[ Upstream commit 22650f148126571be1098d34160eb4931fc77241 ]
The generic/464 xfstest causes kAFS to emit occasional warnings of the
form:
kAFS: vnode modified {100055:8a} 30->31 YFS.StoreData64 (c=6015)
This indicates that the data version received back from the server did not
match the expected value (the DV should be incremented monotonically for
each individual modification op committed to a vnode).
What is happening is that a lookup call is doing a bulk status fetch
speculatively on a bunch of vnodes in a directory besides getting the
status of the vnode it's actually interested in. This is racing with a
StoreData operation (though it could also occur with, say, a MakeDir op).
On the client, a modification operation locks the vnode, but the bulk
status fetch only locks the parent directory, so no ordering is imposed
there (thereby avoiding an avenue to deadlock).
On the server, the StoreData op handler doesn't lock the vnode until it's
received all the request data, and downgrades the lock after committing the
data until it has finished sending change notifications to other clients -
which allows the status fetch to occur before it has finished.
This means that:
- a status fetch can access the target vnode either side of the exclusive
section of the modification
- the status fetch could start before the modification, yet finish after,
and vice-versa.
- the status fetch and the modification RPCs can complete in either order.
- the status fetch can return either the before or the after DV from the
modification.
- the status fetch might regress the locally cached DV.
Some of these are handled by the previous fix[1], but that's not sufficient
because it checks the DV it received against the DV it cached at the start
of the op, but the DV might've been updated in the meantime by a locally
generated modification op.
Fix this by the following means:
(1) Keep track of when we're performing a modification operation on a
vnode. This is done by marking vnode parameters with a 'modification'
note that causes the AFS_VNODE_MODIFYING flag to be set on the vnode
for the duration.
(2) Alter the speculation race detection to ignore speculative status
fetches if either the vnode is marked as being modified or the data
version number is not what we expected.
Note that whilst the "vnode modified" warning does get recovered from as it
causes the client to refetch the status at the next opportunity, it will
also invalidate the pagecache, so changes might get lost.
Fixes: a9e5c87ca7 ("afs: Fix speculative status fetch going out of order wrt to modifications")
Reported-by: Marc Dionne <marc.dionne@auristor.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-and-reviewed-by: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
Link: https://lore.kernel.org/r/160605082531.252452.14708077925602709042.stgit@warthog.procyon.org.uk/ [1]
Link: https://lore.kernel.org/linux-fsdevel/161961335926.39335.2552653972195467566.stgit@warthog.procyon.org.uk/ # v1
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit a7889c6320b9200e3fe415238f546db677310fa9 upstream.
afs_listxattr() lists all the available special afs xattrs (i.e. those in
the "afs.*" space), no matter what type of server we're dealing with. But
OpenAFS servers, for example, cannot deal with some of the extra-capable
attributes that AuriStor (YFS) servers provide. Unfortunately, the
presence of the afs.yfs.* attributes causes errors[1] for anything that
tries to read them if the server is of the wrong type.
Fix the problem by removing afs_listxattr() so that none of the special
xattrs are listed (AFS doesn't support xattrs). It does mean, however,
that getfattr won't list them, though they can still be accessed with
getxattr() and setxattr().
This can be tested with something like:
getfattr -d -m ".*" /afs/example.com/path/to/file
With this change, none of the afs.* attributes should be visible.
Changes:
ver #2:
- Hide all of the afs.* xattrs, not just the ACL ones.
Fixes: ae46578b96 ("afs: Get YFS ACLs and information through xattrs")
Reported-by: Gaja Sophie Peters <gaja.peters@math.uni-hamburg.de>
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-by: Gaja Sophie Peters <gaja.peters@math.uni-hamburg.de>
Reviewed-by: Jeffrey Altman <jaltman@auristor.com>
Reviewed-by: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
Link: http://lists.infradead.org/pipermail/linux-afs/2021-March/003502.html [1]
Link: http://lists.infradead.org/pipermail/linux-afs/2021-March/003567.html # v1
Link: http://lists.infradead.org/pipermail/linux-afs/2021-March/003573.html # v2
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When doing a lookup in a directory, the afs filesystem uses a bulk
status fetch to speculatively retrieve the statuses of up to 48 other
vnodes found in the same directory and it will then either update extant
inodes or create new ones - effectively doing 'lookup ahead'.
To avoid the possibility of deadlocking itself, however, the filesystem
doesn't lock all of those inodes; rather just the directory inode is
locked (by the VFS).
When the operation completes, afs_inode_init_from_status() or
afs_apply_status() is called, depending on whether the inode already
exists, to commit the new status.
A case exists, however, where the speculative status fetch operation may
straddle a modification operation on one of those vnodes. What can then
happen is that the speculative bulk status RPC retrieves the old status,
and whilst that is happening, the modification happens - which returns
an updated status, then the modification status is committed, then we
attempt to commit the speculative status.
This results in something like the following being seen in dmesg:
kAFS: vnode modified {100058:861} 8->9 YFS.InlineBulkStatus
showing that for vnode 861 on volume 100058, we saw YFS.InlineBulkStatus
say that the vnode had data version 8 when we'd already recorded version
9 due to a local modification. This was causing the cache to be
invalidated for that vnode when it shouldn't have been. If it happens
on a data file, this might lead to local changes being lost.
Fix this by ignoring speculative status updates if the data version
doesn't match the expected value.
Note that it is possible to get a DV regression if a volume gets
restored from a backup - but we should get a callback break in such a
case that should trigger a recheck anyway. It might be worth checking
the volume creation time in the volsync info and, if a change is
observed in that (as would happen on a restore), invalidate all caches
associated with the volume.
