In btrfs_ref_tree_mod(), 'ref' and 'ra' are allocated through kzalloc() and
kmalloc(), respectively. In the following code, if an error occurs, the
execution will be redirected to 'out' or 'out_unlock' and the function will
be exited. However, on some of the paths, 'ref' and 'ra' are not
deallocated, leading to memory leaks. For example, if 'action' is
BTRFS_ADD_DELAYED_EXTENT, add_block_entry() will be invoked. If the return
value indicates an error, the execution will be redirected to 'out'. But,
'ref' is not deallocated on this path, causing a memory leak.
To fix the above issues, deallocate both 'ref' and 'ra' before exiting from
the function when an error is encountered.
CC: stable@vger.kernel.org # 4.15+
Signed-off-by: Wenwen Wang <wenwen@cs.uga.edu>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Pull dax fixes from Dan Williams:
"A fix for an xfstest failure and some and an update that removes an
fsdax dependency on block devices.
Summary:
- Fix RWF_NOWAIT writes to properly return -EAGAIN
- Clean up an unused helper
- Update dax_writeback_mapping_range to not need a block_device
argument"
* tag 'dax-fixes-5.6-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:
dax: pass NOWAIT flag to iomap_apply
dax: Get rid of fs_dax_get_by_host() helper
dax: Pass dax_dev instead of bdev to dax_writeback_mapping_range()
For the old mount API, the module parameters parseing function will
be called in ceph_mount() and also just after the default posix acl
flag set, so we can control to enable/disable it via the mount option.
But for the new mount API, it will call the module parameters
parseing function before ceph_get_tree(), so the posix acl will always
be enabled.
Fixes: 82995cc6c5 ("libceph, rbd, ceph: convert to use the new mount API")
Signed-off-by: Xiubo Li <xiubli@redhat.com>
Reviewed-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
syzbot reported that 4fbc0c711b ("ceph: remove the extra slashes in
the server path") had caused a regression where an allocation could be
done under a spinlock -- compare_mount_options() is called by sget_fc()
with sb_lock held.
We don't really need the supplied server path, so canonicalize it
in place and compare it directly. To make this work, the leading
slash is kept around and the logic in ceph_real_mount() to skip it
is restored. CEPH_MSG_CLIENT_SESSION now reports the same (i.e.
canonicalized) path, with the leading slash of course.
Fixes: 4fbc0c711b ("ceph: remove the extra slashes in the server path")
Reported-by: syzbot+98704a51af8e3d9425a9@syzkaller.appspotmail.com
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
In O_APPEND & O_DIRECT mode, the data from different writers will
be possibly overlapping each other since they take the shared lock.
For example, both Writer1 and Writer2 are in O_APPEND and O_DIRECT
mode:
Writer1 Writer2
shared_lock() shared_lock()
getattr(CAP_SIZE) getattr(CAP_SIZE)
iocb->ki_pos = EOF iocb->ki_pos = EOF
write(data1)
write(data2)
shared_unlock() shared_unlock()
The data2 will overlap the data1 from the same file offset, the
old EOF.
Switch to exclusive lock instead when O_APPEND is specified.
Signed-off-by: Xiubo Li <xiubli@redhat.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
I have an experimental setup where almost every possible system
service (even early startup ones) runs in separate namespace, using a
dedicated, minimal file system. In process of minimizing the contents
of the file systems with regards to modules and firmware files, I
noticed that in my system, the firmware files are loaded from three
different mount namespaces, those of systemd-udevd, init and
systemd-networkd. The logic of the source namespace is not very clear,
it seems to depend on the driver, but the namespace of the current
process is used.
So, this patch tries to make things a bit clearer and changes the
loading of firmware files only from the mount namespace of init. This
may also improve security, though I think that using firmware files as
attack vector could be too impractical anyway.
Later, it might make sense to make the mount namespace configurable,
for example with a new file in /proc/sys/kernel/firmware_config/. That
would allow a dedicated file system only for firmware files and those
need not be present anywhere else. This configurability would make
more sense if made also for kernel modules and /sbin/modprobe. Modules
are already loaded from init namespace (usermodehelper uses kthreadd
namespace) except when directly loaded by systemd-udevd.
