Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix handling of errors from dmu_write_uio_dbuf() on FreeBSD #12964

Merged
merged 1 commit into from
Jan 21, 2022

Conversation

markjdb
Copy link
Contributor

@markjdb markjdb commented Jan 12, 2022

The change restores error handling in zfs_write() that was lost when the OS-specific implementations were merged in commit
e53d678.

Motivation and Context

The change fixes FreeBSD PR 260453: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=260453

In particular the bug can cause various problems with write(2): for example, it may fail to update the page cache, or with O_APPEND it may cause write(2) to report that it wrote n bytes while only extending the size of the file by m, with n > m.

Description

The change adds special handling of EFAULT from dmu_write_uio_dbuf(). This does not quite match the old implementation and should result in no change of behaviour on Linux.

How Has This Been Tested?

Manual testing using the reproduction scenario from the FreeBSD PR.

Types of changes

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Performance enhancement (non-breaking change which improves efficiency)
  • Code cleanup (non-breaking change which makes code smaller or more readable)
  • Breaking change (fix or feature that would cause existing functionality to change)
  • Library ABI change (libzfs, libzfs_core, libnvpair, libuutil and libzfsbootenv)
  • Documentation (a change to man pages or other documentation)

Checklist:

FreeBSD's implementation of zfs_uio_fault_move() returns EFAULT when a
page fault occurs while copying data in or out of user buffers.  The VFS
treats such errors specially and will retry the I/O operation (which may
have made some partial progress).

When the FreeBSD and Linux implementations of zfs_write() were merged,
the handling of errors from dmu_write_uio_dbuf() changed such that
EFAULT is not handled as a partial write.  For example, when appending
to a file, the z_size field of the znode is not updated after a partial
write resulting in EFAULT.

Restore the old handling of errors from dmu_write_uio_dbuf() to fix
this.  This should have no impact on Linux, which has special handling
for EFAULT already.

Sponsored-by: The FreeBSD Foundation
Signed-off-by: Mark Johnston <markj@FreeBSD.org>
@behlendorf behlendorf added the Status: Accepted Ready to integrate (reviewed, tested) label Jan 13, 2022
@behlendorf behlendorf merged commit 063daa8 into openzfs:master Jan 21, 2022
markjdb added a commit to markjdb/zfs that referenced this pull request Feb 1, 2022
FreeBSD's implementation of zfs_uio_fault_move() returns EFAULT when a
page fault occurs while copying data in or out of user buffers.  The VFS
treats such errors specially and will retry the I/O operation (which may
have made some partial progress).

When the FreeBSD and Linux implementations of zfs_write() were merged,
the handling of errors from dmu_write_uio_dbuf() changed such that
EFAULT is not handled as a partial write.  For example, when appending
to a file, the z_size field of the znode is not updated after a partial
write resulting in EFAULT.

Restore the old handling of errors from dmu_write_uio_dbuf() to fix
this.  This should have no impact on Linux, which has special handling
for EFAULT already.

Reviewed-by: Andriy Gapon <avg@FreeBSD.org>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes openzfs#12964
tonyhutter pushed a commit that referenced this pull request Feb 3, 2022
FreeBSD's implementation of zfs_uio_fault_move() returns EFAULT when a
page fault occurs while copying data in or out of user buffers.  The VFS
treats such errors specially and will retry the I/O operation (which may
have made some partial progress).

When the FreeBSD and Linux implementations of zfs_write() were merged,
the handling of errors from dmu_write_uio_dbuf() changed such that
EFAULT is not handled as a partial write.  For example, when appending
to a file, the z_size field of the znode is not updated after a partial
write resulting in EFAULT.

Restore the old handling of errors from dmu_write_uio_dbuf() to fix
this.  This should have no impact on Linux, which has special handling
for EFAULT already.

Reviewed-by: Andriy Gapon <avg@FreeBSD.org>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes #12964
nicman23 pushed a commit to nicman23/zfs that referenced this pull request Aug 22, 2022
FreeBSD's implementation of zfs_uio_fault_move() returns EFAULT when a
page fault occurs while copying data in or out of user buffers.  The VFS
treats such errors specially and will retry the I/O operation (which may
have made some partial progress).

When the FreeBSD and Linux implementations of zfs_write() were merged,
the handling of errors from dmu_write_uio_dbuf() changed such that
EFAULT is not handled as a partial write.  For example, when appending
to a file, the z_size field of the znode is not updated after a partial
write resulting in EFAULT.

Restore the old handling of errors from dmu_write_uio_dbuf() to fix
this.  This should have no impact on Linux, which has special handling
for EFAULT already.

Reviewed-by: Andriy Gapon <avg@FreeBSD.org>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes openzfs#12964
nicman23 pushed a commit to nicman23/zfs that referenced this pull request Aug 22, 2022
FreeBSD's implementation of zfs_uio_fault_move() returns EFAULT when a
page fault occurs while copying data in or out of user buffers.  The VFS
treats such errors specially and will retry the I/O operation (which may
have made some partial progress).

When the FreeBSD and Linux implementations of zfs_write() were merged,
the handling of errors from dmu_write_uio_dbuf() changed such that
EFAULT is not handled as a partial write.  For example, when appending
to a file, the z_size field of the znode is not updated after a partial
write resulting in EFAULT.

Restore the old handling of errors from dmu_write_uio_dbuf() to fix
this.  This should have no impact on Linux, which has special handling
for EFAULT already.

Reviewed-by: Andriy Gapon <avg@FreeBSD.org>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes openzfs#12964
bsdjhb pushed a commit to CTSRD-CHERI/zfs that referenced this pull request Jul 13, 2023
This is not associated with a specific upstream commit but apparently
a local diff applied as part of:

commit e92ffd9b626833ebdbf2742c8ffddc6cd94b963e
Merge: 3c3df3660072 17b2ae0
Author: Martin Matuska <mm@FreeBSD.org>
Date:   Sat Jan 22 23:05:15 2022 +0100

    zfs: merge openzfs/zfs@17b2ae0b2 (master) into main

    Notable upstream pull request merges:
      openzfs#12766 Fix error propagation from lzc_send_redacted
      openzfs#12805 Updated the lz4 decompressor
      openzfs#12851 FreeBSD: Provide correct file generation number
      openzfs#12857 Verify dRAID empty sectors
      openzfs#12874 FreeBSD: Update argument types for VOP_READDIR
      openzfs#12896 Reduce number of arc_prune threads
      openzfs#12934 FreeBSD: Fix zvol_*_open() locking
      openzfs#12947 lz4: Cherrypick fix for CVE-2021-3520
      openzfs#12961 FreeBSD: Fix leaked strings in libspl mnttab
      openzfs#12964 Fix handling of errors from dmu_write_uio_dbuf() on FreeBSD
      openzfs#12981 Introduce a flag to skip comparing the local mac when raw sending
      openzfs#12985 Avoid memory allocations in the ARC eviction thread

    Obtained from:  OpenZFS
    OpenZFS commit: 17b2ae0
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Status: Accepted Ready to integrate (reviewed, tested)
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants