Josef Bacik [Fri, 6 Dec 2019 14:37:18 +0000 (09:37 -0500)]
btrfs: do not leak reloc root if we fail to read the fs root
If we fail to read the fs root corresponding with a reloc root we'll
just break out and free the reloc roots. But we remove our current
reloc_root from this list higher up, which means we'll leak this
reloc_root. Fix this by adding ourselves back to the reloc_roots list
so we are properly cleaned up.
Josef Bacik [Fri, 6 Dec 2019 14:37:17 +0000 (09:37 -0500)]
btrfs: skip log replay on orphaned roots
My fsstress modifications coupled with generic/475 uncovered a failure
to mount and replay the log if we hit a orphaned root. We do not want
to replay the log for an orphan root, but it's completely legitimate to
have an orphaned root with a log attached. Fix this by simply skipping
replaying the log. We still need to pin it's root node so that we do
not overwrite it while replaying other logs, as we re-read the log root
at every stage of the replay.
Josef Bacik [Fri, 6 Dec 2019 16:39:00 +0000 (11:39 -0500)]
btrfs: handle ENOENT in btrfs_uuid_tree_iterate
If we get an -ENOENT back from btrfs_uuid_iter_rem when iterating the
uuid tree we'll just continue and do btrfs_next_item(). However we've
done a btrfs_release_path() at this point and no longer have a valid
path. So increment the key and go back and do a normal search.
Filipe Manana [Thu, 5 Dec 2019 16:58:41 +0000 (16:58 +0000)]
Btrfs: fix hole extent items with a zero size after range cloning
Normally when cloning a file range if we find an implicit hole at the end
of the range we assume it is because the NO_HOLES feature is enabled.
However that is not always the case. One well known case [1] is when we
have a power failure after mixing buffered and direct IO writes against
the same file.
In such cases we need to punch a hole in the destination file, and if
the NO_HOLES feature is not enabled, we need to insert explicit file
extent items to represent the hole. After commit 690a5dbfc51315
("Btrfs: fix ENOSPC errors, leading to transaction aborts, when cloning
extents"), we started to insert file extent items representing the hole
with an item size of 0, which is invalid and should be 53 bytes (the size
of a btrfs_file_extent_item structure), resulting in all sorts of
corruptions and invalid memory accesses. This is detected by the tree
checker when we attempt to write a leaf to disk.
The problem can be sporadically triggered by test case generic/561 from
fstests. That test case does not exercise power failure and creates a new
filesystem when it starts, so it does not use a filesystem created by any
previous test that tests power failure. However the test does both
buffered and direct IO writes (through fsstress) and it's precisely that
which is creating the implicit holes in files. That happens even before
the commit mentioned earlier. I need to investigate why we get those
implicit holes to check if there is a real problem or not. For now this
change fixes the regression of introducing file extent items with an item
size of 0 bytes.
Fix the issue by calling btrfs_punch_hole_range() without passing a
btrfs_clone_extent_info structure, which ensures file extent items are
inserted to represent the hole with a correct item size. We were passing
a btrfs_clone_extent_info with a value of 0 for its 'item_size' field,
which was causing the insertion of file extent items with an item size
of 0.
Reported-by: David Sterba <[email protected]> Fixes: 690a5dbfc51315 ("Btrfs: fix ENOSPC errors, leading to transaction aborts, when cloning extents") Signed-off-by: Filipe Manana <[email protected]> Reviewed-by: Josef Bacik <[email protected]> Signed-off-by: David Sterba <[email protected]>
Filipe Manana [Fri, 6 Dec 2019 12:27:39 +0000 (12:27 +0000)]
Btrfs: fix removal logic of the tree mod log that leads to use-after-free issues
When a tree mod log user no longer needs to use the tree it calls
btrfs_put_tree_mod_seq() to remove itself from the list of users and
delete all no longer used elements of the tree's red black tree, which
should be all elements with a sequence number less then our equals to
the caller's sequence number. However the logic is broken because it
can delete and free elements from the red black tree that have a
sequence number greater then the caller's sequence number:
1) At a point in time we have sequence numbers 1, 2, 3 and 4 in the
tree mod log;
2) The task which got assigned the sequence number 1 calls
btrfs_put_tree_mod_seq();
3) Sequence number 1 is deleted from the list of sequence numbers;
4) The current minimum sequence number is computed to be the sequence
number 2;
5) A task using sequence number 2 is at tree_mod_log_rewind() and gets
a pointer to one of its elements from the red black tree through
a call to tree_mod_log_search();
6) The task with sequence number 1 iterates the red black tree of tree
modification elements and deletes (and frees) all elements with a
sequence number less then or equals to 2 (the computed minimum sequence
number) - it ends up only leaving elements with sequence numbers of 3
and 4;
7) The task with sequence number 2 now uses the pointer to its element,
already freed by the other task, at __tree_mod_log_rewind(), resulting
in a use-after-free issue. When CONFIG_DEBUG_PAGEALLOC=y it produces
a trace like the following:
Fix this by making btrfs_put_tree_mod_seq() skip deletion of elements that
have a sequence number equals to the computed minimum sequence number, and
not just elements with a sequence number greater then that minimum.
Filipe Manana [Mon, 2 Dec 2019 11:01:03 +0000 (11:01 +0000)]
Btrfs: make tree checker detect checksum items with overlapping ranges
Having checksum items, either on the checksums tree or in a log tree, that
represent ranges that overlap each other is a sign of a corruption. Such
case confuses the checksum lookup code and can result in not being able to
find checksums or find stale checksums.
So add a check for such case.
This is motivated by a recent fix for a case where a log tree had checksum
items covering ranges that overlap each other due to extent cloning, and
resulted in missing checksums after replaying the log tree. It also helps
detect past issues such as stale and outdated checksums due to overlapping,
commit 27b9a8122ff71a ("Btrfs: fix csum tree corruption, duplicate and
outdated checksums").
Filipe Manana [Thu, 5 Dec 2019 16:58:30 +0000 (16:58 +0000)]
Btrfs: fix missing data checksums after replaying a log tree
When logging a file that has shared extents (reflinked with other files or
with itself), we can end up logging multiple checksum items that cover
overlapping ranges. This confuses the search for checksums at log replay
time causing some checksums to never be added to the fs/subvolume tree.
Consider the following example of a file that shares the same extent at
offsets 0 and 256Kb:
[ bytenr 13631488, offset 64Kb, len 192Kb ]
64Kb 256Kb
[ bytenr 13893632, offset 0, len 256Kb ]
256Kb 512Kb
When logging the inode, at tree-log.c:copy_items(), when processing the
file extent item at offset 0, we log a checksum item covering the range 13959168 to 14024704, which corresponds to 13893632 + 64Kb and 13893632 +
64Kb + 64Kb, respectively.
Later when processing the extent item at offset 256K, we log the checksums
for the range from 13893632 to 14155776 (which corresponds to 13893632 +
256Kb). These checksums get merged with the checksum item for the range
from 13631488 to 13893632 (13631488 + 256Kb), logged by a previous fsync.