Instead of using the mount namespace of the current process to load
firmware files, use the mount namespace of init process.
Link: https://lore.kernel.org/lkml/bb46ebae-4746-90d9-ec5b-fce4c9328c86@gmail.com/
Link: https://lore.kernel.org/lkml/0e3f7653-c59d-9341-9db2-c88f5b988c68@gmail.com/
Signed-off-by: Topi Miettinen <toiwoton@gmail.com>
Link: https://lore.kernel.org/r/20200123125839.37168-1-toiwoton@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We want to make sure that we revalidate the dentry if and only if
we've done an OPEN by filename.
In order to avoid races with remote changes to the directory on the
server, we want to save the verifier before calling OPEN. The exception
is if the server returned a delegation with our OPEN, as we then
know that the filename can't have changed on the server.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Reviewed-by: Benjamin Coddington <bcodding@gmail.com>
Tested-by: Benjamin Coddington <bcodding@gmail.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
In order to avoid having our dentry revalidation race with an update
of the directory on the server, we need to store the verifier before
the RPC calls to LOOKUP and READDIR.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Reviewed-by: Benjamin Coddington <bcodding@gmail.com>
Tested-by: Benjamin Coddington <bcodding@gmail.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Fix display for sec=krb5i which was wrongly interleaved by cruid,
resulting in string "sec=krb5,cruid=<...>i" instead of
"sec=krb5i,cruid=<...>".
Fixes: 96281b9e46 ("smb3: for kerberos mounts display the credential uid used")
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Before this patch function check_journal_clean was in ops_fstype.c.
This patch moves it to util.c so we can make use of it elsewhere
in a future patch.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
When a node fails, user space informs dlm of the node failure,
and dlm instructs gfs2 on the surviving nodes to perform journal
recovery. It does this by calling various callback functions in
lock_dlm.c. To mark its progress, it keeps generation numbers
and recover bits in a dlm "control" lock lvb, which is seen by
all nodes to determine which journals need to be replayed.
The gfs2 on all nodes get the same recovery requests from dlm,
so they all try to do the recovery, but only one will be
granted the exclusive lock on the journal. The others fail
with a "Busy" message on their "try lock."
However, when a node is withdrawn, it cannot safely do any
recovery or replay any journals. To make matters worse,
gfs2 might withdraw as a result of attempting recovery. For
example, this might happen if the device goes offline, or if
an hba fails. But in today's gfs2 code, it doesn't check for
being withdrawn at any step in the recovery process. What's
worse is that these callbacks from dlm have no return code,
so there is no way to indicate failure back to dlm. We can
send a "Recovery failed" uevent eventually, but that tells
user space what happened, not dlm's kernel code.
Before this patch, lock_dlm would perform its recovery steps but
ignore the result, and eventually it would still update its
generation number in the lvb, despite the fact that it may have
withdrawn or encountered an error. The other nodes would then
see the newer generation number in the lvb and conclude that
they don't need to do recovery because the generation number
is newer than the last one they saw. They think a different
node has already recovered the journal.
This patch adds checks to several of the callbacks used by dlm
in its recovery state machine so that the functions are ignored
and skipped if an io error has occurred or if the file system
is withdrawn. That prevents the lvb bits from being updated, and
therefore dlm and user space still see the need for recovery to
take place.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Before this patch, all io errors received by the quota daemon or the
logd daemon would cause a complaint message to be issued, such as:
gfs2: fsid=dm-13.0: Error 10 writing to journal, jid=0
This patch changes it so that the error message is only issued the
first time the error is encountered.
Also, before this patch function gfs2_end_log_write did not set the
sd_log_error value, so log errors would not cause the file system to
be withdrawn. This patch sets the error code so the file system is
properly withdrawn if an io error is encountered writing to the journal.
WARNING: This change in function breaks check xfstests generic/441
and causes it to fail: io errors writing to the log should cause a
file system to be withdrawn, and no further operations are tolerated.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Before this patch, gfs2 kept track of journal io errors in two
places sd_log_error and the SDF_AIL1_IO_ERROR flag in sd_flags.
This patch consolidates the two into sd_log_error so that it
reflects the first error encountered writing to the journal.
