| Message ID | 20240328-gewendet-spargel-aa60a030ef74@brauner (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | [v2] fs: claw back a few FMODE_* bits | expand |
Looks good:
Reviewed-by: Christoph Hellwig <hch@lst.de>
On Thu 28-03-24 13:27:24, Christian Brauner wrote: > There's a bunch of flags that are purely based on what the file > operations support while also never being conditionally set or unset. > IOW, they're not subject to change for individual files. Imho, such > flags don't need to live in f_mode they might as well live in the fops > structs itself. And the fops struct already has that lonely > mmap_supported_flags member. We might as well turn that into a generic > fop_flags member and move a few flags from FMODE_* space into FOP_* > space. That gets us four FMODE_* bits back and the ability for new > static flags that are about file ops to not have to live in FMODE_* > space but in their own FOP_* space. It's not the most beautiful thing > ever but it gets the job done. Yes, there'll be an additional pointer > chase but hopefully that won't matter for these flags. > > I suspect there's a few more we can move into there and that we can also > redirect a bunch of new flag suggestions that follow this pattern into > the fop_flags field instead of f_mode. > > (Fwiw, FMODE_NOACCOUNT and FMODE_BACKING could live in fop_flags as > well because they're also completely static but they aren't really > about file operations so they're better suited for FMODE_* imho.) > > Signed-off-by: Christian Brauner <brauner@kernel.org> Looks good. Feel free to add: Reviewed-by: Jan Kara <jack@suse.cz> Honza > --- > block/bdev.c | 2 +- > block/fops.c | 1 + > drivers/dax/device.c | 2 +- > fs/btrfs/file.c | 4 ++-- > fs/ext4/file.c | 6 +++--- > fs/f2fs/file.c | 3 ++- > fs/read_write.c | 2 +- > fs/xfs/xfs_file.c | 8 +++++--- > include/linux/fs.h | 22 ++++++++++++---------- > io_uring/io_uring.c | 2 +- > io_uring/rw.c | 9 +++++---- > mm/mmap.c | 4 +++- > 12 files changed, 37 insertions(+), 28 deletions(-) > > diff --git a/block/bdev.c b/block/bdev.c > index b8e32d933a63..dd26d37356aa 100644 > --- a/block/bdev.c > +++ b/block/bdev.c > @@ -903,7 +903,7 @@ int bdev_open(struct block_device *bdev, blk_mode_t mode, void *holder, > disk_unblock_events(disk); > > bdev_file->f_flags |= O_LARGEFILE; > - bdev_file->f_mode |= FMODE_BUF_RASYNC | FMODE_CAN_ODIRECT; > + bdev_file->f_mode |= FMODE_CAN_ODIRECT; > if (bdev_nowait(bdev)) > bdev_file->f_mode |= FMODE_NOWAIT; > if (mode & BLK_OPEN_RESTRICT_WRITES) > diff --git a/block/fops.c b/block/fops.c > index 679d9b752fe8..af6c244314af 100644 > --- a/block/fops.c > +++ b/block/fops.c > @@ -863,6 +863,7 @@ const struct file_operations def_blk_fops = { > .splice_read = filemap_splice_read, > .splice_write = iter_file_splice_write, > .fallocate = blkdev_fallocate, > + .fop_flags = FOP_BUFFER_RASYNC, > }; > > static __init int blkdev_init(void) > diff --git a/drivers/dax/device.c b/drivers/dax/device.c > index 93ebedc5ec8c..c24ef4d3cf31 100644 > --- a/drivers/dax/device.c > +++ b/drivers/dax/device.c > @@ -377,7 +377,7 @@ static const struct file_operations dax_fops = { > .release = dax_release, > .get_unmapped_area = dax_get_unmapped_area, > .mmap = dax_mmap, > - .mmap_supported_flags = MAP_SYNC, > + .fop_flags = FOP_MMAP_SYNC, > }; > > static void dev_dax_cdev_del(void *cdev) > diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c > index f9d76072398d..1640c46f2153 100644 > --- a/fs/btrfs/file.c > +++ b/fs/btrfs/file.c > @@ -3719,8 +3719,7 @@ static int btrfs_file_open(struct inode *inode, struct file *filp) > { > int ret; > > - filp->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC | FMODE_BUF_WASYNC | > - FMODE_CAN_ODIRECT; > + filp->f_mode |= FMODE_NOWAIT | FMODE_CAN_ODIRECT; > > ret = fsverity_file_open(inode, filp); > if (ret) > @@ -3850,6 +3849,7 @@ const struct file_operations btrfs_file_operations = { > .compat_ioctl = btrfs_compat_ioctl, > #endif > .remap_file_range = btrfs_remap_file_range, > + .fop_flags = FOP_BUFFER_RASYNC | FOP_BUFFER_WASYNC, > }; > > int btrfs_fdatawrite_range(struct inode *inode, loff_t start, loff_t end) > diff --git a/fs/ext4/file.c b/fs/ext4/file.c > index 54d6ff22585c..28c51b0cc4db 100644 > --- a/fs/ext4/file.c > +++ b/fs/ext4/file.c > @@ -885,8 +885,7 @@ static int ext4_file_open(struct inode *inode, struct file *filp) > return ret; > } > > - filp->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC | > - FMODE_DIO_PARALLEL_WRITE; > + filp->f_mode |= FMODE_NOWAIT; > return dquot_file_open(inode, filp); > } > > @@ -938,7 +937,6 @@ const struct file_operations ext4_file_operations = { > .compat_ioctl = ext4_compat_ioctl, > #endif > .mmap = ext4_file_mmap, > - .mmap_supported_flags = MAP_SYNC, > .open = ext4_file_open, > .release = ext4_release_file, > .fsync = ext4_sync_file, > @@ -946,6 +944,8 @@ const struct file_operations ext4_file_operations = { > .splice_read = ext4_file_splice_read, > .splice_write = iter_file_splice_write, > .fallocate = ext4_fallocate, > + .fop_flags = FOP_MMAP_SYNC | FOP_BUFFER_RASYNC | > + FOP_DIO_PARALLEL_WRITE, > }; > > const struct inode_operations ext4_file_inode_operations = { > diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c > index 1761ad125f97..2b65e09822d4 100644 > --- a/fs/f2fs/file.c > +++ b/fs/f2fs/file.c > @@ -569,7 +569,7 @@ static int f2fs_file_open(struct inode *inode, struct file *filp) > if (err) > return err; > > - filp->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC; > + filp->f_mode |= FMODE_NOWAIT; > filp->f_mode |= FMODE_CAN_ODIRECT; > > return dquot_file_open(inode, filp); > @@ -5045,4 +5045,5 @@ const struct file_operations f2fs_file_operations = { > .splice_read = f2fs_file_splice_read, > .splice_write = iter_file_splice_write, > .fadvise = f2fs_file_fadvise, > + .fop_flags = FOP_BUFFER_RASYNC, > }; > diff --git a/fs/read_write.c b/fs/read_write.c > index d4c036e82b6c..2115d1f40bd5 100644 > --- a/fs/read_write.c > +++ b/fs/read_write.c > @@ -1685,7 +1685,7 @@ int generic_write_checks_count(struct kiocb *iocb, loff_t *count) > > if ((iocb->ki_flags & IOCB_NOWAIT) && > !((iocb->ki_flags & IOCB_DIRECT) || > - (file->f_mode & FMODE_BUF_WASYNC))) > + (file->f_op->fop_flags & FOP_BUFFER_WASYNC))) > return -EINVAL; > > return generic_write_check_limits(iocb->ki_filp, iocb->ki_pos, count); > diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c > index 632653e00906..147439ad3581 100644 > --- a/fs/xfs/xfs_file.c > +++ b/fs/xfs/xfs_file.c > @@ -1230,8 +1230,7 @@ xfs_file_open( > { > if (xfs_is_shutdown(XFS_M(inode->i_sb))) > return -EIO; > - file->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC | FMODE_BUF_WASYNC | > - FMODE_DIO_PARALLEL_WRITE | FMODE_CAN_ODIRECT; > + file->f_mode |= FMODE_NOWAIT | FMODE_CAN_ODIRECT; > return generic_file_open(inode, file); > } > > @@ -1490,7 +1489,6 @@ const struct file_operations xfs_file_operations = { > .compat_ioctl = xfs_file_compat_ioctl, > #endif > .mmap = xfs_file_mmap, > - .mmap_supported_flags = MAP_SYNC, > .open = xfs_file_open, > .release = xfs_file_release, > .fsync = xfs_file_fsync, > @@ -1498,6 +1496,8 @@ const struct file_operations xfs_file_operations = { > .fallocate = xfs_file_fallocate, > .fadvise = xfs_file_fadvise, > .remap_file_range = xfs_file_remap_range, > + .fop_flags = FOP_MMAP_SYNC | FOP_BUFFER_RASYNC | FOP_BUFFER_WASYNC | > + FOP_DIO_PARALLEL_WRITE, > }; > > const struct file_operations xfs_dir_file_operations = { > @@ -1510,4 +1510,6 @@ const struct file_operations xfs_dir_file_operations = { > .compat_ioctl = xfs_file_compat_ioctl, > #endif > .fsync = xfs_dir_fsync, > + .fop_flags = FOP_MMAP_SYNC | FOP_BUFFER_RASYNC | FOP_BUFFER_WASYNC | > + FOP_DIO_PARALLEL_WRITE, > }; > diff --git a/include/linux/fs.h b/include/linux/fs.h > index 8dfd53b52744..ece6e681ec77 100644 > --- a/include/linux/fs.h > +++ b/include/linux/fs.h > @@ -165,9 +165,6 @@ typedef int (dio_iodone_t)(struct kiocb *iocb, loff_t offset, > > #define FMODE_NOREUSE ((__force fmode_t)0x800000) > > -/* File supports non-exclusive O_DIRECT writes from multiple threads */ > -#define FMODE_DIO_PARALLEL_WRITE ((__force fmode_t)0x1000000) > - > /* File is embedded in backing_file object */ > #define FMODE_BACKING ((__force fmode_t)0x2000000) > > @@ -183,12 +180,6 @@ typedef int (dio_iodone_t)(struct kiocb *iocb, loff_t offset, > /* File does not contribute to nr_files count */ > #define FMODE_NOACCOUNT ((__force fmode_t)0x20000000) > > -/* File supports async buffered reads */ > -#define FMODE_BUF_RASYNC ((__force fmode_t)0x40000000) > - > -/* File supports async nowait buffered writes */ > -#define FMODE_BUF_WASYNC ((__force fmode_t)0x80000000) > - > /* > * Attribute flags. These should be or-ed together to figure out what > * has been changed! > @@ -2003,8 +1994,11 @@ struct iov_iter; > struct io_uring_cmd; > struct offset_ctx; > > +typedef unsigned int __bitwise fop_flags_t; > + > struct file_operations { > struct module *owner; > + fop_flags_t fop_flags; > loff_t (*llseek) (struct file *, loff_t, int); > ssize_t (*read) (struct file *, char __user *, size_t, loff_t *); > ssize_t (*write) (struct file *, const char __user *, size_t, loff_t *); > @@ -2017,7 +2011,6 @@ struct file_operations { > long (*unlocked_ioctl) (struct file *, unsigned int, unsigned