The patch titled
Subject: mm-madvise-fix-madvise_pageout-for-private-file-mappings-v2
has been added to the -mm mm-unstable branch. Its filename is
mm-madvise-fix-madvise_pageout-for-private-file-mappings-v2.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/mm-madvise-fix-madvise_pageout-for-private-file-mappings-v2.patch
This patch will later appear in the mm-unstable branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next via the mm-everything
branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there every 2-3 working days
------------------------------------------------------
From: Pavankumar Kondeti <quic_pkondeti@xxxxxxxxxxx>
Subject: mm-madvise-fix-madvise_pageout-for-private-file-mappings-v2
Date: Fri, 2 Dec 2022 11:59:57 +0530
- As per David's suggestion, removed new argument introduced in
madvise_walk_private struct and directly call can_do_file_pageout()
- As per Mark's suggestions, optimized PageAnon() checks. Also bail
out early if the file mapping is not private.
Link: https://lkml.kernel.org/r/1669962597-27724-1-git-send-email-quic_pkondeti@xxxxxxxxxxx
Signed-off-by: Pavankumar Kondeti <quic_pkondeti@xxxxxxxxxxx>
Cc: Charan Teja Kalla <quic_charante@xxxxxxxxxxx>
Cc: Minchan Kim <minchan@xxxxxxxxxx>
Cc: Suren Baghdasaryan <surenb@xxxxxxxxxx>
Cc: David Hildenbrand <david@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/madvise.c | 57 ++++++++++++++++++++++++-------------------------
1 file changed, 29 insertions(+), 28 deletions(-)
--- a/mm/madvise.c~mm-madvise-fix-madvise_pageout-for-private-file-mappings-v2
+++ a/mm/madvise.c
@@ -40,7 +40,6 @@
struct madvise_walk_private {
struct mmu_gather *tlb;
bool pageout;
- bool can_pageout_file;
};
/*
@@ -319,6 +318,21 @@ static long madvise_willneed(struct vm_a
return 0;
}
+static inline bool can_do_file_pageout(struct vm_area_struct *vma)
+{
+ if (!vma->vm_file)
+ return false;
+ /*
+ * paging out pagecache only for non-anonymous mappings that correspond
+ * to the files the calling process could (if tried) open for writing;
+ * otherwise we'd be including shared non-exclusive mappings, which
+ * opens a side channel.
+ */
+ return inode_owner_or_capable(&init_user_ns,
+ file_inode(vma->vm_file)) ||
+ file_permission(vma->vm_file, MAY_WRITE) == 0;
+}
+
static int madvise_cold_or_pageout_pte_range(pmd_t *pmd,
unsigned long addr, unsigned long end,
struct mm_walk *walk)
@@ -326,17 +340,20 @@ static int madvise_cold_or_pageout_pte_r
struct madvise_walk_private *private = walk->private;
struct mmu_gather *tlb = private->tlb;
bool pageout = private->pageout;
- bool pageout_anon_only = pageout && !private->can_pageout_file;
struct mm_struct *mm = tlb->mm;
struct vm_area_struct *vma = walk->vma;
pte_t *orig_pte, *pte, ptent;
spinlock_t *ptl;
struct page *page = NULL;
LIST_HEAD(page_list);
+ bool pageout_anon_only_filter;
if (fatal_signal_pending(current))
return -EINTR;
+ pageout_anon_only_filter = pageout && !vma_is_anonymous(vma) &&
+ !can_do_file_pageout(vma);
+
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
if (pmd_trans_huge(*pmd)) {
pmd_t orig_pmd;
@@ -363,7 +380,7 @@ static int madvise_cold_or_pageout_pte_r
if (page_mapcount(page) != 1)
goto huge_unlock;
- if (pageout_anon_only && !PageAnon(page))
+ if (pageout_anon_only_filter && !PageAnon(page))
goto huge_unlock;
if (next - addr != HPAGE_PMD_SIZE) {
@@ -434,7 +451,7 @@ regular_page:
if (PageTransCompound(page)) {
if (page_mapcount(page) != 1)
break;
- if (pageout_anon_only && !PageAnon(page))
+ if (pageout_anon_only_filter && !PageAnon(page))
break;
get_page(page);
if (!trylock_page(page)) {
@@ -463,7 +480,7 @@ regular_page:
if (!PageLRU(page) || page_mapcount(page) != 1)
continue;
- if (pageout_anon_only && !PageAnon(page))
+ if (pageout_anon_only_filter && !PageAnon(page))
continue;
VM_BUG_ON_PAGE(PageTransCompound(page), page);
@@ -548,13 +565,11 @@ static long madvise_cold(struct vm_area_
static void madvise_pageout_page_range(struct mmu_gather *tlb,
struct vm_area_struct *vma,
- unsigned long addr, unsigned long end,
- bool can_pageout_file)
+ unsigned long addr, unsigned long end)
{
struct madvise_walk_private walk_private = {
.pageout = true,
.tlb = tlb,
- .can_pageout_file = can_pageout_file,
};
tlb_start_vma(tlb, vma);
@@ -562,28 +577,12 @@ static void madvise_pageout_page_range(s
tlb_end_vma(tlb, vma);
}
-static inline bool can_do_file_pageout(struct vm_area_struct *vma)
-{
- if (!vma->vm_file)
- return false;
- /*
- * paging out pagecache only for non-anonymous mappings that correspond
- * to the files the calling process could (if tried) open for writing;
- * otherwise we'd be including shared non-exclusive mappings, which
- * opens a side channel.
- */
- return inode_owner_or_capable(&init_user_ns,
- file_inode(vma->vm_file)) ||
- file_permission(vma->vm_file, MAY_WRITE) == 0;
-}
-
static long madvise_pageout(struct vm_area_struct *vma,
struct vm_area_struct **prev,
unsigned long start_addr, unsigned long end_addr)
{
struct mm_struct *mm = vma->vm_mm;
struct mmu_gather tlb;
- bool can_pageout_file;
*prev = vma;
if (!can_madv_lru_vma(vma))
@@ -592,14 +591,16 @@ static long madvise_pageout(struct vm_ar
/*
* If the VMA belongs to a private file mapping, there can be private
* dirty pages which can be paged out if even this process is neither
- * owner nor write capable of the file. Cache the file access check
- * here and use it later during page walk.
+ * owner nor write capable of the file. We allow private file mappings
+ * further to pageout dirty anon pages.
*/
- can_pageout_file = can_do_file_pageout(vma);
+ if (!vma_is_anonymous(vma) && (!can_do_file_pageout(vma) &&
+ (vma->vm_flags & VM_MAYSHARE)))
+ return 0;
lru_add_drain();
tlb_gather_mmu(&tlb, mm);
- madvise_pageout_page_range(&tlb, vma, start_addr, end_addr, can_pageout_file);
+ madvise_pageout_page_range(&tlb, vma, start_addr, end_addr);
tlb_finish_mmu(&tlb);
return 0;
_
Patches currently in -mm which might be from quic_pkondeti@xxxxxxxxxxx are
mm-madvise-fix-madvise_pageout-for-private-file-mappings.patch
mm-madvise-fix-madvise_pageout-for-private-file-mappings-v2.patch
