mm/madvise: break reclaim when lock race
Subsystems (2)
PATCH 1/1 mm/madvise: break reclaim when lock race
From: xiaofeng <[email protected]>
When the process_madvise() system call is being used, it takes mmap_lock and blocks the application from allocating memory, leading to unreasonable delays. This patchset aims to fix it.
--- mm/madvise.c | 3 +++ 1 file changed, 3 insertions(+)
mm/madvise.c +3 -0
@@ -322,6 +322,9 @@ static int madvise_cold_or_pageout_pte_range(pmd_t *pmd,if (fatal_signal_pending (current))return -EINTR ; +if (mmap_lock_is_contended (mm)) +return -EINTR ; +#ifdef CONFIG_TRANSPARENT_HUGEPAGE if (pmd_trans_huge (*pmd)) {pmd_t orig_pmd;
Signature · 1 line
2.17.1
Quoted 4 lines · On Mon, Dec 6, 2021 at 7:22 PM <[email protected]> wrote:
From: xiaofeng <[email protected]>
When the
process_madvise()system call is being used, it takesmmap_lockand blocks the application from allocating memory, leading to unreasonable delays. This patchset aims to fix it.
Please use checkpatch.pl script before posting (see this blog for
example: http://nickdesaulniers.github.io/blog/2017/05/16/submitting-your-first-patch-to-the-linux-kernel-and-responding-to-feedback/).
Could you please describe your usecase a bit more? Which MADV is being used specifically (MADV_COLD or MADV_PAGEOUT)?
CC'ing Minchan as he might be interested in learning more about this.
Quoted 23 lines
Signed-off-by xiaofeng <[email protected]>--- mm/madvise.c | 3 +++ 1 file changed, 3 insertions(+)mm/madvise.c +3
-1@@ -322,6 +322,9 @@ static int madvise_cold_or_pageout_pte_range(pmd_t *pmd,if (fatal_signal_pending (current))return -EINTR ; +if (mmap_lock_is_contended (mm)) +return -EINTR ; +#ifdef CONFIG_TRANSPARENT_HUGEPAGE if (pmd_trans_huge (*pmd)) {pmd_t orig_pmd; --2.17.1
Quoted 3 lines · On Tue, Dec 07, 2021 at 11:22:02AM +0800, [email protected] wrote:
From: xiaofeng <[email protected]>
When the
process_madvise()system call is being used, it takesmmap_lockand blocks the application from allocating memory, leading to unreasonable delays. This patchset aims to fix it.
This means that any page fault will cause madvise() to fail. NACK.