Fixes: 5cf9dd55a0 ("afs: Prospectively look up extra files when doing a single lookup")
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The dirty region bounds stored in page->private on an afs page are 15 bits
on a 32-bit box and can, at most, represent a range of up to 32K within a
32K page with a resolution of 1 byte. This is a problem for powerpc32 with
64K pages enabled.
Further, transparent huge pages may get up to 2M, which will be a problem
for the afs filesystem on all 32-bit arches in the future.
Fix this by decreasing the resolution. For the moment, a 64K page will
have a resolution determined from PAGE_SIZE. In the future, the page will
need to be passed in to the helper functions so that the page size can be
assessed and the resolution determined dynamically.
Note that this might not be the ideal way to handle this, since it may
allow some leakage of undirtied zero bytes to the server's copy in the case
of a 3rd-party conflict. Fixing that would require a separately allocated
record and is a more complicated fix.
Fixes: 4343d00872 ("afs: Get rid of the afs_writeback record")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Fix afs_invalidatepage() to adjust the dirty region recorded in
page->private when truncating a page. If the dirty region is entirely
removed, then the private data is cleared and the page dirty state is
cleared.
Without this, if the page is truncated and then expanded again by truncate,
zeros from the expanded, but no-longer dirty region may get written back to
the server if the page gets laundered due to a conflicting 3rd-party write.
It mustn't, however, shorten the dirty region of the page if that page is
still mmapped and has been marked dirty by afs_page_mkwrite(), so a flag is
stored in page->private to record this.
Fixes: 4343d00872 ("afs: Get rid of the afs_writeback record")
Signed-off-by: David Howells <dhowells@redhat.com>
Currently, page->private on an afs page is used to store the range of
dirtied data within the page, where the range includes the lower bound, but
excludes the upper bound (e.g. 0-1 is a range covering a single byte).
This, however, requires a superfluous bit for the last-byte bound so that
on a 4KiB page, it can say 0-4096 to indicate the whole page, the idea
being that having both numbers the same would indicate an empty range.
This is unnecessary as the PG_private bit is clear if it's an empty range
(as is PG_dirty).
Alter the way the dirty range is encoded in page->private such that the
upper bound is reduced by 1 (e.g. 0-0 is then specified the same single
byte range mentioned above).
Applying this to both bounds frees up two bits, one of which can be used in
a future commit.
This allows the afs filesystem to be compiled on ppc32 with 64K pages;
without this, the following warnings are seen:
../fs/afs/internal.h: In function 'afs_page_dirty_to':
../fs/afs/internal.h:881:15: warning: right shift count >= width of type [-Wshift-count-overflow]
881 | return (priv >> __AFS_PAGE_PRIV_SHIFT) & __AFS_PAGE_PRIV_MASK;
| ^~
../fs/afs/internal.h: In function 'afs_page_dirty':
../fs/afs/internal.h:886:28: warning: left shift count >= width of type [-Wshift-count-overflow]
886 | return ((unsigned long)to << __AFS_PAGE_PRIV_SHIFT) | from;
| ^~
Fixes: 4343d00872 ("afs: Get rid of the afs_writeback record")
Signed-off-by: David Howells <dhowells@redhat.com>
The afs filesystem uses page->private to store the dirty range within a
page such that in the event of a conflicting 3rd-party write to the server,
we write back just the bits that got changed locally.
However, there are a couple of problems with this:
(1) I need a bit to note if the page might be mapped so that partial
invalidation doesn't shrink the range.
(2) There aren't necessarily sufficient bits to store the entire range of
data altered (say it's a 32-bit system with 64KiB pages or transparent
huge pages are in use).
So wrap the accesses in inline functions so that future commits can change
how this works.
Also move them out of the tracing header into the in-directory header.
There's not really any need for them to be in the tracing header.
Signed-off-by: David Howells <dhowells@redhat.com>
Fix afs_launder_page() to not clear PG_writeback on the page it is
laundering as the flag isn't set in this case.
Fixes: 4343d00872 ("afs: Get rid of the afs_writeback record")
Signed-off-by: David Howells <dhowells@redhat.com>
Add a tracepoint to log the cell refcount and active user count and pass in
a reason code through various functions that manipulate these counters.
Additionally, a helper function, afs_see_cell(), is provided to log
interesting places that deal with a cell without actually doing any
accounting directly.
Signed-off-by: David Howells <dhowells@redhat.com>
Management of the lifetime of afs_cell struct has some problems due to the
usage counter being used to determine whether objects of that type are in
use in addition to whether anyone might be interested in the structure.
This is made trickier by cell objects being cached for a period of time in
case they're quickly reused as they hold the result of a setup process that
may be slow (DNS lookups, AFS RPC ops).
Problems include the cached root volume from alias resolution pinning its
parent cell record, rmmod occasionally hanging and occasionally producing
assertion failures.
Fix this by splitting the count of active users from the struct reference
count. Things then work as follows:
(1) The cell cache keeps +1 on the cell's activity count and this has to
be dropped before the cell can be removed. afs_manage_cell() tries to
exchange the 1 to a 0 with the cells_lock write-locked, and if
successful, the record is removed from the net->cells.
(2) One struct ref is 'owned' by the activity count. That is put when the
active count is reduced to 0 (final_destruction label).
(3) A ref can be held on a cell whilst it is queued for management on a
work queue without confusing the active count. afs_queue_cell() is
added to wrap this.
(4) The queue's ref is dropped at the end of the management. This is
split out into a separate function, afs_manage_cell_work().
(5) The root volume record is put after a cell is removed (at the
final_destruction label) rather then in the RCU destruction routine.
(6) Volumes hold struct refs, but aren't active users.
(7) Both counts are displayed in /proc/net/afs/cells.
There are some management function changes:
(*) afs_put_cell() now just decrements the refcount and triggers the RCU
destruction if it becomes 0. It no longer sets a timer to have the
manager do this.