So after this we get the two following checksum items in the log tree:
(...)
item 6 key (EXTENT_CSUM EXTENT_CSUM 13631488) itemoff 3095 itemsize 512
range start 13631488 end 14155776 length 524288
item 7 key (EXTENT_CSUM EXTENT_CSUM 13959168) itemoff 3031 itemsize 64
range start 13959168 end 14024704 length 65536
The first one covers the range from the second one, they overlap.
So far this does not cause a problem after replaying the log, because
when replaying the file extent item for offset 256K, we copy all the
checksums for the extent 13893632 from the log tree to the fs/subvolume
tree, since searching for an checksum item for bytenr 13893632 leaves us
at the first checksum item, which covers the whole range of the extent.
However if we write 64Kb to file offset 256Kb for example, we will
not be able to find and copy the checksums for the last 128Kb of the
extent at bytenr 13893632, referenced by the file range 384Kb to 512Kb.
After writing 64Kb into file offset 256Kb we get the following extent
layout for our file:
[ bytenr 13631488, offset 64Kb, len 192Kb ]
64Kb 256Kb
[ bytenr 14155776, offset 0, len 64Kb ]
256Kb 320Kb
[ bytenr 13893632, offset 64Kb, len 192Kb ]
320Kb 512Kb
After fsync'ing the file, if we have a power failure and then mount
the filesystem to replay the log, the following happens:
1) When replaying the file extent item for file offset 320Kb, we
lookup for the checksums for the extent range from 13959168
(13893632 + 64Kb) to 14155776 (13893632 + 256Kb), through a call
to btrfs_lookup_csums_range();
2) btrfs_lookup_csums_range() finds the checksum item that starts
precisely at offset 13959168 (item 7 in the log tree, shown before);
3) However that checksum item only covers 64Kb of data, and not 192Kb
of data;
4) As a result only the checksums for the first 64Kb of data referenced
by the file extent item are found and copied to the fs/subvolume tree.
The remaining 128Kb of data, file range 384Kb to 512Kb, doesn't get
the corresponding data checksums found and copied to the fs/subvolume
tree.
5) After replaying the log userspace will not be able to read the file
range from 384Kb to 512Kb, because the checksums are missing and
resulting in an -EIO error.
The following steps reproduce this scenario:
$ mkfs.btrfs -f /dev/sdc
$ mount /dev/sdc /mnt/sdc
$ dmesg | tail
[165305.003464] BTRFS info (device sdc): no csum found for inode 257 start 401408
[165305.004014] BTRFS info (device sdc): no csum found for inode 257 start 405504
[165305.004559] BTRFS info (device sdc): no csum found for inode 257 start 409600
[165305.005101] BTRFS info (device sdc): no csum found for inode 257 start 413696
[165305.005627] BTRFS info (device sdc): no csum found for inode 257 start 417792
[165305.006134] BTRFS info (device sdc): no csum found for inode 257 start 421888
[165305.006625] BTRFS info (device sdc): no csum found for inode 257 start 425984
[165305.007278] BTRFS info (device sdc): no csum found for inode 257 start 430080
[165305.008248] BTRFS warning (device sdc): csum failed root 5 ino 257 off 393216 csum 0x1337385e expected csum 0x00000000 mirror 1
[165305.009550] BTRFS warning (device sdc): csum failed root 5 ino 257 off 393216 csum 0x1337385e expected csum 0x00000000 mirror 1
Fix this simply by deleting first any checksums, from the log tree, for the
range of the extent we are logging at copy_items(). This ensures we do not
get checksum items in the log tree that have overlapping ranges.
This is a long time issue that has been present since we have the clone
(and deduplication) ioctl, and can happen both when an extent is shared
between different files and within the same file.
Dan Carpenter [Tue, 3 Dec 2019 11:24:58 +0000 (14:24 +0300)]
btrfs: return error pointer from alloc_test_extent_buffer
Callers of alloc_test_extent_buffer have not correctly interpreted the
return value as error pointer, as alloc_test_extent_buffer should behave
as alloc_extent_buffer. The self-tests were unaffected but
btrfs_find_create_tree_block could call both functions and that would
cause problems up in the call chain.
David Sterba [Wed, 27 Nov 2019 15:10:54 +0000 (16:10 +0100)]
btrfs: fix devs_max constraints for raid1c3 and raid1c4
The value 0 for devs_max means to spread the allocated chunks over all
available devices, eg. stripe for RAID0 or RAID5. This got mistakenly
copied to the RAID1C3/4 profiles. The intention is to have exactly 3 and
4 copies respectively.
Fixes: 47e6f7423b91 ("btrfs: add support for 3-copy replication (raid1c3)") Fixes: 8d6fac0087e5 ("btrfs: add support for 4-copy replication (raid1c4)") Signed-off-by: David Sterba <[email protected]>
Andreas Färber [Fri, 8 Nov 2019 21:38:52 +0000 (22:38 +0100)]
btrfs: tree-checker: Fix error format string for size_t
Argument BTRFS_FILE_EXTENT_INLINE_DATA_START is defined as offsetof(),
which returns type size_t, so we need %zu instead of %lu.
This fixes a build warning on 32-bit ARM:
../fs/btrfs/tree-checker.c: In function 'check_extent_data_item':
../fs/btrfs/tree-checker.c:230:43: warning: format '%lu' expects argument of type 'long unsigned int', but argument 5 has type 'unsigned int' [-Wformat=]
230 | "invalid item size, have %u expect [%lu, %u)",
| ~~^
| long unsigned int
| %u
Fixes: 153a6d299956 ("btrfs: tree-checker: Check item size before reading file extent type") Acked-by: Geert Uytterhoeven <[email protected]> Signed-off-by: Andreas Färber <[email protected]> Reviewed-by: David Sterba <[email protected]> Signed-off-by: David Sterba <[email protected]>
Josef Bacik [Tue, 19 Nov 2019 18:59:00 +0000 (13:59 -0500)]
btrfs: handle error in btrfs_cache_block_group
We have a BUG_ON(ret < 0) in find_free_extent from
btrfs_cache_block_group. If we fail to allocate our ctl we'll just
panic, which is not good. Instead just go on to another block group.
If we fail to find a block group we don't want to return ENOSPC, because
really we got a ENOMEM and that's the root of the problem. Save our
return from btrfs_cache_block_group(), and then if we still fail to make
our allocation return that ret so we get the right error back.
Josef Bacik [Tue, 19 Nov 2019 18:59:35 +0000 (13:59 -0500)]
btrfs: do not call synchronize_srcu() in inode_tree_del
Testing with the new fsstress uncovered a pretty nasty deadlock with
lookup and snapshot deletion.
