Now that its two callers have their own version-specific instance of
this function, do_nfsd_create() is no longer used.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
The NFSv3 CREATE and NFSv4 OPEN(CREATE) use cases are about to
diverge such that it makes sense to split do_nfsd_create() into one
version for NFSv3 and one for NFSv4.
As a first step, copy do_nfsd_create() to nfs3proc.c and remove
NFSv4-specific logic.
One immediate legibility benefit is that the logic for handling
NFSv3 createhow is now quite straightforward. NFSv4 createhow
has some subtleties that IMO do not belong in generic code.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
I'd like to move do_nfsd_create() out of vfs.c. Therefore
nfsd_create_setattr() needs to be made publicly visible.
Note that both call sites in vfs.c commit both the new object and
its parent directory, so just combine those common metadata commits
into nfsd_create_setattr().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Clean up: The "out" label already invokes fh_drop_write().
Note that fh_drop_write() is already careful not to invoke
mnt_drop_write() if either it has already been done or there is
nothing to drop. Therefore no change in behavior is expected.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
As near as I can tell, mode bit masking and setting S_IFREG is
already done by do_nfsd_create() and vfs_create(). The NFSv4 path
(do_open_lookup), for example, does not bother with this special
processing.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Technically speaking, objects allocated out of a specific slab cache are
supposed to be freed to that slab cache. The popular slab backends will
take care of this for us, but SLOB famously doesn't. Fix this, even if
slob + xfs are not that common of a combination.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
When we're validating a recovered xattr log item during log recovery, we
should check the name before starting to allocate resources. This isn't
strictly necessary on its own, but it means that we won't bother with
huge memory allocations during recovery if the attr name is garbage,
which will simplify the changes in the next patch.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Make sure we screen the "attr flags" field of recovered xattr intent log
items to reject flag bits that we don't know about. This is really the
attr *filter* field from xfs_da_args, so rename the field and create
a mask to make checking for invalid bits easier.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Make sure we screen the op flags field of recovered xattr intent log
items to reject flag bits that we don't know about.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
If a setxattr operation finds an xattr structure in leaf format, adding
the attr can fail due to lack of space and hence requires an upgrade to
node format. After this happens, we'll roll the transaction and
re-enter the state machine, at which time we need to perform a second
lookup of the attribute name to find its new location. This lookup
attaches a new da state structure to the xfs_attr_item but doesn't free
the old one (from the leaf lookup) and leaks it. Fix that.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
kmemleak reported that we lost an xfs_da_state while removing xattrs in
generic/020:
unreferenced object 0xffff88801c0e4b40 (size 480):
comm "attr", pid 30515, jiffies 4294931061 (age 5.960s)
hex dump (first 32 bytes):
78 bc 65 07 00 c9 ff ff 00 30 60 1c 80 88 ff ff x.e......0`.....
02 00 00 00 00 00 00 00 80 18 83 4e 80 88 ff ff ...........N....
backtrace:
[<ffffffffa023ef4a>] xfs_da_state_alloc+0x1a/0x30 [xfs]
[<ffffffffa021b6f3>] xfs_attr_node_hasname+0x23/0x90 [xfs]
[<ffffffffa021c6f1>] xfs_attr_set_iter+0x441/0xa30 [xfs]
[<ffffffffa02b5104>] xfs_xattri_finish_update+0x44/0x80 [xfs]
[<ffffffffa02b515e>] xfs_attr_finish_item+0x1e/0x40 [xfs]
[<ffffffffa0244744>] xfs_defer_finish_noroll+0x184/0x740 [xfs]
[<ffffffffa02a6473>] __xfs_trans_commit+0x153/0x3e0 [xfs]
[<ffffffffa021d149>] xfs_attr_set+0x469/0x7e0 [xfs]
[<ffffffffa02a78d9>] xfs_xattr_set+0x89/0xd0 [xfs]
[<ffffffff812e6512>] __vfs_removexattr+0x52/0x70
[<ffffffff812e6a08>] __vfs_removexattr_locked+0xb8/0x150
[<ffffffff812e6af6>] vfs_removexattr+0x56/0x100
[<ffffffff812e6bf8>] removexattr+0x58/0x90
[<ffffffff812e6cce>] path_removexattr+0x9e/0xc0
[<ffffffff812e6d44>] __x64_sys_lremovexattr+0x14/0x20
[<ffffffff81786b35>] do_syscall_64+0x35/0x80
I think this is a consequence of xfs_attr_node_removename_setup
attaching a new da(btree) state to xfs_attr_item and never freeing it.
I /think/ it's the case that the remove paths could detach the da state
earlier in the remove state machine since nothing else accesses the
state. However, let's future-proof the new xattr code by adding a
catch-all when we free the xfs_attr_item to make sure we never leak the
da state.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Combination of LOOKUP_IS_SCOPED and NULL nd->root.mnt is impossible
after successful path_init(). All places where ->root.mnt might
become NULL do that only if LOOKUP_IS_SCOPED is not there and
path_init() itself can return success without setting nd->root
only if ND_ROOT_PRESET had been set (in which case nd->root
had been set by caller and never changed) or if the name had
been a relative one *and* none of the bits in LOOKUP_IS_SCOPED
had been present.
Since all calls of legitimize_root() must be downstream of successful
path_init(), the check for !nd->root.mnt && (nd->flags & LOOKUP_IS_SCOPED)
is pure paranoia.
FWIW, it had been discussed (and agreed upon) with Aleksa back when
scoped lookups had been merged; looks like that had fallen through the
cracks back then.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
!foo() != 0 is a strange way to spell !foo(); fallout from
"fs: make unlazy_walk() error handling consistent"...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
In f2fs_read_inline_data(), it is confused with checking of
inline_data flag, as we checked it before calling. So this
patch add some comments for f2fs_has_inline_data().
Signed-off-by: Chao Liu <liuchao@coolpad.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
The variable idx is assigned a value and is never read. The variable is
not used and is redundant, remove it.
Cleans up clang scan build warning:
warning: Although the value stored to 'idx' is used in the enclosing
expression, the value is never actually read from 'idx'
[deadcode.DeadStores]
Link: https://lkml.kernel.org/r/20220517093646.93628-2-colin.i.king@gmail.com
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Reviewed-by: Anton Altaparmakov <anton@tuxera.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
All the timestamps in vfat_create() and vfat_mkdir() come from
fat_time_fat2unix() which ensures time granularity. We don't need to
truncate them to fit FAT's format.
Moreover, fat_truncate_crtime() and fat_timespec64_trunc_10ms() are also
removed because there is no caller anymore.
Link: https://lkml.kernel.org/r/20220503152536.2503003-4-cccheng@synology.com
Signed-off-by: Chung-Chiang Cheng <cccheng@synology.com>
Acked-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
creation time is no longer mixed with change time. Add an in-memory field
for it, and report it in statx if supported.
Link: https://lkml.kernel.org/r/20220503152536.2503003-3-cccheng@synology.com
Signed-off-by: Chung-Chiang Cheng <cccheng@synology.com>
Acked-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
FAT supports creation time but not change time, and there was no
corresponding timestamp for creation time in previous VFS. The original
implementation took the compromise of saving the in-memory change time
into the on-disk creation time field, but this would lead to compatibility
issues with non-linux systems.
To address this issue, this patch changes the behavior of ctime. It will
no longer be loaded and stored from the creation time on disk. Instead of
that, it'll be consistent with the in-memory mtime and share the same
on-disk field. All updates to mtime will also be applied to ctime in
memory, while all updates to ctime will be ignored.
Link: https://lkml.kernel.org/r/20220503152536.2503003-2-cccheng@synology.com
Signed-off-by: Chung-Chiang Cheng <cccheng@synology.com>
Acked-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Separate fat_truncate_time() to each timestamps for later creation time
work.
This patch does not introduce any functional changes, it's merely
refactoring change.
Link: https://lkml.kernel.org/r/20220503152536.2503003-1-cccheng@synology.com
Signed-off-by: Chung-Chiang Cheng <cccheng@synology.com>
Acked-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Currently it requires poking at debugfs to figure out the size and
population of the zswap cache on a host. There are no counters for reads
and writes against the cache. As a result, it's difficult to understand
zswap behavior on production systems.
Print zswap memory consumption and how many pages are zswapped out in
/proc/meminfo. Count zswapouts and zswapins in /proc/vmstat.
Link: https://lkml.kernel.org/r/20220510152847.230957-6-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: David Hildenbrand <david@redhat.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Previously the protection of kernfs_pr_cont_buf was piggy backed by
rename_lock, which means that pr_cont() needs to be protected under
rename_lock. This can cause potential circular lock dependencies.
If there is an OOM, we have the following call hierarchy:
-> cpuset_print_current_mems_allowed()
-> pr_cont_cgroup_name()
-> pr_cont_kernfs_name()
pr_cont_kernfs_name() will grab rename_lock and call printk. So we have
the following lock dependencies:
kernfs_rename_lock -> console_sem
Sometimes, printk does a wakeup before releasing console_sem, which has
the dependence chain:
console_sem -> p->pi_lock -> rq->lock
Now, imagine one wants to read cgroup_name under rq->lock, for example,
printing cgroup_name in a tracepoint in the scheduler code. They will
be holding rq->lock and take rename_lock:
rq->lock -> kernfs_rename_lock
Now they will deadlock.
A prevention to this circular lock dependency is to separate the
protection of pr_cont_buf from rename_lock. In principle, rename_lock
is to protect the integrity of cgroup name when copying to buf. Once
pr_cont_buf has got its content, rename_lock can be dropped. So it's
safe to drop rename_lock after kernfs_name_locked (and
kernfs_path_from_node_locked) and rely on a dedicated pr_cont_lock
to protect pr_cont_buf.
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Hao Luo <haoluo@google.com>
Link: https://lore.kernel.org/r/20220516190951.3144144-1-haoluo@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Update client_info_show to show state of courtesy client
and seconds since last renew.
Reviewed-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
This patch allows expired client with lock state to be in COURTESY
state. Lock conflict with COURTESY client is resolved by the fs/lock
code using the lm_lock_expirable and lm_expire_lock callback in the
struct lock_manager_operations.
If conflict client is in COURTESY state, set it to EXPIRABLE and
schedule the laundromat to run immediately to expire the client. The
callback lm_expire_lock waits for the laundromat to flush its work
queue before returning to caller.
Reviewed-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Add 2 new callbacks, lm_lock_expirable and lm_expire_lock, to
lock_manager_operations to allow the lock manager to take appropriate
action to resolve the lock conflict if possible.
A new field, lm_mod_owner, is also added to lock_manager_operations.
The lm_mod_owner is used by the fs/lock code to make sure the lock
manager module such as nfsd, is not freed while lock conflict is being
resolved.
lm_lock_expirable checks and returns true to indicate that the lock
conflict can be resolved else return false. This callback must be
called with the flc_lock held so it can not block.
lm_expire_lock is called to resolve the lock conflict if the returned
value from lm_lock_expirable is true. This callback is called without
the flc_lock held since it's allowed to block. Upon returning from
this callback, the lock conflict should be resolved and the caller is
expected to restart the conflict check from the beginnning of the list.
Lock manager, such as NFSv4 courteous server, uses this callback to
resolve conflict by destroying lock owner, or the NFSv4 courtesy client
(client that has expired but allowed to maintains its states) that owns
the lock.
Reviewed-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Add helper locks_owner_has_blockers to check if there is any blockers
for a given lockowner.
Reviewed-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
This patch moves create/destroy of laundry_wq from nfs4_state_start
and nfs4_state_shutdown_net to init_nfsd and exit_nfsd to prevent
the laundromat from being freed while a thread is processing a
conflicting lock.
Reviewed-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
This patch allows expired client with open state to be in COURTESY
state. Share/access conflict with COURTESY client is resolved by
setting COURTESY client to EXPIRABLE state, schedule laundromat
to run and returning nfserr_jukebox to the request client.
Reviewed-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
This patch provides courteous server support for delegation only.
Only expired client with delegation but no conflict and no open
or lock state is allowed to be in COURTESY state.
Delegation conflict with COURTESY/EXPIRABLE client is resolved by
setting it to EXPIRABLE, queue work for the laundromat and return
delay to the caller. Conflict is resolved when the laudromat runs
and expires the EXIRABLE client while the NFS client retries the
OPEN request. Local thread request that gets conflict is doing the
retry in _break_lease.
Client in COURTESY or EXPIRABLE state is allowed to reconnect and
continues to have access to its state. Access to the nfs4_client by
the reconnecting thread and the laundromat is serialized via the
client_lock.
Reviewed-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
nfsd_splice_actor() checks that the page being spliced does not
match the previous element in the svc_rqst::rq_pages array. We
believe this is to prevent a double put_page() in cases where the
READ payload is partially contained in the xdr_buf's head buffer.
However, the NFSD READ proc functions no longer place any part of
the READ payload in the head buffer, in order to properly support
NFS/RDMA READ with Write chunks. Therefore, simplify the logic in
nfsd_splice_actor() to remove this unnecessary check.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Yanming reported a kernel bug in Bugzilla kernel [1], which can be
reproduced. The bug message is:
The kernel message is shown below:
kernel BUG at fs/inode.c:611!
Call Trace:
evict+0x282/0x4e0
__dentry_kill+0x2b2/0x4d0
dput+0x2dd/0x720
do_renameat2+0x596/0x970
__x64_sys_rename+0x78/0x90
do_syscall_64+0x3b/0x90
[1] https://bugzilla.kernel.org/show_bug.cgi?id=215895
The bug is due to fuzzed inode has both inline_data and encrypted flags.
During f2fs_evict_inode(), as the inode was deleted by rename(), it
will cause inline data conversion due to conflicting flags. The page
cache will be polluted and the panic will be triggered in clear_inode().
Try fixing the bug by doing more sanity checks for inline data inode in
sanity_check_inode().
Cc: stable@vger.kernel.org
Reported-by: Ming Yan <yanming@tju.edu.cn>
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This patch tries to fix permission consistency issue as all other
mainline filesystems.
Since the initial introduction of (posix) fallocate back at the turn of
the century, it has been possible to use this syscall to change the
user-visible contents of files. This can happen by extending the file
size during a preallocation, or through any of the newer modes (punch,
zero, collapse, insert range). Because the call can be used to change
file contents, we should treat it like we do any other modification to a
file -- update the mtime, and drop set[ug]id privileges/capabilities.
The VFS function file_modified() does all this for us if pass it a
locked inode, so let's make fallocate drop permissions correctly.
Cc: stable@kernel.org
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
It's nonsensical to register a provided buffer ring, if a classic
provided buffer group with the same ID exists. Depending on the order of
which we decide what type to pick, the other type will never get used.
Explicitly disallow it and return an error if this is attempted.
Fixes: c7fb19428d ("io_uring: add support for ring mapped supplied buffers")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We use ->buf_pages != 0 to tell if this is a shared buffer ring or a
classic provided buffer group. If we unregister the shared ring and
then attempt to use it, buf_pages is zero yet the classic list head
isn't properly initialized. This causes io_buffer_select() to think
that we have classic buffers available, but then we crash when we try
and get one from the list.
Just initialize the list if we unregister a shared buffer ring, leaving
it in a sane state for either re-registration or for attempting to use
it. And do the same for the initial setup from the classic path.
Fixes: c7fb19428d ("io_uring: add support for ring mapped supplied buffers")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The parameter desc_size in fsverity_create_info() is useless and it is
not referenced anywhere. The greatest meaning of desc_size here is to
indecate the size of struct fsverity_descriptor and futher calculate the
size of signature. However, the desc->sig_size can do it also and it is
indeed, so remove it.
Therefore, it is no need to acquire desc_size by fsverity_get_descriptor()
in ensure_verity_info(), so remove the parameter desc_ret in
fsverity_get_descriptor() too.
Signed-off-by: Zhang Jianhua <chris.zjh@huawei.com>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Link: https://lore.kernel.org/r/20220518132256.2297655-1-chris.zjh@huawei.com
If processing the on-disk mount options fails after any memory was
allocated in the ext4_fs_context, e.g. s_qf_names, then this memory is
leaked. Fix this by calling ext4_fc_free() instead of kfree() directly.
Reproducer:
mkfs.ext4 -F /dev/vdc
tune2fs /dev/vdc -E mount_opts=usrjquota=file
echo clear > /sys/kernel/debug/kmemleak
mount /dev/vdc /vdc
echo scan > /sys/kernel/debug/kmemleak
sleep 5
echo scan > /sys/kernel/debug/kmemleak
cat /sys/kernel/debug/kmemleak
Fixes: 7edfd85b1f ("ext4: Completely separate options parsing and sb setup")
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers <ebiggers@google.com>
Tested-by: Ritesh Harjani <ritesh.list@gmail.com>
Link: https://lore.kernel.org/r/20220513231605.175121-2-ebiggers@kernel.org
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The 'commit' option is only applicable for ext3 and ext4 filesystems,
and has never been accepted by the ext2 filesystem driver, so the ext4
driver shouldn't allow it on ext2 filesystems.
This fixes a failure in xfstest ext4/053.
Fixes: 8dc0aa8cf0 ("ext4: check incompatible mount options while mounting ext2/3")
Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Ritesh Harjani <ritesh.list@gmail.com>
Reviewed-by: Lukas Czerner <lczerner@redhat.com>
Link: https://lore.kernel.org/r/20220510183232.172615-1-ebiggers@kernel.org
Fix following includecheck warning:
./fs/ext4/inode.c: linux/dax.h is included more than once.
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220504225025.44753-1-yang.lee@linux.alibaba.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
When I wrote the freevxfs driver I had some odd choice of licensing
statements, the options are either GPL (without version) or an odd
BSD-ish licensense with advertising clause.
The GPL vs always meant to be the same as the kernel, that is version
2 only, and the odd BSD-ish license doesn't make much sense. Add
a GPL2.0-only SPDX tag to make the GPL intentions clear and drop the
bogus BSD license.
Acked-by: Krzysztof Błaszkowski <kb@sysmikro.com.pl>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The logic for handling events on child in groups that have a mark on
the parent inode, but without FS_EVENT_ON_CHILD flag in the mask is
duplicated in several places and inconsistent.
Move the logic into the preparation of mark type iterator, so that the
parent mark type will be excluded from all mark type iterations in that
case.
This results in several subtle changes of behavior, hopefully all
desired changes of behavior, for example:
- Group A has a mount mark with FS_MODIFY in mask
- Group A has a mark with ignore mask that does not survive FS_MODIFY
and does not watch children on directory D.