(*) afs_use_cell() and afs_unuse_cell() are added to increase and decrease
the active count. afs_unuse_cell() sets the management timer.
(*) afs_queue_cell() is added to queue a cell with approprate refs.
There are also some other fixes:
(*) Don't let /proc/net/afs/cells access a cell's vllist if it's NULL.
(*) Make sure that candidate cells in lookups are properly destroyed
rather than being simply kfree'd. This ensures the bits it points to
are destroyed also.
(*) afs_dec_cells_outstanding() is now called in cell destruction rather
than at "final_destruction". This ensures that cell->net is still
valid to the end of the destructor.
(*) As a consequence of the previous two changes, move the increment of
net->cells_outstanding that was at the point of insertion into the
tree to the allocation routine to correctly balance things.
Fixes: 989782dcdc ("afs: Overhaul cell database management")
Signed-off-by: David Howells <dhowells@redhat.com>
There are a number of problems that are being seen by the rapidly mounting
and unmounting an afs dynamic root with an explicit cell and volume
specified (which should probably be rejected, but that's a separate issue):
What the tests are doing is to look up/create a cell record for the name
given and then tear it down again without actually using it to try to talk
to a server. This is repeated endlessly, very fast, and the new cell
collides with the old one if it's not quick enough to reuse it.
It appears (as suggested by Hillf Danton) that the search through the RB
tree under a read_seqbegin_or_lock() under RCU conditions isn't safe and
that it's not blocking the write_seqlock(), despite taking two passes at
it. He suggested that the code should take a ref on the cell it's
attempting to look at - but this shouldn't be necessary until we've
compared the cell names. It's possible that I'm missing a barrier
somewhere.
However, using an RCU search for this is overkill, really - we only need to
access the cell name in a few places, and they're places where we're may
end up sleeping anyway.
Fix this by switching to an R/W semaphore instead.
Additionally, draw the down_read() call inside the function (renamed to
afs_find_cell()) since all the callers were taking the RCU read lock (or
should've been[*]).
[*] afs_probe_cell_name() should have been, but that doesn't appear to be
involved in the bug reports.
The symptoms of this look like:
general protection fault, probably for non-canonical address 0xf27d208691691fdb: 0000 [#1] PREEMPT SMP KASAN
KASAN: maybe wild-memory-access in range [0x93e924348b48fed8-0x93e924348b48fedf]
...
RIP: 0010:strncasecmp lib/string.c:52 [inline]
RIP: 0010:strncasecmp+0x5f/0x240 lib/string.c:43
afs_lookup_cell_rcu+0x313/0x720 fs/afs/cell.c:88
afs_lookup_cell+0x2ee/0x1440 fs/afs/cell.c:249
afs_parse_source fs/afs/super.c:290 [inline]
...
Fixes: 989782dcdc ("afs: Overhaul cell database management")
Reported-by: syzbot+459a5dce0b4cb70fd076@syzkaller.appspotmail.com
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Hillf Danton <hdanton@sina.com>
cc: syzkaller-bugs@googlegroups.com
The afs filesystem has a lock[*] that it uses to serialise I/O operations
going to the server (vnode->io_lock), as the server will only perform one
modification operation at a time on any given file or directory. This
prevents the the filesystem from filling up all the call slots to a server
with calls that aren't going to be executed in parallel anyway, thereby
allowing operations on other files to obtain slots.
[*] Note that is probably redundant for directories at least since
i_rwsem is used to serialise directory modifications and
lookup/reading vs modification. The server does allow parallel
non-modification ops, however.
When a file truncation op completes, we truncate the in-memory copy of the
file to match - but we do it whilst still holding the io_lock, the idea
being to prevent races with other operations.
However, if writeback starts in a worker thread simultaneously with
truncation (whilst notify_change() is called with i_rwsem locked, writeback
pays it no heed), it may manage to set PG_writeback bits on the pages that
will get truncated before afs_setattr_success() manages to call
truncate_pagecache(). Truncate will then wait for those pages - whilst
still inside io_lock:
# cat /proc/8837/stack
[<0>] wait_on_page_bit_common+0x184/0x1e7
[<0>] truncate_inode_pages_range+0x37f/0x3eb
[<0>] truncate_pagecache+0x3c/0x53
[<0>] afs_setattr_success+0x4d/0x6e
[<0>] afs_wait_for_operation+0xd8/0x169
[<0>] afs_do_sync_operation+0x16/0x1f
[<0>] afs_setattr+0x1fb/0x25d
[<0>] notify_change+0x2cf/0x3c4
[<0>] do_truncate+0x7f/0xb2
[<0>] do_sys_ftruncate+0xd1/0x104
[<0>] do_syscall_64+0x2d/0x3a
[<0>] entry_SYSCALL_64_after_hwframe+0x44/0xa9
The writeback operation, however, stalls indefinitely because it needs to
get the io_lock to proceed:
# cat /proc/5940/stack
[<0>] afs_get_io_locks+0x58/0x1ae
[<0>] afs_begin_vnode_operation+0xc7/0xd1
[<0>] afs_store_data+0x1b2/0x2a3
[<0>] afs_write_back_from_locked_page+0x418/0x57c
[<0>] afs_writepages_region+0x196/0x224
[<0>] afs_writepages+0x74/0x156
[<0>] do_writepages+0x2d/0x56
[<0>] __writeback_single_inode+0x84/0x207
[<0>] writeback_sb_inodes+0x238/0x3cf
[<0>] __writeback_inodes_wb+0x68/0x9f
[<0>] wb_writeback+0x145/0x26c
[<0>] wb_do_writeback+0x16a/0x194
[<0>] wb_workfn+0x74/0x177
[<0>] process_one_work+0x174/0x264
[<0>] worker_thread+0x117/0x1b9
[<0>] kthread+0xec/0xf1
[<0>] ret_from_fork+0x1f/0x30
and thus deadlock has occurred.
