[PATCH v2] f2fs: remove mappedtodisk flag usage in mmap write path

[PATCH v2] f2fs: remove mappedtodisk flag usage in mmap write path

[Nanzhe Zhao]

PG_mappedtodisk was inherited from the nobh/buffer-head era.  However,
f2fs has no buffer heads and the flag is no longer meaningful.  The only
user of the flag is the "no holes" shortcut in f2fs_vm_page_mkwrite(),
which makes mmap write faults skip marking the folio dirty.  Chao Yu
pointed out that this is not only a cleanup: it causes the following
race condition with filesystem freezing:

 CPU0 (write fault)     CPU1 (freeze_super)
 ----                   ----
 f2fs_vm_page_mkwrite()
 sb_start_pagefault()
 /* shortcut: no dirty */
 sb_end_pagefault()

                        sb_wait_write(SB_FREEZE_PAGEFAULT)
                        /* no waiter, passes */
                        sync_filesystem()
                        /* folio not dirty, skipped */

 finish_fault()
 /* PTE writable */
 fault_dirty_shared_page()
 folio_mark_dirty()
 /* dirty folio on a frozen fs */

It will also skip APP_MAPPED_IO iostat/REQ_TIME updates and post-EOF
folio zeroing.

Remove this shortcut in f2fs_vm_page_mkwrite() and stop setting it in
f2fs_read_single_page().

Reported-by: Chao Yu <chao@kernel.org>
Link: https://lore.kernel.org/linux-f2fs-devel/496b913c-eb81-4fa8-be72-c6a06a10aba2@kernel.org/
Fixes: fbfa2cc58d53 ("f2fs: add file operations")
Cc: stable@vger.kernel.org
Signed-off-by: Nanzhe Zhao <zhaonanzhe@xiaomi.com>
---
 fs/f2fs/data.c | 1 -
 fs/f2fs/file.c | 6 ------
 2 files changed, 7 deletions(-)

diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 21f396ebe22c..35d9ddaf78f4 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -2260,7 +2260,6 @@ static int f2fs_read_single_page(struct inode *inode, struct fsverity_info *vi,
 got_it:
 	if ((map->m_flags & F2FS_MAP_MAPPED)) {
 		block_nr = map->m_pblk + block_in_file - map->m_lblk;
-		folio_set_mappedtodisk(folio);
 
 		if (!f2fs_is_valid_blkaddr(F2FS_I_SB(inode), block_nr,
 						DATA_GENERIC_ENHANCE_READ)) {
diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index edc352569e87..2d763e156d06 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -217,12 +217,6 @@ static vm_fault_t f2fs_vm_page_mkwrite(struct vm_fault *vmf)
 	/* wait for GCed page writeback via META_MAPPING */
 	f2fs_wait_on_block_writeback(inode, dn.data_blkaddr);
 
-	/*
-	 * check to see if the page is mapped already (no holes)
-	 */
-	if (folio_test_mappedtodisk(folio))
-		goto out_sem;
-
 	/* page is wholly or partially inside EOF */
 	if (((loff_t)(folio->index + 1) << PAGE_SHIFT) >
 						i_size_read(inode)) {
-- 
2.43.0

Re: [PATCH v2] f2fs: remove mappedtodisk flag usage in mmap write path

[Chao Yu]

On 10/11/26 12:02, Nanzhe Zhao wrote:
> PG_mappedtodisk was inherited from the nobh/buffer-head era.  However,
> f2fs has no buffer heads and the flag is no longer meaningful.  The only
> user of the flag is the "no holes" shortcut in f2fs_vm_page_mkwrite(),
> which makes mmap write faults skip marking the folio dirty.  Chao Yu
> pointed out that this is not only a cleanup: it causes the following
> race condition with filesystem freezing:
> 
>   CPU0 (write fault)     CPU1 (freeze_super)
>   ----                   ----
>   f2fs_vm_page_mkwrite()
>   sb_start_pagefault()
>   /* shortcut: no dirty */
>   sb_end_pagefault()
> 
>                          sb_wait_write(SB_FREEZE_PAGEFAULT)
>                          /* no waiter, passes */
>                          sync_filesystem()
>                          /* folio not dirty, skipped */
> 
>   finish_fault()
>   /* PTE writable */
>   fault_dirty_shared_page()
>   folio_mark_dirty()
>   /* dirty folio on a frozen fs */
> 
> It will also skip APP_MAPPED_IO iostat/REQ_TIME updates and post-EOF
> folio zeroing.
> 
> Remove this shortcut in f2fs_vm_page_mkwrite() and stop setting it in
> f2fs_read_single_page().
> 
> Reported-by: Chao Yu <chao@kernel.org>
> Link: https://lore.kernel.org/linux-f2fs-devel/496b913c-eb81-4fa8-be72-c6a06a10aba2@kernel.org/
> Fixes: fbfa2cc58d53 ("f2fs: add file operations")
> Cc: stable@vger.kernel.org
> Signed-off-by: Nanzhe Zhao <zhaonanzhe@xiaomi.com>

Reviewed-by: Chao Yu <chao@kernel.org>

Thanks,

Re: [PATCH v2] f2fs: remove mappedtodisk flag usage in mmap write path

[Barry Song]

On Sun, Oct 11, 2026 at 12:02 PM Nanzhe Zhao <zhaonanzhe@xiaomi.com> wrote:
>
> PG_mappedtodisk was inherited from the nobh/buffer-head era.  However,
> f2fs has no buffer heads and the flag is no longer meaningful.  The only
> user of the flag is the "no holes" shortcut in f2fs_vm_page_mkwrite(),
> which makes mmap write faults skip marking the folio dirty.  Chao Yu
> pointed out that this is not only a cleanup: it causes the following
> race condition with filesystem freezing:

Somehow, I feel that "Remove this shortcut in
`f2fs_vm_page_mkwrite()` and stop setting it in
`f2fs_read_single_page()`" should come before "Chao Yu
pointed out that this is not only a cleanup: it causes
the following".

We're saying that removing this shortcut is both a cleanup
and a fix for the race, right?

>
>  CPU0 (write fault)     CPU1 (freeze_super)
>  ----                   ----
>  f2fs_vm_page_mkwrite()
>  sb_start_pagefault()
>  /* shortcut: no dirty */
>  sb_end_pagefault()
>
>                         sb_wait_write(SB_FREEZE_PAGEFAULT)
>                         /* no waiter, passes */
>                         sync_filesystem()
>                         /* folio not dirty, skipped */
>
>  finish_fault()
>  /* PTE writable */
>  fault_dirty_shared_page()
>  folio_mark_dirty()
>  /* dirty folio on a frozen fs */
>
> It will also skip APP_MAPPED_IO iostat/REQ_TIME updates and post-EOF
> folio zeroing.
>
> Remove this shortcut in f2fs_vm_page_mkwrite() and stop setting it in
> f2fs_read_single_page().

I mean, this sentence should be moved earlier, or the
preceding sentences should be reworded.

>
> Reported-by: Chao Yu <chao@kernel.org>
> Link: https://lore.kernel.org/linux-f2fs-devel/496b913c-eb81-4fa8-be72-c6a06a10aba2@kernel.org/
> Fixes: fbfa2cc58d53 ("f2fs: add file operations")
> Cc: stable@vger.kernel.org
> Signed-off-by: Nanzhe Zhao <zhaonanzhe@xiaomi.com>
> ---

Reviewed-by: Barry Song <baohua@kernel.org>