- Group B has a mark with FS_MODIFY in mask that does watch children
on directory D.
- FS_MODIFY event on file D/foo should not clear the ignore mask of
group A, but before this change it does
And if group A ignore mask was set to survive FS_MODIFY:
- FS_MODIFY event on file D/foo should be reported to group A on account
of the mount mark, but before this change it is wrongly ignored
Fixes: 2f02fd3fa1 ("fanotify: fix ignore mask logic for events on child and on dir")
Reported-by: Jan Kara <jack@suse.com>
Link: https://lore.kernel.org/linux-fsdevel/20220314113337.j7slrb5srxukztje@quack3.lan/
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20220511190213.831646-3-amir73il@gmail.com
fsnotify_foreach_iter_mark_type() is used to reduce boilerplate code
of iterating all marks of a specific group interested in an event
by consulting the iterator report_mask.
Use an open coded version of that iterator in fsnotify_iter_next()
that collects all marks of the current iteration group without
consulting the iterator report_mask.
At the moment, the two iterator variants are the same, but this
decoupling will allow us to exclude some of the group's marks from
reporting the event, for example for event on child and inode marks
on parent did not request to watch events on children.
Fixes: 2f02fd3fa1 ("fanotify: fix ignore mask logic for events on child and on dir")
Reported-by: Jan Kara <jack@suse.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20220511190213.831646-2-amir73il@gmail.com
POLL* are unannotated values for the userspace ABI, while everything
in-kernel should use EPOLL* and the __poll_t type.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220518084005.3255380-6-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_file_get_normal isn't marked inline, so don't claim it as such in the
forward declaration.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220518084005.3255380-4-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
ERR_PTR abuses the high bits of a pointer to transport error information.
This is only safe for kernel pointers and not user pointers. Fix
io_buffer_select and its helpers to just return NULL for failure and get
rid of this abuse.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220518084005.3255380-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Provided buffers allow an application to supply io_uring with buffers
that can then be grabbed for a read/receive request, when the data
source is ready to deliver data. The existing scheme relies on using
IORING_OP_PROVIDE_BUFFERS to do that, but it can be difficult to use
in real world applications. It's pretty efficient if the application
is able to supply back batches of provided buffers when they have been
consumed and the application is ready to recycle them, but if
fragmentation occurs in the buffer space, it can become difficult to
supply enough buffers at the time. This hurts efficiency.
Add a register op, IORING_REGISTER_PBUF_RING, which allows an application
to setup a shared queue for each buffer group of provided buffers. The
application can then supply buffers simply by adding them to this ring,
and the kernel can consume then just as easily. The ring shares the head
with the application, the tail remains private in the kernel.
Provided buffers setup with IORING_REGISTER_PBUF_RING cannot use
IORING_OP_{PROVIDE,REMOVE}_BUFFERS for adding or removing entries to the
ring, they must use the mapped ring. Mapped provided buffer rings can
co-exist with normal provided buffers, just not within the same group ID.
To gauge overhead of the existing scheme and evaluate the mapped ring
approach, a simple NOP benchmark was written. It uses a ring of 128
entries, and submits/completes 32 at the time. 'Replenish' is how
many buffers are provided back at the time after they have been
consumed:
Test Replenish NOPs/sec
================================================================
No provided buffers NA ~30M
Provided buffers 32 ~16M
Provided buffers 1 ~10M
Ring buffers 32 ~27M
Ring buffers 1 ~27M
The ring mapped buffers perform almost as well as not using provided
buffers at all, and they don't care if you provided 1 or more back at
the same time. This means application can just replenish as they go,
rather than need to batch and compact, further reducing overhead in the
application. The NOP benchmark above doesn't need to do any compaction,
so that overhead isn't even reflected in the above test.
Co-developed-by: Dylan Yudaken <dylany@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Abstract this out from io_sqe_buffer_register() so we can use it
elsewhere too without duplicating this code.
No intended functional changes in this patch.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Obviously not really useful since it's not transferring data, but it
is helpful in benchmarking overhead of provided buffers.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_provided_buffer_select() must drop the submit lock, if needed, even
in the error handling case. Failure to do so will leave us with the
ctx->uring_lock held, causing spew like:
====================================
WARNING: iou-wrk-366/368 still has locks held!
5.18.0-rc6-00294-gdf8dc7004331 #994 Not tainted
------------------------------------
1 lock held by iou-wrk-366/368:
#0: ffff0000c72598a8 (&ctx->uring_lock){+.+.}-{3:3}, at: io_ring_submit_lock+0x20/0x48
stack backtrace:
CPU: 4 PID: 368 Comm: iou-wrk-366 Not tainted 5.18.0-rc6-00294-gdf8dc7004331 #994
Hardware name: linux,dummy-virt (DT)
Call trace:
dump_backtrace.part.0+0xa4/0xd4
show_stack+0x14/0x5c
dump_stack_lvl+0x88/0xb0
dump_stack+0x14/0x2c
debug_check_no_locks_held+0x84/0x90
try_to_freeze.isra.0+0x18/0x44
get_signal+0x94/0x6ec
io_wqe_worker+0x1d8/0x2b4
ret_from_fork+0x10/0x20
and triggering later hangs off get_signal() because we attempt to
re-grab the lock.
Reported-by: syzbot+987d7bb19195ae45208c@syzkaller.appspotmail.com
Fixes: 149c69b04a ("io_uring: abstract out provided buffer list selection")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pass updated i_size in fscache_unuse_cookie() when called
from nfs_fscache_release_file(), which ensures the size of
an fscache object gets written to the cache storage. Failing
to do so results in unnessary reads from the NFS server, even
when the data is cached, due to a cachefiles object coherency
check failing with a trace similar to the following:
cachefiles_coherency: o=0000000e BAD osiz B=afbb3 c=0
This problem can be reproduced as follows:
#!/bin/bash
v=4.2; NFS_SERVER=127.0.0.1
set -e; trap cleanup EXIT; rc=1
function cleanup {
umount /mnt/nfs > /dev/null 2>&1
RC_STR="TEST PASS"
[ $rc -eq 1 ] && RC_STR="TEST FAIL"
echo "$RC_STR on $(uname -r) with NFSv$v and server $NFS_SERVER"
}
mount -o vers=$v,fsc $NFS_SERVER:/export /mnt/nfs
rm -f /mnt/nfs/file1.bin > /dev/null 2>&1
dd if=/dev/zero of=/mnt/nfs/file1.bin bs=4096 count=1 > /dev/null 2>&1
echo 3 > /proc/sys/vm/drop_caches
echo Read file 1st time from NFS server into fscache
dd if=/mnt/nfs/file1.bin of=/dev/null > /dev/null 2>&1
umount /mnt/nfs && mount -o vers=$v,fsc $NFS_SERVER:/export /mnt/nfs
echo 3 > /proc/sys/vm/drop_caches
echo Read file 2nd time from fscache
dd if=/mnt/nfs/file1.bin of=/dev/null > /dev/null 2>&1
echo Check mountstats for NFS read
grep -q "READ: 0" /proc/self/mountstats # (1st number) == 0
[ $? -eq 0 ] && rc=0
Fixes: a6b5a28eb5 "nfs: Convert to new fscache volume/cookie API"
Signed-off-by: Dave Wysochanski <dwysocha@redhat.com>
Tested-by: Daire Byrne <daire@dneg.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
NFSv4 can lose locks if, for example there is a network partition for
longer than the lease period. When this happens a warning message
NFS: __nfs4_reclaim_open_state: Lock reclaim failed!
is generated, possibly once for each lock (though rate limited).
This is potentially misleading as is can be read as suggesting that lock
reclaim was attempted. However the default behaviour is to not attempt
to recover locks (except due to server report).
This patch changes the reporting to produce at most one message for each
attempt to recover all state from a given server. The message reports
the server name and the number of locks lost if that number is non-zero.
It reports that locks were lost and give no suggestion as to whether
there was an attempt or not.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
There is a race condition in smb2_compound_op:
after_close:
num_rqst++;
if (cfile) {
cifsFileInfo_put(cfile); // sends SMB2_CLOSE to the server
cfile = NULL;
This is triggered by smb2_query_path_info operation that happens during
revalidate_dentry. In smb2_query_path_info, get_readable_path is called to
load the cfile, increasing the reference counter. If in the meantime, this
reference becomes the very last, this call to cifsFileInfo_put(cfile) will
trigger a SMB2_CLOSE request sent to the server just before sending this compound
request – and so then the compound request fails either with EBADF/EIO depending
on the timing at the server, because the handle is already closed.
In the first scenario, the race seems to be happening between smb2_query_path_info
triggered by the rename operation, and between “cleanup” of asynchronous writes – while
fsync(fd) likely waits for the asynchronous writes to complete, releasing the writeback
structures can happen after the close(fd) call. So the EBADF/EIO errors will pop up if
the timing is such that:
1) There are still outstanding references after close(fd) in the writeback structures
2) smb2_query_path_info successfully fetches the cfile, increasing the refcounter by 1
3) All writeback structures release the same cfile, reducing refcounter to 1
4) smb2_compound_op is called with that cfile
In the second scenario, the race seems to be similar – here open triggers the
smb2_query_path_info operation, and if all other threads in the meantime decrease the
refcounter to 1 similarly to the first scenario, again SMB2_CLOSE will be sent to the
server just before issuing the compound request. This case is harder to reproduce.
See https://bugzilla.samba.org/show_bug.cgi?id=15051
Cc: stable@vger.kernel.org
Fixes: 8de9e86c67 ("cifs: create a helper to find a writeable handle by path name")
Signed-off-by: Ondrej Hubsch <ohubsch@purestorage.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: Steve French <stfrench@microsoft.com>
We gate whether to IOPOLL for a request on whether the opcode is allowed
on a ring setup for IOPOLL and if it's got a file assigned. MSG_RING
is the only one that allows a file yet isn't pollable, it's merely
supported to allow communication on an IOPOLL ring, not because we can
poll for completion of it.
Put the assigned file early and clear it, so we don't attempt to poll
for it.
Reported-by: syzbot+1a0a53300ce782f8b3ad@syzkaller.appspotmail.com
Fixes: 3f1d52abf0 ("io_uring: defer msg-ring file validity check until command issue")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Tryng to rename a directory that has all following properties fails with
EINVAL and triggers the 'WARN_ON_ONCE(!fscrypt_has_encryption_key(dir))'
in f2fs_match_ci_name():
- The directory is casefolded
- The directory is encrypted
- The directory's encryption key is not yet set up
- The parent directory is *not* encrypted
The problem is incorrect handling of the lookup of ".." to get the
parent reference to update. fscrypt_setup_filename() treats ".." (and
".") specially, as it's never encrypted. It's passed through as-is, and
setting up the directory's key is not attempted. As the name isn't a
no-key name, f2fs treats it as a "normal" name and attempts a casefolded
comparison. That breaks the assumption of the WARN_ON_ONCE() in
f2fs_match_ci_name() which assumes that for encrypted directories,
casefolded comparisons only happen when the directory's key is set up.
We could just remove this WARN_ON_ONCE(). However, since casefolding is
always a no-op on "." and ".." anyway, let's instead just not casefold
these names. This results in the standard bytewise comparison.
Fixes: 7ad08a58bf ("f2fs: Handle casefolding with Encryption")
Cc: <stable@vger.kernel.org> # v5.11+
Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Gabriel Krisman Bertazi <krisman@collabora.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
The f2fs_gc uses a bitmap to indicate pinned sections, but when disabling
chckpoint, we call f2fs_gc() with NULL_SEGNO which selects the same dirty
segment as a victim all the time, resulting in checkpoint=disable failure,
for example. Let's pick another one, if we fail to collect it.
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Hulk Robot reported a BUG_ON:
==================================================================
EXT4-fs error (device loop3): ext4_mb_generate_buddy:805: group 0,
block bitmap and bg descriptor inconsistent: 25 vs 31513 free clusters
kernel BUG at fs/ext4/ext4_jbd2.c:53!
invalid opcode: 0000 [#1] SMP KASAN PTI
CPU: 0 PID: 25371 Comm: syz-executor.3 Not tainted 5.10.0+ #1
RIP: 0010:ext4_put_nojournal fs/ext4/ext4_jbd2.c:53 [inline]
RIP: 0010:__ext4_journal_stop+0x10e/0x110 fs/ext4/ext4_jbd2.c:116
[...]
Call Trace:
ext4_write_inline_data_end+0x59a/0x730 fs/ext4/inline.c:795
generic_perform_write+0x279/0x3c0 mm/filemap.c:3344
ext4_buffered_write_iter+0x2e3/0x3d0 fs/ext4/file.c:270
ext4_file_write_iter+0x30a/0x11c0 fs/ext4/file.c:520
do_iter_readv_writev+0x339/0x3c0 fs/read_write.c:732
do_iter_write+0x107/0x430 fs/read_write.c:861
vfs_writev fs/read_write.c:934 [inline]
do_pwritev+0x1e5/0x380 fs/read_write.c:1031
[...]
==================================================================
Above issue may happen as follows:
cpu1 cpu2
__________________________|__________________________
do_pwritev
vfs_writev
do_iter_write
ext4_file_write_iter
ext4_buffered_write_iter
generic_perform_write
ext4_da_write_begin
vfs_fallocate
ext4_fallocate
ext4_convert_inline_data
ext4_convert_inline_data_nolock
ext4_destroy_inline_data_nolock
clear EXT4_STATE_MAY_INLINE_DATA
ext4_map_blocks
ext4_ext_map_blocks
ext4_mb_new_blocks
ext4_mb_regular_allocator
ext4_mb_good_group_nolock
ext4_mb_init_group
ext4_mb_init_cache
ext4_mb_generate_buddy --> error
ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)
ext4_restore_inline_data
set EXT4_STATE_MAY_INLINE_DATA
ext4_block_write_begin
ext4_da_write_end
ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)
ext4_write_inline_data_end
handle=NULL
ext4_journal_stop(handle)
__ext4_journal_stop
ext4_put_nojournal(handle)
ref_cnt = (unsigned long)handle
BUG_ON(ref_cnt == 0) ---> BUG_ON
The lock held by ext4_convert_inline_data is xattr_sem, but the lock
held by generic_perform_write is i_rwsem. Therefore, the two locks can
be concurrent.
To solve above issue, we add inode_lock() for ext4_convert_inline_data().
At the same time, move ext4_convert_inline_data() in front of
ext4_punch_hole(), remove similar handling from ext4_punch_hole().
Fixes: 0c8d414f16 ("ext4: let fallocate handle inline data correctly")
Cc: stable@vger.kernel.org
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Baokun Li <libaokun1@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20220428134031.4153381-1-libaokun1@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Symlink's external data block is one kind of metadata block, and now
that almost all ext4 metadata block's page cache (e.g. directory blocks,
quota blocks...) belongs to bdev backing inode except the symlink. It
is essentially worked in data=journal mode like other regular file's
data block because probably in order to make it simple for generic VFS
code handling symlinks or some other historical reasons, but the logic
of creating external data block in ext4_symlink() is complicated. and it
also make things confused if user do not want to let the filesystem
worked in data=journal mode. This patch convert the final exceptional
case and make things clean, move the mapping of the symlink's external
data block to bdev like any other metadata block does.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Link: https://lore.kernel.org/r/20220424140936.1898920-3-yi.zhang@huawei.com
Current ext4_getblk() might sleep if some resources are not valid or
could be race with a concurrent extents modifing procedure. So we
cannot call ext4_getblk() and ext4_map_blocks() to get map blocks in
the atomic context in some fast path (e.g. the upcoming procedure of
getting symlink external block in the RCU context), even if the map
extents have already been check and cached.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Link: https://lore.kernel.org/r/20220424140936.1898920-2-yi.zhang@huawei.com
In __ext4_super() we always overwrote the user specified journal_ioprio
value with a default value, expecting parse_apply_sb_mount_options() to
later correctly set ctx->journal_ioprio to the user specified value.
However, if parse_apply_sb_mount_options() returned early because of
empty sbi->es_s->s_mount_opts, the correct journal_ioprio value was
never set.
This patch fixes __ext4_super() to only use the default value if the
user has not specified any value for journal_ioprio.
Similarly, the remount behavior was to either use journal_ioprio
value specified during initial mount, or use the default value
irrespective of the journal_ioprio value specified during remount.
This patch modifies this to first check if a new value for ioprio
has been passed during remount and apply it. If no new value is
passed, use the value specified during initial mount.
Signed-off-by: Ojaswin Mujoo <ojaswin@linux.ibm.com>
Reviewed-by: Ritesh Harjani <riteshh@linux.ibm.com>
Tested-by: Ritesh Harjani <riteshh@linux.ibm.com>
Link: https://lore.kernel.org/r/20220418083545.45778-1-ojaswin@linux.ibm.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
Otherwise nonaligned fstrim calls will works inconveniently for iterative
scanners, for example:
// trim [0,16MB] for group-1, but mark full group as trimmed
fstrim -o $((1024*1024*128)) -l $((1024*1024*16)) ./m
// handle [16MB,16MB] for group-1, do nothing because group already has the flag.
fstrim -o $((1024*1024*144)) -l $((1024*1024*16)) ./m
[ Update function documentation for ext4_trim_all_free -- TYT ]
Signed-off-by: Dmitry Monakhov <dmtrmonakhov@yandex-team.ru>
Link: https://lore.kernel.org/r/1650214995-860245-1-git-send-email-dmtrmonakhov@yandex-team.ru
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
Zoned devices are expected to have zone sizes in the range of 1-2GB for
ZNS SSDs and SMR HDDs have zone sizes of 256MB, so there is no need to
allow arbitrarily small zone sizes on btrfs.
But for testing purposes with emulated devices it is sometimes desirable
to create devices with as small as 4MB zone size to uncover errors.
So use 4MB as the smallest possible zone size and reject mounts of devices
with a smaller zone size.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Btrfs defaults to max_inline=2K to make small writes inlined into
metadata.
The default value is always a win, as even DUP/RAID1/RAID10 doubles the
metadata usage, it should still cause less physical space used compared
to a 4K regular extents.
But since the introduction of RAID1C3 and RAID1C4 it's no longer the case,
users may find inlined extents causing too much space wasted, and want
to convert those inlined extents back to regular extents.
Unfortunately defrag will unconditionally skip all inline extents, no
matter if the user is trying to converting them back to regular extents.
So this patch will add a small exception for defrag_collect_targets() to
allow defragging inline extents, if and only if the inlined extents are
larger than max_inline, allowing users to convert them to regular ones.