Note that whilst afs_setattr() calls filemap_write_and_wait(), the fact
that the caller is holding i_rwsem doesn't preclude more pages being
dirtied through an mmap'd region.
Fix this by:
(1) Use the vnode validate_lock to mediate access between afs_setattr()
and afs_writepages():
(a) Exclusively lock validate_lock in afs_setattr() around the whole
RPC operation.
(b) If WB_SYNC_ALL isn't set on entry to afs_writepages(), trying to
shared-lock validate_lock and returning immediately if we couldn't
get it.
(c) If WB_SYNC_ALL is set, wait for the lock.
The validate_lock is also used to validate a file and to zap its cache
if the file was altered by a third party, so it's probably a good fit
for this.
(2) Move the truncation outside of the io_lock in setattr, using the same
hook as is used for local directory editing.
This requires the old i_size to be retained in the operation record as
we commit the revised status to the inode members inside the io_lock
still, but we still need to know if we reduced the file size.
Fixes: d2ddc776a4 ("afs: Overhaul volume and server record caching and fileserver rotation")
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Don't use the running state for VL server probes to make decisions about
which server to use as the state is cleared at the start of a probe and
intermediate values might also be misleading.
Instead, add a separate 'latest known' rtt in the afs_vlserver struct and a
flag to indicate if the server is known to be responding and update these
as and when we know what to change them to.
Fixes: 3bf0fb6f33 ("afs: Probe multiple fileservers simultaneously")
Signed-off-by: David Howells <dhowells@redhat.com>
Convert various bitfields in afs_vlserver::probe to a mask and then expose
this and some other bits of information through /proc/net/afs/<cell>/vlservers
to make it easier to debug VL server communication issues.
Signed-off-by: David Howells <dhowells@redhat.com>
The cell name stored in the afs_cell struct is a 64-char + NUL buffer -
when it needs to be able to handle up to AFS_MAXCELLNAME (256 chars) + NUL.
Fix this by changing the array to a pointer and allocating the string.
Found using Coverity.
Fixes: 989782dcdc ("afs: Overhaul cell database management")
Reported-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The fileserver probe timer, net->fs_probe_timer, isn't cancelled when
the kafs module is being removed and so the count it holds on
net->servers_outstanding doesn't get dropped..
This causes rmmod to wait forever. The hung process shows a stack like:
afs_purge_servers+0x1b5/0x23c [kafs]
afs_net_exit+0x44/0x6e [kafs]
ops_exit_list+0x72/0x93
unregister_pernet_operations+0x14c/0x1ba
unregister_pernet_subsys+0x1d/0x2a
afs_exit+0x29/0x6f [kafs]
__do_sys_delete_module.isra.0+0x1a2/0x24b
do_syscall_64+0x51/0x95
entry_SYSCALL_64_after_hwframe+0x44/0xa9
Fix this by:
(1) Attempting to cancel the probe timer and, if successful, drop the
count that the timer was holding.
(2) Make the timer function just drop the count and not schedule the
prober if the afs portion of net namespace is being destroyed.
Also, whilst we're at it, make the following changes:
(3) Initialise net->servers_outstanding to 1 and decrement it before
waiting on it so that it doesn't generate wake up events by being
decremented to 0 until we're cleaning up.
(4) Switch the atomic_dec() on ->servers_outstanding for ->fs_timer in
afs_purge_servers() to use the helper function for that.
Fixes: f6cbb368bc ("afs: Actively poll fileservers to maintain NAT or firewall openings")
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Fix AFS's silly rename by the following means:
(1) Set the destination directory in afs_do_silly_rename() so as to avoid
misbehaviour and indicate that the directory data version will
increment by 1 so as to avoid warnings about unexpected changes in the
DV. Also indicate that the ctime should be updated to avoid xfstest
grumbling.
(2) Note when the server indicates that a directory changed more than we
expected (AFS_OPERATION_DIR_CONFLICT), indicating a conflict with a
third party change, checking on successful completion of unlink and
rename.
The problem is that the FS.RemoveFile RPC op doesn't report the status
of the unlinked file, though YFS.RemoveFile2 does. This can be
mitigated by the assumption that if the directory DV cranked by
exactly 1, we can be sure we removed one link from the file; further,
ordinarily in AFS, files cannot be hardlinked across directories, so
if we reduce nlink to 0, the file is deleted.
However, if the directory DV jumps by more than 1, we cannot know if a
third party intervened by adding or removing a link on the file we
just removed a link from.
The same also goes for any vnode that is at the destination of the
FS.Rename RPC op.
(3) Make afs_vnode_commit_status() apply the nlink drop inside the cb_lock
section along with the other attribute updates if ->op_unlinked is set
on the descriptor for the appropriate vnode.
(4) Issue a follow up status fetch to the unlinked file in the event of a
third party conflict that makes it impossible for us to know if we
actually deleted the file or not.
(5) Provide a flag, AFS_VNODE_SILLY_DELETED, to make afs_getattr() lie to
the user about the nlink of a silly deleted file so that it appears as
0, not 1.
Found with the generic/035 and generic/084 xfstests.
Fixes: e49c7b2f6d ("afs: Build an abstraction around an "operation" concept")
Reported-by: Marc Dionne <marc.dionne@auristor.com>
Signed-off-by: David Howells <dhowells@redhat.com>
afs_check_for_remote_deletion() checks to see if error ENOENT is returned
by the server in response to an operation and, if so, marks the primary
vnode as having been deleted as the FID is no longer valid.
However, it's being called from the operation success functions, where no
abort has happened - and if an inline abort is recorded, it's handled by
afs_vnode_commit_status().