Process A
unlink
-> final iput
-> inode_tree_del
-> synchronize_srcu(subvol_srcu)
Process B
btrfs_lookup <- srcu_read_lock() acquired here
-> btrfs_iget
-> find inode that has I_FREEING set
-> __wait_on_freeing_inode()
We're holding the srcu_read_lock() while doing the iget in order to make
sure our fs root doesn't go away, and then we are waiting for the inode
to finish freeing. However because the free'ing process is doing a
synchronize_srcu() we deadlock.
Fix this by dropping the synchronize_srcu() in inode_tree_del(). We
don't need people to stop accessing the fs root at this point, we're
only adding our empty root to the dead roots list.
A larger much more invasive fix is forthcoming to address how we deal
with fs roots, but this fixes the immediate problem.
Since commit 0e4a459f56c32d3e ("tracing: Remove unnecessary DEBUG_FS
dependency"), CONFIG_DEBUG_FS is no longer auto-enabled. This breaks
booting Debian 9, as systemd needs debugfs:
[FAILED] Failed to mount /sys/kernel/debug.
See 'systemctl status sys-kernel-debug.mount' for details.
[DEPEND] Dependency failed for Local File Systems.
...
You are in emergGive root password for maintenance
(or press Control-D to continue):
Fix this by enabling CONFIG_DEBUG_FS explicitly.
See also commit 18977008f44c66bd ("ARM: multi_v7_defconfig: Restore
debugfs support").
Filipe Manana [Thu, 5 Dec 2019 16:57:39 +0000 (16:57 +0000)]
Btrfs: fix cloning range with a hole when using the NO_HOLES feature
When using the NO_HOLES feature if we clone a range that contains a hole
and a temporary ENOSPC happens while dropping extents from the target
inode's range, we can end up failing and aborting the transaction with
-EEXIST or with a corrupt file extent item, that has a length greater
than it should and overlaps with other extents. For example when cloning
the following range from inode A to inode B:
Target range: [1MB, 6MB) (same as source, to make it easier to explain)
The following can happen:
1) btrfs_punch_hole_range() gets -ENOSPC from __btrfs_drop_extents();
2) At that point, 'cur_offset' is set to 1MB and __btrfs_drop_extents()
set 'drop_end' to 2MB, meaning it was able to drop only extent B2;
3) We then compute 'clone_len' as 'drop_end' - 'cur_offset' = 2MB - 1MB =
1MB;
4) We then attempt to insert a file extent item at inode B with a file
offset of 5MB, which is the value of clone_info->file_offset. This
fails with error -EEXIST because there's already an extent at that
offset (extent B4);
5) We abort the current transaction with -EEXIST and return that error
to user space as well.
Target range: [1MB, 12MB) (same as source, to make it easier to explain)
1) btrfs_punch_hole_range() gets -ENOSPC from __btrfs_drop_extents();
2) At that point, 'cur_offset' is set to 1MB and __btrfs_drop_extents()
set 'drop_end' to 5MB, meaning it was able to drop only extent B2;
3) We then compute 'clone_len' as 'drop_end' - 'cur_offset' = 5MB - 1MB =
4MB;
4) We then insert a file extent item at inode B with a file offset of 11MB
which is the value of clone_info->file_offset, and a length of 4MB (the
value of 'clone_len'). So we get 2 extents items with ranges that
overlap and an extent length of 4MB, larger then the extent A2 from
inode A (1MB length);
5) After that we end the transaction, balance the btree dirty pages and
then start another or join the previous transaction. It might happen
that the transaction which inserted the incorrect extent was committed
by another task so we end up with extent corruption if a power failure
happens.
So fix this by making sure we attempt to insert the extent to clone at
the destination inode only if we are past dropping the sub-range that
corresponds to a hole.
Fixes: 690a5dbfc51315 ("Btrfs: fix ENOSPC errors, leading to transaction aborts, when cloning extents") Signed-off-by: Filipe Manana <[email protected]> Signed-off-by: David Sterba <[email protected]>
Nikolay Borisov [Mon, 18 Nov 2019 12:16:44 +0000 (14:16 +0200)]
btrfs: Fix error messages in qgroup_rescan_init
The branch of qgroup_rescan_init which is executed from the mount
path prints wrong errors messages. The textual print out in case
BTRFS_QGROUP_STATUS_FLAG_RESCAN/BTRFS_QGROUP_STATUS_FLAG_ON are not
set are transposed. Fix it by exchanging their place.
Mike Rapoport [Wed, 4 Dec 2019 12:35:24 +0000 (14:35 +0200)]
powerpc: Ensure that swiotlb buffer is allocated from low memory
Some powerpc platforms (e.g. 85xx) limit DMA-able memory way below 4G.
If a system has more physical memory than this limit, the swiotlb
buffer is not addressable because it is allocated from memblock using
top-down mode.
Force memblock to bottom-up mode before calling swiotlb_init() to
ensure that the swiotlb buffer is DMA-able.
powerpc/vcpu: Assume dedicated processors as non-preempt
With commit 247f2f6f3c70 ("sched/core: Don't schedule threads on
pre-empted vCPUs"), the scheduler avoids preempted vCPUs to schedule
tasks on wakeup. This leads to wrong choice of CPU, which in-turn
leads to larger wakeup latencies. Eventually, it leads to performance
regression in latency sensitive benchmarks like soltp, schbench etc.
On Powerpc, vcpu_is_preempted() only looks at yield_count. If the
yield_count is odd, the vCPU is assumed to be preempted. However
yield_count is increased whenever the LPAR enters CEDE state (idle).
So any CPU that has entered CEDE state is assumed to be preempted.
Even if vCPU of dedicated LPAR is preempted/donated, it should have
right of first-use since they are supposed to own the vCPU.
On a Power9 System with 32 cores:
# lscpu
Architecture: ppc64le
Byte Order: Little Endian
CPU(s): 128
On-line CPU(s) list: 0-127
Thread(s) per core: 8
Core(s) per socket: 1
Socket(s): 16
NUMA node(s): 2
Model: 2.2 (pvr 004e 0202)
Model name: POWER9 (architected), altivec supported
Hypervisor vendor: pHyp
Virtualization type: para
L1d cache: 32K
L1i cache: 32K
L2 cache: 512K
L3 cache: 10240K
NUMA node0 CPU(s): 0-63
NUMA node1 CPU(s): 64-127
Rahul Tanwar [Thu, 5 Dec 2019 03:01:31 +0000 (11:01 +0800)]
pinctrl: Modify Kconfig to fix linker error
Fix below linker error
ld: drivers/pinctrl/pinctrl-equilibrium.o: in function
`pinconf_generic_dt_node_to_map_all':
pinctrl-equilibrium.c:(.text+0xb): undefined reference
to `pinconf_generic_dt_node_to_map'
Caused by below commit
1948d5c51dba ("pinctrl: Add pinmux & GPIO controller driver for a new SoC")
by adding 'depends on OF' in Kconfig driver entry.