In future patches, we will take advantage of this by checking
this value rather than having to check both when reacting to
io errors.
In addition, this fixes a tight loop in unmount: If buffers
get on the ail1 list and an io error occurs elsewhere, the
ail1 list would never be cleared because they were always busy.
So unmount would hang, waiting for the ail1 list to empty.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Before this patch, the rgrp code had a serious problem related to
how it managed buffer_heads for resource groups. The problem caused
file system corruption, especially in cases of journal replay.
When an rgrp glock was demoted to transfer ownership to a
different cluster node, do_xmote() first calls rgrp_go_sync and then
rgrp_go_inval, as expected. When it calls rgrp_go_sync, that called
gfs2_rgrp_brelse() that dropped the buffer_head reference count.
In most cases, the reference count went to zero, which is right.
However, there were other places where the buffers are handled
differently.
After rgrp_go_sync, do_xmote called rgrp_go_inval which called
gfs2_rgrp_brelse a second time, then rgrp_go_inval's call to
truncate_inode_pages_range would get rid of the pages in memory,
but only if the reference count drops to 0.
Unfortunately, gfs2_rgrp_brelse was setting bi->bi_bh = NULL.
So when rgrp_go_sync called gfs2_rgrp_brelse, it lost the pointer
to the buffer_heads in cases where the reference count was still 1.
Therefore, when rgrp_go_inval called gfs2_rgrp_brelse a second time,
it failed the check for "if (bi->bi_bh)" and thus failed to call
brelse a second time. Because of that, the reference count on those
buffers sometimes failed to drop from 1 to 0. And that caused
function truncate_inode_pages_range to keep the pages in page cache
rather than freeing them.
The next time the rgrp glock was acquired, the metadata read of
the rgrp buffers re-used the pages in memory, which were now
wrong because they were likely modified by the other node who
acquired the glock in EX (which is why we demoted the glock).
This re-use of the page cache caused corruption because changes
made by the other nodes were never seen, so the bitmaps were
inaccurate.
For some reason, the problem became most apparent when journal
replay forced the replay of rgrps in memory, which caused newer
rgrp data to be overwritten by the older in-core pages.
A big part of the problem was that the rgrp buffer were released
in multiple places: The go_unlock function would release them when
the glock was released rather than when the glock is demoted,
which is clearly wrong because our intent was to cache them until
the glock is demoted from SH or EX.
This patch attempts to clean up the mess and make one consistent
and centralized mechanism for managing the rgrp buffer_heads by
implementing several changes:
1. It eliminates the call to gfs2_rgrp_brelse() from rgrp_go_sync.
We don't want to release the buffers or zero the pointers when
syncing for the reasons stated above. It only makes sense to
release them when the glock is actually invalidated (go_inval).
And when we do, then we set the bh pointers to NULL.
2. The go_unlock function (which was only used for rgrps) is
eliminated, as we've talked about doing many times before.
The go_unlock function was called too early in the glock dq
process, and should not happen until the glock is invalidated.
3. It also eliminates the call to rgrp_brelse in gfs2_clear_rgrpd.
That will now happen automatically when the rgrp glocks are
demoted, and shouldn't happen any sooner or later than that.
Instead, function gfs2_clear_rgrpd has been modified to demote
the rgrp glocks, and therefore, free those pages, before the
remaining glocks are culled by gfs2_gl_hash_clear. This
prevents the gl_object from hanging around when the glocks are
culled.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
This patch fixes a bug in which function gfs2_log_flush can get into
an infinite loop when a gfs2 file system is withdrawn. The problem
is the infinite loop "for (;;)" in gfs2_log_flush which would never
finish because the io error and subsequent withdraw prevented the
items from being taken off the ail list.
This patch tries to clean up the mess by allowing withdraw situations
to move not-in-flight buffer_heads to the ail2 list, where they will
be dealt with later.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
File system withdraws can be delayed when inconsistencies are
discovered when we cannot withdraw immediately, for example, when
critical spin_locks are held. But delaying the withdraw can cause
gfs2 to ignore the error and keep running for a short period of time.
For example, an rgrp glock may be dequeued and demoted while there
are still buffers that haven't been properly revoked, due to io
errors writing to the journal.