Fix this by actually calling the operation aborted method if provided and
having that point to afs_check_for_remote_deletion().
Fixes: e49c7b2f6d ("afs: Build an abstraction around an "operation" concept")
Signed-off-by: David Howells <dhowells@redhat.com>
Fix the following issues:
(1) Fix writeback to reduce the size of a store operation to i_size,
effectively discarding the extra data.
The problem comes when afs_page_mkwrite() records that a page is about
to be modified by mmap(). It doesn't know what bits of the page are
going to be modified, so it records the whole page as being dirty
(this is stored in page->private as start and end offsets).
Without this, the marshalling for the store to the server extends the
size of the file to the end of the page (in afs_fs_store_data() and
yfs_fs_store_data()).
(2) Fix setattr to actually truncate the pagecache, thereby clearing
the discarded part of a file.
(3) Fix setattr to check that the new size is okay and to disable
ATTR_SIZE if i_size wouldn't change.
(4) Force i_size to be updated as the result of a truncate.
(5) Don't truncate if ATTR_SIZE is not set.
(6) Call pagecache_isize_extended() if the file was enlarged.
Note that truncate_set_size() isn't used because the setting of i_size is
done inside afs_vnode_commit_status() under the vnode->cb_lock.
Found with the generic/029 and generic/393 xfstests.
Fixes: 31143d5d51 ("AFS: implement basic file write support")
Fixes: 4343d00872 ("afs: Get rid of the afs_writeback record")
Signed-off-by: David Howells <dhowells@redhat.com>
The in-kernel afs filesystem ignores ctime because the AFS fileserver
protocol doesn't support ctimes. This, however, causes various xfstests to
fail.
Work around this by:
(1) Setting ctime to attr->ia_ctime in afs_setattr().
(2) Not ignoring ATTR_MTIME_SET, ATTR_TIMES_SET and ATTR_TOUCH settings.
(3) Setting the ctime from the server mtime when on the target file when
creating a hard link to it.
(4) Setting the ctime on directories from their revised mtimes when
renaming/moving a file.
Found by the generic/221 and generic/309 xfstests.
Signed-off-by: David Howells <dhowells@redhat.com>
Fix AFS file locking to use the correct vnode pointer and remove a member
of the afs_operation struct that is never set, but it is read and followed,
causing an oops.
This can be triggered by:
flock -s /afs/example.com/foo sleep 1
when it calls the kernel to get a file lock.
Fixes: e49c7b2f6d ("afs: Build an abstraction around an "operation" concept")
Reported-by: Dave Botsch <botsch@cnf.cornell.edu>
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-by: Dave Botsch <botsch@cnf.cornell.edu>
Adjust the fileserver rotation algorithm so that if we've tried all the
addresses on a server (cumulatively over multiple operations) until we've
run out of untried addresses, immediately reprobe all that server's
interfaces and retry the op at least once before we move onto the next
server.
Signed-off-by: David Howells <dhowells@redhat.com>
Display more information about the state of a server record, including the
flags, rtt and break counter plus the probe state for each server in
/proc/net/afs/servers.
Rearrange the server flags a bit to make them easier to read at a glance in
the proc file.
Signed-off-by: David Howells <dhowells@redhat.com>
Don't use the running state for fileserver probes to make decisions about
which server to use as the state is cleared at the start of a probe and
also intermediate values might be misleading.
Instead, add a separate 'latest known' rtt in the afs_server struct and a
flag to indicate if the server is known to be responding and update these
as and when we know what to change them to.
Signed-off-by: David Howells <dhowells@redhat.com>
Whilst it shouldn't happen, it is possible for multiple fileservers to
share a UUID, particularly if an entire cell has been duplicated, UUIDs and
all. In such a case, it's not necessarily possible to map the effect of
the CB.InitCallBackState3 incoming RPC to a specific server unambiguously
by UUID and thus to a specific cell.
Indeed, there's a problem whereby multiple server records may need to
occupy the same spot in the rb_tree rooted in the afs_net struct.
Fix this by allowing servers to form a list, with the head of the list in
the tree. When the front entry in the list is removed, the second in the
list just replaces it. afs_init_callback_state() then just goes down the
line, poking each server in the list.
This means that some servers will be unnecessarily poked, unfortunately.
An alternative would be to route by call parameters.
Reported-by: Jeffrey Altman <jaltman@auristor.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Fixes: d2ddc776a4 ("afs: Overhaul volume and server record caching and fileserver rotation")
Reorganise afs_volume objects such that they're in a tree keyed on volume
ID, rooted at on an afs_cell object rather than being in multiple trees,
each of which is rooted on an afs_server object.
afs_server structs become per-cell and acquire a pointer to the cell.
The process of breaking a callback then starts with finding the server by
its network address, following that to the cell and then looking up each
volume ID in the volume tree.
This is simpler than the afs_vol_interest/afs_cb_interest N:M mapping web
and allows those structs and the code for maintaining them to be simplified
or removed.
It does make a couple of things a bit more tricky, though:
(1) Operations now start with a volume, not a server, so there can be more
than one answer as to whether or not the server we'll end up using
supports the FS.InlineBulkStatus RPC.
(2) CB RPC operations that specify the server UUID. There's still a tree
of servers by UUID on the afs_net struct, but the UUIDs in it aren't
guaranteed unique.
Signed-off-by: David Howells <dhowells@redhat.com>
Put in the first phase of cell alias detection. This part handles alias
detection for cells that have root.cell volumes (which is expected to be
likely).
When a cell becomes newly active, it is probed for its root.cell volume,
and if it has one, this volume is compared against other root.cell volumes
to find out if the list of fileserver UUIDs have any in common - and if
that's the case, do the address lists of those fileservers have any
addresses in common. If they do, the new cell is adjudged to be an alias
of the old cell and the old cell is used instead.