Linus Walleij [Fri, 13 Dec 2019 10:01:10 +0000 (11:01 +0100)]
Merge tag 'intel-pinctrl-v5.5-2' of git://git.kernel.org/pub/scm/linux/kernel/git/pinctrl/intel into fixes
intel-pinctrl for v5.5-2
* Fix Baytrail silicon issue by using a global lock
* Fix North community pin names that user will assume their functions
* Convert Cherryview and Baytrail to pass IRQ chip along with GPIO one
The following is an automated git shortlog grouped by driver:
baytrail:
- Pass irqchip when adding gpiochip
- Add GPIO <-> pin mapping ranges via callback
- Update North Community pin list
- Really serialize all register accesses
cherryview:
- Pass irqchip when adding gpiochip
- Add GPIO <-> pin mapping ranges via callback
- Split out irq hw-init into a separate helper function
pinctrl: pinmux: fix a possible null pointer in pinmux_can_be_used_for_gpio
This commit adds a check on ops pointer to avoid a kernel panic when
ops->strict is used. Indeed, on some pinctrl driver (at least for
pinctrl-stmfx) the pinmux ops is not implemented. Let's assume than gpio
can be used in this case.
Instead of just having an airtime flag in debugfs, turn AQL into a proper
NL80211_EXT_FEATURE, so drivers can turn it on when they are ready, and so
we also expose the presence of the feature to userspace.
This also has the effect of flipping the default, so drivers have to opt in
to using AQL instead of getting it by default with TXQs. To keep
functionality the same as pre-patch, we set this feature for ath10k (which
is where it is needed the most).
While we're at it, split out the debugfs interface so AQL gets its own
per-station debugfs file instead of using the 'airtime' file.
[Johannes:]
This effectively disables AQL for iwlwifi, where it fixes a number of
issues:
* TSO in iwlwifi is causing underflows and associated warnings in AQL
* HE (802.11ax) rates aren't reported properly so at HE rates, AQL could
never have a valid estimate (it'd use 6 Mbps instead of up to 2400!)
* pm-cpuidle:
cpuidle: Drop unnecessary type cast in cpuidle_poll_time()
cpuidle: Fix cpuidle_driver_state_disabled()
cpuidle: use first valid target residency as poll time
* acpi-pm:
ACPI: PM: Avoid attaching ACPI PM domain to certain devices
Dan Carpenter [Tue, 26 Nov 2019 12:09:39 +0000 (15:09 +0300)]
mac80211: airtime: Fix an off by one in ieee80211_calc_rx_airtime()
This code was copied from mt76 and inherited an off by one bug from
there. The > should be >= so that we don't read one element beyond
the end of the array.
Stefan Bühler [Tue, 26 Nov 2019 10:05:44 +0000 (11:05 +0100)]
cfg80211: fix double-free after changing network namespace
If wdev->wext.keys was initialized it didn't get reset to NULL on
unregister (and it doesn't get set in cfg80211_init_wdev either), but
wdev is reused if unregister was triggered through
cfg80211_switch_netns.
The next unregister (for whatever reason) will try to free
wdev->wext.keys again.
Fredrik Olofsson [Tue, 19 Nov 2019 13:34:51 +0000 (14:34 +0100)]
mac80211: fix TID field in monitor mode transmit
Fix overwriting of the qos_ctrl.tid field for encrypted frames injected on
a monitor interface. While qos_ctrl.tid is not encrypted, it's used as an
input into the encryption algorithm so it's protected, and thus cannot be
modified after encryption. For injected frames, the encryption may already
have been done in userspace, so we cannot change any fields.
Before passing the frame to the driver, the qos_ctrl.tid field is updated
from skb->priority. Prior to dbd50a851c50 skb->priority was updated in
ieee80211_select_queue_80211(), but this function is no longer always
called.
Update skb->priority in ieee80211_monitor_start_xmit() so that the value
is stored, and when later code 'modifies' the TID it really sets it to
the same value as before, preserving the encryption.
Juergen Gross [Thu, 12 Dec 2019 14:17:50 +0000 (15:17 +0100)]
xen/balloon: fix ballooned page accounting without hotplug enabled
When CONFIG_XEN_BALLOON_MEMORY_HOTPLUG is not defined
reserve_additional_memory() will set balloon_stats.target_pages to a
wrong value in case there are still some ballooned pages allocated via
alloc_xenballooned_pages().
This will result in balloon_process() no longer be triggered when
ballooned pages are freed in batches.
Paul Durrant [Tue, 10 Dec 2019 14:53:05 +0000 (14:53 +0000)]
xen-blkback: prevent premature module unload
Objects allocated by xen_blkif_alloc come from the 'blkif_cache' kmem
cache. This cache is destoyed when xen-blkif is unloaded so it is
necessary to wait for the deferred free routine used for such objects to
complete. This necessity was missed in commit 14855954f636 "xen-blkback:
allow module to be cleanly unloaded". This patch fixes the problem by
taking/releasing extra module references in xen_blkif_alloc/free()
respectively.
Pavel Shilovsky [Tue, 10 Dec 2019 19:44:52 +0000 (11:44 -0800)]
CIFS: Close cached root handle only if it has a lease
SMB2_tdis() checks if a root handle is valid in order to decide
whether it needs to close the handle or not. However if another
thread has reference for the handle, it may end up with putting
the reference twice. The extra reference that we want to put
during the tree disconnect is the reference that has a directory
lease. So, track the fact that we have a directory lease and
close the handle only in that case.
Steve French [Tue, 10 Dec 2019 04:34:10 +0000 (22:34 -0600)]
SMB3: Fix crash in SMB2_open_init due to uninitialized field in compounding path
Ran into an intermittent crash in
SMB2_open_init+0x2f6/0x970
due to oparms.cifs_sb not being initialized when called from:
smb2_compound_op+0x45d/0x1690
Zero the whole oparms struct in the compounding path before setting up the
oparms so we don't risk any uninitialized fields.
Fixes: fdef665ba44a ("smb3: fix mode passed in on create for modetosid mount option") Signed-off-by: Steve French <[email protected]> Acked-by: Ronnie Sahlberg <[email protected]> Reviewed-by: Pavel Shilovsky <[email protected]>
Dave Airlie [Fri, 13 Dec 2019 04:50:01 +0000 (14:50 +1000)]
Merge tag 'drm-fixes-5.5-2019-12-12' of git://people.freedesktop.org/~agd5f/linux into drm-fixes
drm-fixes-5.5-2019-12-12:
amdgpu:
- DC fixes for renoir
- Gfx8 fence flush align with mesa
- Power profile fix for arcturus
- Freesync fix
- DC I2c over aux fix
- DC aux defer fix
- GPU reset fix
- GPUVM invalidation semaphore fixes for PCO and SR-IOV
- Golden settings updates for gfx10
Dave Airlie [Fri, 13 Dec 2019 04:44:08 +0000 (14:44 +1000)]
Merge tag 'drm-intel-fixes-2019-12-12' of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes
- Fix user reported issue #673: GPU hang on transition to idle
- Avoid corruption on the top of the screen on GLK+ by disabling FBC
- Fix non-privileged access to OA on Tigerlake
- Fix HDCP code not to touch global state when just computing commit
- Fix CI splat by saving irqstate around virtual_context_destroy
- Serialise context retirement possibly on another CPU
Tina Zhang [Fri, 13 Dec 2019 03:23:14 +0000 (11:23 +0800)]
drm/i915/gvt: Pin vgpu dma address before using
Dma-buf display uses the vgpu dma address saved in the guest part GGTT
table which is updated by vCPU thread. In host side, when the dma
address is used by qemu ui thread, gvt-g must make sure the dma address
is validated before letting it go to the HW. Invalid guest dma address
will easily cause DMA fault and make GPU hang.
