fuse: keep inode->i_blkbits constant
Subsystems (2)
PATCH v2 1/1 fuse: keep inode->i_blkbits constant
With fuse now using iomap for writeback handling, inode blkbits changes are problematic because iomap relies on inode->i_blkbits for its internal bitmap logic. Currently we change inode->i_blkbits in fuse to match the attr->blksize value passed in by the server.
This commit keeps inode->i_blkbits constant in fuse. Any attr->blksize values passed in by the server will not update inode->i_blkbits. The client-side behavior for stat is unaffected, stat will still reflect the blocksize passed in by the server.
---
Changelog:
v1: https://lore.kernel.org/linux-fsdevel/[email protected]/#t
v1 -> v2:
* Remove warning and keep stat() behavior unchanged (Miklos)
* Dropped 2nd patch
--- fs/fuse/inode.c | 5 ----- 1 file changed, 5 deletions(-)
fs/fuse/inode.c +0 -5
@@ -285,11 +285,6 @@ void fuse_change_attributes_common(struct inode *inode, struct fuse_attr *attr, } } -if (attr->blksize !=0 ) - inode->i_blkbits =ilog2 (attr->blksize); -else - inode->i_blkbits = inode->i_sb->s_blocksize_bits; -/* * Don't set the sticky bit in i_mode, unless we want the VFS * to check permissions. This prevents failures due to the
Signature · 1 line
2.47.3
Quoted 11 lines · Joanne Koong
With fuse now using iomap for writeback handling, inode blkbits changes are problematic because iomap relies on
inode->i_blkbitsfor its internal bitmap logic. Currently we changeinode->i_blkbitsin fuse to match theattr->blksizevalue passed in by the server.This commit keeps
inode->i_blkbitsconstant in fuse. Anyattr->blksizevalues passed in by the server will not updateinode->i_blkbits. The client-side behavior for stat is unaffected, stat will still reflect the blocksize passed in by the server.[...]
Applied to the vfs.fixes branch of the vfs/vfs.git tree. Patches in the vfs.fixes branch should appear in linux-next soon.
Please report any outstanding bugs that were missed during review in a new review to the original patch series allowing us to drop it.
It's encouraged to provide Acked-bys and Reviewed-bys even though the patch has now been applied. If possible patch trailers will be updated.
Note that commit hashes shown below are subject to change due to rebase, trailer updates or similar. If in doubt, please check the listed branch.
tree: https://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs.git branch: vfs.fixes
[1/1] fuse: keep inode->i_blkbits constant
https://git.kernel.org/vfs/vfs/c/0110b6d5545d
Miklos Szeredi
maintainerQuoted 10 lines · Joanne Koong
With fuse now using iomap for writeback handling, inode blkbits changes are problematic because iomap relies on
inode->i_blkbitsfor its internal bitmap logic. Currently we changeinode->i_blkbitsin fuse to match theattr->blksizevalue passed in by the server.This commit keeps
inode->i_blkbitsconstant in fuse. Anyattr->blksizevalues passed in by the server will not updateinode->i_blkbits. The client-side behavior for stat is unaffected, stat will still reflect the blocksize passed in by the server.
Not quite. You also need to save ilog2(attr->blksize) in fi->orig_i_blkbits and restore it after calling generic_fillattr() in fuse_update_get_attr() just like it's done for i_mode and i_ino.
Thanks,
Miklos
Quoted 8 lines · Miklos Szeredi
Quoted 6 lines · Joanne Koong
With fuse now using iomap for writeback handling, inode blkbits changes are problematic because iomap relies on
inode->i_blkbitsfor its internal bitmap logic. Currently we changeinode->i_blkbitsin fuse to match theattr->blksizevalue passed in by the server.
Quoted 9 lines
Quoted 4 lines
This commit keeps
inode->i_blkbitsconstant in fuse. Anyattr->blksizevalues passed in by the server will not updateinode->i_blkbits. The client-side behavior for stat is unaffected, stat will still reflect the blocksize passed in by the server.Not quite. You also need to save
ilog2(attr->blksize)infi->orig_i_blkbitsand restore it after callinggeneric_fillattr()infuse_update_get_attr()just like it's done fori_modeandi_ino.
Oh I see, thanks. I'll resubmit this. And will do the same logic for the fuseblk i_blkbits patch I submitted yesterday.