Comparing is aided by the server list in struct afs_server_list being
sorted in UUID order and the addresses in the fileserver address lists
being sorted in address order.
The cell then retains the afs_volume object for the root.cell volume, even
if it's not mounted for future alias checking.
This necessary because:
(1) Whilst fileservers have UUIDs that are meant to be globally unique, in
practice they are not because cells get cloned without changing the
UUIDs - so afs_server records need to be per cell.
(2) Sometimes the DNS is used to make cell aliases - but if we don't know
they're the same, we may end up with multiple superblocks and multiple
afs_server records for the same thing, impairing our ability to
deliver callback notifications of third party changes
(3) The fileserver RPC API doesn't contain the cell name, so it can't tell
us which cell it's notifying and can't see that a change made to to
one cell should notify the same client that's also accessed as the
other cell.
Reported-by: Jeffrey Altman <jaltman@auristor.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Implement client support for the YFSVL.GetCellName RPC operation by which
YFS permits the canonical cell name to be queried from a VL server.
Signed-off-by: David Howells <dhowells@redhat.com>
Save more bits from the volume location database record obtained for a
server so that we can use this information in cell alias detection.
Signed-off-by: David Howells <dhowells@redhat.com>
Don't get the epoch from a server, particularly one that we're looking up
by UUID, as UUIDs may be ambiguous and may map to more than one server - so
we can't draw any conclusions from it.
Reported-by: Jeffrey Altman <jaltman@auristor.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Turn the afs_operation struct into the main way that most fileserver
operations are managed. Various things are added to the struct, including
the following:
(1) All the parameters and results of the relevant operations are moved
into it, removing corresponding fields from the afs_call struct.
afs_call gets a pointer to the op.
(2) The target volume is made the main focus of the operation, rather than
the target vnode(s), and a bunch of op->vnode->volume are made
op->volume instead.
(3) Two vnode records are defined (op->file[]) for the vnode(s) involved
in most operations. The vnode record (struct afs_vnode_param)
contains:
- The vnode pointer.
- The fid of the vnode to be included in the parameters or that was
returned in the reply (eg. FS.MakeDir).
- The status and callback information that may be returned in the
reply about the vnode.
- Callback break and data version tracking for detecting
simultaneous third-parth changes.
(4) Pointers to dentries to be updated with new inodes.
(5) An operations table pointer. The table includes pointers to functions
for issuing AFS and YFS-variant RPCs, handling the success and abort
of an operation and handling post-I/O-lock local editing of a
directory.
To make this work, the following function restructuring is made:
(A) The rotation loop that issues calls to fileservers that can be found
in each function that wants to issue an RPC (such as afs_mkdir()) is
extracted out into common code, in a new file called fs_operation.c.
(B) The rotation loops, such as the one in afs_mkdir(), are replaced with
a much smaller piece of code that allocates an operation, sets the
parameters and then calls out to the common code to do the actual
work.
(C) The code for handling the success and failure of an operation are
moved into operation functions (as (5) above) and these are called
from the core code at appropriate times.
(D) The pseudo inode getting stuff used by the dynamic root code is moved
over into dynroot.c.
(E) struct afs_iget_data is absorbed into the operation struct and
afs_iget() expects to be given an op pointer and a vnode record.
(F) Point (E) doesn't work for the root dir of a volume, but we know the
FID in advance (it's always vnode 1, unique 1), so a separate inode
getter, afs_root_iget(), is provided to special-case that.
(G) The inode status init/update functions now also take an op and a vnode
record.
(H) The RPC marshalling functions now, for the most part, just take an
afs_operation struct as their only argument. All the data they need
is held there. The result delivery functions write their answers
there as well.
(I) The call is attached to the operation and then the operation core does
the waiting.
And then the new operation code is, for the moment, made to just initialise
the operation, get the appropriate vnode I/O locks and do the same rotation
loop as before.
This lays the foundation for the following changes in the future:
(*) Overhauling the rotation (again).
(*) Support for asynchronous I/O, where the fileserver rotation must be
done asynchronously also.
Signed-off-by: David Howells <dhowells@redhat.com>
As a prelude to implementing asynchronous fileserver operations in the afs
filesystem, rename struct afs_fs_cursor to afs_operation.
This struct is going to form the core of the operation management and is
going to acquire more members in later.
Signed-off-by: David Howells <dhowells@redhat.com>
Set a flag in the call struct to indicate an unmarshalling error rather
than return and handle an error from the decoding of file statuses. This
flag is checked on a successful return from the delivery function.
Signed-off-by: David Howells <dhowells@redhat.com>
afs_vol_interest objects represent the volume IDs currently being accessed
from a fileserver. These hold lists of afs_cb_interest objects that
repesent the superblocks using that volume ID on that server.
When a callback notification from the server telling of a modification by
another client arrives, the volume ID specified in the notification is
looked up in the server's afs_vol_interest list. Through the
afs_cb_interest list, the relevant superblocks can be iterated over and the
specific inode looked up and marked in each one.
Make the following efficiency improvements:
(1) Hold rcu_read_lock() over the entire processing rather than locking it
each time.
(2) Do all the callbacks for each vid together rather than individually.
Each volume then only needs to be looked up once.
(3) afs_vol_interest objects are now stored in an rb_tree rather than a
flat list to reduce the lookup step count.
(4) afs_vol_interest lookup is now done with RCU, but because it's in an
rb_tree which may rotate under us, a seqlock is used so that if it
changes during the walk, we repeat the walk with a lock held.
With this and the preceding patch which adds RCU-based lookups in the inode
cache, target volumes/vnodes can be taken without the need to take any
locks, except on the target itself.