Zhenyu Wang [Thu, 12 Dec 2019 08:46:14 +0000 (16:46 +0800)]
drm/i915/gvt: set guest display buffer as readonly
We shouldn't allow write for exposed guest display buffer which
doesn't make sense. So explicitly set read only flag for display
dmabuf allocated object.
Stephen Boyd [Fri, 13 Dec 2019 03:00:01 +0000 (19:00 -0800)]
Merge tag 'imx-clk-fixes-5.5' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into clk-fixes
Pull i.MX clk fixes from Shawn Guo:
- Add missing lock to divider in the composite driver for exclusive
register access
- Add missing sentinel for ulp_div_table in clk-imx7ulp driver
- Fix clk_pll14xx_wait_lock() function which calls into
readl_poll_timeout() with incorrect parameter
* tag 'imx-clk-fixes-5.5' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
clk: imx: pll14xx: fix clk_pll14xx_wait_lock
clk: imx: clk-imx7ulp: Add missing sentinel of ulp_div_table
clk: imx: clk-composite-8m: add lock to gate/mux
Jerome Brunet [Tue, 3 Dec 2019 08:08:05 +0000 (09:08 +0100)]
clk: walk orphan list on clock provider registration
So far, we walked the orphan list every time a new clock was registered
in CCF. This was fine since the clocks were only referenced by name.
Now that the clock can be referenced through DT, it is not enough:
* Controller A register first a reference clocks from controller B
through DT.
* Controller B register all its clocks then register the provider.
Each time controller B registers a new clock, the orphan list is walked
but it can't match since the provider is registered yet. When the
provider is finally registered, the orphan list is not walked unless
another clock is registered afterward.
This can lead to situation where some clocks remain orphaned even if
the parent is available.
Walking the orphan list on provider registration solves the problem.
Saravana Kannan [Mon, 9 Dec 2019 19:31:19 +0000 (11:31 -0800)]
of/platform: Unconditionally pause/resume sync state during kernel init
Commit 5e6669387e22 ("of/platform: Pause/resume sync state during init
and of_platform_populate()") paused/resumed sync state during init only
if Linux had parsed and populated a devicetree.
However, the check for that (of_have_populated_dt()) can change after
of_platform_default_populate_init() executes. One example of this is
when devicetree unittests are enabled. This causes an unmatched
pause/resume of sync state. To avoid this, just unconditionally
pause/resume sync state during init.
Fixes: 5e6669387e22 ("of/platform: Pause/resume sync state during init and of_platform_populate()") Reported-by: kernel test robot <[email protected]> Signed-off-by: Saravana Kannan <[email protected]> Reviewed-by: Frank Rowand <[email protected]> Signed-off-by: Rob Herring <[email protected]>
Rob Herring [Wed, 11 Dec 2019 14:51:01 +0000 (08:51 -0600)]
dt-bindings: memory-controllers: tegra: Fix type references
Json-schema requires a $ref to be under an 'allOf' if there are
additional constraints otherwise the additional constraints are
ignored. (Note that this behavior will be changed in draft8.)
Yishai Hadas [Thu, 12 Dec 2019 10:02:37 +0000 (12:02 +0200)]
IB/mlx5: Fix device memory flows
Fix device memory flows so that only once there will be no live mmaped
VA to a given allocation the matching object will be destroyed.
This prevents a potential scenario that existing VA that was mmaped by
one process might still be used post its deallocation despite that it's
owned now by other process.
The above is achieved by integrating with IB core APIs to manage
mmap/munmap. Only once the refcount will become 0 the DM object and its
underlay area will be freed.
Eric Biggers [Wed, 9 Oct 2019 23:04:43 +0000 (16:04 -0700)]
KEYS: remove CONFIG_KEYS_COMPAT
KEYS_COMPAT now always takes the value of COMPAT && KEYS. But the
security/keys/ directory is only compiled if KEYS is enabled, so in
practice KEYS_COMPAT is the same as COMPAT. Therefore, remove the
unnecessary KEYS_COMPAT and just use COMPAT directly.
PCI: rockchip: Fix IO outbound ATU register number
Since 62240a88004b ("PCI: rockchip: Drop storing driver private outbound
resource data), the offset calculation is wrong to access the register
number to program the IO outbound ATU.
Fix this by computing the ATU IO register number based on the number of MEM
registers, not the size of the IO region.
This causes 'synchronous external aborts' like the following:
changzhu [Tue, 10 Dec 2019 14:00:59 +0000 (22:00 +0800)]
drm/amdgpu: add invalidate semaphore limit for SRIOV and picasso in gmc9
It may fail to load guest driver in round 2 or cause Xstart problem
when using invalidate semaphore for SRIOV or picasso. So it needs avoid
using invalidate semaphore for SRIOV and picasso.
Mark Zhang [Thu, 12 Dec 2019 09:12:12 +0000 (11:12 +0200)]
RDMA/counter: Prevent auto-binding a QP which are not tracked with res
Some QPs (e.g. XRC QP) are not tracked in kernel, in this case they have
an invalid res and should not be bound to any dynamically-allocated
counter in auto mode.
Manish Chopra [Thu, 12 Dec 2019 14:49:28 +0000 (06:49 -0800)]
qede: Fix multicast mac configuration
Driver doesn't accommodate the configuration for max number
of multicast mac addresses, in such particular case it leaves
the device with improper/invalid multicast configuration state,
causing connectivity issues (in lacp bonding like scenarios).
Lan78xx driver accesses the PHY registers through MDIO bus over USB
connection. When performing a suspend/resume, the PHY registers can be
accessed before the USB connection is resumed. This will generate an
error and will prevent the device to resume correctly.
This patch adds the dependency between the MDIO bus and USB device to
allow correct handling of suspend/resume.
Fixes: ce85e13ad6ef ("lan78xx: Update to use phylib instead of mii_if_info.") Signed-off-by: Cristian Birsan <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Linus Torvalds [Thu, 12 Dec 2019 18:56:37 +0000 (10:56 -0800)]
Merge tag 'ceph-for-5.5-rc2' of git://github.com/ceph/ceph-client
Pull ceph fixes from Ilya Dryomov:
"A fix to avoid a corner case when scheduling cap reclaim in batches
from Xiubo, a patch to add some observability into cap waiters from
Jeff and a couple of cleanups"
* tag 'ceph-for-5.5-rc2' of git://github.com/ceph/ceph-client:
ceph: add more debug info when decoding mdsmap
ceph: switch to global cap helper
ceph: trigger the reclaim work once there has enough pending caps
ceph: show tasks waiting on caps in debugfs caps file
ceph: convert int fields in ceph_mount_options to unsigned int
ksys_dup() is used only at one place in the kernel, namely to duplicate
fd 0 of /dev/console to stdout and stderr. The same functionality can be
achieved by using functions already available within the kernel namespace.