This also allows us to defrag extents like the following:
item 6 key (257 EXTENT_DATA 0) itemoff 15794 itemsize 69
generation 7 type 0 (inline)
inline extent data size 48 ram_bytes 4096 compression 1 (zlib)
item 7 key (257 EXTENT_DATA 4096) itemoff 15741 itemsize 53
generation 7 type 1 (regular)
extent data disk byte 13631488 nr 4096
extent data offset 0 nr 16384 ram 16384
extent compression 1 (zlib)
Previously we're unable to do any defrag, since the first extent is
inlined, and the second one has no extent to merge.
Now we can defrag it to just one single extent, saving 48 bytes metadata
space.
item 6 key (257 EXTENT_DATA 0) itemoff 15810 itemsize 53
generation 8 type 1 (regular)
extent data disk byte 13635584 nr 4096
extent data offset 0 nr 20480 ram 20480
extent compression 1 (zlib)
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The following error message lack the "0x" obviously:
cannot mount because of unsupported optional features (4000)
Add the prefix to make it less confusing. This can happen on older
kernels that try to mount a filesystem with newer features so it makes
sense to backport to older trees.
CC: stable@vger.kernel.org # 4.14+
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When reserving metadata units for creating an inode, we don't need to
reserve one extra unit for the inode ref item because when creating the
inode, at btrfs_create_new_inode(), we always insert the inode item and
the inode ref item in a single batch (a single btree insert operation,
and both ending up in the same leaf).
As we have accounted already one unit for the inode item, the extra unit
for the inode ref item is superfluous, it only makes us reserve more
metadata than necessary and often adding more reclaim pressure if we are
low on available metadata space.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The block_group->alloc_offset is an offset from the start of the block
group. OTOH, the ->meta_write_pointer is an address in the logical
space. So, we should compare the alloc_offset shifted with the
block_group->start.
Fixes: afba2bc036 ("btrfs: zoned: implement active zone tracking")
CC: stable@vger.kernel.org # 5.16+
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
A send operation reads extent data using the buffered IO path for getting
extent data to send in write commands and this is both because it's simple
and to make use of the generic readahead infrastructure, which results in
a massive speedup.
However this fills the page cache with data that, most of the time, is
really only used by the send operation - once the write commands are sent,
it's not useful to have the data in the page cache anymore. For large
snapshots, bringing all data into the page cache eventually leads to the
need to evict other data from the page cache that may be more useful for
applications (and kernel subsystems).
Even if extents are shared with the subvolume on which a snapshot is based
on and the data is currently on the page cache due to being read through
the subvolume, attempting to read the data through the snapshot will
always result in bringing a new copy of the data into another location in
the page cache (there's currently no shared memory for shared extents).
So make send evict the data it has read before if when it first opened
the inode, its mapping had no pages currently loaded: when
inode->i_mapping->nr_pages has a value of 0. Do this instead of deciding
based on the return value of filemap_range_has_page() before reading an
extent because the generic readahead mechanism may read pages beyond the
range we request (and it very often does it), which means a call to
filemap_range_has_page() will return true due to the readahead that was
triggered when processing a previous extent - we don't have a simple way
to distinguish this case from the case where the data was brought into
the page cache through someone else. So checking for the mapping number
of pages being 0 when we first open the inode is simple, cheap and it
generally accomplishes the goal of not trashing the page cache - the
only exception is if part of data was previously loaded into the page
cache through the snapshot by some other process, in that case we end
up not evicting any data send brings into the page cache, just like
before this change - but that however is not the common case.
Example scenario, on a box with 32G of RAM:
$ btrfs subvolume create /mnt/sv1
$ xfs_io -f -c "pwrite 0 4G" /mnt/sv1/file1
$ btrfs subvolume snapshot -r /mnt/sv1 /mnt/snap1
$ free -m
total used free shared buff/cache available
Mem: 31937 186 26866 0 4883 31297
Swap: 8188 0 8188
# After this we get less 4G of free memory.
$ btrfs send /mnt/snap1 >/dev/null
$ free -m
total used free shared buff/cache available
Mem: 31937 186 22814 0 8935 31297
Swap: 8188 0 8188
The same, obviously, applies to an incremental send.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Enable access to the NFSv4 acl via the NFSv4.1 'dacl' and 'sacl'
attributes.
This allows the server to authenticate the DACL and the SACL operations
separately, since reading and/or editing the SACL is usually considered
to be a privileged operation.
It also allows the propagation of automatic inheritance information that
was not supported by the NFSv4.0 'acl' attribute.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Add the ability to set or retrieve the acl using the NFSv4.1 'dacl' and
'sacl' attributes to the NFSv4 xdr encoders/decoders.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When caching a NFSv4 ACL, we want to specify whether we are caching an
NFSv4.0 type acl, the NFSv4.1 dacl or the NFSv4.1 sacl.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When doing layoutget as part of the open() compound, we have to be
careful to release the layout locks before we can call any further RPC
calls, such as setattr(). The reason is that those calls could trigger
a recall, which could deadlock.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Only report the error when the server is returning a fatal error, such
as ESTALE, EIO, etc...
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When we handle an error by redirtying the page, we're not corrupting the
mapping, so we don't want the error to be recorded in the mapping.
If the caller has specified a sync_mode of WB_SYNC_NONE, we can just
return AOP_WRITEPAGE_ACTIVATE. However if we're dealing with
WB_SYNC_ALL, we need to ensure that retries happen when the errors are
non-fatal.
Reported-by: Olga Kornievskaia <aglo@umich.edu>
Fixes: 8fc75bed96 ("NFS: Fix up return value on fatal errors in nfs_page_async_flush()")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Commit 587f03deb6 caused pnfs_update_layout() to stop returning ENOMEM
when the memory allocation fails, and hence causes it to fall back to
trying to do I/O through the MDS. There is no guarantee that this will
fare any better. If we're failing the pNFS layout allocation, then we
should just redirty the page and retry later.
Reported-by: Olga Kornievskaia <aglo@umich.edu>
Fixes: 587f03deb6 ("pnfs: refactor send_layoutget")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
We need to filter out ENOMEM in nfs_error_is_fatal_on_server(), because
running out of memory on our client is not a server error.
Reported-by: Olga Kornievskaia <aglo@umich.edu>
Fixes: 2dc23afffb ("NFS: ENOMEM should also be a fatal error.")
Cc: stable@vger.kernel.org
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When "-o device" mount option is not specified, scan the device table
and instantiate the devices if there's any in the device table. In this
case, the tag field of each device slot uniquely specifies a device.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220512055601.106109-1-jefflexu@linux.alibaba.com
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Use asynchronous io to read data from fscache may greatly improve IO
bandwidth for sequential buffered read scenario.
Change erofs_fscache_read_folios to erofs_fscache_read_folios_async,
and read data from fscache asynchronously.
Make .readpage()/.readahead() to use this new helper.
Signed-off-by: Xin Yin <yinxin.x@bytedance.com>
Reviewed-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220509074028.74954-23-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
[ Gao Xiang: minor styling changes. ]
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Introduce 'fsid' mount option to enable on-demand read sementics, in
which case, erofs will be mounted from data blobs. Users could specify
the name of primary data blob by this mount option.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-22-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Tested-by: Zichen Tian <tianzichen@kuaishou.com>
Tested-by: Jia Zhu <zhujia.zj@bytedance.com>
Tested-by: Yan Song <yansong.ys@antgroup.com>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Implement the data plane of reading data from data blobs over fscache
for inline layout.
For the heading non-inline part, the data plane for non-inline layout is
reused, while only the tail packing part needs special handling.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-20-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Implement the data plane of reading data from data blobs over fscache
for non-inline layout.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-19-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Similar to the multi-device mode, erofs could be mounted from one
primary data blob (mandatory) and multiple extra data blobs (optional).
Register fscache context for each extra data blob.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-17-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Registers fscache context for primary data blob. Also move the
initialization of s_op and related fields forward, since anonymous
inode will be allocated under the super block when registering the
fscache context.
Something worth mentioning about the cleanup routine.
1. The fscache context will instantiate anonymous inodes under the super
block. Release these anonymous inodes when .put_super() is called, or
we'll get "VFS: Busy inodes after unmount." warning.
2. The fscache context is initialized prior to the root inode. If
.kill_sb() is called when mount failed, .put_super() won't be called
when root inode has not been initialized yet. Thus .kill_sb() shall
also contain the cleanup routine.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-16-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Add erofs_fscache_read_folios() helper reading from fscache. It supports
on-demand read semantics. That is, it will make the backend prepare for
the data when cache miss. Once data ready, it will read from the cache.
This helper can then be used to implement .readpage()/.readahead() of
on-demand read semantics.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-15-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Introduce one anonymous inode for data blobs so that erofs can cache
metadata directly within such anonymous inode.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-14-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Introduce a context structure for managing data blobs, and helper
functions for initializing and cleaning up this context structure.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-13-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
A new fscache based mode is going to be introduced for erofs, in which
case on-demand read semantics is implemented through fscache.
As the first step, register fscache volume for each erofs filesystem.
That means, data blobs can not be shared among erofs filesystems. In the
following iteration, we are going to introduce the domain semantics, in
which case several erofs filesystems can belong to one domain, and data
blobs can be shared among these erofs filesystems of one domain.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-12-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Until then erofs is exactly blockdev based filesystem.
A new fscache-based mode is going to be introduced for erofs to support
scenarios where on-demand read semantics is needed, e.g. container
image distribution. In this case, erofs could be mounted from data blobs
through fscache.
Add a helper checking which mode erofs works in, and twist the code in
preparation for the upcoming fscache mode.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-11-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
... so that it can be used in the following introduced fscache mode.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220425122143.56815-10-jefflexu@linux.alibaba.com
Acked-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Add tracepoints for on-demand read mode. Currently following tracepoints
are added:
OPEN request / COPEN reply
CLOSE request
READ request / CREAD reply
write through anonymous fd
release of anonymous fd
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Acked-by: David Howells <dhowells@redhat.com>
Link: https://lore.kernel.org/r/20220425122143.56815-8-jefflexu@linux.alibaba.com
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Enable on-demand read mode by adding an optional parameter to the "bind"
command.
On-demand mode will be turned on when this parameter is "ondemand", i.e.
"bind ondemand". Otherwise cachefiles will work in the original mode.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220509074028.74954-7-jefflexu@linux.alibaba.com
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Implement the data plane of on-demand read mode.
The early implementation [1] place the entry to
cachefiles_ondemand_read() in fscache_read(). However, fscache_read()
can only detect if the requested file range is fully cache miss, whilst
we need to notify the user daemon as long as there's a hole inside the
requested file range.
Thus the entry is now placed in cachefiles_prepare_read(). When working
in on-demand read mode, once a hole detected, the read routine will send
a READ request to the user daemon. The user daemon needs to fetch the
data and write it to the cache file. After sending the READ request, the
read routine will hang there, until the READ request is handled by the
user daemon. Then it will retry to read from the same file range. If no
progress encountered, the read routine will fail then.
A new NETFS_SREQ_ONDEMAND flag is introduced to indicate that on-demand
read should be done when a cache miss encountered.
[1] https://lore.kernel.org/all/20220406075612.60298-6-jefflexu@linux.alibaba.com/ #v8
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Acked-by: David Howells <dhowells@redhat.com>
Link: https://lore.kernel.org/r/20220425122143.56815-6-jefflexu@linux.alibaba.com
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Notify the user daemon that cookie is going to be withdrawn, providing a
hint that the associated anonymous fd can be closed.
Be noted that this is only a hint. The user daemon may close the
associated anonymous fd when receiving the CLOSE request, then it will
receive another anonymous fd when the cookie gets looked up. Or it may
ignore the CLOSE request, and keep writing data through the anonymous
fd. However the next time the cookie gets looked up, the user daemon
will still receive another new anonymous fd.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Acked-by: David Howells <dhowells@redhat.com>
Link: https://lore.kernel.org/r/20220425122143.56815-5-jefflexu@linux.alibaba.com
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Add a refcount to avoid the deadlock in on-demand read mode. The
on-demand read mode will pin the corresponding cachefiles object for
each anonymous fd. The cachefiles object is unpinned when the anonymous
fd gets closed. When the user daemon exits and the fd of
"/dev/cachefiles" device node gets closed, it will wait for all
cahcefiles objects getting withdrawn. Then if there's any anonymous fd
getting closed after the fd of the device node, the user daemon will
hang forever, waiting for all objects getting withdrawn.
To fix this, add a refcount indicating if there's any object pinned by
anonymous fds. The cachefiles cache gets unbound and withdrawn when the
refcount is decreased to 0. It won't change the behaviour of the
original mode, in which case the cachefiles cache gets unbound and
withdrawn as long as the fd of the device node gets closed.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220509074028.74954-4-jefflexu@linux.alibaba.com
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Fscache/CacheFiles used to serve as a local cache for a remote
networking fs. A new on-demand read mode will be introduced for
CacheFiles, which can boost the scenario where on-demand read semantics
are needed, e.g. container image distribution.
The essential difference between these two modes is seen when a cache
miss occurs: In the original mode, the netfs will fetch the data from
the remote server and then write it to the cache file; in on-demand
read mode, fetching the data and writing it into the cache is delegated
to a user daemon.
As the first step, notify the user daemon when looking up cookie. In
this case, an anonymous fd is sent to the user daemon, through which the
user daemon can write the fetched data to the cache file. Since the user
daemon may move the anonymous fd around, e.g. through dup(), an object
ID uniquely identifying the cache file is also attached.
Also add one advisory flag (FSCACHE_ADV_WANT_CACHE_SIZE) suggesting that
the cache file size shall be retrieved at runtime. This helps the
scenario where one cache file contains multiple netfs files, e.g. for
the purpose of deduplication. In this case, netfs itself has no idea the
size of the cache file, whilst the user daemon should give the hint on
it.
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Link: https://lore.kernel.org/r/20220509074028.74954-3-jefflexu@linux.alibaba.com
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Extract the generic routine of writing data to cache files, and make it
generally available.
This will be used by the following patch implementing on-demand read
mode. Since it's called inside CacheFiles module, make the interface
generic and unrelated to netfs_cache_resources.
It is worth noting that, ki->inval_counter is not initialized after
this cleanup. It shall not make any visible difference, since
inval_counter is no longer used in the write completion routine, i.e.
cachefiles_write_complete().
Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Acked-by: David Howells <dhowells@redhat.com>
Link: https://lore.kernel.org/r/20220425122143.56815-2-jefflexu@linux.alibaba.com
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Since errors from nfs_pageio_complete() are already being reported
through nfs_async_write_error(), we should not be returning them to the
callers of do_writepages() as well. They will end up being reported
through the generic mechanism instead.
Fixes: 6fbda89b25 ("NFS: Replace custom error reporting mechanism with generic one")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
If we do flush cached writebacks in nfs_write_end() due to the imminent
expiration of an RPCSEC_GSS session, then we should defer reporting any
resulting errors until the calls to file_check_and_advance_wb_err() in
nfs_file_write() and nfs_file_fsync().
Fixes: 6fbda89b25 ("NFS: Replace custom error reporting mechanism with generic one")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Any errors reported by the write() system call need to be cleared from
the file descriptor's error tracking. The current call to nfs_wb_all()
causes the error to be reported, but since it doesn't call
file_check_and_advance_wb_err(), we can end up reporting the same error
a second time when the application calls fsync().
Note that since Linux 4.13, the rule is that EIO may be reported for
write(), but it must be reported by a subsequent fsync(), so let's just
drop reporting it in write.
The check for nfs_ctx_key_to_expire() is just a duplicate to the one
already in nfs_write_end(), so let's drop that too.
Reported-by: ChenXiaoSong <chenxiaosong2@huawei.com>
Fixes: ce368536dd ("nfs: nfs_file_write() should check for writeback errors")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
If the commit to disk is interrupted, we should still first check for
filesystem errors so that we can report them in preference to the error
due to the signal.
Fixes: 2197e9b06c ("NFS: Fix up fsync() when the server rebooted")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
If the attempt to flush data was interrupted due to a local signal, then
just requeue the writes back for I/O.
Fixes: 6fbda89b25 ("NFS: Replace custom error reporting mechanism with generic one")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
This patch enables idmapped mounts for erofs, since all dedicated helpers
for this functionality existsm, so, in this patch we just pass down the
user_namespace argument from the VFS methods to the relevant helpers.
Simple idmap example on erofs image:
1. mkdir dir
2. touch dir/file
3. mkfs.erofs erofs.img dir
4. mount -t erofs -o loop erofs.img /mnt/erofs/
5. ls -ln /mnt/erofs/
total 0
-rw-rw-r-- 1 1000 1000 0 May 17 15:26 file
6. mount-idmapped --map-mount b:1000:1001:1 /mnt/erofs/ /mnt/scratch_erofs/
7. ls -ln /mnt/scratch_erofs/
total 0
-rw-rw-r-- 1 1001 1001 0 May 17 15:26 file
Reviewed-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Link: https://lore.kernel.org/r/20220517104103.3570721-1-chao@kernel.org
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Implement export operations in order to make EROFS support accessing
inodes with filehandles so that it can be exported via NFS and used
by overlayfs.
Without this patch, 'exportfs -rv' will report:
exportfs: /root/erofs_mp does not support NFS export
Also tested with unionmount-testsuite and the testcase below passes now:
./run --ov --erofs --verify hard-link
For more details about the testcase, see:
https://github.com/amir73il/unionmount-testsuite/pull/6
Signed-off-by: Hongnan Li <hongnan.li@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Link: https://lore.kernel.org/r/20220425040712.91685-1-hongnan.li@linux.alibaba.com
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
I got some KASAN report as below:
[ 46.959738] ==================================================================
[ 46.960430] BUG: KASAN: use-after-free in z_erofs_shifted_transform+0x2bd/0x370
[ 46.960430] Read of size 4074 at addr ffff8880300c2f8e by task fssum/188
...
[ 46.960430] Call Trace:
[ 46.960430] <TASK>
[ 46.960430] dump_stack_lvl+0x41/0x5e
[ 46.960430] print_report.cold+0xb2/0x6b7
[ 46.960430] ? z_erofs_shifted_transform+0x2bd/0x370
[ 46.960430] kasan_report+0x8a/0x140
[ 46.960430] ? z_erofs_shifted_transform+0x2bd/0x370
[ 46.960430] kasan_check_range+0x14d/0x1d0
[ 46.960430] memcpy+0x20/0x60
[ 46.960430] z_erofs_shifted_transform+0x2bd/0x370
[ 46.960430] z_erofs_decompress_pcluster+0xaae/0x1080
The root cause is that the tail pcluster won't be a complete filesystem
block anymore. So if ztailpacking is used, the second part of an
uncompressed tail pcluster may not be ``rq->pageofs_out``.