This patch introduces a new concept of a pending withdraw, which
means an inconsistency has been discovered and we need to withdraw
at the earliest possible opportunity. In these cases, we aren't
quite withdrawn yet, but we still need to not dequeue glocks and
other critical things. If we dequeue the glocks and the withdraw
results in our journal being replayed, the replay could overwrite
data that's been modified by a different node that acquired the
glock in the meantime.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
The gfs2_assert functions only print messages when the filesystem hasn't been
withdrawn yet, and they indicate whether or not they've printed something in
their return value. However, none of the callers use that information, so
simply return whether or not the assert has failed.
(The gfs2_assert functions are still backwards; they return false when an
assertion is true.)
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Change the various gfs2_consist functions to return void.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
These arguments are always passed as 0, and they are never evaluated.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
In gfs2_rgrp_verify and compute_bitstructs, make sure to report errors before
withdrawing the filesystem: otherwise, when we withdraw first and withdraw is
configured to panic, we'll never get to the error reporting.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Split gfs2_lm_withdraw into a function that prints an error message and a
function that withdraws the filesystem.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Pull more Kbuild updates from Masahiro Yamada:
- fix randconfig to generate a sane .config
- rename hostprogs-y / always to hostprogs / always-y, which are more
natual syntax.
- optimize scripts/kallsyms
- fix yes2modconfig and mod2yesconfig
- make multiple directory targets ('make foo/ bar/') work
* tag 'kbuild-v5.6-2' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kbuild: make multiple directory targets work
kconfig: Invalidate all symbols after changing to y or m.
kallsyms: fix type of kallsyms_token_table[]
scripts/kallsyms: change table to store (strcut sym_entry *)
scripts/kallsyms: rename local variables in read_symbol()
kbuild: rename hostprogs-y/always to hostprogs/always-y
kbuild: fix the document to use extra-y for vmlinux.lds
kconfig: fix broken dependency in randconfig-generated .config
Pull new zonefs file system from Damien Le Moal:
"Zonefs is a very simple file system exposing each zone of a zoned
block device as a file.
Unlike a regular file system with native zoned block device support
(e.g. f2fs or the on-going btrfs effort), zonefs does not hide the
sequential write constraint of zoned block devices to the user. As a
result, zonefs is not a POSIX compliant file system. Its goal is to
simplify the implementation of zoned block devices support in
applications by replacing raw block device file accesses with a richer
file based API, avoiding relying on direct block device file ioctls
which may be more obscure to developers.
One example of this approach is the implementation of LSM
(log-structured merge) tree structures (such as used in RocksDB and
LevelDB) on zoned block devices by allowing SSTables to be stored in a
zone file similarly to a regular file system rather than as a range of
sectors of a zoned device. The introduction of the higher level
construct "one file is one zone" can help reducing the amount of
changes needed in the application while at the same time allowing the
use of zoned block devices with various programming languages other
than C.
Zonefs IO management implementation uses the new iomap generic code.
Zonefs has been successfully tested using a functional test suite
(available with zonefs userland format tool on github) and a prototype
implementation of LevelDB on top of zonefs"
* tag 'zonefs-5.6-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/zonefs:
zonefs: Add documentation
fs: New zonefs file system
Pull cifs fixes from Steve French:
"13 cifs/smb3 patches, most from testing at the SMB3 plugfest this week:
- Important fix for multichannel and for modefromsid mounts.
- Two reconnect fixes
- Addition of SMB3 change notify support
- Backup tools fix
- A few additional minor debug improvements (tracepoints and
additional logging found useful during testing this week)"
* tag '5.6-rc-smb3-plugfest-patches' of git://git.samba.org/sfrench/cifs-2.6:
smb3: Add defines for new information level, FileIdInformation
smb3: print warning once if posix context returned on open
smb3: add one more dynamic tracepoint missing from strict fsync path
cifs: fix mode bits from dir listing when mounted with modefromsid
cifs: fix channel signing
cifs: add SMB3 change notification support
cifs: make multichannel warning more visible
cifs: fix soft mounts hanging in the reconnect code
cifs: Add tracepoints for errors on flush or fsync
cifs: log warning message (once) if out of disk space
cifs: fail i/o on soft mounts if sessionsetup errors out
smb3: fix problem with null cifs super block with previous patch
SMB3: Backup intent flag missing from some more ops
Pull vboxfs from Al Viro:
"This is the VirtualBox guest shared folder support by Hans de Goede,
with fixups for fs_parse folded in to avoid bisection hazards from
those API changes..."
