zram: use copy_page for full page copy
Subsystems (2)
PATCH 1/1 zram: use copy_page for full page copy
commit 42e99bd975fd ("zram: optimize memory operations with clear_page()/copy_page()") optimize page copy/clean operations, but then commit d72e9a7a93e4 ("zram: do not use copy_page with non-page aligned address") removes the optimization because there's memory corruption at that time, the reason was well explained. But after commit 1f7319c74275 ("zram: partial IO refactoring"), partial IO uses alloc_page() instead of kmalloc to allocate a page, so we can bring back the optimization.
commit 80ba4caf8ba9 ("zram: use copy_page for full page copy") brings back partial optimization, missed one point in zram_write_page(). optimize the full page copying in zram_write_page() with copy_page()
--- drivers/block/zram/zram_drv.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/block/zram/zram_drv.c +5 -3
@@ -1478,11 +1478,13 @@ static int zram_write_page(struct zram *zram, struct page *page, u32 index) dst =zs_map_object (zram->mem_pool, handle,ZS_MM_WO ); src = zstrm->buffer; -if (comp_len ==PAGE_SIZE ) +if (comp_len ==PAGE_SIZE ) { src =kmap_local_page (page); -memcpy (dst, src, comp_len); -if (comp_len ==PAGE_SIZE ) +copy_page (dst, src);kunmap_local (src); + }else { +memcpy (dst, src, comp_len); + }zcomp_stream_put (zram->comps[ZRAM_PRIMARY_COMP ]);zs_unmap_object (zram->mem_pool, handle);
Signature · 1 line
2.43.0
Sergey Senozhatsky
maintainerQuoted 9 lines · Jisheng Zhang
commit
42e99bd975fd("zram: optimize memory operations with clear_page()/copy_page()") optimize page copy/clean operations, but then commitd72e9a7a93e4("zram: do not usecopy_pagewith non-page aligned address") removes the optimization because there's memory corruption at that time, the reason was well explained. But after commit1f7319c74275("zram: partial IO refactoring"), partial IO usesalloc_page()instead of kmalloc to allocate a page, so we can bring back the optimization.
Quoted 3 lines
commit
80ba4caf8ba9("zram: usecopy_pagefor full page copy") brings back partial optimization, missed one point inzram_write_page(). optimize the full page copying inzram_write_page()withcopy_page()
Is copy_page() really more optimal than memcpy(PAGE_SIZE)?
Quoted 10 lines · Sergey Senozhatsky
Quoted 9 lines · Jisheng Zhang
commit
42e99bd975fd("zram: optimize memory operations with clear_page()/copy_page()") optimize page copy/clean operations, but then commitd72e9a7a93e4("zram: do not usecopy_pagewith non-page aligned address") removes the optimization because there's memory corruption at that time, the reason was well explained. But after commit1f7319c74275("zram: partial IO refactoring"), partial IO usesalloc_page()instead of kmalloc to allocate a page, so we can bring back the optimization.
Quoted 5 lines
Quoted 3 lines
commit
80ba4caf8ba9("zram: usecopy_pagefor full page copy") brings back partial optimization, missed one point inzram_write_page(). optimize the full page copying inzram_write_page()withcopy_page()Is
copy_page()really more optimal thanmemcpy(PAGE_SIZE)?
I think yes copy_page performs better than memcpy(PAGE_SIZE) commit afb2d666d025 ("zsmalloc: use copy_page for full page copy") also shows the result.
Quoted 26 lines · Jisheng Zhang
commit
42e99bd975fd("zram: optimize memory operations with clear_page()/copy_page()") optimize page copy/clean operations, but then commitd72e9a7a93e4("zram: do not usecopy_pagewith non-page aligned address") removes the optimization because there's memory corruption at that time, the reason was well explained. But after commit1f7319c74275("zram: partial IO refactoring"), partial IO usesalloc_page()instead of kmalloc to allocate a page, so we can bring back the optimization.commit
80ba4caf8ba9("zram: usecopy_pagefor full page copy") brings back partial optimization, missed one point inzram_write_page(). optimize the full page copying inzram_write_page()withcopy_page()Signed-off-by Jisheng Zhang <[email protected]>--- drivers/block/zram/zram_drv.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)drivers/block/zram/zram_drv.c +0
-0@@ -1478,11 +1478,13 @@ static int zram_write_page(struct zram *zram, struct page *page, u32 index) dst =zs_map_object (zram->mem_pool, handle,ZS_MM_WO ); src = zstrm->buffer;
drivers/block/zram/zram_drv.c static int zram_write_page(struct zram *zram, struct page *page, u32 index) · line 1487
Quoted 10 lines
drivers/block/zram/zram_drv.c +5
-3-if (comp_len ==PAGE_SIZE ) +if (comp_len ==PAGE_SIZE ) { src =kmap_local_page (page); -memcpy (dst, src, comp_len); -if (comp_len ==PAGE_SIZE ) +copy_page (dst, src);kunmap_local (src); + }else { +memcpy (dst, src, comp_len); + }
I know this is pre-existing code, but why do we need to kmap for comp_len == PAGE_SIZE and not for the other cases? Something feels really obsfucated here.
Sergey Senozhatsky
maintainerQuoted 26 lines · Christoph Hellwig
Quoted 25 lines · Jisheng Zhang
commit
42e99bd975fd("zram: optimize memory operations with clear_page()/copy_page()") optimize page copy/clean operations, but then commitd72e9a7a93e4("zram: do not usecopy_pagewith non-page aligned address") removes the optimization because there's memory corruption at that time, the reason was well explained. But after commit1f7319c74275("zram: partial IO refactoring"), partial IO usesalloc_page()instead of kmalloc to allocate a page, so we can bring back the optimization.commit
80ba4caf8ba9("zram: usecopy_pagefor full page copy") brings back partial optimization, missed one point inzram_write_page(). optimize the full page copying inzram_write_page()withcopy_page()Signed-off-by Jisheng Zhang <[email protected]>--- drivers/block/zram/zram_drv.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)drivers/block/zram/zram_drv.c +0
-0@@ -1478,11 +1478,13 @@ static int zram_write_page(struct zram *zram, struct page *page, u32 index) dst =zs_map_object (zram->mem_pool, handle,ZS_MM_WO );
Quoted 15 lines
drivers/block/zram/zram_drv.c static int zram_write_page(struct zram *zram, struct page *page, u32 index) · line 1487
Quoted 11 lines
drivers/block/zram/zram_drv.c +5
-3src = zstrm->buffer; -if (comp_len ==PAGE_SIZE ) +if (comp_len ==PAGE_SIZE ) { src =kmap_local_page (page); -memcpy (dst, src, comp_len); -if (comp_len ==PAGE_SIZE ) +copy_page (dst, src);kunmap_local (src); + }else { +memcpy (dst, src, comp_len); + }I know this is pre-existing code, but why do we need to kmap for
comp_len==PAGE_SIZEand not for the other cases? Something feels really obsfucated here.
It is tricky a little.
If we managed to compress page (size < zsmalloc uncompressible watermark) then src is per-CPU buffer with compressed data. Otherwise src is original page (with uncompressed data).