2. It is initialized in page_pool_set_pp_info(). > 3. It's pp_frag_count is set to one. > 4. It is passed to driver by page pool. > 5. It is passed to stack by the driver. > 6. It is passed back to the page pool by the stack, depending on the > pp_frag_count and page_ref_count() checking: > 6.1 pp_frag_count and page_ref_count() being one, the page is now > owned by the page pool, and may be passed to the driver by > going to step 3. > 6.2 otherwise the page is released by page pool doing resoure > cleaning like dma mapping and put_page(). > > Aren't we dirtying the cache line of 'struct page' everytime the page > is recycled? Or did I miss something obvious here? What you are stating makes sense. So we would probably want to keep the pp_frag_count at 1 when it is sitting in the pool. > For my implementation, for a specific page without splitting, it may > go through: > 1. It is allocated from the page allocator. > 2. It is initialized in page_pool_set_pp_info() and it's pp_frag_count > set to one. > 3. It is passed to driver by page pool. > 4. It is passed to stack by the driver. > 5. It is passed back to the page pool by the stack, depending on the > page_ref_count() checking: > 5.1 pp_frag_count and page_ref_count() being one, the page is now > owned by the page pool, and as the optimization by not updating > page->pp_frag_count in page_pool_defrag_page() for the last > frag user, it can be passed to the driver by going to step 3 > without resetting the pp_frag_count to 1, which may dirty > the cache line of 'struct page'. > 5.2 otherwise the page is released by page pool doing resoure > cleaning like dma mapping and put_page(). > > Does it make any sense, or it doesn't really matter we are dirtying > the cache line of 'struct page' whenever a page without splitted is > recycled? No, that makes sense. No point in dirtying a cache line if we don't have to. > >> > >> Does 'defragging a fragmented page' mean doing decrementing pp_frag_count? > >> "freeing it" mean calling put_page()? What does 'combined' really means > >> here? > > > > The change is that the code would do the subtraction and if it hit 0 > > it was freeing the page. That is the one piece that gets more > > complicated because we really should be hitting 1. So we may be adding > > a few more operations to that case. > > > >>> > > I am not sure I understand it. Does 'gets more complicated' means doing > some optimization like page_pool_defrag_page() does? > https://elixir.bootlin.com/linux/v6.4-rc5/source/include/net/page_pool.h#L314 Our standard case is to decrement by 1. We will need to for the code that is doing your step 5.1 to handle a case where we are removing multiple frag references. That is what I was getting at with the "more complicated" comment. Basically if we push it to 0 then we either have to free or recycle the page by resetting the fragments. > > > >>> > >>> With that you could then let drivers like the Mellanox one handle its > >>> own fragmenting knowing it has to return things to a mono-frag in > >>> order for it to be working correctly. > >> > >> I still really don't how it will be better for mlx5 to handle its > >> own fragmenting yet? > >> > >> +cc Dragos & Saeed to share some info here, so that we can see > >> if page pool learn from it. > > > > It has more to do with the fact that the driver knows what it is going > > to do beforehand. In many cases it can look at the page and know that > > it isn't going to reuse it again so it can just report the truesize > > being the length from the current pointer to the end of the page. > > > > You can think of it as the performance advantage of a purpose built > > ASIC versus a general purpose CPU. The fact is we are able to cut out > > much of the unnecessary overhead if we know exactly how we are going > > to use the memory in the driver versus having to guess at it in the > > page pool API. > > In general, I would agree with that. > But for the specific case with mlx5, I am not sure about that, that's > why I am curious about what is the exact reason about it doing the > complicated frag_count handing in the driver instead of improving > the page pool to support it's usecase, if it is about the last frag > truesize problem here, we can do something like virtio_net do in the > page pool too. I suspect it has to do with their hardware doing the fragmentation of the page. As I recall some of the Mellanox parts support Rx packing so it is likely that their hardware is packing multiple frames into a single page and so they are marking it pre-fragmented, and then when the hardware completes the DMA they go through and record the offsets for the individual fragments and pass them up the stack.