long); > long (*compat_ioctl) (struct file *, unsigned int, unsigned long); > int (*mmap) (struct file *, struct vm_area_struct *); > - unsigned long mmap_supported_flags; > int (*open) (struct inode *, struct file *); > int (*flush) (struct file *, fl_owner_t id); > int (*release) (struct inode *, struct file *); > @@ -2048,6 +2041,15 @@ struct file_operations { > unsigned int poll_flags); > } __randomize_layout; > > +/* Supports async buffered reads */ > +#define FOP_BUFFER_RASYNC ((__force fop_flags_t)(1 << 0)) > +/* Supports async buffered writes */ > +#define FOP_BUFFER_WASYNC ((__force fop_flags_t)(1 << 1)) > +/* Supports synchronous page faults for mappings */ > +#define FOP_MMAP_SYNC ((__force fop_flags_t)(1 << 2)) > +/* Supports non-exclusive O_DIRECT writes from multiple threads */ > +#define FOP_DIO_PARALLEL_WRITE ((__force fop_flags_t)(1 << 3)) > + > /* Wrap a directory iterator that needs exclusive inode access */ > int wrap_directory_iterator(struct file *, struct dir_context *, > int (*) (struct file *, struct dir_context *)); > diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c > index 5d4b448fdc50..d73c9ad2d2f8 100644 > --- a/io_uring/io_uring.c > +++ b/io_uring/io_uring.c > @@ -471,7 +471,7 @@ static void io_prep_async_work(struct io_kiocb *req) > > /* don't serialize this request if the fs doesn't need it */ > if (should_hash && (req->file->f_flags & O_DIRECT) && > - (req->file->f_mode & FMODE_DIO_PARALLEL_WRITE)) > + (req->file->f_op->fop_flags & FOP_DIO_PARALLEL_WRITE)) > should_hash = false; > if (should_hash || (ctx->flags & IORING_SETUP_IOPOLL)) > io_wq_hash_work(&req->work, file_inode(req->file)); > diff --git a/io_uring/rw.c b/io_uring/rw.c > index 0585ebcc9773..d9dfde1142a1 100644 > --- a/io_uring/rw.c > +++ b/io_uring/rw.c > @@ -683,7 +683,8 @@ static bool io_rw_should_retry(struct io_kiocb *req) > * just use poll if we can, and don't attempt if the fs doesn't > * support callback based unlocks > */ > - if (io_file_can_poll(req) || !(req->file->f_mode & FMODE_BUF_RASYNC)) > + if (io_file_can_poll(req) || > + !(req->file->f_op->fop_flags & FOP_BUFFER_RASYNC)) > return false; > > wait->wait.func = io_async_buf_func; > @@ -1022,10 +1023,10 @@ int io_write(struct io_kiocb *req, unsigned int issue_flags) > if (unlikely(!io_file_supports_nowait(req))) > goto copy_iov; > > - /* File path supports NOWAIT for non-direct_IO only for block devices. */ > + /* Check if we can support NOWAIT. */ > if (!(kiocb->ki_flags & IOCB_DIRECT) && > - !(kiocb->ki_filp->f_mode & FMODE_BUF_WASYNC) && > - (req->flags & REQ_F_ISREG)) > + !(req->file->f_op->fop_flags & FOP_BUFFER_WASYNC) && > + (req->flags & REQ_F_ISREG)) > goto copy_iov; > > kiocb->ki_flags |= IOCB_NOWAIT; > diff --git a/mm/mmap.c b/mm/mmap.c > index 6dbda99a47da..3490af70f259 100644 > --- a/mm/mmap.c > +++ b/mm/mmap.c > @@ -1294,7 +1294,9 @@ unsigned long do_mmap(struct file *file, unsigned long addr, > if (!file_mmap_ok(file, inode, pgoff, len)) > return -EOVERFLOW; > > - flags_mask = LEGACY_MAP_MASK | file->f_op->mmap_supported_flags; > + flags_mask = LEGACY_MAP_MASK; > + if (file->f_op->fop_flags & FOP_MMAP_SYNC) > + flags_mask |= MAP_SYNC; > > switch (flags & MAP_TYPE) { > case MAP_SHARED: > -- > 2.43.0 >
On 3/28/24 6:27 AM, Christian Brauner wrote: > There's a bunch of flags that are purely based on what the file > operations support while also never being conditionally set or unset. > IOW, they're not subject to change for individual files. Imho, such > flags don't need to live in f_mode they might as well live in the fops > structs itself. And the fops struct already has that lonely > mmap_supported_flags member. We might as well turn that into a generic > fop_flags member and move a few flags from FMODE_* space into FOP_* > space. That gets us four FMODE_* bits back and the ability for new > static flags that are about file ops to not have to live in FMODE_* > space but in their own FOP_* space. It's not the most beautiful thing > ever but it gets the job done. Yes, there'll be an additional pointer > chase but hopefully that won't matter for these flags. > > I suspect there's a few more we can move into there and that we can also > redirect a bunch of new flag suggestions that follow this pattern into > the fop_flags field instead of f_mode. > > (Fwiw, FMODE_NOACCOUNT and FMODE_BACKING could live in fop_flags as > well because they're also completely static but they aren't really > about file operations so they're better suited for FMODE_* imho.) Reviewed-by: Jens Axboe <axboe@kernel.dk> As you know, this is going to cause conflicts. Wondering if it's worth doing anything about that...
On Thu, Mar 28, 2024 at 01:27:24PM +0100, Christian Brauner wrote: > There's a bunch of flags that are purely based on what the file > operations support while also never being conditionally set or unset. > IOW, they're not subject to change for individual files. Imho, such > flags don't need to live in f_mode they might as well live in the fops > structs itself. And the fops struct already has that lonely > mmap_supported_flags member. We might as well turn that into a generic > fop_flags member and move a few flags from FMODE_* space into FOP_* > space. That gets us four FMODE_* bits back and the ability for new > static flags that are about file ops to not have to live in FMODE_* > space but in their own FOP_* space. It's not the most beautiful thing > ever but it gets the job done. Yes, there'll be an additional pointer > chase but hopefully that won't matter for these flags. For the past couple of days several LTP tests (open_by_handle_at0[12] and name_to_handle_at01) have been failing on all the arm64 platforms we're running these tests on. I ran a bisect which came back to this commit which is in -next as 80a07849c0b8d8a2e839c8. I didn't verify a revert against -next due to conflicts and it being pretty late but I'm a little suspicous this report is misdrected, the commit doesn't look obviously relevant but equally I'm totally unfamiliar with this bit of the code. We use a NFS root for our testing. The last pass was next-20240328 and while the bisect was run on yesterday's -next things still look bad today. Bisect log: git bisect start # bad: [c0b832517f627ead3388c6f0c74e8ac10ad5774b] Add linux-next specific files for 20240402 git bisect bad c0b832517f627ead3388c6f0c74e8ac10ad5774b # good: [0fc83069bcaee78f60b8511d9453a9441963a072] Merge branch 'for-linux-next-fixes' of https://gitlab.freedesktop.org/drm/misc/kernel.git git bisect good 0fc83069bcaee78f60b8511d9453a9441963a072 # bad: [784b758e641c4b36be7ef8ab585bea834099b030] Merge branch 'for-linux-next' of https://gitlab.freedesktop.org/drm/misc/kernel.git git bisect bad 784b758e641c4b36be7ef8ab585bea834099b030 # good: [8b8b4dca2ddd82d3ae7e2a6a2fc7d49e511ceae7] Merge branch 'dev' of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/exfat.git git bisect good 8b8b4dca2ddd82d3ae7e2a6a2fc7d49e511ceae7 # bad: [2c20b30ed316f5cb8773e5f99c02cd997f2374b7] Merge branch 'main' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git git bisect bad 2c20b30ed316f5cb8773e5f99c02cd997f2374b7 # bad: [bbe99111e156e6838845511debc03f4ef5041f1a] Merge branch 'linux-next' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git git bisect bad bbe99111e156e6838845511debc03f4ef5041f1a # bad: [7e4f9465775837c3af79501698b0bb5ce28ffb11] Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/printk/linux.git git bisect bad 7e4f9465775837c3af79501698b0bb5ce28ffb11 # good: [a2b4cab9da7746c42f87c13721d305baf0085a20] Merge branch 'for-6.10' into for-next git bisect good a2b4cab9da7746c42f87c13721d305baf0085a20 # bad: [35c44ac8370af87614db9d0727da06a07d5436a2] Merge branch 'vfs.mount.api' into vfs.all git bisect bad 35c44ac8370af87614db9d0727da06a07d5436a2 # good: [16634c0975ba6569991274e7a4ccbb15766e31d3] Merge patch series 'fs: aio: more folio conversion' of https://lore.kernel.org/r/20240321131640.948634-1-wangkefeng.wang@huawei.com git bisect good 16634c0975ba6569991274e7a4ccbb15766e31d3 # bad: [276832a7ef840783bf867abe680db0eb6b3d4655] Merge branch 'vfs.misc' into vfs.all git bisect bad 276832a7ef840783bf867abe680db0eb6b3d4655 # good: [1b43c4629756a2c4bbbe4170eea1cc869fd8cb91] fs: Annotate struct file_handle with __counted_by() and use struct_size() git bisect good 1b43c4629756a2c4bbbe4170eea1cc869fd8cb91 # good: [22650a99821dda3d05f1c334ea90330b4982de56] fs,block: yield devices early git bisect good 22650a99821dda3d05f1c334ea90330b4982de56 # bad: [80a07849c0b8d8a2e839c83cea939c2e1cd7800b] fs: claw back a few FMODE_* bits git bisect bad 80a07849c0b8d8a2e839c83cea939c2e1cd7800b # first bad commit: [80a07849c0b8d8a2e839c83cea939c2e1cd7800b] fs: claw back a few FMODE_* bits
On Thu, Mar 28, 2024 at 01:27:24PM +0100, Christian Brauner wrote: > (Fwiw, FMODE_NOACCOUNT and FMODE_BACKING could live in fop_flags as > well because they're also completely static but they aren't really > about file operations so they're better suited for FMODE_* imho.) ???? FMODE_BACKING is set for files opened by e.g. overlayfs in the underlying layers. They bloody well can share file_operations with the files opened in the same underlying filesystem the usual way - you wouldn't be able to store that in file_operations, simply because the instances with identical ->f_op might differ in that flag. The same goes for FMODE_NOACCOUNT - kernel_file_open() vs. open(2) can easily yield struct file instances that differ in that flag, but have the same ->f_op.