Fixes: ab749badf9 ("erofs: support unaligned data decompression")
Fixes: cecf864d3d ("erofs: support inline data decompression")
Reviewed-by: Yue Hu <huyue2@coolpad.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Link: https://lore.kernel.org/r/20220512115833.24175-1-hsiangkao@linux.alibaba.com
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Some comments haven't been useful anymore since the code updated.
Let's drop them instead.
Link: https://lore.kernel.org/r/20220506194612.117120-2-hsiangkao@linux.alibaba.com
Reviewed-by: Yue Hu <huyue2@coolpad.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
The big pcluster feature has been merged for a year, it has been mostly
stable now.
Signed-off-by: Yue Hu <huyue2@coolpad.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Link: https://lore.kernel.org/r/20220407050505.12683-1-huyue2@coolpad.com
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
XFS has the unique behavior (as compared to the other Linux filesystems)
that on writeback errors it will completely invalidate the affected
folio and force the page cache to reread the contents from disk. All
other filesystems leave the page mapped and up to date.
This is a rude awakening for user programs, since (in the case where
write fails but reread doesn't) file contents will appear to revert to
old disk contents with no notification other than an EIO on fsync. This
might have been annoying back in the days when iomap dealt with one page
at a time, but with multipage folios, we can now throw away *megabytes*
worth of data for a single write error.
On *most* Linux filesystems, a program can respond to an EIO on write by
redirtying the entire file and scheduling it for writeback. This isn't
foolproof, since the page that failed writeback is no longer dirty and
could be evicted, but programs that want to recover properly *also*
have to detect XFS and regenerate every write they've made to the file.
When running xfs/314 on arm64, I noticed a UAF when xfs_discard_folio
invalidates multipage folios that could be undergoing writeback. If,
say, we have a 256K folio caching a mix of written and unwritten
extents, it's possible that we could start writeback of the first (say)
64K of the folio and then hit a writeback error on the next 64K. We
then free the iop attached to the folio, which is really bad because
writeback completion on the first 64k will trip over the "blocks per
folio > 1 && !iop" assertion.
This can't be fixed by only invalidating the folio if writeback fails at
the start of the folio, since the folio is marked !uptodate, which trips
other assertions elsewhere. Get rid of the whole behavior entirely.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Introduce dax_recovery_write() operation. The function is used to
recover a dax range that contains poison. Typical use case is when
a user process receives a SIGBUS with si_code BUS_MCEERR_AR
indicating poison(s) in a dax range, in response, the user process
issues a pwrite() to the page-aligned dax range, thus clears the
poison and puts valid data in the range.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jane Chu <jane.chu@oracle.com>
Link: https://lore.kernel.org/r/20220422224508.440670-6-jane.chu@oracle.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Up till now, dax_direct_access() is used implicitly for normal
access, but for the purpose of recovery write, dax range with
poison is requested. To make the interface clear, introduce
enum dax_access_mode {
DAX_ACCESS,
DAX_RECOVERY_WRITE,
}
where DAX_ACCESS is used for normal dax access, and
DAX_RECOVERY_WRITE is used for dax recovery write.
Suggested-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Jane Chu <jane.chu@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Cc: Mike Snitzer <snitzer@redhat.com>
Reviewed-by: Vivek Goyal <vgoyal@redhat.com>
Link: https://lore.kernel.org/r/165247982851.52965.11024212198889762949.stgit@dwillia2-desk3.amr.corp.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Every time we send a write command, we open the inode, read some data to
a buffer and then close the inode. The amount of data we read for each
write command is at most 48K, returned by max_send_read_size(), and that
corresponds to: BTRFS_SEND_BUF_SIZE - 16K = 48K. In practice this does
not add any significant overhead, because the time elapsed between every
close (iput()) and open (btrfs_iget()) is very short, so the inode is kept
in the VFS's cache after the iput() and it's still there by the time we
do the next btrfs_iget().
As between processing extents of the current inode we don't do anything
else, it makes sense to keep the inode open after we process its first
extent that needs to be sent and keep it open until we start processing
the next inode. This serves to facilitate the next change, which aims
to avoid having send operations trash the page cache with data extents.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Create a new bio_set that contains all the per-bio private data needed
by btrfs for direct I/O and tell the iomap code to use that instead
of separately allocation the btrfs_dio_private structure.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The btrfs_dio_private structure is only used in inode.c, so move the
definition there.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This field is never used, so remove it. Last use was probably in
23ea8e5a07 ("Btrfs: load checksum data once when submitting a direct
read io").
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Make use of the new iomap_iter->private field to avoid a memory
allocation per iomap range.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Allow the file system to keep state for all iterations. For now only
wire it up for direct I/O as there is an immediate need for it there.
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Allow the file system to provide a specific bio_set for allocating
direct I/O bios. This will allow file systems that use the
->submit_io hook to stash away additional information for file system
use.
To make use of this additional space for information in the completion
path, the file system needs to override the ->bi_end_io callback and
then call back into iomap, so export iomap_dio_bio_end_io for that.
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Add a wrapper around iomap_dio_rw that keeps the direct I/O internals
isolated in inode.c.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
While the active zones within an active block group are reset, and their
active resource is released, the block group itself is kept in the active
block group list and marked as active. As a result, the list will contain
more than max_active_zones block groups. That itself is not fatal for the
device as the zones are properly reset.
However, that inflated list is, of course, strange. Also, a to-appear
patch series, which deactivates an active block group on demand, gets
confused with the wrong list.
So, fix the issue by finishing the unused block group once it gets
read-only, so that we can release the active resource in an early stage.
Fixes: be1a1d7a5d ("btrfs: zoned: finish fully written block group")
CC: stable@vger.kernel.org # 5.16+
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Commit be1a1d7a5d ("btrfs: zoned: finish fully written block group")
introduced zone finishing code both for data and metadata end_io path.
However, the metadata side is not working as it should. First, it
compares logical address (eb->start + eb->len) with offset within a
block group (cache->zone_capacity) in submit_eb_page(). That essentially
disabled zone finishing on metadata end_io path.
Furthermore, fixing the issue above revealed we cannot call
btrfs_zone_finish_endio() in end_extent_buffer_writeback(). We cannot
call btrfs_lookup_block_group() which require spin lock inside end_io
context.
Introduce btrfs_schedule_zone_finish_bg() to wait for the extent buffer
writeback and do the zone finish IO in a workqueue.
Also, drop EXTENT_BUFFER_ZONE_FINISH as it is no longer used.
Fixes: be1a1d7a5d ("btrfs: zoned: finish fully written block group")
CC: stable@vger.kernel.org # 5.16+
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently, btrfs_zone_finish_endio() finishes a block group only when the
written region reaches the end of the block group. We can also finish the
block group when no more allocation is possible.
Fixes: be1a1d7a5d ("btrfs: zoned: finish fully written block group")
CC: stable@vger.kernel.org # 5.16+
Reviewed-by: Pankaj Raghav <p.raghav@samsung.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_zone_finish() and btrfs_zone_finish_endio() have similar code.
Introduce do_zone_finish() to factor out the common code.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Introduce a wrapper to check if all the space in a block group is
allocated or not.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When iterating the backrefs in an extent item if the ptr to the
'current' backref record goes beyond the extent item a warning is
generated and -ENOENT is returned. However what's more appropriate to
debug such cases would be to return EUCLEAN and also print identifying
information about the performed search as well as the current content of
the leaf containing the possibly corrupted extent item.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The bio_ctrl is the last use of bio_flags that has been converted to
compress type everywhere else.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Several functions take parameter bio_flags that was simplified to just
compress type, unify it and change the type accordingly.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The bio_flags is now used to store unchanged compress type, so unify
that.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The helpers extent_set_compress_type and extent_compress_type have
become trivial after previous cleanups and can be removed.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The bio_flags are used only to encode the compression and there are no
other EXTENT_BIO_* flags, so the compress type can be stored directly.
The struct member name is left unchanged and will be cleaned in later
patches.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The helper used to do more with the wbc state but now it's just one
subtraction, no need to have a special helper.
It became trivial in a91326679f ("Btrfs: make mapping->writeback_index
point to the last written page").
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The value of btrfs_delayed_extent_op::is_data is always false, we can
cascade the change and simplify code that depends on it, removing the
structure member eventually.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The parameter has been added in 2009 in the infamous monster commit
5d4f98a28c ("Btrfs: Mixed back reference (FORWARD ROLLING FORMAT
CHANGE)") but not used ever since. We can sink it and allow further
simplifications.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When reserving data space for a direct IO write we can end up deadlocking
if we have multiple tasks attempting a write to the same file range, there
are multiple extents covered by that file range, we are low on available
space for data and the writes don't expand the inode's i_size.
The deadlock can happen like this:
1) We have a file with an i_size of 1M, at offset 0 it has an extent with
a size of 128K and at offset 128K it has another extent also with a
size of 128K;
2) Task A does a direct IO write against file range [0, 256K), and because
the write is within the i_size boundary, it takes the inode's lock (VFS
level) in shared mode;
3) Task A locks the file range [0, 256K) at btrfs_dio_iomap_begin(), and
then gets the extent map for the extent covering the range [0, 128K).
At btrfs_get_blocks_direct_write(), it creates an ordered extent for
that file range ([0, 128K));
4) Before returning from btrfs_dio_iomap_begin(), it unlocks the file
range [0, 256K);
5) Task A executes btrfs_dio_iomap_begin() again, this time for the file
range [128K, 256K), and locks the file range [128K, 256K);
6) Task B starts a direct IO write against file range [0, 256K) as well.
It also locks the inode in shared mode, as it's within the i_size limit,
and then tries to lock file range [0, 256K). It is able to lock the
subrange [0, 128K) but then blocks waiting for the range [128K, 256K),
as it is currently locked by task A;
7) Task A enters btrfs_get_blocks_direct_write() and tries to reserve data
space. Because we are low on available free space, it triggers the
async data reclaim task, and waits for it to reserve data space;
8) The async reclaim task decides to wait for all existing ordered extents
to complete (through btrfs_wait_ordered_roots()).
It finds the ordered extent previously created by task A for the file
range [0, 128K) and waits for it to complete;
9) The ordered extent for the file range [0, 128K) can not complete
because it blocks at btrfs_finish_ordered_io() when trying to lock the
file range [0, 128K).
This results in a deadlock, because:
- task B is holding the file range [0, 128K) locked, waiting for the
range [128K, 256K) to be unlocked by task A;
- task A is holding the file range [128K, 256K) locked and it's waiting
for the async data reclaim task to satisfy its space reservation
request;
- the async data reclaim task is waiting for ordered extent [0, 128K)
to complete, but the ordered extent can not complete because the
file range [0, 128K) is currently locked by task B, which is waiting
on task A to unlock file range [128K, 256K) and task A waiting
on the async data reclaim task.
This results in a deadlock between 4 task: task A, task B, the async
data reclaim task and the task doing ordered extent completion (a work
queue task).
This type of deadlock can sporadically be triggered by the test case
generic/300 from fstests, and results in a stack trace like the following:
[12084.033689] INFO: task kworker/u16:7:123749 blocked for more than 241 seconds.
[12084.034877] Not tainted 5.18.0-rc2-btrfs-next-115 #1
[12084.035562] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[12084.036548] task:kworker/u16:7 state:D stack: 0 pid:123749 ppid: 2 flags:0x00004000
[12084.036554] Workqueue: btrfs-flush_delalloc btrfs_work_helper [btrfs]
[12084.036599] Call Trace:
[12084.036601] <TASK>
[12084.036606] __schedule+0x3cb/0xed0
[12084.036616] schedule+0x4e/0xb0
[12084.036620] btrfs_start_ordered_extent+0x109/0x1c0 [btrfs]
[12084.036651] ? prepare_to_wait_exclusive+0xc0/0xc0
[12084.036659] btrfs_run_ordered_extent_work+0x1a/0x30 [btrfs]
[12084.036688] btrfs_work_helper+0xf8/0x400 [btrfs]
[12084.036719] ? lock_is_held_type+0xe8/0x140
[12084.036727] process_one_work+0x252/0x5a0
[12084.036736] ? process_one_work+0x5a0/0x5a0
[12084.036738] worker_thread+0x52/0x3b0
[12084.036743] ? process_one_work+0x5a0/0x5a0
[12084.036745] kthread+0xf2/0x120
[12084.036747] ? kthread_complete_and_exit+0x20/0x20
[12084.036751] ret_from_fork+0x22/0x30
[12084.036765] </TASK>
[12084.036769] INFO: task kworker/u16:11:153787 blocked for more than 241 seconds.
[12084.037702] Not tainted 5.18.0-rc2-btrfs-next-115 #1
[12084.038540] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[12084.039506] task:kworker/u16:11 state:D stack: 0 pid:153787 ppid: 2 flags:0x00004000
[12084.039511] Workqueue: events_unbound btrfs_async_reclaim_data_space [btrfs]
[12084.039551] Call Trace:
[12084.039553] <TASK>
[12084.039557] __schedule+0x3cb/0xed0
[12084.039566] schedule+0x4e/0xb0
[12084.039569] schedule_timeout+0xed/0x130
[12084.039573] ? mark_held_locks+0x50/0x80
[12084.039578] ? _raw_spin_unlock_irq+0x24/0x50
[12084.039580] ? lockdep_hardirqs_on+0x7d/0x100
[12084.039585] __wait_for_common+0xaf/0x1f0
[12084.039587] ? usleep_range_state+0xb0/0xb0
[12084.039596] btrfs_wait_ordered_extents+0x3d6/0x470 [btrfs]
[12084.039636] btrfs_wait_ordered_roots+0x175/0x240 [btrfs]
[12084.039670] flush_space+0x25b/0x630 [btrfs]
[12084.039712] btrfs_async_reclaim_data_space+0x108/0x1b0 [btrfs]
[12084.039747] process_one_work+0x252/0x5a0
[12084.039756] ? process_one_work+0x5a0/0x5a0
[12084.039758] worker_thread+0x52/0x3b0
[12084.039762] ? process_one_work+0x5a0/0x5a0
[12084.039765] kthread+0xf2/0x120
[12084.039766] ? kthread_complete_and_exit+0x20/0x20
[12084.039770] ret_from_fork+0x22/0x30
[12084.039783] </TASK>
[12084.039800] INFO: task kworker/u16:17:217907 blocked for more than 241 seconds.
[12084.040709] Not tainted 5.18.0-rc2-btrfs-next-115 #1
[12084.041398] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[12084.042404] task:kworker/u16:17 state:D stack: 0 pid:217907 ppid: 2 flags:0x00004000
[12084.042411] Workqueue: btrfs-endio-write btrfs_work_helper [btrfs]
[12084.042461] Call Trace:
[12084.042463] <TASK>
[12084.042471] __schedule+0x3cb/0xed0
[12084.042485] schedule+0x4e/0xb0
[12084.042490] wait_extent_bit.constprop.0+0x1eb/0x260 [btrfs]
[12084.042539] ? prepare_to_wait_exclusive+0xc0/0xc0
[12084.042551] lock_extent_bits+0x37/0x90 [btrfs]
[12084.042601] btrfs_finish_ordered_io.isra.0+0x3fd/0x960 [btrfs]
[12084.042656] ? lock_is_held_type+0xe8/0x140
[12084.042667] btrfs_work_helper+0xf8/0x400 [btrfs]
[12084.042716] ? lock_is_held_type+0xe8/0x140
[12084.042727] process_one_work+0x252/0x5a0
[12084.042742] worker_thread+0x52/0x3b0
[12084.042750] ? process_one_work+0x5a0/0x5a0
[12084.042754] kthread+0xf2/0x120
[12084.042757] ? kthread_complete_and_exit+0x20/0x20
[12084.042763] ret_from_fork+0x22/0x30
[12084.042783] </TASK>
[12084.042798] INFO: task fio:234517 blocked for more than 241 seconds.
[12084.043598] Not tainted 5.18.0-rc2-btrfs-next-115 #1
[12084.044282] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[12084.045244] task:fio state:D stack: 0 pid:234517 ppid:234515 flags:0x00004000
[12084.045248] Call Trace:
[12084.045250] <TASK>
[12084.045254] __schedule+0x3cb/0xed0
[12084.045263] schedule+0x4e/0xb0
[12084.045266] wait_extent_bit.constprop.0+0x1eb/0x260 [btrfs]
[12084.045298] ? prepare_to_wait_exclusive+0xc0/0xc0
[12084.045306] lock_extent_bits+0x37/0x90 [btrfs]
[12084.045336] btrfs_dio_iomap_begin+0x336/0xc60 [btrfs]
[12084.045370] ? lock_is_held_type+0xe8/0x140
[12084.045378] iomap_iter+0x184/0x4c0
[12084.045383] __iomap_dio_rw+0x2c6/0x8a0
[12084.045406] iomap_dio_rw+0xa/0x30
[12084.045408] btrfs_do_write_iter+0x370/0x5e0 [btrfs]
[12084.045440] aio_write+0xfa/0x2c0
[12084.045448] ? __might_fault+0x2a/0x70
[12084.045451] ? kvm_sched_clock_read+0x14/0x40
[12084.045455] ? lock_release+0x153/0x4a0
[12084.045463] io_submit_one+0x615/0x9f0
[12084.045467] ? __might_fault+0x2a/0x70
[12084.045469] ? kvm_sched_clock_read+0x14/0x40
[12084.045478] __x64_sys_io_submit+0x83/0x160
[12084.045483] ? syscall_enter_from_user_mode+0x1d/0x50
[12084.045489] do_syscall_64+0x3b/0x90
[12084.045517] entry_SYSCALL_64_after_hwframe+0x44/0xae
[12084.045521] RIP: 0033:0x7fa76511af79
[12084.045525] RSP: 002b:00007ffd6d6b9058 EFLAGS: 00000246 ORIG_RAX: 00000000000000d1
[12084.045530] RAX: ffffffffffffffda RBX: 00007fa75ba6e760 RCX: 00007fa76511af79
[12084.045532] RDX: 0000557b304ff3f0 RSI: 0000000000000001 RDI: 00007fa75ba4c000
[12084.045535] RBP: 00007fa75ba4c000 R08: 00007fa751b76000 R09: 0000000000000330
[12084.045537] R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000001
[12084.045540] R13: 0000000000000000 R14: 0000557b304ff3f0 R15: 0000557b30521eb0
[12084.045561] </TASK>
Fix this issue by always reserving data space before locking a file range
at btrfs_dio_iomap_begin(). If we can't reserve the space, then we don't
error out immediately - instead after locking the file range, check if we
can do a NOCOW write, and if we can we don't error out since we don't need
to allocate a data extent, however if we can't NOCOW then error out with
-ENOSPC. This also implies that we may end up reserving space when it's
not needed because the write will end up being done in NOCOW mode - in that
case we just release the space after we noticed we did a NOCOW write - this
is the same type of logic that is done in the path for buffered IO writes.