Olof Johansson [Thu, 12 Dec 2019 17:38:22 +0000 (09:38 -0800)]
Merge tag 'imx-fixes-5.5' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into arm/fixes
i.MX fixes for 5.5:
- Add missing jedec,spi-nor compatible for imx6ul-14x14-evk board,
so that SPI NOR device can be probed.
- Fix power button of E60K02 board by removing LDORTC2 regulator.
- A couple of fixes on serial number support of i.MX6ULL/ULZ SoCs to
remove the boot regression caused by 8267ff89b713 ("ARM: imx: Add
serial number support for i.MX6/7 SoCs").
- A couple of fixes on LS1028A SoC TMU regarding to calibration data
and reboot register configuration.
- Fix a regression seen on imx6ul-evk board by marking always-on for
the regulator that is shared by many peripherals.
- Explicitly restore CONFIG_DEBUG_FS in imx_v6_v7_defconfig.
* tag 'imx-fixes-5.5' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
ARM: imx: Fix boot crash if ocotp is not found
ARM: imx_v6_v7_defconfig: Explicitly restore CONFIG_DEBUG_FS
ARM: dts: imx6ul-evk: Fix peripheral regulator
arm64: dts: ls1028a: fix reboot node
arm64: dts: ls1028a: fix typo in TMU calibration data
ARM: imx: Correct ocotp id for serial number support of i.MX6ULL/ULZ SoCs
ARM: dts: e60k02: fix power button
ARM: dts: imx6ul: imx6ul-14x14-evk.dtsi: Fix SPI NOR probing
cpufreq: Avoid leaving stale IRQ work items during CPU offline
The scheduler code calling cpufreq_update_util() may run during CPU
offline on the target CPU after the IRQ work lists have been flushed
for it, so the target CPU should be prevented from running code that
may queue up an IRQ work item on it at that point.
Unfortunately, that may not be the case if dvfs_possible_from_any_cpu
is set for at least one cpufreq policy in the system, because that
allows the CPU going offline to run the utilization update callback
of the cpufreq governor on behalf of another (online) CPU in some
cases.
If that happens, the cpufreq governor callback may queue up an IRQ
work on the CPU running it, which is going offline, and the IRQ work
may not be flushed after that point. Moreover, that IRQ work cannot
be flushed until the "offlining" CPU goes back online, so if any
other CPU calls irq_work_sync() to wait for the completion of that
IRQ work, it will have to wait until the "offlining" CPU is back
online and that may not happen forever. In particular, a system-wide
deadlock may occur during CPU online as a result of that.
The failing scenario is as follows. CPU0 is the boot CPU, so it
creates a cpufreq policy and becomes the "leader" of it
(policy->cpu). It cannot go offline, because it is the boot CPU.
Next, other CPUs join the cpufreq policy as they go online and they
leave it when they go offline. The last CPU to go offline, say CPU3,
may queue up an IRQ work while running the governor callback on
behalf of CPU0 after leaving the cpufreq policy because of the
dvfs_possible_from_any_cpu effect described above. Then, CPU0 is
the only online CPU in the system and the stale IRQ work is still
queued on CPU3. When, say, CPU1 goes back online, it will run
irq_work_sync() to wait for that IRQ work to complete and so it
will wait for CPU3 to go back online (which may never happen even
in principle), but (worse yet) CPU0 is waiting for CPU1 at that
point too and a system-wide deadlock occurs.
To address this problem notice that CPUs which cannot run cpufreq
utilization update code for themselves (for example, because they
have left the cpufreq policies that they belonged to), should also
be prevented from running that code on behalf of the other CPUs that
belong to a cpufreq policy with dvfs_possible_from_any_cpu set and so
in that case the cpufreq_update_util_data pointer of the CPU running
the code must not be NULL as well as for the CPU which is the target
of the cpufreq utilization update in progress.
Accordingly, change cpufreq_this_cpu_can_update() into a regular
function in kernel/sched/cpufreq.c (instead of a static inline in a
header file) and make it check the cpufreq_update_util_data pointer
of the local CPU if dvfs_possible_from_any_cpu is set for the target
cpufreq policy.
Also update the schedutil governor to do the
cpufreq_this_cpu_can_update() check in the non-fast-switch
case too to avoid the stale IRQ work issues.
Marc Zyngier [Wed, 11 Dec 2019 16:56:48 +0000 (16:56 +0000)]
KVM: arm/arm64: Properly handle faulting of device mappings
A device mapping is normally always mapped at Stage-2, since there
is very little gain in having it faulted in.
Nonetheless, it is possible to end-up in a situation where the device
mapping has been removed from Stage-2 (userspace munmaped the VFIO
region, and the MMU notifier did its job), but present in a userspace
mapping (userpace has mapped it back at the same address). In such
a situation, the device mapping will be demand-paged as the guest
performs memory accesses.
This requires to be careful when dealing with mapping size, cache
management, and to handle potential execution of a device mapping.
Tony Lindgren [Thu, 12 Dec 2019 16:20:10 +0000 (08:20 -0800)]
bus: ti-sysc: Fix missing reset delay handling
We have dts property for "ti,sysc-delay-us", and we're using it, but the
wait after OCP softreset only happens if devices are probed in legacy mode.
Let's add a delay after writing the OCP softreset when specified.
Fixes: e0db94fe87da ("bus: ti-sysc: Make OCP reset work for sysstatus and sysconfig reset bits") Cc: Keerthy <[email protected]> Signed-off-by: Tony Lindgren <[email protected]>
Early revisions of the AST2600 datasheet are conflicted about the state
of the LPC/eSPI strapping bit (SCU510[6]). Conversations with ASPEED
determined that the reference pinmux configuration tables were in error
and the SCU documentation contained the correct configuration. Update
the driver to reflect the state described in the SCU documentation.
Logan Gunthorpe [Tue, 10 Dec 2019 18:47:04 +0000 (11:47 -0700)]
block: fix NULL pointer dereference in account statistics with IDE
The IDE driver creates some passthru requests which never get
submitted to the block layer in such a way that blk_account_io_start()
gets called. However, the driver still calls __blk_mq_end_request() in
ide_end_rq() which will call blk_account_io_completion() which tries
to dereferences req->part which is never set. See ide_prep_sense() for
an example of where these requests come from.