* 'work.vboxsf' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
fs: Add VirtualBox guest shared folder (vboxsf) support
Jonas reports that he sometimes sees -97/-22 error returns from
sendmsg, if it gets punted async. This is due to not retaining the
sockaddr_storage between calls. Include that in the state we copy when
going async.
Cc: stable@vger.kernel.org # 5.3+
Reported-by: Jonas Bonn <jonas@norrbonn.se>
Tested-by: Jonas Bonn <jonas@norrbonn.se>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Normally we cancel all work we track, but for untracked work we could
leave the async worker behind until that work completes. This is totally
fine, but does leave resources pending after the task is gone until that
work completes.
Cancel work that this task queued up when it goes away.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a helper that allows the caller to cancel work based on what mm
it belongs to. This allows io_uring to cancel work from a given
task or thread when it exits.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We want to use the cancel functionality for canceling based on not
just the work itself. Instead of matching on the work address
manually, allow a match handler to tell us if we found the right work
item or not.
No functional changes in this patch.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
VirtualBox hosts can share folders with guests, this commit adds a
VFS driver implementing the Linux-guest side of this, allowing folders
exported by the host to be mounted under Linux.
This driver depends on the guest <-> host IPC functions exported by
the vboxguest driver.
Acked-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Pull compat-ioctl fix from Arnd Bergmann:
"One patch in the compat-ioctl series broke 32-bit rootfs for multiple
people testing on 64-bit kernels. Let's fix it in -rc1 before others
run into the same issue"
* tag 'compat-ioctl-fix' of git://git.kernel.org:/pub/scm/linux/kernel/git/arnd/playground:
compat_ioctl: fix FIONREAD on devices
Pull vfs file system parameter updates from Al Viro:
"Saner fs_parser.c guts and data structures. The system-wide registry
of syntax types (string/enum/int32/oct32/.../etc.) is gone and so is
the horror switch() in fs_parse() that would have to grow another case
every time something got added to that system-wide registry.
New syntax types can be added by filesystems easily now, and their
namespace is that of functions - not of system-wide enum members. IOW,
they can be shared or kept private and if some turn out to be widely
useful, we can make them common library helpers, etc., without having
to do anything whatsoever to fs_parse() itself.
And we already get that kind of requests - the thing that finally
pushed me into doing that was "oh, and let's add one for timeouts -
things like 15s or 2h". If some filesystem really wants that, let them
do it. Without somebody having to play gatekeeper for the variants
blessed by direct support in fs_parse(), TYVM.
Quite a bit of boilerplate is gone. And IMO the data structures make a
lot more sense now. -200LoC, while we are at it"
* 'merge.nfs-fs_parse.1' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (25 commits)
tmpfs: switch to use of invalfc()
cgroup1: switch to use of errorfc() et.al.
procfs: switch to use of invalfc()
hugetlbfs: switch to use of invalfc()
cramfs: switch to use of errofc() et.al.
gfs2: switch to use of errorfc() et.al.
fuse: switch to use errorfc() et.al.
ceph: use errorfc() and friends instead of spelling the prefix out
prefix-handling analogues of errorf() and friends
turn fs_param_is_... into functions
fs_parse: handle optional arguments sanely
fs_parse: fold fs_parameter_desc/fs_parameter_spec
fs_parser: remove fs_parameter_description name field
add prefix to fs_context->log
ceph_parse_param(), ceph_parse_mon_ips(): switch to passing fc_log
new primitive: __fs_parse()
switch rbd and libceph to p_log-based primitives
struct p_log, variants of warnf() et.al. taking that one instead
teach logfc() to handle prefices, give it saner calling conventions
get rid of cg_invalf()
...