Fixes: f0bfa76a11 ("btrfs: fix ENOSPC failure when attempting direct IO write into NOCOW range")
CC: stable@vger.kernel.org # 5.17+
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Derive the compression type from extent map as opposed to the bio flags
passed. This makes it more precise and not reliant on function
parameters.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[SUSPICIOUS CODE]
When refactoring scrub code, I noticed a very strange behavior around
scrub_remap_extent():
if (sctx->is_dev_replace)
scrub_remap_extent(fs_info, cur_logical, scrub_len,
&cur_physical, &target_dev, &cur_mirror);
As replace target is a 1:1 copy of the source device, thus physical
offset inside the target should be the same as physical inside source,
thus this remap call makes no sense to me.
[REAL FUNCTIONALITY]
After more investigation, the function name scrub_remap_extent()
doesn't tell anything of the truth, nor does its if () condition.
The real story behind this function is that, for scrub_pages() we never
expect missing device, even for replacing missing device.
What scrub_remap_extent() is really doing is to find a live mirror, and
make later scrub_pages() to read data from the good copy, other than
from the missing device and increase error counters unnecessarily.
[IMPROVEMENT]
We have no need to bother scrub_remap_extent() in scrub_simple_mirror()
at all, we only need to call it before we call scrub_pages().
And rename the function to scrub_find_live_copy(), add extra comments on
them.
By this we can remove one parameter from scrub_extent(), and reduce the
unnecessary calls to scrub_remap_extent() for regular replace.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since we have find_first_extent_item() to iterate the extent items of a
certain range, there is no need to use the open-coded version.
Replace the final scrub call site with find_first_extent_item().
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently scrub_raid56_parity() has a large double loop, handling the
following things at the same time:
- Iterate each data stripe
- Iterate each extent item in one data stripe
Refactor this by:
- Introduce a new helper to handle data stripe iteration
The new helper is scrub_raid56_data_stripe_for_parity(), which
only has one while() loop handling the extent items inside the
data stripe.
The code is still mostly the same as the old code.
- Call cond_resched() for each extent
Previously we only call cond_resched() under a complex if () check.
I see no special reason to do that, and for other scrub functions,
like scrub_simple_mirror() we're already doing the same cond_resched()
after scrubbing one extent.
- Add more comments
Please note that, this patch is only to address the double loop, there
are incoming patches to do extra cleanup.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Although RAID56 has complex repair mechanism, which involves reading the
whole full stripe, but inside one data stripe, it's in fact no different
than SINGLE/RAID1.
The point here is, for data stripe we just check the csum for each
extent we hit. Only for csum mismatch case, our repair paths divide.
So we can still reuse scrub_simple_mirror() for RAID56 data stripes,
which saves quite some code.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since we have moved all other profiles handling into their own
functions, now the main body of scrub_stripe() is just handling RAID56
profiles.
There is no need to address other profiles in the main loop of
scrub_stripe(), so we can remove those dead branches.
Since we're here, also slightly change the timing of initialization of
variables like @offset, @increment and @logical.
Especially for @logical, we don't really need to initialize it for
btrfs_extent_root()/btrfs_csum_root(), we can use bg->start for that
purpose.
Now those variables are only initialize for RAID56 branches.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The new entrance will iterate through each data stripe which belongs to
the target device.
And since inside each data stripe, RAID0 is just SINGLE, while RAID10 is
just RAID1, we can reuse scrub_simple_mirror() to do the scrub properly.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The new helper, scrub_simple_mirror(), will scrub all extents inside a
range which only has simple mirror based duplication.
This covers every range of SINGLE/DUP/RAID1/RAID1C*, and inside each
data stripe for RAID0/RAID10.
Currently we will use this function to scrub SINGLE/DUP/RAID1/RAID1C*
profiles. As one can see, the new entrance for those simple-mirror
based profiles can be small enough (with comments, just reach 100
lines).
This function will be the basis for the incoming scrub refactor.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The new helper, find_first_extent_item(), will locate an extent item
(either EXTENT_ITEM or METADATA_ITEM) which covers any byte of the
search range.
This helper will later be used to refactor scrub code.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The variable @physical_end is the exclusive stripe end, currently it's
calculated using @physical + @dev_extent_len / map->stripe_len *
map->stripe_len.
And since at allocation time we ensured dev_extent_len is stripe_len
aligned, the result is the same as @physical + @dev_extent_len.
So this patch will just assign @physical and @physical_end early,
without using @nstripes.
This is especially helpful for any possible out: label user, as now we
only need to initialize @offset before going to out: label.
Since we're here, also make @physical_end constant.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
… rename it to simply fs_roots and adjust all usages of this object to use
the XArray API, because it is notionally easier to use and understand, as
it provides array semantics, and also takes care of locking for us,
further simplifying the code.
Also do some refactoring, esp. where the API change requires largely
rewriting some functions, anyway.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
… named 'extent_buffers'. Also adjust all usages of this object to use
the XArray API, which greatly simplifies the code as it takes care of
locking and is generally easier to use and understand, providing
notionally simpler array semantics.
Also perform some light refactoring.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
… and adjust all usages of this object to use the XArray API for the sake
of consistency.
XArray API provides array semantics, so it is notionally easier to use and
understand, and it also takes care of locking for us.
None of this makes a real difference in this particular patch, but it does
in other places where similar replacements are or have been made and we
want to be consistent in our usage of data structures in btrfs.
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
… in the btrfs_root struct and adjust all usages of this object to use
the XArray API, because it is notionally easier to use and understand,
as it provides array semantics, and also takes care of locking for us,
further simplifying the code.
Also use the opportunity to do some light refactoring.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In function btrfs_bg_flags_to_raid_index(), we use quite some if () to
convert the BTRFS_BLOCK_GROUP_* bits to a index number.
But the truth is, there is really no such need for so many branches at
all.
Since all BTRFS_BLOCK_GROUP_* flags are just one single bit set inside
BTRFS_BLOCK_GROUP_PROFILES_MASK, we can easily use ilog2() to calculate
their values.
This calculation has an anchor point, the lowest PROFILE bit, which is
RAID0.
Even it's fixed on-disk format and should never change, here I added
extra compile time checks to make it super safe:
1. Make sure RAID0 is always the lowest bit in PROFILE_MASK
This is done by finding the first (least significant) bit set of
RAID0 and PROFILE_MASK & ~RAID0.
2. Make sure RAID0 bit set beyond the highest bit of TYPE_MASK
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
It's only internally used as another way to represent btrfs profiles,
it's not exposed through any on-disk format, in fact this
btrfs_raid_types is diverted from the on-disk format values.
Furthermore, since it's internal structure, its definition can change in
the future.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
rmw_workers doesn't need ordered execution or thread disabling threshold
(as the thresh parameter is less than DFT_THRESHOLD).
Just switch to the normal workqueues that use a lot less resources,
especially in the work_struct vs btrfs_work structures.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
All three scrub workqueues don't need ordered execution or thread
disabling threshold (as the thresh parameter is less than DFT_THRESHOLD).
Just switch to the normal workqueues that use a lot less resources,
especially in the work_struct vs btrfs_work structures.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Just let the one caller that wants optional WQ_HIGHPRI handling allocate
a separate btrfs_workqueue for that. This allows to rename struct
__btrfs_workqueue to btrfs_workqueue, remove a pointer indirection and
separate allocation for all btrfs_workqueue users and generally simplify
the code.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now the btrfs RAID56 infrastructure has migrated to use sector_ptr
interface, it should be safe to enable subpage support for RAID56.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The non-compatible part is only the bitmap iteration part, now the
bitmap size is extended to rbio::stripe_nsectors, not the old
rbio::stripe_npages.
Since we're here, also slightly improve the function by:
- Rename @i to @stripe
- Rename @bit to @sectornr
- Move @page and @index into the inner loop
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Function steal_rbio() will take all the uptodate pages from the source
rbio to destination rbio.
With the new stripe_sectors[] array, we also need to do the extra check:
- Check sector::flags to make sure the full page is uptodate
Now we don't use PageUptodate flag for subpage cases to indicate
if the page is uptodate.
Instead we need to check all the sectors belong to the page to be sure
about whether it's full page uptodate.
So here we introduce a new helper, full_page_sectors_uptodate() to do
the check.
- Update rbio::stripe_sectors[] to use the new page pointer
We only need to change the page pointer, no need to change anything
else.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Unlike previous code, we can not directly set PageUptodate for stripe
pages now. Instead we have to iterate through all the sectors and set
SECTOR_UPTODATE flag there.
Introduce a new helper find_stripe_sector(), to do the work.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The functionality is completely replaced by the new bio_sectors member,
now it's time to remove the old member.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This requires one extra parameter @pgoff for the function.
In the current code base, scrub is still one page per sector, thus the
new parameter will always be 0.
It needs the extra subpage scrub optimization code to fully take
advantage.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There is only one caller for that helper now, and we're definitely fine
to open-code it.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
With this function converted to subpage compatible sector interfaces,
the following helper functions can be removed:
- rbio_stripe_page()
- rbio_pstripe_page()
- rbio_qstripe_page()
- page_in_rbio()
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This involves:
- Use sector_ptr interface to grab the pointers
- Add sector->pgoff to pointers[]
- Rebuild data using sectorsize instead of PAGE_SIZE
- Use memcpy() to replace copy_page()
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The core is to convert direct page usage into sector_ptr usage, and
use memcpy() to replace copy_page().
For pointers usage, we need to convert it to kmap_local_page() +
sector->pgoff.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Make rbio_add_io_page() subpage compatible, which involves:
- Rename rbio_add_io_page() to rbio_add_io_sector()
Although we still rely on PAGE_SIZE == sectorsize, so add a new
ASSERT() inside rbio_add_io_sector() to make sure all pgoff is 0.
- Introduce rbio_stripe_sector() helper
The equivalent of rbio_stripe_page().
This new helper has extra ASSERT()s to validate the stripe and sector
number.
- Introduce sector_in_rbio() helper
The equivalent of page_in_rbio().
- Rename @pagenr variables to @sectornr
- Use rbio::stripe_nsectors when iterating the bitmap
Please note that, this only changes the interface, the bios are still
using full page for IO.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This new member is going to fully replace bio_pages in the future, but
for now let's keep them co-exist, until the full switch is done.
Currently cache_rbio_pages() and index_rbio_pages() will also populate
the new array.
And cache_rbio_pages() need to record which sectors are uptodate, so we
also need to introduce sector_ptr::uptodate bit.
To avoid extra memory usage, we let the new @uptodate bit to share bits
with @pgoff. Now pgoff only has at most 31 bits, which is already more
than enough, as even for 256K page size, we only need 18 bits.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The new member is an array of sector_ptr pointers, they will represent
all sectors inside a full stripe (including P/Q).
They co-operate with btrfs_raid_bio::stripe_pages:
stripe_pages: | Page 0, range [0, 64K) | Page 1 ...
stripe_sectors: | | | ... | |
| | \- sector 15, page 0, pgoff=60K
| \- sector 1, page 0, pgoff=4K
\---- sector 0, page 0, pfoff=0
With such structure, we can represent subpage sectors without using
extra pages.
Here we introduce a new helper, index_stripe_sectors(), to update
stripe_sectors[] to point to correct page and pgoff.
So every time rbio::stripe_pages[] pointer gets updated, the new helper
should be called.
The following functions have to call the new helper:
- steal_rbio()
- alloc_rbio_pages()
- alloc_rbio_parity_pages()
- alloc_rbio_essential_pages()
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The new members are all related to number of sectors, but the existing
number of pages members are kept as is:
- nr_sectors
Total sectors of the full stripe including P/Q.
- stripe_nsectors
The sectors of a single stripe.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are a lot of members using much larger type in btrfs_raid_bio than
necessary, like nr_pages which represents the total number of a full
stripe.
Instead of int (which is at least 32bits), u16 is already enough
(max stripe length will be 256MiB, already beyond current RAID56 device
number limit).
So this patch will reduce the width of the following members:
- stripe_len to u32
- nr_pages to u16
- nr_data to u8
- real_stripes to u8
- scrubp to u8
- faila/b to s8
As -1 is used to indicate no corruption
This will slightly reduce the size of btrfs_raid_bio from 272 bytes to
256 bytes, reducing 16 bytes usage.
But please note that, when using btrfs_raid_bio, we allocate extra space
for it to cover various pointer array, so the reduce memory is not
really a big saving overall.
As we're here modifying the comments already, update existing comments
to current code standard.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The function rbio_nr_pages() is only called once inside alloc_rbio(),
there is no reason to make it dedicated helper.
Furthermore, the return type doesn't match, the function return "unsigned
long" which may not be necessary, while the only caller only uses "int".
Since we're doing cleaning up here, also fix the type to "const unsigned
int" for all involved local variables.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently btrfs uses fixed stripe length (64K), thus u32 is wide enough
for the usage.
Furthermore, even in the future we choose to enlarge stripe length to
larger values, I don't believe we would want stripe as large as 4G or
larger.
So this patch will reduce the width for all in-memory structures and
parameters, this involves:
- RAID56 related function argument lists
This allows us to do direct division related to stripe_len.
Although we will use bits shift to replace the division anyway.
- btrfs_io_geometry structure
This involves one change to simplify the calculation of both @stripe_nr
and @stripe_offset, using div64_u64_rem().
And add extra sanity check to make sure @stripe_offset is always small
enough for u32.
This saves 8 bytes for the structure.
- map_lookup structure
This convert @stripe_len to u32, which saves 8 bytes. (saved 4 bytes,
and removed a 4-bytes hole)
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Both btrfs_repair_one_sector and submit_bio_one as the direct caller of
one of the instances ignore errors as they expect the methods themselves
to call ->bi_end_io on error. Remove the unused and dangerous return
value.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_submit_compressed_read already calls ->bi_end_io on error and
the caller must ignore the return value, so remove it.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_submit_metadata_bio already calls ->bi_end_io on error and the
caller must ignore the return value, so remove it.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This argument is unused since commit 953651eb30 ("btrfs: factor out
helper adding a page to bio") and commit 1b36294a6c ("btrfs: call
submit_bio_hook directly for metadata pages") reworked the way metadata
bio submission is handled.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Keep btrfs_readpage next to btrfs_do_readpage and the other address
space operations. This allows to keep submit_one_bio and
struct btrfs_bio_ctrl file local in extent_io.c.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
There is a report that a btrfs has a bad super block num devices.
This makes btrfs to reject the fs completely.
BTRFS error (device sdd3): super_num_devices 3 mismatch with num_devices 2 found here
BTRFS error (device sdd3): failed to read chunk tree: -22
BTRFS error (device sdd3): open_ctree failed
[CAUSE]
During btrfs device removal, chunk tree and super block num devs are
updated in two different transactions:
btrfs_rm_device()
|- btrfs_rm_dev_item(device)
| |- trans = btrfs_start_transaction()
| | Now we got transaction X
| |
| |- btrfs_del_item()
| | Now device item is removed from chunk tree
| |
| |- btrfs_commit_transaction()
| Transaction X got committed, super num devs untouched,
| but device item removed from chunk tree.
| (AKA, super num devs is already incorrect)
|
|- cur_devices->num_devices--;
|- cur_devices->total_devices--;
|- btrfs_set_super_num_devices()
All those operations are not in transaction X, thus it will
only be written back to disk in next transaction.
So after the transaction X in btrfs_rm_dev_item() committed, but before
transaction X+1 (which can be minutes away), a power loss happen, then
we got the super num mismatch.
This has been fixed by commit bbac58698a ("btrfs: remove device item
and update super block in the same transaction").
[FIX]
Make the super_num_devices check less strict, converting it from a hard
error to a warning, and reset the value to a correct one for the current
or next transaction commit.
As the number of device items is the critical information where the
super block num_devices is only a cached value (and also useful for
cross checking), it's safe to automatically update it. Other device
related problems like missing device are handled after that and may
require other means to resolve, like degraded mount. With this fix,
potentially affected filesystems won't fail mount and require the manual
repair by btrfs check.
Reported-by: Luca Béla Palkovics <luca.bela.palkovics@gmail.com>
Link: https://lore.kernel.org/linux-btrfs/CA+8xDSpvdm_U0QLBAnrH=zqDq_cWCOH5TiV46CKmp3igr44okQ@mail.gmail.com/
CC: stable@vger.kernel.org # 4.14+
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Parameter struct compressed_bio is not used by the function
submit_compressed_bio(). Remove it.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When doing a NOCOW write, either through direct IO or buffered IO, we do
two lookups for the block group that contains the target extent: once
when we call btrfs_inc_nocow_writers() and then later again when we call
btrfs_dec_nocow_writers() after creating the ordered extent.
The lookups require taking a lock and navigating the red black tree used
to track all block groups, which can take a non-negligible amount of time
for a large filesystem with thousands of block groups, as well as lock
contention and cache line bouncing.
Improve on this by having a single block group search: making
btrfs_inc_nocow_writers() return the block group to its caller and then
have the caller pass that block group to btrfs_dec_nocow_writers().