On Wed 03-04-24 22:12:45, Mark Brown wrote: > On Thu, Mar 28, 2024 at 01:27:24PM +0100, Christian Brauner wrote: > > There's a bunch of flags that are purely based on what the file > > operations support while also never being conditionally set or unset. > > IOW, they're not subject to change for individual files. Imho, such > > flags don't need to live in f_mode they might as well live in the fops > > structs itself. And the fops struct already has that lonely > > mmap_supported_flags member. We might as well turn that into a generic > > fop_flags member and move a few flags from FMODE_* space into FOP_* > > space. That gets us four FMODE_* bits back and the ability for new > > static flags that are about file ops to not have to live in FMODE_* > > space but in their own FOP_* space. It's not the most beautiful thing > > ever but it gets the job done. Yes, there'll be an additional pointer > > chase but hopefully that won't matter for these flags. > > For the past couple of days several LTP tests (open_by_handle_at0[12] > and name_to_handle_at01) have been failing on all the arm64 platforms > we're running these tests on. I ran a bisect which came back to this Do you have some LTP logs / kernel logs available for the failing runs? Honza
On Thu, Apr 04, 2024 at 11:12:15AM +0200, Jan Kara wrote: > On Wed 03-04-24 22:12:45, Mark Brown wrote: > > For the past couple of days several LTP tests (open_by_handle_at0[12] > > and name_to_handle_at01) have been failing on all the arm64 platforms > > we're running these tests on. I ran a bisect which came back to this > Do you have some LTP logs / kernel logs available for the failing runs? Actually it looks like the issue went away with today's -next, but FWIW the logging for the open_by_handle_at0[12] failures was: tst_test.c:1690: TINFO: LTP version: 20230929 tst_test.c:1574: TINFO: Timeout per run is 0h 01m 30s tst_buffers.c:56: TINFO: Test is using guarded buffers name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (0): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (1): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (2): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (3): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (4): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (5): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (6): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (7): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (8): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (9): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (10): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (11): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (12): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (13): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (14): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (15): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (16): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (17): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (18): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (19): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (20): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (21): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (22): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (23): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (24): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (25): EFAULT (14) name_to_handle_at01.c:94: TFAIL: open_by_handle_at() failed (26): EFAULT (14) Summary: passed 0 failed 27 broken 0 skipped 0 warnings 0 tst_test.c:1690: TINFO: LTP version: 20230929 tst_test.c:1574: TINFO: Timeout per run is 0h 01m 30s tst_buffers.c:56: TINFO: Test is using guarded buffers open_by_handle_at02.c:92: TFAIL: invalid-dfd: open_by_handle_at() should fail with EBADF: EFAULT (14) open_by_handle_at02.c:92: TFAIL: stale-dfd: open_by_handle_at() should fail with ESTALE: EFAULT (14) open_by_handle_at02.c:98: TPASS: invalid-file-handle: open_by_handle_at() failed as expected: EFAULT (14) open_by_handle_at02.c:98: TPASS: high-file-handle-size: open_by_handle_at() failed as expected: EINVAL (22) open_by_handle_at02.c:98: TPASS: zero-file-handle-size: open_by_handle_at() failed as expected: EINVAL (22) tst_capability.c:29: TINFO: Dropping CAP_DAC_READ_SEARCH(2) tst_capability.c:41: TINFO: Permitting CAP_DAC_READ_SEARCH(2) open_by_handle_at02.c:98: TPASS: no-capability: open_by_handle_at() failed as expected: EPERM (1) open_by_handle_at02.c:92: TFAIL: symlink: open_by_handle_at() should fail with ELOOP: EFAULT (14) Summary: passed 4 failed 3 broken 0 skipped 0 warnings 0
> in the same underlying filesystem the usual way - you wouldn't be able > to store that in file_operations, simply because the instances with > identical ->f_op might differ in that flag. Yeah, good point. I never tried to actually do it otherwise it would've been quite obvious. Thanks!
On Thu, Apr 04, 2024 at 12:43:30PM +0100, Mark Brown wrote: > On Thu, Apr 04, 2024 at 11:12:15AM +0200, Jan Kara wrote: > > On Wed 03-04-24 22:12:45, Mark Brown wrote: > > > > For the past couple of days several LTP tests (open_by_handle_at0[12] > > > and name_to_handle_at01) have been failing on all the arm64 platforms > > > we're running these tests on. I ran a bisect which came back to this > > > Do you have some LTP logs / kernel logs available for the failing runs? > > Actually it looks like the issue went away with today's -next, but FWIW > the logging for the open_by_handle_at0[12] failures was: The bug was with: fs: Annotate struct file_handle with __counted_by() and use struct_size() and was reported and fixed in the thread around: https://lore.kernel.org/r/20240403110316.qtmypq2rtpueloga@quack3 so this is really unrelated.
On Fri, Apr 05, 2024 at 12:27:41PM +0200, Christian Brauner wrote: > On Thu, Apr 04, 2024 at 12:43:30PM +0100, Mark Brown wrote: > > Actually it looks like the issue went away with today's -next, but FWIW > > the logging for the open_by_handle_at0[12] failures was: > The bug was with: > fs: Annotate struct file_handle with __counted_by() and use struct_size() > and was reported and fixed in the thread around: > https://lore.kernel.org/r/20240403110316.qtmypq2rtpueloga@quack3 > so this is really unrelated. Ah, good - I did actually get to that one on my original bisect but reran the final section of the bisect since it looked like a false positive and wanted to confirm. Not sure why it glitched on the second run.
On Thu, Mar 28, 2024 at 01:27:24PM +0100, Christian Brauner wrote: > There's a bunch of flags that are purely based on what the file > operations support while also never being conditionally set or unset. > IOW, they're not subject to change for individual files. Imho, such > flags don't need to live in f_mode they might as well live in the fops > structs itself. And the fops struct already has that lonely > mmap_supported_flags member. We might as well turn that into a generic > fop_flags member and move a few flags from FMODE_* space into FOP_* > space. That gets us four FMODE_* bits back and the ability for new > static flags that are about file ops to not have to live in FMODE_* > space but in their own FOP_* space. It's not the most beautiful thing > ever but it gets the job done. Yes, there'll be an additional pointer > chase but hopefully that won't matter for these flags. > > I suspect there's a few more we can move into there and that we can also > redirect a bunch of new flag suggestions that follow this pattern into > the fop_flags field instead of f_mode. Looks sane; one suggestion, though - if we are going to try and free bits, etc., it might be a good idea to use e.g. #define FMODE_NOACCOUNT ((__force fmode_t)BIT(29)) instead of hex constants. IME it's easier to keep track of, especially if we have comments between the definitions.
