| Message ID | 20211013160034.3472923-6-kent.overstreet@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Minor mm/struct page work | expand |
On Wed, Oct 13, 2021 at 12:00:34PM -0400, Kent Overstreet wrote: > As part of the struct page cleanups underway, we want to remove as much > usage of page->mapping and page->index as possible, as frequently they > are known from context. > > In the brd code, we're never actually reading from page->index except in > assertions, so references to it can be safely deleted. > > Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com> More than that ... this is essentially asserting that the radix tree code works, and we have a test suite to ensure that. Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
On 13.10.21 18:00, Kent Overstreet wrote: > As part of the struct page cleanups underway, we want to remove as much > usage of page->mapping and page->index as possible, as frequently they > are known from context. > > In the brd code, we're never actually reading from page->index except in > assertions, so references to it can be safely deleted. > > Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com> > --- > drivers/block/brd.c | 4 ---- > 1 file changed, 4 deletions(-) > > diff --git a/drivers/block/brd.c b/drivers/block/brd.c > index 58ec167aa0..0a55aed832 100644 > --- a/drivers/block/brd.c > +++ b/drivers/block/brd.c > @@ -72,8 +72,6 @@ static struct page *brd_lookup_page(struct brd_device *brd, sector_t sector) > page = radix_tree_lookup(&brd->brd_pages, idx); > rcu_read_unlock(); > > - BUG_ON(page && page->index != idx); > - > return page; > } > > @@ -108,12 +106,10 @@ static struct page *brd_insert_page(struct brd_device *brd, sector_t sector) > > spin_lock(&brd->brd_lock); > idx = sector >> PAGE_SECTORS_SHIFT; > - page->index = idx; > if (radix_tree_insert(&brd->brd_pages, idx, page)) { > __free_page(page); > page = radix_tree_lookup(&brd->brd_pages, idx); > BUG_ON(!page); > - BUG_ON(page->index != idx); > } else { > brd->brd_nr_pages++; > } > Reviewed-by: David Hildenbrand <david@redhat.com>
diff --git a/drivers/block/brd.c b/drivers/block/brd.c index 58ec167aa0..0a55aed832 100644 --- a/drivers/block/brd.c +++ b/drivers/block/brd.c @@ -72,8 +72,6 @@ static struct page *brd_lookup_page(struct brd_device *brd, sector_t sector) page = radix_tree_lookup(&brd->brd_pages, idx); rcu_read_unlock(); - BUG_ON(page && page->index != idx); - return page; } @@ -108,12 +106,10 @@ static struct page *brd_insert_page(struct brd_device *brd, sector_t sector) spin_lock(&brd->brd_lock); idx = sector >> PAGE_SECTORS_SHIFT; - page->index = idx; if (radix_tree_insert(&brd->brd_pages, idx, page)) { __free_page(page); page = radix_tree_lookup(&brd->brd_pages, idx); BUG_ON(!page); - BUG_ON(page->index != idx); } else { brd->brd_nr_pages++; }
As part of the struct page cleanups underway, we want to remove as much usage of page->mapping and page->index as possible, as frequently they are known from context. In the brd code, we're never actually reading from page->index except in assertions, so references to it can be safely deleted. Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com> --- drivers/block/brd.c | 4 ---- 1 file changed, 4 deletions(-)