This is part of a patchset comprised of the following patches:
btrfs: remove search start argument from first_logical_byte()
btrfs: use rbtree with leftmost node cached for tracking lowest block group
btrfs: use a read/write lock for protecting the block groups tree
btrfs: return block group directly at btrfs_next_block_group()
btrfs: avoid double search for block group during NOCOW writes
The following test was used to test these changes from a performance
perspective:
$ cat test.sh
#!/bin/bash
modprobe null_blk nr_devices=0
NULL_DEV_PATH=/sys/kernel/config/nullb/nullb0
mkdir $NULL_DEV_PATH
if [ $? -ne 0 ]; then
echo "Failed to create nullb0 directory."
exit 1
fi
echo 2 > $NULL_DEV_PATH/submit_queues
echo 16384 > $NULL_DEV_PATH/size # 16G
echo 1 > $NULL_DEV_PATH/memory_backed
echo 1 > $NULL_DEV_PATH/power
DEV=/dev/nullb0
MNT=/mnt/nullb0
LOOP_MNT="$MNT/loop"
MOUNT_OPTIONS="-o ssd -o nodatacow"
MKFS_OPTIONS="-R free-space-tree -O no-holes"
cat <<EOF > /tmp/fio-job.ini
[io_uring_writes]
rw=randwrite
fsync=0
fallocate=posix
group_reporting=1
direct=1
ioengine=io_uring
iodepth=64
bs=64k
filesize=1g
runtime=300
time_based
directory=$LOOP_MNT
numjobs=8
thread
EOF
echo performance | \
tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
echo
echo "Using config:"
echo
cat /tmp/fio-job.ini
echo
umount $MNT &> /dev/null
mkfs.btrfs -f $MKFS_OPTIONS $DEV &> /dev/null
mount $MOUNT_OPTIONS $DEV $MNT
mkdir $LOOP_MNT
truncate -s 4T $MNT/loopfile
mkfs.btrfs -f $MKFS_OPTIONS $MNT/loopfile &> /dev/null
mount $MOUNT_OPTIONS $MNT/loopfile $LOOP_MNT
# Trigger the allocation of about 3500 data block groups, without
# actually consuming space on underlying filesystem, just to make
# the tree of block group large.
fallocate -l 3500G $LOOP_MNT/filler
fio /tmp/fio-job.ini
umount $LOOP_MNT
umount $MNT
echo 0 > $NULL_DEV_PATH/power
rmdir $NULL_DEV_PATH
The test was run on a non-debug kernel (Debian's default kernel config),
the result were the following.
Before patchset:
WRITE: bw=1455MiB/s (1526MB/s), 1455MiB/s-1455MiB/s (1526MB/s-1526MB/s), io=426GiB (458GB), run=300006-300006msec
After patchset:
WRITE: bw=1503MiB/s (1577MB/s), 1503MiB/s-1503MiB/s (1577MB/s-1577MB/s), io=440GiB (473GB), run=300006-300006msec
+3.3% write throughput and +3.3% IO done in the same time period.
The test has somewhat limited coverage scope, as with only NOCOW writes
we get less contention on the red black tree of block groups, since we
don't have the extra contention caused by COW writes, namely when
allocating data extents, pinning and unpinning data extents, but on the
hand there's access to tree in the NOCOW path, when incrementing a block
group's number of NOCOW writers.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
At btrfs_next_block_group(), we have this long line with two statements:
cache = btrfs_lookup_first_block_group(...); return cache;
This makes it a bit harder to read due to two statements on the same
line, so change that to directly return the result of the call to
btrfs_lookup_first_block_group().
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently we use a spin lock to protect the red black tree that we use to
track block groups. Most accesses to that tree are actually read only and
for large filesystems, with thousands of block groups, it actually has
a bad impact on performance, as concurrent read only searches on the tree
are serialized.
Read only searches on the tree are very frequent and done when:
1) Pinning and unpinning extents, as we need to lookup the respective
block group from the tree;
2) Freeing the last reference of a tree block, regardless if we pin the
underlying extent or add it back to free space cache/tree;
3) During NOCOW writes, both buffered IO and direct IO, we need to check
if the block group that contains an extent is read only or not and to
increment the number of NOCOW writers in the block group. For those
operations we need to search for the block group in the tree.
Similarly, after creating the ordered extent for the NOCOW write, we
need to decrement the number of NOCOW writers from the same block
group, which requires searching for it in the tree;
4) Decreasing the number of extent reservations in a block group;
5) When allocating extents and freeing reserved extents;
6) Adding and removing free space to the free space tree;
7) When releasing delalloc bytes during ordered extent completion;
8) When relocating a block group;
9) During fitrim, to iterate over the block groups;
10) etc;
Write accesses to the tree, to add or remove block groups, are much less
frequent as they happen only when allocating a new block group or when
deleting a block group.
We also use the same spin lock to protect the list of currently caching
block groups. Additions to this list are made when we need to cache a
block group, because we don't have a free space cache for it (or we have
but it's invalid), and removals from this list are done when caching of
the block group's free space finishes. These cases are also not very
common, but when they happen, they happen only once when the filesystem
is mounted.
So switch the lock that protects the tree of block groups from a spinning
lock to a read/write lock.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We keep track of the start offset of the block group with the lowest start
offset at fs_info->first_logical_byte. This requires explicitly updating
that field every time we add, delete or lookup a block group to/from the
red black tree at fs_info->block_group_cache_tree.
Since the block group with the lowest start address happens to always be
the one that is the leftmost node of the tree, we can use a red black tree
that caches the left most node. Then when we need the start address of
that block group, we can just quickly get the leftmost node in the tree
and extract the start offset of that node's block group. This avoids the
need to explicitly keep track of that address in the dedicated member
fs_info->first_logical_byte, and it also allows the next patch in the
series to switch the lock that protects the red black tree from a spin
lock to a read/write lock - without this change it would be tricky
because block group searches also update fs_info->first_logical_byte.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The search start argument passed to first_logical_byte() is always 0, as
we always want to get the logical start address of the block group with
the lowest logical start address. So remove it, as not only it is not
necessary, it also makes the following patches that change the lock that
protects the red black tree of block groups from a spin lock to a
read/write lock.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
If we hit an error from submit_extent_page() inside
__extent_writepage_io(), we could still return 0 to the caller, and
even trigger the warning in btrfs_page_assert_not_dirty().
[CAUSE]
In __extent_writepage_io(), if we hit an error from
submit_extent_page(), we will just clean up the range and continue.
This is completely fine for regular PAGE_SIZE == sectorsize, as we can
only hit one sector in one page, thus after the error we're ensured to
exit and @ret will be saved.
But for subpage case, we may have other dirty subpage range in the page,
and in the next loop, we may succeeded submitting the next range.
In that case, @ret will be overwritten, and we return 0 to the caller,
while we have hit some error.
[FIX]
Introduce @has_error and @saved_ret to record the first error we hit, so
we will never forget what error we hit.
CC: stable@vger.kernel.org # 5.15+
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
Test case generic/475 have a very high chance (almost 100%) to hit a fs
hang, where a data page will never be unlocked and hang all later
operations.
[CAUSE]
In btrfs_do_readpage(), if we hit an error from submit_extent_page() we
will try to do the cleanup for our current io range, and exit.
This works fine for PAGE_SIZE == sectorsize cases, but not for subpage.
For subpage btrfs_do_readpage() will lock the full page first, which can
contain several different sectors and extents:
btrfs_do_readpage()
|- begin_page_read()
| |- btrfs_subpage_start_reader();
| Now the page will have PAGE_SIZE / sectorsize reader pending,
| and the page is locked.
|
|- end_page_read() for different branches
| This function will reduce subpage readers, and when readers
| reach 0, it will unlock the page.
But when submit_extent_page() failed, we only cleanup the current
io range, while the remaining io range will never be cleaned up, and the
page remains locked forever.
[FIX]
Update the error handling of submit_extent_page() to cleanup all the
remaining subpage range before exiting the loop.
Please note that, now submit_extent_page() can only fail due to
sanity check in alloc_new_bio().
Thus regular IO errors are impossible to trigger the error path.
CC: stable@vger.kernel.org # 5.15+
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
When running generic/475 with 64K page size and 4K sector size, it has a
very high chance (almost 100%) to hang, with mostly data page locked but
no one is going to unlock it.
[CAUSE]
With commit 1784b7d502 ("btrfs: handle csum lookup errors properly on
reads"), if we failed to lookup checksum due to metadata IO error, we
will return error for btrfs_submit_data_bio().
This will cause the page to be unlocked twice in btrfs_do_readpage():
btrfs_do_readpage()
|- submit_extent_page()
| |- submit_one_bio()
| |- btrfs_submit_data_bio()
| |- if (ret) {
| |- bio->bi_status = ret;
| |- bio_endio(bio); }
| In the endio function, we will call end_page_read()
| and unlock_extent() to cleanup the subpage range.
|
|- if (ret) {
|- unlock_extent(); end_page_read() }
Here we unlock the extent and cleanup the subpage range
again.
For unlock_extent(), it's mostly double unlock safe.
But for end_page_read(), it's not, especially for subpage case,
as for subpage case we will call btrfs_subpage_end_reader() to reduce
the reader number, and use that to number to determine if we need to
unlock the full page.
If double accounted, it can underflow the number and leave the page
locked without anyone to unlock it.
[FIX]
The commit 1784b7d502 ("btrfs: handle csum lookup errors properly on
reads") itself is completely fine, it's our existing code not properly
handling the error from bio submission hook properly.
This patch will make submit_one_bio() to return void so that the callers
will never be able to do cleanup when bio submission hook fails.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This is an optimization for fix fee13fe965 ("btrfs: correct zstd
workspace manager lock to use spin_lock_bh()")
The critical region for wsm.lock is only accessed by the process context and
the softirq context.
Because in the soft interrupt, the critical section will not be
preempted by the soft interrupt again, there is no need to call
spin_lock_bh(&wsm.lock) to turn off the soft interrupt,
spin_lock(&wsm.lock) is enough for this situation.
Signed-off-by: Schspa Shi <schspa@gmail.com>
[ minor comment update ]
Signed-off-by: David Sterba <dsterba@suse.com>
We are still using the magic value of 2 at btrfs_create_new_inode(), but
there's now a constant for that, named BTRFS_DIR_START_INDEX, which was
introduced in commit 528ee69712 ("btrfs: put initial index value of a
directory in a constant"). So change that to use the constant.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Cleanup the function submit_read_repair() by:
- Remove the fixed argument submit_bio_hook()
The function is only called on buffered data read path, so the
@submit_bio_hook argument is always btrfs_submit_data_bio().
Since it's fixed, then there is no need to pass that argument at all.
- Rename the function to submit_data_read_repair()
Just to be more explicit on all the 3 things, data, read and repair.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Reading a value from a different member of a union is not just a great
way to obfuscate code, but also creates an aliasing violation. Switch
btrfs_is_zoned to look at ->zone_size and remove the union.
Note: union was to simplify the detection of zoned filesystem but now
this is wrapped behind btrfs_is_zoned so we can drop the union.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
[ add note ]
Signed-off-by: David Sterba <dsterba@suse.com>
iput() already handles NULL and non-NULL parameter, so it is not needed
to check that. This unifies all iput calls.
Reported-by: Zeal Robot <zealci@zte.com.cn>
Signed-off-by: Lv Ruyi <lv.ruyi@zte.com.cn>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The bios added to ->bio_list are the original bios fed into
btrfs_map_bio, which are never advanced. Just use the iter in the
bio itself.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
All the scrub bios go straight to the block device or the raid56 code,
none of which looks at the btrfs_bio.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Except for the spurious initialization of ->device just after allocation
nothing uses the btrfs_bio, so just allocate a normal bio without extra
data.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Prepare for further refactoring by moving this initialization to a
single place instead of setting it in the callers.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Pass the block_device to bio_alloc_clone instead of setting it later.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Prepare for additional refactoring, btrfs_map_bio is direct caller of
submit_stripe_bio.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The I/O in repair_io_failue is synchronous and doesn't need a btrfs_bio,
so just use an on-stack bio.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The I/O in repair_io_failue is synchronous and doesn't need a btrfs_bio,
so just use an on-stack bio.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The I/O in repair_io_failue is synchronous and doesn't need a btrfs_bio,
so just use an on-stack bio. Also cleanup the error handling to use goto
labels and not discard the actual return values.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfsic_read_block does not need the btrfs_bio structure, so switch to
plain bio_alloc (that also does not fail as it's backed by a bioset).
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Require a separate call to the integrity checking helpers from the
actual bio submission.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Split out two helpers to make __btrfsic_submit_bio more readable.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The current auto-reclaim algorithm starts reclaiming all block groups
with a zone_unusable value above a configured threshold. This is causing
a lot of reclaim IO even if there would be enough free zones on the
device.
Instead of only accounting a block groups zone_unusable value, also take
the ratio of free and not usable (written as well as zone_unusable)
bytes a device has into account.
Tested-by: Pankaj Raghav <p.raghav@samsung.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
For the non-zoned case we may want to set the threshold for reclaim to
something below 50%. Change the acceptable threshold from 50-100 to
0-100.
Tested-by: Pankaj Raghav <p.raghav@samsung.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This will allow us to set a threshold for block groups to be
automatically relocated even if we don't have zoned devices.
We have found this feature invaluable at Facebook due to how our
workload interacts with the allocator. We have been using this in
production for months with only a single problem that has already been
fixed.
Tested-by: Pankaj Raghav <p.raghav@samsung.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
For non-zoned file systems it's useful to have the auto reclaim feature,
however there are different use cases for non-zoned, for example we may
not want to reclaim metadata chunks ever, only data chunks. Move this
sysfs flag to per-space_info. This won't affect current users because
this tunable only ever did anything for zoned, and that is currently
hidden behind BTRFS_CONFIG_DEBUG.
Tested-by: Pankaj Raghav <p.raghav@samsung.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
[ jth restore global bg_reclaim_threshold ]
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When checking if we can do a NOCOW write against a range covered by a file
extent item, we do a quick a check to determine if the inode's root was
snapshotted in a generation older than the generation of the file extent
item or not. This is to quickly determine if the extent is likely shared
and avoid the expensive check for cross references (this was added in
commit 78d4295b1e ("btrfs: lift some btrfs_cross_ref_exist checks in
nocow path").
We restrict that check to the case where the inode is not a free space
inode (since commit 27a7ff554e ("btrfs: skip file_extent generation
check for free_space_inode in run_delalloc_nocow")). That is because when
we had the inode cache feature, inode caches were backed by a free space
inode that belonged to the inode's root.
However we don't have support for the inode cache feature since kernel
5.11, so we don't need this check anymore since free space inodes are
now always related to free space caches, which are always associated to
the root tree (which can't be snapshotted, and its last_snapshot field
is always 0).
So remove that condition.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Verifying if we can do a NOCOW write against a range fully or partially
covered by a file extent item requires verifying several constraints, and
these are currently duplicated at two different places: can_nocow_extent()
and run_delalloc_nocow().
This change moves those checks into a common helper function to avoid
duplication. It adds some comments and also preserves all existing
behaviour like for example can_nocow_extent() treating errors from the
calls to btrfs_cross_ref_exist() and csum_exist_in_range() as meaning
we can not NOCOW, instead of propagating the error back to the caller.
That specific behaviour is questionable but also reasonable to some
degree.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When allocating memory in a loop, each iteration should call
memalloc_retry_wait() in order to prevent starving memory-freeing
processes (and to mark where allocation loops are). Other filesystems do
that as well.
The bulk page allocation is the only place in btrfs with an allocation
retry loop, so add an appropriate call to it.
Signed-off-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
While calling alloc_page() in a loop is an effective way to populate an
array of pages, the MM subsystem provides a method to allocate pages in
bulk. alloc_pages_bulk_array() populates the NULL slots in a page
array, trying to grab more than one page at a time.
Unfortunately, it doesn't guarantee allocating all slots in the array,
but it's easy to call it in a loop and return an error if no progress
occurs.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Several functions currently populate an array of page pointers one
allocated page at a time. Factor out the common code so as to allow
improvements to all of the sites at once.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Explicit type casts are not necessary when it's void* to another pointer
type.
Signed-off-by: Yu Zhe <yuzhe@nfschina.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
With the recent change in metadata handling, we can handle metadata in
the following cases:
- nodesize < PAGE_SIZE and sectorsize < PAGE_SIZE
Go subpage routine for both metadata and data.
- nodesize < PAGE_SIZE and sectorsize >= PAGE_SIZE
Invalid case for now. As we require nodesize >= sectorsize.
- nodesize >= PAGE_SIZE and sectorsize < PAGE_SIZE
Go subpage routine for data, but regular page routine for metadata.
- nodesize >= PAGE_SIZE and sectorsize >= PAGE_SIZE
Go regular page routine for both metadata and data.
Now we can handle any sectorsize < PAGE_SIZE, plus the existing
sectorsize == PAGE_SIZE support.
But here we introduce an artificial limit, any PAGE_SIZE > 4K case, we
will only support 4K and PAGE_SIZE as sector size.
The idea here is to reduce the test combinations, and push 4K as the
default standard in the future.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The reason why we only support 64K page size for subpage is, for 64K
page size we can ensure no matter what the nodesize is, we can fit it
into one page.
When other page size come, especially like 16K, the limitation is a bit
limiting.
To remove such limitation, we allow nodesize >= PAGE_SIZE case to go the
non-subpage routine. By this, we can allow 4K sectorsize on 16K page
size.
Although this introduces another smaller limitation, the metadata can
not cross page boundary, which is already met by most recent mkfs.
Another small improvement is, we can avoid the overhead for metadata if
nodesize >= PAGE_SIZE.
For 4K sector size and 64K page size/node size, or 4K sector size and
16K page size/node size, we don't need to allocate extra memory for the
metadata pages.
Please note that, this patch will not yet enable other page size support
yet.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In function btrfs_read_sys_array(), we allocate a real extent buffer
using btrfs_find_create_tree_block().
Such extent buffer will be even cached into buffer_radix tree, and using
btree inode address space.
However we only use such extent buffer to enable the accessors, thus we
don't even need to bother using real extent buffer, a dummy one is
what we really need.
And for dummy extent buffer, we no longer need to do any special
handling for the first page, as subpage helper is already doing it
properly.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Relocation of a data block group creates ordered extents. They can cause
a hang when a process is trying to thaw the filesystem.
We should have called sb_start_write(), so the filesystem is not being
frozen. Add an ASSERT to check it is protected.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Move code in btrfs_ioctl_balance to simplify its flow. This is
possible thanks to the removal of balance v1 ioctl and ensuring 'arg'
argument is always present. First move the code duplicating the
userspace arg to the kernel 'barg'. This makes the out_unlock label
redundant. Secondly, check the validity of bargs::flags before copying
to the dynamically allocated 'bctl'. This removes the need for the
out_bctl label.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
With the removal of balance v1 ioctl the 'arg' argument is guaranteed to
be present so simply remove all conditional code which checks for its
presence.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The original code resets the page to 0x1 for not apparent reason, it's
been like that since the initial 2007 code added in commit 07157aacb1
("Btrfs: Add file data csums back in via hooks in the extent map code").
It could mean that a failed buffer can be detected from the data but
that's just a guess and any value is good.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ update changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
When doing a NOWAIT direct IO write, if we can NOCOW then it means we can
proceed with the non-blocking, NOWAIT path. However reserving the metadata
space and qgroup meta space can often result in blocking - flushing
delalloc, wait for ordered extents to complete, trigger transaction
commits, etc, going against the semantics of a NOWAIT write.