On Sat, Apr 06, 2024 at 07:10:02AM +0100, Al Viro wrote: > On Thu, Mar 28, 2024 at 01:27:24PM +0100, Christian Brauner wrote: > > There's a bunch of flags that are purely based on what the file > > operations support while also never being conditionally set or unset. > > IOW, they're not subject to change for individual files. Imho, such > > flags don't need to live in f_mode they might as well live in the fops > > structs itself. And the fops struct already has that lonely > > mmap_supported_flags member. We might as well turn that into a generic > > fop_flags member and move a few flags from FMODE_* space into FOP_* > > space. That gets us four FMODE_* bits back and the ability for new > > static flags that are about file ops to not have to live in FMODE_* > > space but in their own FOP_* space. It's not the most beautiful thing > > ever but it gets the job done. Yes, there'll be an additional pointer > > chase but hopefully that won't matter for these flags. > > > > I suspect there's a few more we can move into there and that we can also > > redirect a bunch of new flag suggestions that follow this pattern into > > the fop_flags field instead of f_mode. > > Looks sane; one suggestion, though - if we are going to try and free > bits, etc., it might be a good idea to use e.g. > #define FMODE_NOACCOUNT ((__force fmode_t)BIT(29)) > instead of hex constants. IME it's easier to keep track of, especially > if we have comments between the definitions. ... or (1u << 29), for that matter; the point is that counting zeroes visually is error-prone, so seeing the binary logarithm of the value somewhere would be a good idea.
On Sat, Apr 06, 2024 at 07:16:04AM +0100, Al Viro wrote: > On Sat, Apr 06, 2024 at 07:10:02AM +0100, Al Viro wrote: > > On Thu, Mar 28, 2024 at 01:27:24PM +0100, Christian Brauner wrote: > > > There's a bunch of flags that are purely based on what the file > > > operations support while also never being conditionally set or unset. > > > IOW, they're not subject to change for individual files. Imho, such > > > flags don't need to live in f_mode they might as well live in the fops > > > structs itself. And the fops struct already has that lonely > > > mmap_supported_flags member. We might as well turn that into a generic > > > fop_flags member and move a few flags from FMODE_* space into FOP_* > > > space. That gets us four FMODE_* bits back and the ability for new > > > static flags that are about file ops to not have to live in FMODE_* > > > space but in their own FOP_* space. It's not the most beautiful thing > > > ever but it gets the job done. Yes, there'll be an additional pointer > > > chase but hopefully that won't matter for these flags. > > > > > > I suspect there's a few more we can move into there and that we can also > > > redirect a bunch of new flag suggestions that follow this pattern into > > > the fop_flags field instead of f_mode. > > > > Looks sane; one suggestion, though - if we are going to try and free > > bits, etc., it might be a good idea to use e.g. > > #define FMODE_NOACCOUNT ((__force fmode_t)BIT(29)) > > instead of hex constants. IME it's easier to keep track of, especially > > if we have comments between the definitions. > > ... or (1u << 29), for that matter; the point is that counting zeroes > visually is error-prone, so seeing the binary logarithm of the value > somewhere would be a good idea. Sounds good. I've converted all FMODE_* flags to use <<.
diff --git a/block/bdev.c b/block/bdev.c index b8e32d933a63..dd26d37356aa 100644 --- a/block/bdev.c +++ b/block/bdev.c @@ -903,7 +903,7 @@ int bdev_open(struct block_device *bdev, blk_mode_t mode, void *holder, disk_unblock_events(disk); bdev_file->f_flags |= O_LARGEFILE; - bdev_file->f_mode |= FMODE_BUF_RASYNC | FMODE_CAN_ODIRECT; + bdev_file->f_mode |= FMODE_CAN_ODIRECT; if (bdev_nowait(bdev)) bdev_file->f_mode |= FMODE_NOWAIT; if (mode & BLK_OPEN_RESTRICT_WRITES) diff --git a/block/fops.c b/block/fops.c index 679d9b752fe8..af6c244314af 100644 --- a/block/fops.c +++ b/block/fops.c @@ -863,6 +863,7 @@ const struct file_operations def_blk_fops = { .splice_read = filemap_splice_read, .splice_write = iter_file_splice_write, .fallocate = blkdev_fallocate, + .fop_flags = FOP_BUFFER_RASYNC, }; static __init int blkdev_init(void) diff --git a/drivers/dax/device.c b/drivers/dax/device.c index 93ebedc5ec8c..c24ef4d3cf31 100644 --- a/drivers/dax/device.c +++ b/drivers/dax/device.c @@ -377,7 +377,7 @@ static const struct file_operations dax_fops = { .release = dax_release, .get_unmapped_area = dax_get_unmapped_area, .mmap = dax_mmap, - .mmap_supported_flags = MAP_SYNC, + .fop_flags = FOP_MMAP_SYNC, }; static void dev_dax_cdev_del(void *cdev) diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index f9d76072398d..1640c46f2153 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -3719,8 +3719,7 @@ static int btrfs_file_open(struct inode *inode, struct file *filp) { int ret; - filp->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC | FMODE_BUF_WASYNC | - FMODE_CAN_ODIRECT; + filp->f_mode |= FMODE_NOWAIT | FMODE_CAN_ODIRECT; ret = fsverity_file_open(inode, filp); if (ret) @@ -3850,6 +3849,7 @@ const struct file_operations btrfs_file_operations = { .compat_ioctl = btrfs_compat_ioctl, #endif .remap_file_range = btrfs_remap_file_range, + .fop_flags = FOP_BUFFER_RASYNC | FOP_BUFFER_WASYNC, }; int btrfs_fdatawrite_range(struct inode *inode, loff_t start, loff_t end) diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 54d6ff22585c..28c51b0cc4db 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -885,8 +885,7 @@ static int ext4_file_open(struct inode *inode, struct file *filp) return ret; } - filp->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC | - FMODE_DIO_PARALLEL_WRITE; + filp->f_mode |= FMODE_NOWAIT; return dquot_file_open(inode, filp); } @@ -938,7 +937,6 @@ const struct file_operations ext4_file_operations = { .compat_ioctl = ext4_compat_ioctl, #endif .mmap = ext4_file_mmap, - .mmap_supported_flags = MAP_SYNC, .open = ext4_file_open, .release = ext4_release_file, .fsync = ext4_sync_file, @@ -946,6 +944,8 @@ const struct file_operations ext4_file_operations = { .splice_read = ext4_file_splice_read, .splice_write = iter_file_splice_write, .fallocate = ext4_fallocate, + .fop_flags = FOP_MMAP_SYNC | FOP_BUFFER_RASYNC | + FOP_DIO_PARALLEL_WRITE, }; const struct inode_operations ext4_file_inode_operations = { diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 1761ad125f97..2b65e09822d4 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -569,7 +569,7 @@ static int f2fs_file_open(struct inode *inode, struct file *filp) if (err) return err; - filp->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC; + filp->f_mode |= FMODE_NOWAIT; filp->f_mode |= FMODE_CAN_ODIRECT; return dquot_file_open(inode, filp); @@ -5045,4 +5045,5 @@ const struct file_operations f2fs_file_operations = { .splice_read = f2fs_file_splice_read, .splice_write = iter_file_splice_write, .fadvise = f2fs_file_fadvise, + .fop_flags = FOP_BUFFER_RASYNC, }; diff --git a/fs/read_write.c b/fs/read_write.c index d4c036e82b6c..2115d1f40bd5 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -1685,7 +1685,7 @@ int generic_write_checks_count(struct kiocb *iocb, loff_t *count) if ((iocb->ki_flags & IOCB_NOWAIT) && !((iocb->ki_flags & IOCB_DIRECT) || - (file->f_mode & FMODE_BUF_WASYNC))) + (file->f_op->fop_flags & FOP_BUFFER_WASYNC))) return -EINVAL; return generic_write_check_limits(iocb->ki_filp, iocb->ki_pos, count); diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 632653e00906..147439ad3581 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -1230,8 +1230,7 @@ xfs_file_open( { if (xfs_is_shutdown(XFS_M(inode->i_sb))) return -EIO; - file->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC | FMODE_BUF_WASYNC | - FMODE_DIO_PARALLEL_WRITE | FMODE_CAN_ODIRECT; + file->f_mode |= FMODE_NOWAIT | FMODE_CAN_ODIRECT; return generic_file_open(inode, file); } @@ -1490,7 +1489,6 @@ const struct file_operations xfs_file_operations = { .compat_ioctl = xfs_file_compat_ioctl, #endif .mmap = xfs_file_mmap, - .mmap_supported_flags = MAP_SYNC, .open = xfs_file_open, .release = xfs_file_release, .fsync = xfs_file_fsync, @@ -1498,6 +1496,8 @@ const struct file_operations xfs_file_operations = { .fallocate = xfs_file_fallocate, .fadvise = xfs_file_fadvise, .remap_file_range = xfs_file_remap_range, + .fop_flags = FOP_MMAP_SYNC | FOP_BUFFER_RASYNC | FOP_BUFFER_WASYNC | + FOP_DIO_PARALLEL_WRITE, }; const struct file_operations xfs_dir_file_operations = { @@ -1510,4 +1510,6 @@ const struct file_operations xfs_dir_file_operations = { .compat_ioctl = xfs_file_compat_ioctl, #endif .fsync = xfs_dir_fsync, + .fop_flags = FOP_MMAP_SYNC | FOP_BUFFER_RASYNC | FOP_BUFFER_WASYNC | + FOP_DIO_PARALLEL_WRITE, }; diff --git a/include/linux/fs.h b/include/linux/fs.h index 8dfd53b52744..ece6e681ec77 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -165,9 +165,6 @@ typedef int (dio_iodone_t)(struct kiocb *iocb, loff_t offset, #define FMODE_NOREUSE ((__force fmode_t)0x800000) -/* File supports non-exclusive O_DIRECT writes from multiple threads */ -#define FMODE_DIO_PARALLEL_WRITE ((__force fmode_t)0x1000000) - /* File is embedded in backing_file object */ #define FMODE_BACKING ((__force fmode_t)0x2000000) @@ -183,12 +180,6 @@ typedef int (dio_iodone_t)(struct kiocb *iocb, loff_t offset, /* File does not contribute to nr_files count */ #define FMODE_NOACCOUNT ((__force fmode_t)0x20000000) -/* File supports async buffered reads */ -#define FMODE_BUF_RASYNC ((__force fmode_t)0x40000000) - -/* File supports async nowait buffered writes */ -#define FMODE_BUF_WASYNC ((__force fmode_t)0x80000000) - /* * Attribute flags. These should be or-ed together to figure out what * has been changed! @@ -2003,8 +1994,11 @@ struct iov_iter; struct io_uring_cmd; struct offset_ctx; +typedef unsigned int __bitwise fop_flags_t; + struct file_operations { struct module *owner; + fop_flags_t