Thanks,
Joanne
Quoted 2 lines
Thanks,
Miklos
Quoted 7 lines · Miklos Szeredi
Quoted 6 lines · Joanne Koong
With fuse now using iomap for writeback handling, inode blkbits changes are problematic because iomap relies on
inode->i_blkbitsfor its internal bitmap logic. Currently we changeinode->i_blkbitsin fuse to match theattr->blksizevalue passed in by the server.
Quoted 8 lines
Quoted 4 lines
This commit keeps
inode->i_blkbitsconstant in fuse. Anyattr->blksizevalues passed in by the server will not updateinode->i_blkbits. The client-side behavior for stat is unaffected, stat will still reflect the blocksize passed in by the server.Not quite. You also need to save
ilog2(attr->blksize)infi->orig_i_blkbitsand restore it after callinggeneric_fillattr()infuse_update_get_attr()just like it's done fori_modeandi_ino.
Why is that? Is the goal here that fstat() should always return the most recent blocksize the fuse server reported, even if the pagecache accounting still uses the blocksize first reported when the kernel created the incore struct inode?
--D
Quoted 14 lines · Darrick J. Wong
Quoted 12 lines · Miklos Szeredi
Quoted 10 lines · Joanne Koong
With fuse now using iomap for writeback handling, inode blkbits changes are problematic because iomap relies on
inode->i_blkbitsfor its internal bitmap logic. Currently we changeinode->i_blkbitsin fuse to match theattr->blksizevalue passed in by the server.This commit keeps
inode->i_blkbitsconstant in fuse. Anyattr->blksizevalues passed in by the server will not updateinode->i_blkbits. The client-side behavior for stat is unaffected, stat will still reflect the blocksize passed in by the server.
Quoted 8 lines
Quoted 3 lines
Not quite. You also need to save
ilog2(attr->blksize)infi->orig_i_blkbitsand restore it after callinggeneric_fillattr()infuse_update_get_attr()just like it's done fori_modeandi_ino.Why is that? Is the goal here that
fstat()should always return the most recent blocksize the fuse server reported, even if the pagecache accounting still uses the blocksize first reported when the kernel created the incore struct inode?
My understanding is that it's because in that path it uses cached stat values instead of fetching with another statx call to the server so it has to reflect the blocksize the server previously set. It took me a while to realize that the blocksize the server reports to the client is unrelated to whatever blocksize the kernel internally uses for the inode since the kernel doesn't do any block i/o for fuse; the commit message in commit 0e9663ee452ff ("fuse: add blksize field to fuse_attr") says the blocksize attribute is if "the filesystem might want to give a hint to the app about the optimal I/O size".
Thanks,
Joanne
Quoted 2 lines
--D
Miklos Szeredi
maintainerQuoted 9 lines · Joanne Koong
My understanding is that it's because in that path it uses cached stat values instead of fetching with another statx call to the server so it has to reflect the blocksize the server previously set. It took me a while to realize that the blocksize the server reports to the client is unrelated to whatever blocksize the kernel internally uses for the inode since the kernel doesn't do any block i/o for fuse; the commit message in commit
0e9663ee452ff("fuse: add blksize field tofuse_attr") says the blocksize attribute is if "the filesystem might want to give a hint to the app about the optimal I/O size".
Right, that's what POSIX says:
st_blksize A file system-specific preferred I/O block size for
this object. In some file system types, this may vary from file to file.
Thanks,
Miklos
Quoted 12 lines · Miklos Szeredi
Quoted 9 lines · Joanne Koong
My understanding is that it's because in that path it uses cached stat values instead of fetching with another statx call to the server so it has to reflect the blocksize the server previously set. It took me a while to realize that the blocksize the server reports to the client is unrelated to whatever blocksize the kernel internally uses for the inode since the kernel doesn't do any block i/o for fuse; the commit message in commit
0e9663ee452ff("fuse: add blksize field tofuse_attr") says the blocksize attribute is if "the filesystem might want to give a hint to the app about the optimal I/O size".
Quoted 5 lines
Right, that's what POSIX says:
st_blksizeA file system-specific preferred I/O block size for
this object. In some file system types, this may vary from file to file.
Ahahaha yes, I forgot about that. Regular filesystems can set i_blkbits to whatever granularity they want and it persists until eviction, and if they want to trick userspace they set st_blksize to something else. Kind of like how XFS rounds that up to PAGE_SIZE for 1k fsblock filesystems.
Now I see what Joanne meant about it taking a while to wrap her head around i_blkbits vs. st_blksize. Thanks for pointing that out. :)
--D
Quoted 3 lines
Thanks,
Miklos