So make the NOWAIT write path to try to reserve all the metadata it needs
without resulting in a blocking behaviour - if we get -ENOSPC or -EDQUOT
then return -EAGAIN to make the caller fallback to a blocking direct IO
write.
This is part of a patchset comprised of the following patches:
btrfs: avoid blocking on page locks with nowait dio on compressed range
btrfs: avoid blocking nowait dio when locking file range
btrfs: avoid double nocow check when doing nowait dio writes
btrfs: stop allocating a path when checking if cross reference exists
btrfs: free path at can_nocow_extent() before checking for checksum items
btrfs: release path earlier at can_nocow_extent()
btrfs: avoid blocking when allocating context for nowait dio read/write
btrfs: avoid blocking on space revervation when doing nowait dio writes
The following test was run before and after applying this patchset:
$ cat io-uring-nodatacow-test.sh
#!/bin/bash
DEV=/dev/sdc
MNT=/mnt/sdc
MOUNT_OPTIONS="-o ssd -o nodatacow"
MKFS_OPTIONS="-R free-space-tree -O no-holes"
NUM_JOBS=4
FILE_SIZE=8G
RUN_TIME=300
cat <<EOF > /tmp/fio-job.ini
[io_uring_rw]
rw=randrw
fsync=0
fallocate=posix
group_reporting=1
direct=1
ioengine=io_uring
iodepth=64
bssplit=4k/20:8k/20:16k/20:32k/10:64k/10:128k/5:256k/5:512k/5:1m/5
filesize=$FILE_SIZE
runtime=$RUN_TIME
time_based
filename=foobar
directory=$MNT
numjobs=$NUM_JOBS
thread
EOF
echo performance | \
tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
umount $MNT &> /dev/null
mkfs.btrfs -f $MKFS_OPTIONS $DEV &> /dev/null
mount $MOUNT_OPTIONS $DEV $MNT
fio /tmp/fio-job.ini
umount $MNT
The test was run a 12 cores box with 64G of ram, using a non-debug kernel
config (Debian's default config) and a spinning disk.
Result before the patchset:
READ: bw=407MiB/s (427MB/s), 407MiB/s-407MiB/s (427MB/s-427MB/s), io=119GiB (128GB), run=300175-300175msec
WRITE: bw=407MiB/s (427MB/s), 407MiB/s-407MiB/s (427MB/s-427MB/s), io=119GiB (128GB), run=300175-300175msec
Result after the patchset:
READ: bw=436MiB/s (457MB/s), 436MiB/s-436MiB/s (457MB/s-457MB/s), io=128GiB (137GB), run=300044-300044msec
WRITE: bw=435MiB/s (456MB/s), 435MiB/s-435MiB/s (456MB/s-456MB/s), io=128GiB (137GB), run=300044-300044msec
That's about +7.2% throughput for reads and +6.9% for writes.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When doing a NOWAIT direct IO read/write, we allocate a context object
(struct btrfs_dio_data) with GFP_NOFS, which can result in blocking
waiting for memory allocation (GFP_NOFS is __GFP_RECLAIM | __GFP_IO).
This is undesirable for the NOWAIT semantics, so do the allocation with
GFP_NOWAIT if we are serving a NOWAIT request and if the allocation fails
return -EAGAIN, so that the caller can fallback to a blocking context and
retry with a non-blocking write.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
At can_nocow_extent(), we are releasing the path only after checking if
the block group that has the target extent is read only, and after
checking if there's delalloc in the range in case our extent is a
preallocated extent. The read only extent check can be expensive if we
have a very large filesystem with many block groups, as well as the
check for delalloc in the inode's io_tree in case the io_tree is big
due to IO on other file ranges.
Our path is holding a read lock on a leaf and there's no need to keep
the lock while doing those two checks, so release the path before doing
them, immediately after the last use of the leaf.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When we look for checksum items, through csum_exist_in_range(), at
can_nocow_extent(), we no longer need the path that we have previously
allocated. Through csum_exist_in_range() -> btrfs_lookup_csums_range(),
we also end up allocating a path, so we are adding unnecessary extra
memory usage. So free the path before calling csum_exist_in_range().
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
At btrfs_cross_ref_exist() we always allocate a path, but we really don't
need to because all its callers (only 2) already have an allocated path
that is not being used when they call btrfs_cross_ref_exist(). So change
btrfs_cross_ref_exist() to take a path as an argument and update both
its callers to pass in the unused path they have when they call
btrfs_cross_ref_exist().
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When doing a NOWAIT direct IO write we are checking twice if we can COW
into the target file range using can_nocow_extent() - once at the very
beginning of the write path, at btrfs_write_check() via
check_nocow_nolock(), and later again at btrfs_get_blocks_direct_write().
The can_nocow_extent() function does a lot of expensive things - searching
for the file extent item in the inode's subvolume tree, searching for the
extent item in the extent tree, checking delayed references, etc, so it
isn't a very cheap call.
We can remove the first check at btrfs_write_check(), and add there a
quick check to verify if the inode has the NODATACOW or PREALLOC flags,
and quickly bail out if it doesn't have neither of those flags, as that
means we have to COW and therefore can't comply with the NOWAIT semantics.
After this we do only one call to can_nocow_extent(), while we are at
btrfs_get_blocks_direct_write(), where we have already locked the file
range and we did a try lock on the range before, at
btrfs_dio_iomap_begin() (since the previous patch in the series).
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
If we are doing a NOWAIT direct IO read/write, we can block when locking
the file range at btrfs_dio_iomap_begin(), as it's possible the range (or
a part of it) is already locked by another task (mmap writes, another
direct IO read/write racing with us, fiemap, etc). We are also waiting for
completion of any ordered extent we find in the range, which also can
block us for a significant amount of time.
There's also the incorrect fallback to buffered IO (returning -ENOTBLK)
when we are dealing with a NOWAIT request and we can't proceed. In this
case we should be returning -EAGAIN, as falling back to buffered IO can
result in blocking for many different reasons, so that the caller can
delegate a retry to a context where blocking is more acceptable.
Fix these cases by:
1) Doing a try lock on the file range and failing with -EAGAIN if we
can not lock right away;
2) Fail with -EAGAIN if we find an ordered extent;
3) Return -EAGAIN instead of -ENOTBLK when we need to fallback to
buffered IO and we have a NOWAIT request.
This will also allow us to avoid a duplicated check that verifies if we
are able to do a NOCOW write for NOWAIT direct IO writes, done in the
next patch.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
If we are doing NOWAIT direct IO read/write and our inode has compressed
extents, we call filemap_fdatawrite_range() against the range in order
to wait for compressed writeback to complete, since the generic code at
iomap_dio_rw() calls filemap_write_and_wait_range() once, which is not
enough to wait for compressed writeback to complete.
This call to filemap_fdatawrite_range() can block on page locks, since
the first writepages() on a range that we will try to compress results
only in queuing a work to compress the data while holding the pages
locked.
Even though the generic code at iomap_dio_rw() will do the right thing
and return -EAGAIN for NOWAIT requests in case there are pages in the
range, we can still end up at btrfs_dio_iomap_begin() with pages in the
range because either of the following can happen:
1) Memory mapped writes, as we haven't locked the range yet;
2) Buffered reads might have started, which lock the pages, and we do
the filemap_fdatawrite_range() call before locking the file range.
So don't call filemap_fdatawrite_range() at btrfs_dio_iomap_begin() if we
are doing a NOWAIT read/write. Instead call filemap_range_needs_writeback()
to check if there are any locked, dirty, or under writeback pages, and
return -EAGAIN if that's the case.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In order for end users to quickly react to new issues that come up in
production, it is proving useful to leverage this printk indexing
system. This printk index enables kernel developers to use calls to
printk() with changeable ad-hoc format strings, while still enabling end
users to detect changes and develop a semi-stable interface for
detecting and parsing these messages.
So that detailed Btrfs messages are captured by this printk index, this
patch wraps btrfs_printk and btrfs_handle_fs_error with macros.
Example of the generated list:
https://lore.kernel.org/lkml/12588e13d51a9c3bf59467d3fc1ac2162f1275c1.1647539056.git.jof@thejof.com
Signed-off-by: Jonathan Lassoff <jof@thejof.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Btrfs doesn't check whether the tree block respects the root owner.
This means, if a tree block referred by a parent in extent tree, but has
owner of 5, btrfs can still continue reading the tree block, as long as
it doesn't trigger other sanity checks.
Normally this is fine, but combined with the empty tree check in
check_leaf(), if we hit an empty extent tree, but the root node has
csum tree owner, we can let such extent buffer to sneak in.
Shrink the hole by:
- Do extra eb owner check at tree read time
- Make sure the root owner extent buffer exactly matches the root id.
Unfortunately we can't yet completely patch the hole, there are several
call sites can't pass all info we need:
- For reloc/log trees
Their owner is key::offset, not key::objectid.
We need the full root key to do that accurate check.
For now, we just skip the ownership check for those trees.
- For add_data_references() of relocation
That call site doesn't have any parent/ownership info, as all the
bytenrs are all from btrfs_find_all_leafs().
- For direct backref items walk
Direct backref items records the parent bytenr directly, thus unlike
indirect backref item, we don't do a full tree search.
Thus in that case, we don't have full parent owner to check.
For the later two cases, they all pass 0 as @owner_root, thus we can
skip those cases if @owner_root is 0.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We have four different scenarios where we don't expect to find ordered
extents after locking a file range:
1) During plain fallocate;
2) During hole punching;
3) During zero range;
4) During reflinks (both cloning and deduplication).
This is because in all these cases we follow the pattern:
1) Lock the inode's VFS lock in exclusive mode;
2) Lock the inode's i_mmap_lock in exclusive node, to serialize with
mmap writes;
3) Flush delalloc in a file range and wait for all ordered extents
to complete - both done through btrfs_wait_ordered_range();
4) Lock the file range in the inode's io_tree.
So add a helper that asserts that we don't have ordered extents for a
given range. Make the four scenarios listed above use this helper after
locking the respective file range.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
For hole punching and zero range we have this loop that checks if we have
ordered extents after locking the file range, and if so unlock the range,
wait for ordered extents, and retry until we don't find more ordered
extents.
This logic was needed in the past because:
1) Direct IO writes within the i_size boundary did not take the inode's
VFS lock. This was because that lock used to be a mutex, then some
years ago it was switched to a rw semaphore (commit 9902af79c0
("parallel lookups: actual switch to rwsem")), and then btrfs was
changed to take the VFS inode's lock in shared mode for writes that
don't cross the i_size boundary (commit e9adabb971 ("btrfs: use
shared lock for direct writes within EOF"));
2) We could race with memory mapped writes, because memory mapped writes
don't acquire the inode's VFS lock. We don't have that race anymore,
as we have a rw semaphore to synchronize memory mapped writes with
fallocate (and reflinking too). That change happened with commit
8d9b4a162a ("btrfs: exclude mmap from happening during all
fallocate operations").
So stop looking for ordered extents after locking the file range when
doing hole punching and zero range operations.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When doing hole punching we are flushing delalloc and waiting for ordered
extents to complete before locking the inode (VFS lock and the btrfs
specific i_mmap_lock). This is fine because even if a write happens after
we call btrfs_wait_ordered_range() and before we lock the inode (call
btrfs_inode_lock()), we will notice the write at
btrfs_punch_hole_lock_range() and flush delalloc and wait for its ordered
extent.
We can however make this simpler by locking first the inode an then call
btrfs_wait_ordered_range(), which will allow us to remove the ordered
extent lookup logic from btrfs_punch_hole_lock_range() in the next patch.
It also makes the behaviour the same as plain fallocate, hole punching
and reflinks.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
For fallocate() we have this loop that checks if we have ordered extents
after locking the file range, and if so unlock the range, wait for ordered
extents, and retry until we don't find more ordered extents.
This logic was needed in the past because:
1) Direct IO writes within the i_size boundary did not take the inode's
VFS lock. This was because that lock used to be a mutex, then some
years ago it was switched to a rw semaphore (commit 9902af79c0
("parallel lookups: actual switch to rwsem")), and then btrfs was
changed to take the VFS inode's lock in shared mode for writes that
don't cross the i_size boundary (commit e9adabb971 ("btrfs: use
shared lock for direct writes within EOF"));
2) We could race with memory mapped writes, because memory mapped writes
don't acquire the inode's VFS lock. We don't have that race anymore,
as we have a rw semaphore to synchronize memory mapped writes with
fallocate (and reflinking too). That change happened with commit
8d9b4a162a ("btrfs: exclude mmap from happening during all
fallocate operations").
So stop looking for ordered extents after locking the file range when
doing a plain fallocate.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When starting a reflink operation we have these calls to inode_dio_wait()
which used to be needed because direct IO writes that don't cross the
i_size boundary did not take the inode's VFS lock, so we could race with
them and end up with ordered extents in target range after calling
btrfs_wait_ordered_range().
However that is not the case anymore, because the inode's VFS lock was
changed from a mutex to a rw semaphore, by commit 9902af79c0
("parallel lookups: actual switch to rwsem"), and several years later we
started to lock the inode's VFS lock in shared mode for direct IO writes
that don't cross the i_size boundary (commit e9adabb971 ("btrfs: use
shared lock for direct writes within EOF")).
So remove those inode_dio_wait() calls.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When starting a fallocate zero range operation, before getting the first
extent map for the range, we make a call to inode_dio_wait().
This logic was needed in the past because direct IO writes within the
i_size boundary did not take the inode's VFS lock. This was because that
lock used to be a mutex, then some years ago it was switched to a rw
semaphore (by commit 9902af79c0 ("parallel lookups: actual switch to
rwsem")), and then btrfs was changed to take the VFS inode's lock in
shared mode for writes that don't cross the i_size boundary (done in
commit e9adabb971 ("btrfs: use shared lock for direct writes within
EOF")). The lockless direct IO writes could result in a race with the
zero range operation, resulting in the later getting a stale extent
map for the range.
So remove this no longer needed call to inode_dio_wait(), as fallocate
takes the inode's VFS lock in exclusive mode and direct IO writes within
i_size take that same lock in shared mode.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
During a plain fallocate, we always start by reserving an amount of data
space that matches the length of the range passed to fallocate. When we
already have extents allocated in that range, we may end up trying to
reserve a lot more data space then we need, which can result in several
undesired behaviours:
1) We fail with -ENOSPC. For example the passed range has a length
of 1G, but there's only one hole with a size of 1M in that range;
2) We temporarily reserve excessive data space that could be used by
other operations happening concurrently;
3) By reserving much more data space then we need, we can end up
doing expensive things like triggering dellaloc for other inodes,
waiting for the ordered extents to complete, trigger transaction
commits, allocate new block groups, etc.
Example:
$ cat test.sh
#!/bin/bash
DEV=/dev/sdj
MNT=/mnt/sdj
mkfs.btrfs -f -b 1g $DEV
mount $DEV $MNT
# Create a file with a size of 600M and two holes, one at [200M, 201M[
# and another at [401M, 402M[
xfs_io -f -c "pwrite -S 0xab 0 200M" \
-c "pwrite -S 0xcd 201M 200M" \
-c "pwrite -S 0xef 402M 198M" \
$MNT/foobar
# Now call fallocate against the whole file range, see if it fails
# with -ENOSPC or not - it shouldn't since we only need to allocate
# 2M of data space.
xfs_io -c "falloc 0 600M" $MNT/foobar
umount $MNT
$ ./test.sh
(...)
wrote 209715200/209715200 bytes at offset 0
200 MiB, 51200 ops; 0.8063 sec (248.026 MiB/sec and 63494.5831 ops/sec)
wrote 209715200/209715200 bytes at offset 210763776
200 MiB, 51200 ops; 0.8053 sec (248.329 MiB/sec and 63572.3172 ops/sec)
wrote 207618048/207618048 bytes at offset 421527552
198 MiB, 50688 ops; 0.7925 sec (249.830 MiB/sec and 63956.5548 ops/sec)
fallocate: No space left on device
$
So fix this by not allocating an amount of data space that matches the
length of the range passed to fallocate. Instead allocate an amount of
data space that corresponds to the sum of the sizes of each hole found
in the range. This reservation now happens after we have locked the file
range, which is safe since we know at this point there's no delalloc
in the range because we've taken the inode's VFS lock in exclusive mode,
we have taken the inode's i_mmap_lock in exclusive mode, we have flushed
delalloc and waited for all ordered extents in the range to complete.
This type of failure actually seems to happen in practice with systemd,
and we had at least one report about this in a very long thread which
is referenced by the Link tag below.
Link: https://lore.kernel.org/linux-btrfs/bdJVxLiFr_PyQSXRUbZJfFW_jAjsGgoMetqPHJMbg-hdy54Xt_ZHhRetmnJ6cJ99eBlcX76wy-AvWwV715c3YndkxneSlod11P1hlaADx0s=@protonmail.com/
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
According to the tree checker, "all xattrs with a given objectid follow
the inode with that objectid in the tree" is an invariant. This was
broken by the recent change "btrfs: move common inode creation code into
btrfs_create_new_inode()", which moved acl creation and property
inheritance (stored in xattrs) to before inode insertion into the tree.
As a result, under certain timings, the xattrs could be written to the
tree before the inode, causing the tree checker to report violation of
the invariant.
Move property inheritance and acl creation back to their old ordering
after the inode insertion.
Suggested-by: Omar Sandoval <osandov@osandov.com>
Reported-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: David Sterba <dsterba@suse.com>
All of our inode creation code paths duplicate the calls to
btrfs_init_inode_security() and btrfs_add_link(). Subvolume creation
additionally duplicates property inheritance and the call to
btrfs_set_inode_index(). Fix this by moving the common code into
btrfs_create_new_inode(). This accomplishes a few things at once:
1. It reduces code duplication.
2. It allows us to set up the inode completely before inserting the
inode item, removing calls to btrfs_update_inode().
3. It fixes a leak of an inode on disk in some error cases. For example,
in btrfs_create(), if btrfs_new_inode() succeeds, then we have
inserted an inode item and its inode ref. However, if something after
that fails (e.g., btrfs_init_inode_security()), then we end the
transaction and then decrement the link count on the inode. If the
transaction is committed and the system crashes before the failed
inode is deleted, then we leak that inode on disk. Instead, this
refactoring aborts the transaction when we can't recover more
gracefully.
