Skip to contentSkip to conversation

erofs: use the shared page cache for splice in inode_share mode

PATCH In review Zhan Xusheng 3 participants 7 messages Aug 20 EROFS file system

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 erofs: use the shared page cache for splice in inode_share mode

From: Zhan Xusheng <[email protected]>

From: Zhan Xusheng <[email protected]>

erofs_ishare_fops routes everything that touches the page cache to the backing file in ->private_data: read_iter clones the iocb onto it, mmap does vma_set_file(), fadvise calls vfs_fadvise() on it. splice_read was left as filemap_splice_read(), which works on the user file's own mapping.

filemap_splice_read() does init_sync_kiocb(&iocb, in), and filemap_get_pages() then takes iocb->ki_filp->f_mapping, so splice() and sendfile() populate the per-file page cache from disk instead of using the shared one. The data is correct, since erofs_fill_inode() sets that mapping's a_ops either way, but the same content ends up cached twice, which is what inode_share is there to avoid.

Pass the backing file, as read_iter already does.

Fixes 5ef3208e3be5 ("erofs: introduce the page cache share feature")
---
 fs/erofs/ishare.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
fs/erofs/ishare.c +8 -1
index fa7d4112dec5..01dc53e9e3ad 100644
@@ -148,6 +148,13 @@ static int erofs_ishare_mmap(struct file *file, struct vm_area_struct *vma)
 	return generic_file_readonly_mmap(file, vma);
 }
 
+static ssize_t erofs_ishare_splice_read(struct file *in, loff_t *ppos,
+					struct pipe_inode_info *pipe,
+					size_t len, unsigned int flags)
+{
+	return filemap_splice_read(in->private_data, ppos, pipe, len, flags);
+}
+
 static int erofs_ishare_fadvise(struct file *file, loff_t offset,
 				loff_t len, int advice)
 {
@@ -161,7 +168,7 @@ const struct file_operations erofs_ishare_fops = {
 	.mmap		= erofs_ishare_mmap,
 	.release	= erofs_ishare_file_release,
 	.get_unmapped_area = thp_get_unmapped_area,
-	.splice_read	= filemap_splice_read,
+	.splice_read	= erofs_ishare_splice_read,
 	.fadvise	= erofs_ishare_fadvise,
 };
 
Signature · 1 line
2.43.0
#1 lore raw
Quoted 30 lines · Zhan Xusheng

From: Zhan Xusheng <[email protected]>

From: Zhan Xusheng <[email protected]>

erofs_ishare_fops routes everything that touches the page cache to the backing file in ->private_data: read_iter clones the iocb onto it, mmap does vma_set_file(), fadvise calls vfs_fadvise() on it. splice_read was left as filemap_splice_read(), which works on the user file's own mapping.

filemap_splice_read() does init_sync_kiocb(&iocb, in), and filemap_get_pages() then takes iocb->ki_filp->f_mapping, so splice() and sendfile() populate the per-file page cache from disk instead of using the shared one. The data is correct, since erofs_fill_inode() sets that mapping's a_ops either way, but the same content ends up cached twice, which is what inode_share is there to avoid.

Pass the backing file, as read_iter already does.

Fixes 5ef3208e3be5 ("erofs: introduce the page cache share feature")
---
 fs/erofs/ishare.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
fs/erofs/ishare.c +0 -0
index fa7d4112dec5..01dc53e9e3ad 100644

fs/erofs/ishare.c static int erofs_ishare_mmap(struct file *file, struct vm_area_struct *vma) · line 155

Quoted 8 lines
fs/erofs/ishare.c +5 -0
@@ -148,6 +148,13 @@ static int erofs_ishare_mmap(struct file *file, struct vm_area_struct *vma)
 	return generic_file_readonly_mmap(file, vma);
 }
 
+static ssize_t erofs_ishare_splice_read(struct file *in, loff_t *ppos,
+					struct pipe_inode_info *pipe,
+					size_t len, unsigned int flags)
+{
+	return filemap_splice_read(in->private_data, ppos, pipe, len, flags);

Please refer to backing_file_splice_read() called from ovl_splice_read(), file_accessed() needs to be called on the original file (just as what .read_iter() i.e. filemap_read() does), and the input @ppos needs to be updated accordingly.