Signed-off-by: David Howells <dhowells@redhat.com>
When an AFS client accesses a file, it receives a limited-duration callback
promise that the server will notify it if another client changes a file.
This callback duration can be a few hours in length.
If a client mounts a volume and then an application prevents it from being
unmounted, say by chdir'ing into it, but then does nothing for some time,
the rxrpc_peer record will expire and rxrpc-level keepalive will cease.
If there is NAT or a firewall between the client and the server, the route
back for the server may close after a comparatively short duration, meaning
that attempts by the server to notify the client may then bounce.
The client, however, may (so far as it knows) still have a valid unexpired
promise and will then rely on its cached data and will not see changes made
on the server by a third party until it incidentally rechecks the status or
the promise needs renewal.
To deal with this, the client needs to regularly probe the server. This
has two effects: firstly, it keeps a route open back for the server, and
secondly, it causes the server to disgorge any notifications that got
queued up because they couldn't be sent.
Fix this by adding a mechanism to emit regular probes.
Two levels of probing are made available: Under normal circumstances the
'slow' queue will be used for a fileserver - this just probes the preferred
address once every 5 mins or so; however, if server fails to respond to any
probes, the server will shift to the 'fast' queue from which all its
interfaces will be probed every 30s. When it finally responds, the record
will switch back to the slow queue.
Further notes:
(1) Probing is now no longer driven from the fileserver rotation
algorithm.
(2) Probes are dispatched to all interfaces on a fileserver when that an
afs_server object is set up to record it.
(3) The afs_server object is removed from the probe queues when we start
to probe it. afs_is_probing_server() returns true if it's not listed
- ie. it's undergoing probing.
(4) The afs_server object is added back on to the probe queue when the
final outstanding probe completes, but the probed_at time is set when
we're about to launch a probe so that it's not dependent on the probe
duration.
(5) The timer and the work item added for this must be handed a count on
net->servers_outstanding, which they hand on or release. This makes
sure that network namespace cleanup waits for them.
Fixes: d2ddc776a4 ("afs: Overhaul volume and server record caching and fileserver rotation")
Reported-by: Dave Botsch <botsch@cnf.cornell.edu>
Signed-off-by: David Howells <dhowells@redhat.com>
Split the usage count on the afs_server struct to have an active count that
registers who's actually using it separately from the reference count on
the object.
This allows a future patch to dispatch polling probes without advancing the
"unuse" time into the future each time we emit a probe, which would
otherwise prevent unused server records from expiring.
Included in this:
(1) The latter part of afs_destroy_server() in which the RCU destruction
of afs_server objects is invoked and the outstanding server count is
decremented is split out into __afs_put_server().
(2) afs_put_server() now calls __afs_put_server() rather then setting the
management timer.
(3) The calls begun by afs_fs_give_up_all_callbacks() and
afs_fs_get_capabilities() can now take a ref on the server record, so
afs_destroy_server() can just drop its ref and needn't wait for the
completion of these calls. They'll put the ref when they're done.
(4) Because of (3), afs_fs_probe_done() no longer needs to wake up
afs_destroy_server() with server->probe_outstanding.
(5) afs_gc_servers can be simplified. It only needs to check if
server->active is 0 rather than playing games with the refcount.
(6) afs_manage_servers() can propose a server for gc if usage == 0 rather
than if ref == 1. The gc is effected by (5).
Signed-off-by: David Howells <dhowells@redhat.com>
The U-version VLDB volume record retrieved by the VL.GetEntryByNameU rpc op
carries a change counter (the serverUnique field) for each fileserver
listed in the record as backing that volume. This is incremented whenever
the registration details for a fileserver change (such as its address
list). Note that the same value will be seen in all UVLDB records that
refer to that fileserver.
This should be checked before calling the VL server to re-query the address
list for a fileserver. If it's the same, there's no point doing the query.
Reported-by: Jeffrey Altman <jaltman@auristor.com>
Signed-off-by: David Howells <dhowells@redhat.com>
When an operation is meant to be done uninterruptibly (such as
FS.StoreData), we should not be allowing volume and server record checking
to be interrupted.
Fixes: d2ddc776a4 ("afs: Overhaul volume and server record caching and fileserver rotation")
Signed-off-by: David Howells <dhowells@redhat.com>
Remove three bits:
(1) afs_server::no_epoch is neither set nor used.
(2) afs_server::have_result is set and a wakeup is applied to it, but
nothing looks at it or waits on it.
(3) afs_vl_dump_edestaddrreq() prints afs_addr_list::probed, but nothing
sets it for VL servers.
Signed-off-by: David Howells <dhowells@redhat.com>
Pull networking fixes from David Miller:
1) Fix deadlock in bpf_send_signal() from Yonghong Song.
2) Fix off by one in kTLS offload of mlx5, from Tariq Toukan.
3) Add missing locking in iwlwifi mvm code, from Avraham Stern.
4) Fix MSG_WAITALL handling in rxrpc, from David Howells.
5) Need to hold RTNL mutex in tcindex_partial_destroy_work(), from Cong
Wang.
6) Fix producer race condition in AF_PACKET, from Willem de Bruijn.
7) cls_route removes the wrong filter during change operations, from
Cong Wang.
8) Reject unrecognized request flags in ethtool netlink code, from
Michal Kubecek.
9) Need to keep MAC in reset until PHY is up in bcmgenet driver, from
Doug Berger.
10) Don't leak ct zone template in act_ct during replace, from Paul
Blakey.
11) Fix flushing of offloaded netfilter flowtable flows, also from Paul
Blakey.
12) Fix throughput drop during tx backpressure in cxgb4, from Rahul
Lakkireddy.
13) Don't let a non-NULL skb->dev leave the TCP stack, from Eric
Dumazet.
14) TCP_QUEUE_SEQ socket option has to update tp->copied_seq as well,
also from Eric Dumazet.