4. It exposes various ways that subvolume creation diverges from mkdir
in terms of inheriting flags, properties, permissions, and POSIX
ACLs, a lot of which appears to be accidental. This patch explicitly
does _not_ change the existing non-standard behavior, but it makes
those differences more clear in the code and documents them so that
we can discuss whether they should be changed.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The various inode creation code paths do not account for the compression
property, POSIX ACLs, or the parent inode item when starting a
transaction. Fix it by refactoring all of these code paths to use a new
function, btrfs_new_inode_prepare(), which computes the correct number
of items. To do so, it needs to know whether POSIX ACLs will be created,
so move the ACL creation into that function. To reduce the number of
arguments that need to be passed around for inode creation, define
struct btrfs_new_inode_args containing all of the relevant information.
btrfs_new_inode_prepare() will also be a good place to set up the
fscrypt context and encrypted filename in the future.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_{mknod,create,mkdir}() are now identical other than the inode
initialization and some inconsequential function call order differences.
Factor out the common code to reduce code duplication.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Instead of calling new_inode() and inode_init_owner() inside of
btrfs_new_inode(), do it in the callers. This allows us to pass in just
the inode instead of the mnt_userns and mode and removes the need for
memalloc_nofs_{save,restores}() since we do it before starting a
transaction. In create_subvol(), it also means we no longer have to look
up the inode again to instantiate it. This also paves the way for some
more cleanups in later patches.
This also removes the comments about Smack checking i_op, which are no
longer true since commit 5d6c31910b ("xattr: Add
__vfs_{get,set,remove}xattr helpers"). Now it checks inode->i_opflags &
IOP_XATTR, which is set based on sb->s_xattr.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Although we have btrfs_extent_buffer_leak_debug_check() (enabled by
CONFIG_BTRFS_DEBUG option) to detect and warn QA testers that we have
some extent buffer leakage, it's just pr_err(), not noisy enough for
fstests to cache.
So here we trigger a WARN_ON() if the allocated_ebs list is not empty.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In the function btrfs_dev_replace_finishing, we dereferenced
fs_info->fs_devices 6 times. Use keep local variable for that.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function can be simplified by refactoring to use the new iterator
macro. No functional changes.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There is a common pattern when searching for a key in btrfs:
* Call btrfs_search_slot to find the slot for the key
* Enter an endless loop:
* If the found slot is larger than the no. of items in the current
leaf, check the next leaf
* If it's still not found in the next leaf, terminate the loop
* Otherwise do something with the found key
* Increment the current slot and continue
To reduce code duplication, we can replace this code pattern with an
iterator macro, similar to the existing for_each_X macros found
elsewhere in the kernel. This also makes the code easier to understand
for newcomers by putting a name to the encapsulated functionality.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Gabriel Niebler <gniebler@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since the subpage support for scrub, one page no longer always represents
one sector, thus scrub_bio::pagev and scrub_bio::sector_count are no
longer accurate.
Rename them to scrub_bio::sectors and scrub_bio::sector_count respectively.
This also involves scrub_ctx::pages_per_bio and other macros involved.
Now the renaming of pages involved in scrub is be finished.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since the subpage support of scrub, scrub_sector is in fact just
representing one sector.
Thus the name scrub_page is no longer correct, rename it to
scrub_sector.
This also involves the following renames:
- spage -> sector
Normally we would just replace "page" with "sector" and result
something like "ssector".
But the repeating 's' is not really eye friendly.
So here we just simple use "sector", as there is nothing from MM layer
called "sector" to cause any confusion.
- scrub_parity::spages -> sectors_list
Normally we use plural to indicate an array, not a list.
Rename it to @sectors_list to be more explicit on the list part.
- Also reformat and update comments that get changed
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The following will be renamed in this patch:
- scrub_block::pagev -> sectors
- scrub_block::page_count -> sector_count
- SCRUB_MAX_PAGES_PER_BLOCK -> SCRUB_MAX_SECTORS_PER_BLOCK
- page_num -> sector_num to iterate scrub_block::sectors
For now scrub_page is not yet renamed to keep the patch reasonable and
it will be updated in a followup.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The function btrfs_read_buffer() is useless, it just calls
btree_read_extent_buffer_pages() with exactly the same arguments.
So remove it and rename btree_read_extent_buffer_pages() to
btrfs_read_extent_buffer(), which is a shorter name, has the "btrfs_"
prefix (since it's used outside disk-io.c) and the name is clear enough
about what it does.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The comment at the top of read_block_for_search() is very outdated, as it
refers to the blocking versus spinning path locking modes. We no longer
have these two locking modes after we switched the btree locks from custom
code to rw semaphores. So update the comment to stop referring to the
blocking mode and put it more up to date.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When reading a btree node (or leaf), at read_block_for_search(), if we
can't find its extent buffer in the cache (the fs_info->buffer_radix
radix tree), then we unlock all upper level nodes before reading the
btree node/leaf from disk, to prevent blocking other tasks for too long.
However if we find that the extent buffer is in the cache but it is not
up to date, we don't unlock upper level nodes before reading it from disk,
potentially blocking other tasks on upper level nodes for too long.
Fix this inconsistent behaviour by unlocking upper level nodes if we need
to read a node/leaf from disk because its in-memory extent buffer is not
up to date. If we unlocked upper level nodes then we must return -EAGAIN
to the caller, just like the case where the extent buffer is not cached in
memory. And like that case, we determine if upper level nodes are locked
by checking only if the parent node is locked - if it isn't, then no other
upper level nodes are locked.
This is actually a rare case, as if we have an extent buffer in memory,
it typically has the uptodate flag set and passes all the checks done by
btrfs_buffer_uptodate().
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When reading a btree node, at read_block_for_search(), if we don't find
the node's (or leaf) extent buffer in the cache, we will read it from
disk. Since that requires waiting on IO, we release all upper level nodes
from our path before reading the target node/leaf, and then return -EAGAIN
to the caller, which will make the caller restart the while btree search.
However we are causing the restart of btree search even for cases where
it is not necessary:
1) We have a path with ->skip_locking set to true, typically when doing
a search on a commit root, so we are never holding locks on any node;
2) We are doing a read search (the "ins_len" argument passed to
btrfs_search_slot() is 0), or we are doing a search to modify an
existing key (the "cow" argument passed to btrfs_search_slot() has
a value of 1 and "ins_len" is 0), in which case we never hold locks
for upper level nodes;
3) We are doing a search to insert or delete a key, in which case we may
or may not have upper level nodes locked. That depends on the current
minimum write lock levels at btrfs_search_slot(), if we had to split
or merge parent nodes, if we had to COW upper level nodes and if
we ever visited slot 0 of an upper level node. It's still common to
not have upper level nodes locked, but our current node must be at
least at level 1, for insertions, or at least at level 2 for deletions.
In these cases when we have locks on upper level nodes, they are always
write locks.
These cases where we are not holding locks on upper level nodes far
outweigh the cases where we are holding locks, so it's completely wasteful
to retry the whole search when we have no upper nodes locked.
So change the logic to not return -EAGAIN, and make the caller retry the
search, when we don't have the parent node locked - when it's not locked
it means no other upper level nodes are locked as well.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_new_inode() inherits the inode flags from the parent directory and
the mount options _after_ we fill the inode item. This works because all
of the callers of btrfs_new_inode() make further changes to the inode
and then call btrfs_update_inode(). It'd be better to fully initialize
the inode once to avoid the extra update, so as a first step, set the
inode flags _before_ filling the inode item.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Every call of btrfs_new_inode() is immediately preceded by a call to
btrfs_get_free_objectid(). Since getting an inode number is part of
creating a new inode, this is better off being moved into
btrfs_new_inode(). While we're here, get rid of the comment about
reclaiming inode numbers, since we only did that when using the ino
cache, which was removed by commit 5297199a8b ("btrfs: remove inode
number cache feature").
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
For everything other than a subvolume root inode, we get the parent
objectid from the parent directory. For the subvolume root inode, the
parent objectid is the same as the inode's objectid. We can find this
within btrfs_new_inode() instead of passing it.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Commit 4a8b34afa9 ("btrfs: handle ACLs on idmapped mounts") added this
parameter but didn't use it. __btrfs_set_acl() is the low-level helper
that writes an ACL to disk. The higher-level btrfs_set_acl() is the one
that translates the ACL based on the user namespace.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_new_inode() already returns an inode with nlink set to 1 (via
inode_init_always()). Get rid of the unnecessary set.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
new_inode() always returns an inode with i_blocks and i_bytes set to 0
(via inode_init_always()). Remove the unnecessary call to
inode_set_bytes() in btrfs_new_inode().
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_new_inode() always returns an inode with i_size and disk_i_size
set to 0 (via inode_init_always() and btrfs_alloc_inode(),
respectively). Remove the unnecessary calls to btrfs_i_size_write() in
btrfs_mkdir() and btrfs_create_subvol_root().
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This is a trivial wrapper around btrfs_add_link(). The only thing it
does other than moving arguments around is translating a > 0 return
value to -EEXIST. As far as I can tell, btrfs_add_link() won't return >
0 (and if it did, the existing callsites in, e.g., btrfs_mkdir() would
be broken). The check itself dates back to commit 2c90e5d658 ("Btrfs:
still corruption hunting"), so it's probably left over from debugging.
Let's just get rid of btrfs_add_nondir().
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When btrfs_qgroup_inherit(), btrfs_alloc_tree_block, or
btrfs_insert_root() fail in create_subvol(), we return without freeing
anon_dev. Reorganize the error handling in create_subvol() to fix this.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_rename() and btrfs_rename_exchange() don't account for enough
items. Replace the incorrect explanations with a specific breakdown of
the number of items and account them accurately.
Note that this glosses over RENAME_WHITEOUT because the next commit is
going to rework that, too.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
__btrfs_unlink_inode() calls btrfs_update_inode() on the parent
directory in order to update its size and sequence number. Make sure we
account for it.
Reviewed-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The m68k architecture code is capable of supporting the binfmt_elf_fdpic
loader, so allow it to be configured. It is restricted to nommu
configurations at this time due to the MMU context structures/code not
supporting everything elf_fdpic needs when MMU is enabled.
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
Currently we have 3 primitives for removing an opened file from descriptor
table - pick_file(), __close_fd_get_file() and close_fd_get_file(). Their
calling conventions are rather odd and there's a code duplication for no
good reason. They can be unified -
1) have __range_close() cap max_fd in the very beginning; that way
we don't need separate way for pick_file() to report being past the end
of descriptor table.
2) make {__,}close_fd_get_file() return file (or NULL) directly, rather
than returning it via struct file ** argument. Don't bother with
(bogus) return value - nobody wants that -ENOENT.
3) make pick_file() return NULL on unopened descriptor - the only caller
that used to care about the distinction between descriptor past the end
of descriptor table and finding NULL in descriptor table doesn't give
a damn after (1).
4) lift ->files_lock out of pick_file()
That actually simplifies the callers, as well as the primitives themselves.
Code duplication is also gone...
Reviewed-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
These two interface were added in 091141a42 commit,
but now there is no place to call them.
The only user of fput/fget_many() was removed in commit
62906e89e6 ("io_uring: remove file batch-get optimisation").
A user of get_file_rcu_many() were removed in commit
f073531070 ("init: add an init_dup helper").
And replace atomic_long_sub/add to atomic_long_dec/inc
can improve performance.
Here are the test results of unixbench:
Cmd: ./Run -c 64 context1
Without patch:
System Benchmarks Partial Index BASELINE RESULT INDEX
Pipe-based Context Switching 4000.0 2798407.0 6996.0
========
System Benchmarks Index Score (Partial Only) 6996.0
With patch:
System Benchmarks Partial Index BASELINE RESULT INDEX
Pipe-based Context Switching 4000.0 3486268.8 8715.7
========
System Benchmarks Index Score (Partial Only) 8715.7
Signed-off-by: Gou Hao <gouhao@uniontech.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Refactor io_accept() to support multishot mode.
theoretical analysis:
1) when connections come in fast
- singleshot:
add accept sqe(userspace) --> accept inline
^ |
|-----------------|
- multishot:
add accept sqe(userspace) --> accept inline
^ |
|--*--|
we do accept repeatedly in * place until get EAGAIN
2) when connections come in at a low pressure
similar thing like 1), we reduce a lot of userspace-kernel context
switch and useless vfs_poll()
tests:
Did some tests, which goes in this way:
server client(multiple)
accept connect
read write
write read
close close
Basically, raise up a number of clients(on same machine with server) to
connect to the server, and then write some data to it, the server will
write those data back to the client after it receives them, and then
close the connection after write return. Then the client will read the
data and then close the connection. Here I test 10000 clients connect
one server, data size 128 bytes. And each client has a go routine for
it, so they come to the server in short time.
test 20 times before/after this patchset, time spent:(unit cycle, which
is the return value of clock())
before:
1930136+1940725+1907981+1947601+1923812+1928226+1911087+1905897+1941075
+1934374+1906614+1912504+1949110+1908790+1909951+1941672+1969525+1934984
+1934226+1914385)/20.0 = 1927633.75
after:
1858905+1917104+1895455+1963963+1892706+1889208+1874175+1904753+1874112
+1874985+1882706+1884642+1864694+1906508+1916150+1924250+1869060+1889506
+1871324+1940803)/20.0 = 1894750.45
(1927633.75 - 1894750.45) / 1927633.75 = 1.65%
Signed-off-by: Hao Xu <howeyxu@tencent.com>
Link: https://lore.kernel.org/r/20220514142046.58072-5-haoxu.linux@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
For operations like accept, multishot is a useful feature, since we can
reduce a number of accept sqe. Let's integrate it to fast poll, it may
be good for other operations in the future.
Signed-off-by: Hao Xu <howeyxu@tencent.com>
Link: https://lore.kernel.org/r/20220514142046.58072-4-haoxu.linux@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a flag to indicate multishot mode for fast poll. currently only
accept use it, but there may be more operations leveraging it in the
future. Also add a mask IO_APOLL_MULTI_POLLED which stands for
REQ_F_APOLL_MULTI | REQ_F_POLLED, to make the code short and cleaner.
Signed-off-by: Hao Xu <howeyxu@tencent.com>
Link: https://lore.kernel.org/r/20220514142046.58072-3-haoxu.linux@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
To move the list iterator variable into the list_for_each_entry_*()
macro in the future it should be avoided to use the list iterator
variable after the loop body.
To *never* use the list iterator variable after the loop it was
concluded to use a separate iterator variable instead of a
found boolean [1].
This removes the need to use a found variable and simply checking if
the variable was set, can determine if the break/goto was hit.
Link: https://lore.kernel.org/all/CAHk-=wgRr_D8CB-D9Kg-c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/ [1]
Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
protected_* files have 600 permissions which prevents non-superuser from
reading them.
Container like "AWS greengrass" refuse to launch unless
protected_hardlinks and protected_symlinks are set. When containers like
these run with "userns-remap" or "--user" mapping container's root to
non-superuser on host, they fail to run due to denied read access to these
files.
As these protections are hardly a secret, and do not possess any security
risk, making them world readable.
Though above greengrass usecase needs read access to only
protected_hardlinks and protected_symlinks files, setting all other
protected_* files to 644 to keep consistency.
Link: http://lkml.kernel.org/r/20200709235115.56954-1-jpitti@cisco.com
Fixes: 800179c9b8 ("fs: add link restrictions")
Signed-off-by: Julius Hemanth Pitti <jpitti@cisco.com>
Acked-by: Kees Cook <keescook@chromium.org>
Acked-by: Luis Chamberlain <mcgrof@kernel.org>
Cc: Iurii Zaikin <yzaikin@google.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
- Fix filesystem block deallocation for short writes.
- Stop using glock holder auto-demotion for now.
- Get rid of buffered writes inefficiencies due to page
faults being disabled.
- Minor other cleanups.
-----BEGIN PGP SIGNATURE-----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=FJq+
-----END PGP SIGNATURE-----
Merge tag 'gfs2-v5.18-rc4-fix3' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2
Pull gfs2 fixes from Andreas Gruenbacher:
"We've finally identified commit dc732906c2 ("gfs2: Introduce flag
for glock holder auto-demotion") to be the other cause of the
filesystem corruption we've been seeing. This feature isn't strictly
necessary anymore, so we've decided to stop using it for now.
With this and the gfs_iomap_end rounding fix you've already seen
("gfs2: Fix filesystem block deallocation for short writes" in this
pull request), we're corruption free again now.
- Fix filesystem block deallocation for short writes.
- Stop using glock holder auto-demotion for now.
- Get rid of buffered writes inefficiencies due to page faults being
disabled.
- Minor other cleanups"
* tag 'gfs2-v5.18-rc4-fix3' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2:
gfs2: Stop using glock holder auto-demotion for now
gfs2: buffered write prefaulting
gfs2: Align read and write chunks to the page cache
gfs2: Pull return value test out of should_fault_in_pages
gfs2: Clean up use of fault_in_iov_iter_{read,write}able
gfs2: Variable rename
gfs2: Fix filesystem block deallocation for short writes
Now that we have kernel message at mount time, system administrator
could acquire the mount time, device and options easily. But we don't
have corresponding unmounting message at umount time, so we cannot know
if someone umount a filesystem easily. Some of the modern filesystems
(e.g. xfs) have the umounting kernel message, so add one for ext4
filesystem for convenience.
EXT4-fs (sdb): mounted filesystem with ordered data mode. Quota mode: none.
EXT4-fs (sdb): unmounting filesystem.
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20220412145320.2669897-1-yi.zhang@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
The check for waking up a request compares the poll_t bits, however this
will always contain some common flags so this always wakes up.
For files with single wait queues such as sockets this can cause the
request to be sent to the async worker unnecesarily. Further if it is
non-blocking will complete the request with EAGAIN which is not desired.
Here exclude these common events, making sure to not exclude POLLERR which
might be important.
Fixes: d7718a9d25 ("io_uring: use poll driven retry for files that support it")
Signed-off-by: Dylan Yudaken <dylany@fb.com>
Link: https://lore.kernel.org/r/20220512091834.728610-3-dylany@fb.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We're having unresolved issues with the glock holder auto-demotion mechanism
introduced in commit dc732906c2. This mechanism was assumed to be essential
for avoiding frequent short reads and writes until commit 296abc0d91
("gfs2: No short reads or writes upon glock contention"). Since then,
when the inode glock is lost, it is simply re-acquired and the operation
is resumed. This means that apart from the performance penalty, we
might as well drop the inode glock before faulting in pages, and
re-acquire it afterwards.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Chuck Lever
Gao Xiang