Pull misc vfs updates from Al Viro:
- bmap series from cmaiolino
- getting rid of convolutions in copy_mount_options() (use a couple of
copy_from_user() instead of the __get_user() crap)
* 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
saner copy_mount_options()
fibmap: Reject negative block numbers
fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
ecryptfs: drop direct calls to ->bmap
cachefiles: drop direct usage of ->bmap method.
fs: Enable bmap() function to properly return errors
As in the previous patch, make openat*_prep() and statx_prep() handle
double preparation to avoid resource leakage.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Requests may be prepared multiple times with ->io allocated (i.e. async
prepared). Preparation functions don't handle it and forget about
previously allocated resources. This may happen in case of:
- spurious defer_check
- non-head (i.e. async prepared) request executed in sync (via nxt).
Make the handlers check, whether they already allocated resources, which
is true IFF REQ_F_NEED_CLEANUP is set.
Cc: stable@vger.kernel.org # 5.5
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
First, io_close() misses filp_close() and io_cqring_add_event(), when
f_op->flush is defined. That's because in this case it will
io_queue_async_work() itself not grabbing files, so the corresponding
chunk in io_close_finish() won't be executed.
Second, when submitted through io_wq_submit_work(), it will do
filp_close() and *_add_event() twice: first inline in io_close(),
and the second one in call to io_close_finish() from io_close().
The second one will also fire, because it was submitted async through
generic path, and so have grabbed files.
And the last nice thing is to remove this weird pilgrimage with checking
work/old_work and casting it to nxt. Just use a helper instead.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This passes it in to io-wq, so it assumes the right fs_struct when
executing async work that may need to do lookups.
Cc: stable@vger.kernel.org # 5.3+
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Some work items need this for relative path lookup, make it available
like the other inherited credentials/mm/etc.
Cc: stable@vger.kernel.org # 5.3+
Signed-off-by: Jens Axboe <axboe@kernel.dk>
For non-blocking issue, we set IOCB_NOWAIT in the kiocb. However, on a
raw block device, this yields an -EOPNOTSUPP return, as non-blocking
writes aren't supported. Turn this -EOPNOTSUPP into -EAGAIN, so we retry
from blocking context with IOCB_NOWAIT cleared.
Cc: stable@vger.kernel.org # 5.5
Signed-off-by: Jens Axboe <axboe@kernel.dk>
openat() and statx() may have allocated ->open.filename, which should be
be put. Add cleanup handlers for them.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Allocated iovec is freed only in io_{read,write,send,recv)(), and just
leaves it if an error occured. There are plenty of such cases:
- cancellation of non-head requests
- fail grabbing files in __io_queue_sqe()
- set REQ_F_NOWAIT and returning in __io_queue_sqe()
Add REQ_F_NEED_CLEANUP, which will force such requests with custom
allocated resourses go through cleanup handlers on put.
Cc: stable@vger.kernel.org # 5.5
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
In io_uring_poll() we must flush overflowed CQ events before to
check if there are CQ events available, to avoid missing events.
We call the io_cqring_events() that checks and flushes any overflow
and returns the number of CQ events available.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
All of these opcodes take a directory file descriptor. We can't easily
support fixed files for these operations, and the use case for that
probably isn't all that clear (or sensible) anyway.
Disable IOSQE_FIXED_FILE for these operations.
Reported-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Merge thundering herd avoidance on pipe IO.
This would have been applied for 5.5 already, but got delayed because of
a user-space race condition in the GNU make jobserver code. Now that
there's a new GNU make 4.3 release, and most distributions seem to have
at least applied the (almost three year old) fix for the problem, let's
see if people notice.
And it might have been just bad random timing luck on my machine.
If you do hit the race condition, things will still work, but the
symptom is that you don't get nearly the expected parallelism when using
"make -j<N>".
The jobserver bug can definitely happen without this patch too, but
seems to be easier to trigger when we no longer wake up pipe waiters
unnecessarily.
* pipe-exclusive-wakeup:
pipe: use exclusive waits when reading or writing
This makes the pipe code use separate wait-queues and exclusive waiting
for readers and writers, avoiding a nasty thundering herd problem when
there are lots of readers waiting for data on a pipe (or, less commonly,
lots of writers waiting for a pipe to have space).
