* [dm-devel] [PATCH v2 1/3] brd: extend the rcu regions to cover read and write @ 2023-07-21 13:49 ` Mikulas Patocka 0 siblings, 0 replies; 4+ messages in thread From: Mikulas Patocka @ 2023-07-21 13:49 UTC (permalink / raw) To: Jens Axboe Cc: linux-block, Chaitanya Kulkarni, Christoph Hellwig, Li Nan, dm-devel, Zdenek Kabelac This patch extends the rcu regions, so that lookup followed by a read or write of a page is done inside rcu read lock. This is needed for the following patch that enables discard. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> --- drivers/block/brd.c | 8 ++++++++ 1 file changed, 8 insertions(+) Index: linux-2.6/drivers/block/brd.c =================================================================== --- linux-2.6.orig/drivers/block/brd.c +++ linux-2.6/drivers/block/brd.c @@ -150,23 +150,27 @@ static void copy_to_brd(struct brd_devic size_t copy; copy = min_t(size_t, n, PAGE_SIZE - offset); + rcu_read_lock(); page = brd_lookup_page(brd, sector); BUG_ON(!page); dst = kmap_atomic(page); memcpy(dst + offset, src, copy); kunmap_atomic(dst); + rcu_read_unlock(); if (copy < n) { src += copy; sector += copy >> SECTOR_SHIFT; copy = n - copy; + rcu_read_lock(); page = brd_lookup_page(brd, sector); BUG_ON(!page); dst = kmap_atomic(page); memcpy(dst, src, copy); kunmap_atomic(dst); + rcu_read_unlock(); } } @@ -182,6 +186,7 @@ static void copy_from_brd(void *dst, str size_t copy; copy = min_t(size_t, n, PAGE_SIZE - offset); + rcu_read_lock(); page = brd_lookup_page(brd, sector); if (page) { src = kmap_atomic(page); @@ -189,11 +194,13 @@ static void copy_from_brd(void *dst, str kunmap_atomic(src); } else memset(dst, 0, copy); + rcu_read_unlock(); if (copy < n) { dst += copy; sector += copy >> SECTOR_SHIFT; copy = n - copy; + rcu_read_lock(); page = brd_lookup_page(brd, sector); if (page) { src = kmap_atomic(page); @@ -201,6 +208,7 @@ static void copy_from_brd(void *dst, str kunmap_atomic(src); } else memset(dst, 0, copy); + rcu_read_unlock(); } } -- dm-devel mailing list dm-devel@redhat.com https://listman.redhat.com/mailman/listinfo/dm-devel ^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH v2 1/3] brd: extend the rcu regions to cover read and write @ 2023-07-21 13:49 ` Mikulas Patocka 0 siblings, 0 replies; 4+ messages in thread From: Mikulas Patocka @ 2023-07-21 13:49 UTC (permalink / raw) To: Jens Axboe Cc: Li Nan, Zdenek Kabelac, Christoph Hellwig, Chaitanya Kulkarni, linux-block, dm-devel This patch extends the rcu regions, so that lookup followed by a read or write of a page is done inside rcu read lock. This is needed for the following patch that enables discard. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> --- drivers/block/brd.c | 8 ++++++++ 1 file changed, 8 insertions(+) Index: linux-2.6/drivers/block/brd.c =================================================================== --- linux-2.6.orig/drivers/block/brd.c +++ linux-2.6/drivers/block/brd.c @@ -150,23 +150,27 @@ static void copy_to_brd(struct brd_devic size_t copy; copy = min_t(size_t, n, PAGE_SIZE - offset); + rcu_read_lock(); page = brd_lookup_page(brd, sector); BUG_ON(!page); dst = kmap_atomic(page); memcpy(dst + offset, src, copy); kunmap_atomic(dst); + rcu_read_unlock(); if (copy < n) { src += copy; sector += copy >> SECTOR_SHIFT; copy = n - copy; + rcu_read_lock(); page = brd_lookup_page(brd, sector); BUG_ON(!page); dst = kmap_atomic(page); memcpy(dst, src, copy); kunmap_atomic(dst); + rcu_read_unlock(); } } @@ -182,6 +186,7 @@ static void copy_from_brd(void *dst, str size_t copy; copy = min_t(size_t, n, PAGE_SIZE - offset); + rcu_read_lock(); page = brd_lookup_page(brd, sector); if (page) { src = kmap_atomic(page); @@ -189,11 +194,13 @@ static void copy_from_brd(void *dst, str kunmap_atomic(src); } else memset(dst, 0, copy); + rcu_read_unlock(); if (copy < n) { dst += copy; sector += copy >> SECTOR_SHIFT; copy = n - copy; + rcu_read_lock(); page = brd_lookup_page(brd, sector); if (page) { src = kmap_atomic(page); @@ -201,6 +208,7 @@ static void copy_from_brd(void *dst, str kunmap_atomic(src); } else memset(dst, 0, copy); + rcu_read_unlock(); } } ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [dm-devel] [PATCH v2 1/3] brd: extend the rcu regions to cover read and write 2023-07-21 13:49 ` Mikulas Patocka @ 2023-07-27 8:31 ` Pankaj Raghav -1 siblings, 0 replies; 4+ messages in thread From: Pankaj Raghav @ 2023-07-27 8:31 UTC (permalink / raw) To: Mikulas Patocka Cc: Jens Axboe, Christoph Hellwig, Chaitanya Kulkarni, linux-block, Li Nan, dm-devel, Zdenek Kabelac, p.raghav > Index: linux-2.6/drivers/block/brd.c > =================================================================== > --- linux-2.6.orig/drivers/block/brd.c > +++ linux-2.6/drivers/block/brd.c > @@ -150,23 +150,27 @@ static void copy_to_brd(struct brd_devic > size_t copy; > > copy = min_t(size_t, n, PAGE_SIZE - offset); > + rcu_read_lock(); > page = brd_lookup_page(brd, sector); xa_load() inside brd_lookup_page() also calls rcu read lock. Instead of nesting rcu locks, could we modify the brd_lookup_page to use: diff --git a/drivers/block/brd.c b/drivers/block/brd.c index 970bd6ff38c4..acc37bfdd181 100644 --- a/drivers/block/brd.c +++ b/drivers/block/brd.c @@ -55,7 +55,9 @@ static struct page *brd_lookup_page(struct brd_device *brd, sector_t sector) struct page *page; idx = sector >> PAGE_SECTORS_SHIFT; /* sector to page index */ - page = xa_load(&brd->brd_pages, idx); + + XA_STATE(xas, &brd->brd_pages, idx); + page = xas_load(&xas); BUG_ON(page && page->index != idx); > BUG_ON(!page); > > dst = kmap_atomic(page); > memcpy(dst + offset, src, copy); > kunmap_atomic(dst); -- dm-devel mailing list dm-devel@redhat.com https://listman.redhat.com/mailman/listinfo/dm-devel ^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH v2 1/3] brd: extend the rcu regions to cover read and write @ 2023-07-27 8:31 ` Pankaj Raghav 0 siblings, 0 replies; 4+ messages in thread From: Pankaj Raghav @ 2023-07-27 8:31 UTC (permalink / raw) To: Mikulas Patocka Cc: Jens Axboe, Li Nan, Zdenek Kabelac, Christoph Hellwig, Chaitanya Kulkarni, linux-block, dm-devel, p.raghav > Index: linux-2.6/drivers/block/brd.c > =================================================================== > --- linux-2.6.orig/drivers/block/brd.c > +++ linux-2.6/drivers/block/brd.c > @@ -150,23 +150,27 @@ static void copy_to_brd(struct brd_devic > size_t copy; > > copy = min_t(size_t, n, PAGE_SIZE - offset); > + rcu_read_lock(); > page = brd_lookup_page(brd, sector); xa_load() inside brd_lookup_page() also calls rcu read lock. Instead of nesting rcu locks, could we modify the brd_lookup_page to use: diff --git a/drivers/block/brd.c b/drivers/block/brd.c index 970bd6ff38c4..acc37bfdd181 100644 --- a/drivers/block/brd.c +++ b/drivers/block/brd.c @@ -55,7 +55,9 @@ static struct page *brd_lookup_page(struct brd_device *brd, sector_t sector) struct page *page; idx = sector >> PAGE_SECTORS_SHIFT; /* sector to page index */ - page = xa_load(&brd->brd_pages, idx); + + XA_STATE(xas, &brd->brd_pages, idx); + page = xas_load(&xas); BUG_ON(page && page->index != idx); > BUG_ON(!page); > > dst = kmap_atomic(page); > memcpy(dst + offset, src, copy); > kunmap_atomic(dst); ^ permalink raw reply related [flat|nested] 4+ messages in thread
end of thread, other threads:[~2023-07-27 9:01 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <CGME20230727083159eucas1p1c4b907caac8371200bbd465143ac39fd@eucas1p1.samsung.com>
2023-07-21 13:49 ` [dm-devel] [PATCH v2 1/3] brd: extend the rcu regions to cover read and write Mikulas Patocka
2023-07-21 13:49 ` Mikulas Patocka
2023-07-27 8:31 ` [dm-devel] " Pankaj Raghav
2023-07-27 8:31 ` Pankaj Raghav
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.