To fix this, blk_account_io_completion() and blk_account_io_done()
should do nothing if req->part is not set.
changzhu [Tue, 10 Dec 2019 02:23:09 +0000 (10:23 +0800)]
drm/amdgpu: avoid using invalidate semaphore for picasso
It may cause timeout waiting for sem acquire in VM flush when using
invalidate semaphore for picasso. So it needs to avoid using invalidate
semaphore for piasso.
In prepare_namespace(), do_mount() can be used instead of ksys_mount()
as the first and third argument are const strings in the kernel, the
second and fourth argument are passed through anyway, and the fifth
argument is NULL.
In do_mount_root(), ksys_mount() is called with the first and third
argument being already kernelspace strings, which do not need to be
copied over from userspace to kernelspace (again). The second and
fourth arguments are passed through to do_mount() anyway. The fifth
argument, while already residing in kernelspace, needs to be put into
a page of its own. Then, do_mount() can be used instead of
ksys_mount().
Once this is done, there are no in-kernel users to ksys_mount() left,
which can therefore be removed.
All three calls to ksys_mount() in initrd-related kernel code can
be switched over to do_mount():
- the first and third arguments are const strings in the kernel,
and do not need to be copied over from userspace;
- the fifth argument is NULL, and therefore no page needs to be,
copied over from userspace;
- the second and fourth argument are passed through anyway.
In devtmpfs, do_mount() can be called directly instead of complex wrapping
by ksys_mount():
- the first and third arguments are const strings in the kernel,
and do not need to be copied over from userspace;
- the fifth argument is NULL, and therefore no page needs to be
copied over from userspace;
- the second and fourth argument are passed through anyway.
A break interrupt will be generated if the RX line was pulled low, which
means some abnomal behaviors occurred of the UART. In this case, we still
need to clear this break interrupt status, otherwise it will cause irq
storm to crash the whole system.
Leo Yan [Wed, 27 Nov 2019 14:15:43 +0000 (22:15 +0800)]
tty: serial: msm_serial: Fix lockup for sysrq and oops
As the commit 677fe555cbfb ("serial: imx: Fix recursive locking bug")
has mentioned the uart driver might cause recursive locking between
normal printing and the kernel debugging facilities (e.g. sysrq and
oops). In the commit it gave out suggestion for fixing recursive
locking issue: "The solution is to avoid locking in the sysrq case
and trylock in the oops_in_progress case."
This patch follows the suggestion (also used the exactly same code with
other serial drivers, e.g. amba-pl011.c) to fix the recursive locking
issue, this can avoid stuck caused by deadlock and print out log for
sysrq and oops.
Heikki Krogerus [Thu, 12 Dec 2019 09:37:13 +0000 (12:37 +0300)]
usb: dwc3: pci: add ID for the Intel Comet Lake -H variant
The original ID that was added for Comet Lake PCH was
actually for the -LP (low power) variant even though the
constant for it said CMLH. Changing that while at it.
Will Deacon [Thu, 12 Dec 2019 09:40:49 +0000 (09:40 +0000)]
KVM: arm64: Ensure 'params' is initialised when looking up sys register
Commit 4b927b94d5df ("KVM: arm/arm64: vgic: Introduce find_reg_by_id()")
introduced 'find_reg_by_id()', which looks up a system register only if
the 'id' index parameter identifies a valid system register. As part of
the patch, existing callers of 'find_reg()' were ported over to the new
interface, but this breaks 'index_to_sys_reg_desc()' in the case that the
initial lookup in the vCPU target table fails because we will then call
into 'find_reg()' for the system register table with an uninitialised
'param' as the key to the lookup.
GCC 10 is bright enough to spot this (amongst a tonne of false positives,
but hey!):
| arch/arm64/kvm/sys_regs.c: In function ‘index_to_sys_reg_desc.part.0.isra’:
| arch/arm64/kvm/sys_regs.c:983:33: warning: ‘params.Op2’ may be used uninitialized in this function [-Wmaybe-uninitialized]
| 983 | (u32)(x)->CRn, (u32)(x)->CRm, (u32)(x)->Op2);
| [...]
Revert the hunk of 4b927b94d5df which breaks 'index_to_sys_reg_desc()' so
that the old behaviour of checking the index upfront is restored.
Guenter Roeck [Thu, 12 Dec 2019 08:34:03 +0000 (16:34 +0800)]
nios2: Fix ioremap
Commit 5ace77e0b41a ("nios2: remove __ioremap") removed the following code,
with the argument that cacheflag is always 0 and the expression would
therefore always be false.
crypto: arm/curve25519 - add arch-specific key generation function
Somehow this was forgotten when Zinc was being split into oddly shaped
pieces, resulting in linker errors. The x86_64 glue has a specific key
generation implementation, but the Arm one does not. However, it can
still receive the NEON speedups by calling the ordinary DH function
using the base point.
Jens Axboe [Thu, 12 Dec 2019 03:49:58 +0000 (20:49 -0700)]
Merge branch 'md-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/song/md into for-linus
Pull MD fixes from Song.
* 'md-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/song/md:
md: make sure desc_nr less than MD_SB_DISKS
md: raid1: check rdev before reference in raid1_sync_request func
raid5: need to set STRIPE_HANDLE for batch head
Leonard Crestez [Tue, 10 Dec 2019 21:49:28 +0000 (23:49 +0200)]
ARM: imx: Fix boot crash if ocotp is not found
The imx_soc_device_init functions tries to fetch the ocotp regmap in
order to soc serial number. If regmap fetch fails then a message is
printed but regmap_read is called anyway and the system crashes.
Failing to lookup ocotp regmap shouldn't be a fatal boot error so check
that the pointer is valid.
Only side-effect of ocotp lookup failure now is that serial number will
be reported as all-zeros which is acceptable.
Linus Torvalds [Thu, 12 Dec 2019 02:10:40 +0000 (18:10 -0800)]
Merge tag 'afs-fixes-20191211' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs
Pull AFS fixes from David Howells:
"Fixes for AFS plus one patch to make debugging easier:
- Fix how addresses are matched to server records. This is currently
incorrect which means cache invalidation callbacks from the server
don't necessarily get delivered correctly. This causes stale data
and metadata to be seen under some circumstances.
- Make the dynamic root superblock R/W so that rpm/dnf can reapply
the SELinux label to it when upgrading the Fedora filesystem-afs
package. If the filesystem is R/O, this fails and the upgrade
fails.
It might be better in future to allow setxattr from an LSM to
bypass the R/O protections, if only for pseudo-filesystems.
- Fix the parsing of mountpoint strings. The mountpoint object has to
have a terminal dot, whereas the source/device string passed to
mount should not. This confuses type-forcing suffix detection
leading to the wrong volume variant being mounted.
- Make lookups in the dynamic root superblock for creation events
(such as mkdir) fail with EOPNOTSUPP rather than something like
EEXIST. The dynamic root only allows implicit creation by the
->lookup() method - and only if the target cell exists.
- Fix the looking up of an AFS superblock to include the cell in the
matching key - otherwise all volumes with the same ID number are
treated as the same thing, irrespective of which cell they're in.
