minor cleanups of include/linux/mm.h
Subsystems (2)
Mike Rapoport
reviewercover minor cleanups of include/linux/mm.h
From: Mike Rapoport <[email protected]>
Hi,
While looking at recent page_mapping_file() changes I've noticed that page_rmapping() and page_anon_vma() could be moved from include/linux/mm.h and we don't need two forward declarations of page_mapping().
I've also noticed that except page_mapping_file() we have somewhat similar page_file_mapping(), which seems superfluous, but I didn't dig further to see how one of them can be dropped.
The patches are vs v5.12-rc5-mmots-2021-03-31-22-26.
Mike Rapoport (2):
mm: move page_rmapping() and page_anon_vma() to mm/internal.h include/linux/mm.h: remove duplicated declaration of page_mapping()
include/linux/mm.h | 4 ---- mm/internal.h | 3 +++ 2 files changed, 3 insertions(+), 4 deletions(-)
Signature · 1 line
2.28.0
Mike Rapoport
reviewerPATCH 1/2 mm: move page_rmapping() and page_anon_vma() to mm/internal.h
From: Mike Rapoport <[email protected]>
The functions page_rmapping() and page_anon_vma() are not used outside core mm. Move their declaration from include/linux/mm.h to mm/internal.h
--- include/linux/mm.h | 2 -- mm/internal.h | 3 +++ 2 files changed, 3 insertions(+), 2 deletions(-)
include/linux/mm.h +0 -2
@@ -1619,8 +1619,6 @@ void page_address_init(void);#define page_address_init () do { } while(0)#endif -extern void *page_rmapping (struct page *page); -extern struct anon_vma *page_anon_vma (struct page *page);extern struct address_space *page_mapping (struct page *page);extern struct address_space *__page_file_mapping (struct page *);
mm/internal.h +3 -0
@@ -358,6 +358,9 @@ void __vma_link_list(struct mm_struct *mm, struct vm_area_struct *vma, struct vm_area_struct *prev);void __vma_unlink_list (struct mm_struct *mm,struct vm_area_struct *vma); +void *page_rmapping (struct page *page); +struct anon_vma *page_anon_vma (struct page *page); +#ifdef CONFIG_MMU extern long populate_vma_page_range (struct vm_area_struct *vma,unsigned long start,unsigned long end,int *nonblocking);
Signature · 1 line
2.28.0
Re: [PATCH 1/2] mm: move page_rmapping() and page_anon_vma() to mm/internal.h
Quoted 6 lines · Mike Rapoport
From: Mike Rapoport <[email protected]>
The functions
page_rmapping()andpage_anon_vma()are not used outside core mm. Move their declaration frominclude/linux/mm.htomm/internal.hSigned-off-by Mike Rapoport <[email protected]>
This has purely textual conflicts with the folio series, but no conceptual conflicts. I generally approve of moving things out of mm.h, so
kernel test robot
build botRe: [PATCH 1/2] mm: move page_rmapping() and page_anon_vma() to mm/internal.h
In reply to Mike RapoportI love your patch! Yet something to improve:
Hi Mike,
I love your patch! Yet something to improve:
[auto build test ERROR on hnaz-linux-mm/master]
[also build test ERROR on v5.12-rc6 next-20210401]
[If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Mike-Rapoport/minor-cleanups-of-include-linux-mm-h/20210405-231514 base: https://github.com/hnaz/linux-mm master config: x86_64-randconfig-a004-20210405 (attached as .config) compiler: clang version 13.0.0 (https://github.com/llvm/llvm-project 2760a808b9916a2839513b7fd7314a464f52481e)
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# install x86_64 cross compiling tool for clang build
# apt-get install binutils-x86-64-linux-gnu
# https://github.com/0day-ci/linux/commit/c025240d41e9b6388c57fd4cde4a520074e41539
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Mike-Rapoport/minor-cleanups-of-include-linux-mm-h/20210405-231514 git checkout c025240d41e9b6388c57fd4cde4a520074e41539
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=x86_64
If you fix the issue, kindly add following tag as appropriate
All errors (new ones prefixed by >>):
>> mm/page_idle.c:105:7: error: implicit declaration of function 'page_rmapping' [-Werror,-Wimplicit-function-declaration]
!page_rmapping(page))
^
mm/page_idle.c:105:7: note: did you mean 'page_mapping'?
include/linux/mm.h:1549:23: note: 'page_mapping' declared here
struct address_space *page_mapping(struct page *page);
^1 error generated.
vim +/page_rmapping +105 mm/page_idle.c
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 91
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 92 static void page_idle_clear_pte_refs(struct page *page)
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 93 {
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 94 /*
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 95 * Since rwc.arg is unused, rwc is effectively immutable, so we
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 96 * can make it static const to save some cycles and stack.
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 97 */
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 98 static const struct rmap_walk_control rwc = {
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 99 .rmap_one = page_idle_clear_pte_refs_one,
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 100 .anon_lock = page_lock_anon_vma_read,
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 101 };
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 102 bool need_lock;
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 103
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 104 if (!page_mapped(page) ||
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 @105 !page_rmapping(page))
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 106 return;
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 107
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 108 need_lock = !PageAnon(page) || PageKsm(page);
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 109 if (need_lock && !trylock_page(page))
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 110 return;
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 111
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 112 rmap_walk(page, (struct rmap_walk_control *)&rwc);
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 113
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 114 if (need_lock)
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 115 unlock_page(page);
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 116 }
33c3fc71c8cfa3 Vladimir Davydov 2015-09-09 117 ---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org
Mike Rapoport
reviewerPATCH 2/2 include/linux/mm.h: remove duplicated declaration of page_mapping()
In reply to Mike RapoportFrom: Mike Rapoport <[email protected]>
There are two declarations of page_mapping() in include/linux/mm.h a dozen lines apart.
Remove the older one with unnecessary "extern".
--- include/linux/mm.h | 2 -- 1 file changed, 2 deletions(-)
include/linux/mm.h +0 -2
@@ -1619,8 +1619,6 @@ void page_address_init(void);#define page_address_init () do { } while(0)#endif -extern struct address_space *page_mapping (struct page *page); -extern struct address_space *__page_file_mapping (struct page *); static inline
Signature · 1 line
2.28.0
Re: [PATCH 2/2] include/linux/mm.h: remove duplicated declaration of page_mapping()
Quoted 6 lines · Mike Rapoport
From: Mike Rapoport <[email protected]>
There are two declarations of
page_mapping()ininclude/linux/mm.ha dozen lines apart.Remove the older one with unnecessary "extern".
This conflicts with
https://lore.kernel.org/linux-mm/[email protected]/
Andrew, please can we get the folio patches in?
Quoted 3 lines · Mike Rapoport
I've also noticed that except
page_mapping_file()we have somewhat similarpage_file_mapping(), which seems superfluous, but I didn't dig further to see how one of them can be dropped.
They can't ;-( I looked at it, and they do different things.
+ * folio_file_mapping - Find the mapping this folio belongs to.
+ * @folio: The folio.
+ *
+ * For folios which are in the page cache, return the mapping that this
+ * page belongs to. Folios in the swap cache return the mapping of the
+ * swap file or swap device where the data is stored. This is different
+ * from the mapping returned by folio_mapping(). The only reason to
+ * use it is if, like NFS, you return 0 from ->activate_swapfile.
page_mapping_file() returns NULL for pages which are in the swap cache, as they no longer need the dcache flushed.