15) Restrict macsec to ethernet devices, from Willem de Bruijn.
16) Fix reference leak in some ethtool *_SET handlers, from Michal
Kubecek.
17) Fix accidental disabling of MSI for some r8169 chips, from Heiner
Kallweit.
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (138 commits)
net: Fix CONFIG_NET_CLS_ACT=n and CONFIG_NFT_FWD_NETDEV={y, m} build
net: ena: Add PCI shutdown handler to allow safe kexec
selftests/net/forwarding: define libs as TEST_PROGS_EXTENDED
selftests/net: add missing tests to Makefile
r8169: re-enable MSI on RTL8168c
net: phy: mdio-bcm-unimac: Fix clock handling
cxgb4/ptp: pass the sign of offset delta in FW CMD
net: dsa: tag_8021q: replace dsa_8021q_remove_header with __skb_vlan_pop
net: cbs: Fix software cbs to consider packet sending time
net/mlx5e: Do not recover from a non-fatal syndrome
net/mlx5e: Fix ICOSQ recovery flow with Striding RQ
net/mlx5e: Fix missing reset of SW metadata in Striding RQ reset
net/mlx5e: Enhance ICOSQ WQE info fields
net/mlx5_core: Set IB capability mask1 to fix ib_srpt connection failure
selftests: netfilter: add nfqueue test case
netfilter: nft_fwd_netdev: allow to redirect to ifb via ingress
netfilter: nft_fwd_netdev: validate family and chain type
netfilter: nft_set_rbtree: Detect partial overlaps on insertion
netfilter: nft_set_rbtree: Introduce and use nft_rbtree_interval_start()
netfilter: nft_set_pipapo: Separate partial and complete overlap cases on insertion
...
When an AFS service handler function aborts a call, AF_RXRPC marks the call
as complete - which means that it's not going to get any more packets from
the receiver. This is a problem because reception of the final ACK is what
triggers afs_deliver_to_call() to drop the final ref on the afs_call
object.
Instead, aborted AFS service calls may then just sit around waiting for
ever or until they're displaced by a new call on the same connection
channel or a connection-level abort.
Fix this by calling afs_set_call_complete() to finalise the afs_call struct
representing the call.
However, we then need to drop the ref that stops the call from being
deallocated. We can do this in afs_set_call_complete(), as the work queue
is holding a separate ref of its own, but then we shouldn't do it in
afs_process_async_call() and afs_delete_async_call().
call->drop_ref is set to indicate that a ref needs dropping for a call and
this is dealt with when we transition a call to AFS_CALL_COMPLETE.
But then we also need to get rid of the ref that pins an asynchronous
client call. We can do this by the same mechanism, setting call->drop_ref
for an async client call too.
We can also get rid of call->incoming since nothing ever sets it and only
one thing ever checks it (futilely).
A trace of the rxrpc_call and afs_call struct ref counting looks like:
<idle>-0 [001] ..s5 164.764892: rxrpc_call: c=00000002 SEE u=3 sp=rxrpc_new_incoming_call+0x473/0xb34 a=00000000442095b5
<idle>-0 [001] .Ns5 164.766001: rxrpc_call: c=00000002 QUE u=4 sp=rxrpc_propose_ACK+0xbe/0x551 a=00000000442095b5
<idle>-0 [001] .Ns4 164.766005: rxrpc_call: c=00000002 PUT u=3 sp=rxrpc_new_incoming_call+0xa3f/0xb34 a=00000000442095b5
<idle>-0 [001] .Ns7 164.766433: afs_call: c=00000002 WAKE u=2 o=11 sp=rxrpc_notify_socket+0x196/0x33c
kworker/1:2-1810 [001] ...1 164.768409: rxrpc_call: c=00000002 SEE u=3 sp=rxrpc_process_call+0x25/0x7ae a=00000000442095b5
kworker/1:2-1810 [001] ...1 164.769439: rxrpc_tx_packet: c=00000002 e9f1a7a8:95786a88:00000008:09c5 00000001 00000000 02 22 ACK CallAck
kworker/1:2-1810 [001] ...1 164.769459: rxrpc_call: c=00000002 PUT u=2 sp=rxrpc_process_call+0x74f/0x7ae a=00000000442095b5
kworker/1:2-1810 [001] ...1 164.770794: afs_call: c=00000002 QUEUE u=3 o=12 sp=afs_deliver_to_call+0x449/0x72c
kworker/1:2-1810 [001] ...1 164.770829: afs_call: c=00000002 PUT u=2 o=12 sp=afs_process_async_call+0xdb/0x11e
kworker/1:2-1810 [001] ...2 164.771084: rxrpc_abort: c=00000002 95786a88:00000008 s=0 a=1 e=1 K-1
kworker/1:2-1810 [001] ...1 164.771461: rxrpc_tx_packet: c=00000002 e9f1a7a8:95786a88:00000008:09c5 00000002 00000000 04 00 ABORT CallAbort
kworker/1:2-1810 [001] ...1 164.771466: afs_call: c=00000002 PUT u=1 o=12 sp=SRXAFSCB_ProbeUuid+0xc1/0x106
The abort generated in SRXAFSCB_ProbeUuid(), labelled "K-1", indicates that
the local filesystem/cache manager didn't recognise the UUID as its own.
Fixes: 2067b2b3f4 ("afs: Fix the CB.ProbeUuid service handler to reply correctly")
Signed-off-by: David Howells <dhowells@redhat.com>
afs_put_addrlist() casts kfree() to rcu_callback_t. Apart from being wrong
in theory, this might also blow up when people start enforcing function
types via compiler instrumentation, and it means the rcu_head has to be
first in struct afs_addr_list.
Use kfree_rcu() instead, it's simpler and more correct.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