While this isn't a common occurrence in the traditional "use a pipe as a
data transport" case, where you typically only have a single reader and
a single writer process, there is one common special case: using a pipe
as a source of "locking tokens" rather than for data communication.
In particular, the GNU make jobserver code ends up using a pipe as a way
to limit parallelism, where each job consumes a token by reading a byte
from the jobserver pipe, and releases the token by writing a byte back
to the pipe.
This pattern is fairly traditional on Unix, and works very well, but
will waste a lot of time waking up a lot of processes when only a single
reader needs to be woken up when a writer releases a new token.
A simplified test-case of just this pipe interaction is to create 64
processes, and then pass a single token around between them (this
test-case also intentionally passes another token that gets ignored to
test the "wake up next" logic too, in case anybody wonders about it):
#include <unistd.h>
int main(int argc, char **argv)
{
int fd[2], counters[2];
pipe(fd);
counters[0] = 0;
counters[1] = -1;
write(fd[1], counters, sizeof(counters));
/* 64 processes */
fork(); fork(); fork(); fork(); fork(); fork();
do {
int i;
read(fd[0], &i, sizeof(i));
if (i < 0)
continue;
counters[0] = i+1;
write(fd[1], counters, (1+(i & 1)) *sizeof(int));
} while (counters[0] < 1000000);
return 0;
}
and in a perfect world, passing that token around should only cause one
context switch per transfer, when the writer of a token causes a
directed wakeup of just a single reader.
But with the "writer wakes all readers" model we traditionally had, on
my test box the above case causes more than an order of magnitude more
scheduling: instead of the expected ~1M context switches, "perf stat"
shows
231,852.37 msec task-clock # 15.857 CPUs utilized
11,250,961 context-switches # 0.049 M/sec
616,304 cpu-migrations # 0.003 M/sec
1,648 page-faults # 0.007 K/sec
1,097,903,998,514 cycles # 4.735 GHz
120,781,778,352 instructions # 0.11 insn per cycle
27,997,056,043 branches # 120.754 M/sec
283,581,233 branch-misses # 1.01% of all branches
14.621273891 seconds time elapsed
0.018243000 seconds user
3.611468000 seconds sys
before this commit.
After this commit, I get
5,229.55 msec task-clock # 3.072 CPUs utilized
1,212,233 context-switches # 0.232 M/sec
103,951 cpu-migrations # 0.020 M/sec
1,328 page-faults # 0.254 K/sec
21,307,456,166 cycles # 4.074 GHz
12,947,819,999 instructions # 0.61 insn per cycle
2,881,985,678 branches # 551.096 M/sec
64,267,015 branch-misses # 2.23% of all branches
1.702148350 seconds time elapsed
0.004868000 seconds user
0.110786000 seconds sys
instead. Much better.
[ Note! This kernel improvement seems to be very good at triggering a
race condition in the make jobserver (in GNU make 4.2.1) for me. It's
a long known bug that was fixed back in June 2017 by GNU make commit
b552b0525198 ("[SV 51159] Use a non-blocking read with pselect to
avoid hangs.").
But there wasn't a new release of GNU make until 4.3 on Jan 19 2020,
so a number of distributions may still have the buggy version. Some
have backported the fix to their 4.2.1 release, though, and even
without the fix it's quite timing-dependent whether the bug actually
is hit. ]
Josh Triplett says:
"I've been hammering on your pipe fix patch (switching to exclusive
wait queues) for a month or so, on several different systems, and I've
run into no issues with it. The patch *substantially* improves
parallel build times on large (~100 CPU) systems, both with parallel
make and with other things that use make's pipe-based jobserver.
All current distributions (including stable and long-term stable
distributions) have versions of GNU make that no longer have the
jobserver bug"
Tested-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
My final cleanup patch for sys_compat_ioctl() introduced a regression on
the FIONREAD ioctl command, which is used for both regular and special
files, but only works on regular files after my patch, as I had missed
the warning that Al Viro put into a comment right above it.
Change it back so it can work on any file again by moving the implementation
to do_vfs_ioctl() instead.
Fixes: 77b9040195 ("compat_ioctl: simplify the implementation")
Reported-and-tested-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Reported-and-tested-by: youling257 <youling257@gmail.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
