| Message ID | 20200314215912.1554-1-richard.weiyang@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | mm/swap_state.c: use the same way to count page in [add_to|delete_from]_swap_cache | expand |
On Sat, Mar 14, 2020 at 09:59:12PM +0000, Wei Yang wrote: > Function add_to_swap_cache() and delete_from_swap_cache() are counter > parts, while currently they use different way to count page. > > It doesn't break any thing because we only have two size for PageAnon, > but this is confusing and not a good practice. > > This patch corrects it by both using compound_nr(). You're converting in the wrong direction. hpage_nr_pages() is optimised away when CONFIG_TRANSPARENT_HUGEPAGE is undefined, whereas compound_nr() is not. I also have this patch pending: http://git.infradead.org/users/willy/linux-dax.git/commitdiff/192b635b428ae74f680574cdcc3d5e9d213fcb64
On Sat, Mar 14, 2020 at 05:21:01PM -0700, Matthew Wilcox wrote: >On Sat, Mar 14, 2020 at 09:59:12PM +0000, Wei Yang wrote: >> Function add_to_swap_cache() and delete_from_swap_cache() are counter >> parts, while currently they use different way to count page. >> >> It doesn't break any thing because we only have two size for PageAnon, >> but this is confusing and not a good practice. >> >> This patch corrects it by both using compound_nr(). > >You're converting in the wrong direction. hpage_nr_pages() is optimised >away when CONFIG_TRANSPARENT_HUGEPAGE is undefined, whereas compound_nr() >is not. Ok, I didn't think about the optimisation. Will use hpage_nr_pages() in v2. > >I also have this patch pending: >http://git.infradead.org/users/willy/linux-dax.git/commitdiff/192b635b428ae74f680574cdcc3d5e9d213fcb64
diff --git a/mm/swap_state.c b/mm/swap_state.c index 8e7ce9a9bc5e..51d8884a693a 100644 --- a/mm/swap_state.c +++ b/mm/swap_state.c @@ -158,7 +158,7 @@ int add_to_swap_cache(struct page *page, swp_entry_t entry, gfp_t gfp) void __delete_from_swap_cache(struct page *page, swp_entry_t entry) { struct address_space *address_space = swap_address_space(entry); - int i, nr = hpage_nr_pages(page); + int i, nr = compound_nr(page); pgoff_t idx = swp_offset(entry); XA_STATE(xas, &address_space->i_pages, idx); @@ -251,7 +251,7 @@ void delete_from_swap_cache(struct page *page) xa_unlock_irq(&address_space->i_pages); put_swap_page(page, entry); - page_ref_sub(page, hpage_nr_pages(page)); + page_ref_sub(page, compound_nr(page)); } /*
Function add_to_swap_cache() and delete_from_swap_cache() are counter parts, while currently they use different way to count page. It doesn't break any thing because we only have two size for PageAnon, but this is confusing and not a good practice. This patch corrects it by both using compound_nr(). Signed-off-by: Wei Yang <richard.weiyang@gmail.com> --- mm/swap_state.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)