Signature · 2 lines
Thanks,
Jingbo
#2 lore raw
Quoted 4 lines · Jingbo Xu

Please refer to backing_file_splice_read() called from ovl_splice_read(), file_accessed() needs to be called on the original file (just as what .read_iter() i.e. filemap_read() does), and the input @ppos needs to be updated accordingly.

Taking the file_accessed() one, thanks. filemap_splice_read() calls it at mm/filemap.c:3155 on whatever file it was handed, so on the backing file, whereas backing_file_splice_read() ends in ctx->accessed(iocb->ki_filp), which for ovl_splice_read() is the original. v2 adds file_accessed(in).

@ppos looks already handled to me. filemap_splice_read() takes a loff_t * and advances it itself, at mm/filemap.c:3144; its internal kiocb is seeded from *ppos at 3083 and 3098, not the other way round. ovl_splice_read() has to copy iocb.ki_pos back because backing_file_splice_read() takes a struct kiocb and hands &iocb->ki_pos to vfs_splice_read(). Say if I have that wrong.

One you may want for read_iter too: it clones the kiocb onto the backing file, so filemap_read() marks that one accessed rather than the user's file, which is the shape splice_read had. Neither is observable today, since erofs_fc_fill_super() sets SB_RDONLY | SB_NOATIME and the backing file is opened O_NOATIME, so both reach a no-op. That is why I left read_iter alone here.

Thanks,
Zhan Xusheng

#3 lore raw
Quoted 11 lines · Zhan Xusheng
Quoted 4 lines · Jingbo Xu

Please refer to backing_file_splice_read() called from ovl_splice_read(), file_accessed() needs to be called on the original file (just as what .read_iter() i.e. filemap_read() does), and the input @ppos needs to be updated accordingly.

Taking the file_accessed() one, thanks. filemap_splice_read() calls it at mm/filemap.c:3155 on whatever file it was handed, so on the backing file, whereas backing_file_splice_read() ends in ctx->accessed(iocb->ki_filp), which for ovl_splice_read() is the original. v2 adds file_accessed(in).

Quoted 6 lines

@ppos looks already handled to me. filemap_splice_read() takes a loff_t * and advances it itself, at mm/filemap.c:3144; its internal kiocb is seeded from *ppos at 3083 and 3098, not the other way round. ovl_splice_read() has to copy iocb.ki_pos back because backing_file_splice_read() takes a struct kiocb and hands &iocb->ki_pos to vfs_splice_read(). Say if I have that wrong.

Make sense.

Quoted 7 lines

One you may want for read_iter too: it clones the kiocb onto the backing file, so filemap_read() marks that one accessed rather than the user's file, which is the shape splice_read had. Neither is observable today, since erofs_fc_fill_super() sets SB_RDONLY | SB_NOATIME and the backing file is opened O_NOATIME, so both reach a no-op. That is why I left read_iter alone here.

Okay, it seems that file_accessed() shall also be added to erofs_ishare_file_read_iter()?

Signature · 2 lines
Thanks,
Jingbo
maintainer
#4 lore raw
Quoted 32 lines · Jingbo Xu
Quoted 11 lines · Zhan Xusheng
Quoted 4 lines · Jingbo Xu

Please refer to backing_file_splice_read() called from ovl_splice_read(), file_accessed() needs to be called on the original file (just as what .read_iter() i.e. filemap_read() does), and the input @ppos needs to be updated accordingly.

Taking the file_accessed() one, thanks. filemap_splice_read() calls it at mm/filemap.c:3155 on whatever file it was handed, so on the backing file, whereas backing_file_splice_read() ends in ctx->accessed(iocb->ki_filp), which for ovl_splice_read() is the original. v2 adds file_accessed(in).

Quoted 6 lines

@ppos looks already handled to me. filemap_splice_read() takes a loff_t * and advances it itself, at mm/filemap.c:3144; its internal kiocb is seeded from *ppos at 3083 and 3098, not the other way round. ovl_splice_read() has to copy iocb.ki_pos back because backing_file_splice_read() takes a struct kiocb and hands &iocb->ki_pos to vfs_splice_read(). Say if I have that wrong.