diff --git a/include/net/page_pool.h b/include/net/page_pool.h index c8ec2f34722b..ea7a0c0592a5 100644 --- a/include/net/page_pool.h +++ b/include/net/page_pool.h @@ -50,6 +50,9 @@ PP_FLAG_DMA_SYNC_DEV |\ PP_FLAG_PAGE_FRAG) +#define PAGE_POOL_DMA_USE_PP_FRAG_COUNT \ + (sizeof(dma_addr_t) > sizeof(unsigned long)) + /* * Fast allocation side cache array/stack * @@ -295,13 +298,20 @@ void page_pool_put_defragged_page(struct page_pool *pool, struct page *page, */ static inline void page_pool_fragment_page(struct page *page, long nr) { - atomic_long_set(&page->pp_frag_count, nr); + if (!PAGE_POOL_DMA_USE_PP_FRAG_COUNT) + atomic_long_set(&page->pp_frag_count, nr); } +/* We need to reset frag_count back to 1 for the last user to allow + * only one user in case the page is recycled and allocated as non-frag + * page. + */ static inline long page_pool_defrag_page(struct page *page, long nr) { long ret; + BUILD_BUG_ON(__builtin_constant_p(nr) && nr != 1); + /* If nr == pp_frag_count then we have cleared all remaining * references to the page. No need to actually overwrite it, instead * we can leave this to be overwritten by the calling function. @@ -311,19 +321,36 @@ static inline long page_pool_defrag_page(struct page *page, long nr) * especially when dealing with a page that may be partitioned * into only 2 or 3 pieces. */ - if (atomic_long_read(&page->pp_frag_count) == nr) + if (atomic_long_read(&page->pp_frag_count) == nr) { + /* As we have ensured nr is always one for constant case + * using the BUILD_BUG_ON() as above, only need to handle + * the non-constant case here for frag count draining. + */ + if (!__builtin_constant_p(nr)) + atomic_long_set(&page->pp_frag_count, 1); + return 0; + } ret = atomic_long_sub_return(nr, &page->pp_frag_count); WARN_ON(ret < 0); + + /* Reset frag count back to 1, this should be the rare case when + * two users call page_pool_defrag_page() currently. + */ + if (!ret) + atomic_long_set(&page->pp_frag_count, 1); + return ret; } static inline bool page_pool_is_last_frag(struct page_pool *pool, struct page *page) { - /* If fragments aren't enabled or count is 0 we were the last user */ - return !(pool->p.flags & PP_FLAG_PAGE_FRAG) || + /* When dma_addr_upper is overlapped with pp_frag_count + * or we were the last page frag user. + */ + return PAGE_POOL_DMA_USE_PP_FRAG_COUNT || (page_pool_defrag_page(page, 1) == 0); } @@ -357,9 +384,6 @@ static inline void page_pool_recycle_direct(struct page_pool *pool, page_pool_put_full_page(pool, page, true); } -#define PAGE_POOL_DMA_USE_PP_FRAG_COUNT \ - (sizeof(dma_addr_t) > sizeof(unsigned long)) - static inline dma_addr_t page_pool_get_dma_addr(struct page *page) { dma_addr_t ret = page->dma_addr; diff --git a/net/core/page_pool.c b/net/core/page_pool.c index e212e9d7edcb..0868aa8f6323 100644 --- a/net/core/page_pool.c +++ b/net/core/page_pool.c @@ -334,6 +334,7 @@ static void page_pool_set_pp_info(struct page_pool *pool, { page->pp = pool; page->pp_magic |= PP_SIGNATURE; + page_pool_fragment_page(page, 1); if (pool->p.init_callback) pool->p.init_callback(page, pool->p.init_arg); }
Currently page_pool_dev_alloc_pages() can not be called when PP_FLAG_PAGE_FRAG is set, because it does not use the frag reference counting. As we are already doing a optimization by not updating page->pp_frag_count in page_pool_defrag_page() for the last frag user, and non-frag page only have one user, so we utilize that to unify frag page and non-frag page handling, so that page_pool_dev_alloc_pages() can also be called with PP_FLAG_PAGE_FRAG set. Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com> CC: Lorenzo Bianconi <lorenzo@kernel.org> CC: Alexander Duyck <alexander.duyck@gmail.com> --- include/net/page_pool.h | 38 +++++++++++++++++++++++++++++++------- net/core/page_pool.c | 1 + 2 files changed, 32 insertions(+), 7 deletions(-)