- Show the volume name of each volume in the volume records displayed
in /proc/net/afs/<cell>/volumes. This proved useful in debugging as
it provides a way to map the volume IDs to names, where the names
are what appear in /proc/mounts"
* tag 'afs-fixes-20191211' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs:
afs: Show volume name in /proc/net/afs/<cell>/volumes
afs: Fix missing cell comparison in afs_test_super()
afs: Fix creation calls in the dynamic root to fail with EOPNOTSUPP
afs: Fix mountpoint parsing
afs: Fix SELinux setting security label on /afs
afs: Fix afs_find_server lookups for ipv4 peers
Dan Williams [Thu, 14 Nov 2019 00:22:06 +0000 (16:22 -0800)]
tools/testing/nvdimm: Fix mock support for ioremap
After commit d092a8707326 "arch: rely on asm-generic/io.h for default
ioremap_* definitions" the ioremap_nocache() symbol has been replaced
with ioremap(). Update the mocked symbol list for nvdimm testing.
Daniel T. Lee [Thu, 5 Dec 2019 08:01:14 +0000 (17:01 +0900)]
samples: bpf: fix syscall_tp due to unused syscall
Currently, open() is called from the user program and it calls the syscall
'sys_openat', not the 'sys_open'. This leads to an error of the program
of user side, due to the fact that the counter maps are zero since no
function such 'sys_open' is called.
This commit adds the kernel bpf program which are attached to the
tracepoint 'sys_enter_openat' and 'sys_enter_openat'.
Fixes: 1da236b6be963 ("bpf: add a test case for syscalls/sys_{enter|exit}_* tracepoints") Signed-off-by: Daniel T. Lee <[email protected]> Signed-off-by: Alexei Starovoitov <[email protected]>
Daniel T. Lee [Thu, 5 Dec 2019 08:01:13 +0000 (17:01 +0900)]
samples: bpf: Replace symbol compare of trace_event
Previously, when this sample is added, commit 1c47910ef8013
("samples/bpf: add perf_event+bpf example"), a symbol 'sys_read' and
'sys_write' has been used without no prefixes. But currently there are
no exact symbols with these under kallsyms and this leads to failure.
This commit changes exact compare to substring compare to keep compatible
with exact symbol or prefixed symbol.
bpf: Make BPF trampoline use register_ftrace_direct() API
Make BPF trampoline attach its generated assembly code to kernel functions via
register_ftrace_direct() API. It helps ftrace-based tracers co-exist with BPF
trampoline on the same kernel function. It also switches attaching logic from
arch specific text_poke to generic ftrace that is available on many
architectures. text_poke is still necessary for bpf-to-bpf attach and for
bpf_tail_call optimization.
Jens Axboe [Wed, 11 Dec 2019 22:55:43 +0000 (15:55 -0700)]
io_uring: ensure we return -EINVAL on unknown opcode
If we submit an unknown opcode and have fd == -1, io_op_needs_file()
will return true as we default to needing a file. Then when we go and
assign the file, we find the 'fd' invalid and return -EBADF. We really
should be returning -EINVAL for that case, as we normally do for
unsupported opcodes.
Change io_op_needs_file() to have the following return values:
0 - does not need a file
1 - does need a file
< 0 - error value
and use this to pass back the right value for this invalid case.
Brian Foster [Wed, 11 Dec 2019 21:18:38 +0000 (13:18 -0800)]
xfs: stabilize insert range start boundary to avoid COW writeback race
generic/522 (fsx) occasionally fails with a file corruption due to
an insert range operation. The primary characteristic of the
corruption is a misplaced insert range operation that differs from
the requested target offset. The reason for this behavior is a race
between the extent shift sequence of an insert range and a COW
writeback completion that causes a front merge with the first extent
in the shift.
The shift preparation function flushes and unmaps from the target
offset of the operation to the end of the file to ensure no
modifications can be made and page cache is invalidated before file
data is shifted. An insert range operation then splits the extent at
the target offset, if necessary, and begins to shift the start
offset of each extent starting from the end of the file to the start
offset. The shift sequence operates at extent level and so depends
on the preparation sequence to guarantee no changes can be made to
the target range during the shift. If the block immediately prior to
the target offset was dirty and shared, however, it can undergo
writeback and move from the COW fork to the data fork at any point
during the shift. If the block is contiguous with the block at the
start offset of the insert range, it can front merge and alter the
start offset of the extent. Once the shift sequence reaches the
target offset, it shifts based on the latest start offset and
silently changes the target offset of the operation and corrupts the
file.
To address this problem, update the shift preparation code to
stabilize the start boundary along with the full range of the
insert. Also update the existing corruption check to fail if any
extent is shifted with a start offset behind the target offset of
the insert range. This prevents insert from racing with COW
writeback completion and fails loudly in the event of an unexpected
extent shift.
Linus Torvalds [Wed, 11 Dec 2019 20:25:32 +0000 (12:25 -0800)]
Merge tag 'erofs-for-5.5-rc2-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs
Pull erofs fixes from Gao Xiang:
"Mainly address a regression reported by David recently observed
together with overlayfs due to the improper return value of
listxattr() without xattr. Update outdated expressions in document as
well.
Summary:
- Fix improper return value of listxattr() with no xattr
- Keep up documentation with latest code"
* tag 'erofs-for-5.5-rc2-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs:
erofs: update documentation
erofs: zero out when listxattr is called with no xattr
Linus Torvalds [Wed, 11 Dec 2019 20:22:38 +0000 (12:22 -0800)]
Merge tag 'trace-v5.5-3' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing fixes from Steven Rostedt:
- Remove code I accidentally applied when doing a minor fix up to a
patch, and then using "git commit -a --amend", which pulled in some
other changes I was playing with.
- Remove an used variable in trace_events_inject code
- Fix function graph tracer when it traces a ftrace direct function.
It will now ignore tracing a function that has a ftrace direct
tramploine attached. This is needed for eBPF to use the ftrace direct
code.
* tag 'trace-v5.5-3' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
ftrace: Fix function_graph tracer interaction with BPF trampoline
tracing: remove set but not used variable 'buffer'
module: Remove accidental change of module_enable_x()
Linus Torvalds [Wed, 11 Dec 2019 19:46:19 +0000 (11:46 -0800)]
pipe: simplify signal handling in pipe_read() and add comments
There's no need to separately check for signals while inside the locked
region, since we're going to do "wait_event_interruptible()" right
afterwards anyway, and the error handling is much simpler there.
The check for whether we had already read anything was also redundant,
since we no longer do the odd merging of reads when there are pending
writers.
But perhaps more importantly, this adds commentary about why we still
need to wake up possible writers even though we didn't read any data,
and why we can skip all the finishing touches now if we get a signal (or
had a signal pending) while waiting for more data.
[ This is a split-out cleanup from my "make pipe IO use exclusive wait
queues" thing, which I can't apply because it triggers a nasty bug in
the GNU make jobserver - Linus ]
projects / linux.git / log