Make sense.

Quoted 7 lines

One you may want for read_iter too: it clones the kiocb onto the backing file, so filemap_read() marks that one accessed rather than the user's file, which is the shape splice_read had. Neither is observable today, since erofs_fc_fill_super() sets SB_RDONLY | SB_NOATIME and the backing file is opened O_NOATIME, so both reach a no-op. That is why I left read_iter alone here.

Quoted 2 lines

Okay, it seems that file_accessed() shall also be added to erofs_ishare_file_read_iter()?

I think file_accessed() is a no-op for erofs?

Thanks,
Gao Xiang

Quoted 7 lines
Signature · 3 lines
-- 
Thanks,
Jingbo
#5 lore raw
Quoted 33 lines · Gao Xiang
Quoted 32 lines · Jingbo Xu
Quoted 11 lines · Zhan Xusheng
Quoted 4 lines · Jingbo Xu

Please refer to backing_file_splice_read() called from ovl_splice_read(), file_accessed() needs to be called on the original file (just as what .read_iter() i.e. filemap_read() does), and the input @ppos needs to be updated accordingly.

Taking the file_accessed() one, thanks. filemap_splice_read() calls it at mm/filemap.c:3155 on whatever file it was handed, so on the backing file, whereas backing_file_splice_read() ends in ctx->accessed(iocb->ki_filp), which for ovl_splice_read() is the original. v2 adds file_accessed(in).

Quoted 6 lines

@ppos looks already handled to me. filemap_splice_read() takes a loff_t * and advances it itself, at mm/filemap.c:3144; its internal kiocb is seeded from *ppos at 3083 and 3098, not the other way round. ovl_splice_read() has to copy iocb.ki_pos back because backing_file_splice_read() takes a struct kiocb and hands &iocb->ki_pos to vfs_splice_read(). Say if I have that wrong.

Make sense.

Quoted 7 lines

One you may want for read_iter too: it clones the kiocb onto the backing file, so filemap_read() marks that one accessed rather than the user's file, which is the shape splice_read had. Neither is observable today, since erofs_fc_fill_super() sets SB_RDONLY | SB_NOATIME and the backing file is opened O_NOATIME, so both reach a no-op. That is why I left read_iter alone here.

Quoted 4 lines
Quoted 2 lines

Okay, it seems that file_accessed() shall also be added to erofs_ishare_file_read_iter()?

I think file_accessed() is a no-op for erofs?

Right. erofs unconditionally sets SB_NOATIME (sb->s_flags |= SB_RDONLY | SB_NOATIME). Please ignore the noise..

Signature · 2 lines
Thanks,
Jingbo
maintainer
#6 lore raw
In reply to Zhan Xusheng

Hi Xusheng,

Quoted 4 lines · Zhan Xusheng

From: Zhan Xusheng <[email protected]>

From: Zhan Xusheng <[email protected]>

the subject can be improved as "erofs: support splice() in inode_share mode"

Quoted 12 lines

erofs_ishare_fops routes everything that touches the page cache to the backing file in ->private_data: read_iter clones the iocb onto it, mmap does vma_set_file(), fadvise calls vfs_fadvise() on it. splice_read was left as filemap_splice_read(), which works on the user file's own mapping.

filemap_splice_read() does init_sync_kiocb(&iocb, in), and filemap_get_pages() then takes iocb->ki_filp->f_mapping, so splice() and sendfile() populate the per-file page cache from disk instead of using the shared one. The data is correct, since erofs_fill_inode() sets that mapping's a_ops either way, but the same content ends up cached twice, which is what inode_share is there to avoid.

Quoted 5 lines

Pass the backing file, as read_iter already does.

Fixes 5ef3208e3be5 ("erofs: introduce the page cache share feature")

I observed the sashiko's report too:
https://sashiko.dev/#/patchset/20260818113713.116849-1-jefflexu%40linux.alibaba.com

Regardless of the implementation details, I don't think it is a fix since this feature should be considered as a best-effort approach.

If it's an enhancement, please also write an erofs-utils test and drop the fixes tag.

Thanks,
Gao Xiang