fop_flags; loff_t (*llseek) (struct file *, loff_t, int); ssize_t (*read) (struct file *, char __user *, size_t, loff_t *); ssize_t (*write) (struct file *, const char __user *, size_t, loff_t *); @@ -2017,7 +2011,6 @@ struct file_operations { long (*unlocked_ioctl) (struct file *, unsigned int, unsigned long); long (*compat_ioctl) (struct file *, unsigned int, unsigned long); int (*mmap) (struct file *, struct vm_area_struct *); - unsigned long mmap_supported_flags; int (*open) (struct inode *, struct file *); int (*flush) (struct file *, fl_owner_t id); int (*release) (struct inode *, struct file *); @@ -2048,6 +2041,15 @@ struct file_operations { unsigned int poll_flags); } __randomize_layout; +/* Supports async buffered reads */ +#define FOP_BUFFER_RASYNC ((__force fop_flags_t)(1 << 0)) +/* Supports async buffered writes */ +#define FOP_BUFFER_WASYNC ((__force fop_flags_t)(1 << 1)) +/* Supports synchronous page faults for mappings */ +#define FOP_MMAP_SYNC ((__force fop_flags_t)(1 << 2)) +/* Supports non-exclusive O_DIRECT writes from multiple threads */ +#define FOP_DIO_PARALLEL_WRITE ((__force fop_flags_t)(1 << 3)) + /* Wrap a directory iterator that needs exclusive inode access */ int wrap_directory_iterator(struct file *, struct dir_context *, int (*) (struct file *, struct dir_context *)); diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c index 5d4b448fdc50..d73c9ad2d2f8 100644 --- a/io_uring/io_uring.c +++ b/io_uring/io_uring.c @@ -471,7 +471,7 @@ static void io_prep_async_work(struct io_kiocb *req) /* don't serialize this request if the fs doesn't need it */ if (should_hash && (req->file->f_flags & O_DIRECT) && - (req->file->f_mode & FMODE_DIO_PARALLEL_WRITE)) + (req->file->f_op->fop_flags & FOP_DIO_PARALLEL_WRITE)) should_hash = false; if (should_hash || (ctx->flags & IORING_SETUP_IOPOLL)) io_wq_hash_work(&req->work, file_inode(req->file)); diff --git a/io_uring/rw.c b/io_uring/rw.c index 0585ebcc9773..d9dfde1142a1 100644 --- a/io_uring/rw.c +++ b/io_uring/rw.c @@ -683,7 +683,8 @@ static bool io_rw_should_retry(struct io_kiocb *req) * just use poll if we can, and don't attempt if the fs doesn't * support callback based unlocks */ - if (io_file_can_poll(req) || !(req->file->f_mode & FMODE_BUF_RASYNC)) + if (io_file_can_poll(req) || + !(req->file->f_op->fop_flags & FOP_BUFFER_RASYNC)) return false; wait->wait.func = io_async_buf_func; @@ -1022,10 +1023,10 @@ int io_write(struct io_kiocb *req, unsigned int issue_flags) if (unlikely(!io_file_supports_nowait(req))) goto copy_iov; - /* File path supports NOWAIT for non-direct_IO only for block devices. */ + /* Check if we can support NOWAIT. */ if (!(kiocb->ki_flags & IOCB_DIRECT) && - !(kiocb->ki_filp->f_mode & FMODE_BUF_WASYNC) && - (req->flags & REQ_F_ISREG)) + !(req->file->f_op->fop_flags & FOP_BUFFER_WASYNC) && + (req->flags & REQ_F_ISREG)) goto copy_iov; kiocb->ki_flags |= IOCB_NOWAIT; diff --git a/mm/mmap.c b/mm/mmap.c index 6dbda99a47da..3490af70f259 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1294,7 +1294,9 @@ unsigned long do_mmap(struct file *file, unsigned long addr, if (!file_mmap_ok(file, inode, pgoff, len)) return -EOVERFLOW; - flags_mask = LEGACY_MAP_MASK | file->f_op->mmap_supported_flags; + flags_mask = LEGACY_MAP_MASK; + if (file->f_op->fop_flags & FOP_MMAP_SYNC) + flags_mask |= MAP_SYNC; switch (flags & MAP_TYPE) { case MAP_SHARED:
There's a bunch of flags that are purely based on what the file operations support while also never being conditionally set or unset. IOW, they're not subject to change for individual files. Imho, such flags don't need to live in f_mode they might as well live in the fops structs itself. And the fops struct already has that lonely mmap_supported_flags member. We might as well turn that into a generic fop_flags member and move a few flags from FMODE_* space into FOP_* space. That gets us four FMODE_* bits back and the ability for new static flags that are about file ops to not have to live in FMODE_* space but in their own FOP_* space. It's not the most beautiful thing ever but it gets the job done. Yes, there'll be an additional pointer chase but hopefully that won't matter for these flags. I suspect there's a few more we can move into there and that we can also redirect a bunch of new flag suggestions that follow this pattern into the fop_flags field instead of f_mode. (Fwiw, FMODE_NOACCOUNT and FMODE_BACKING could live in fop_flags as well because they're also completely static but they aren't really about file operations so they're better suited for FMODE_* imho.) Signed-off-by: Christian Brauner <brauner@kernel.org> --- block/bdev.c | 2 +- block/fops.c | 1 + drivers/dax/device.c | 2 +- fs/btrfs/file.c | 4 ++-- fs/ext4/file.c | 6 +++--- fs/f2fs/file.c | 3 ++- fs/read_write.c | 2 +- fs/xfs/xfs_file.c | 8 +++++--- include/linux/fs.h | 22 ++++++++++++---------- io_uring/io_uring.c | 2 +- io_uring/rw.c | 9 +++++---- mm/mmap.c | 4 +++- 12 files changed, 37 insertions(+), 28 deletions(-)