Filipe Manana [Sat, 3 Aug 2019 08:53:16 +0000 (09:53 +0100)]
Btrfs: fix memory leaks in the test test_find_first_clear_extent_bit
The test creates an extent io tree and sets several ranges with the
CHUNK_ALLOCATED and CHUNK_TRIMMED bits, resulting in the allocation of
several extent state structures. However the test never clears those
ranges, resulting in memory leaks of the extent state structures.
This is detected when CONFIG_BTRFS_DEBUG is set once we remove the
btrfs module (rmmod btrfs):
[57399.787918] BTRFS: state leak: start 67108864 end 75497471 state 1 in tree 1 refs 1
[57399.790155] BTRFS: state leak: start 33554432 end 67108863 state 33 in tree 1 refs 1
[57399.791941] BTRFS: state leak: start 1048576 end 4194303 state 33 in tree 1 refs 1
[57399.793753] BTRFS: state leak: start 67108864 end 75497471 state 1 in tree 1 refs 1
[57399.795188] BTRFS: state leak: start 33554432 end 67108863 state 33 in tree 1 refs 1
[57399.796453] BTRFS: state leak: start 1048576 end 4194303 state 33 in tree 1 refs 1
[57399.797765] BTRFS: state leak: start 67108864 end 75497471 state 1 in tree 1 refs 1
[57399.799049] BTRFS: state leak: start 33554432 end 67108863 state 33 in tree 1 refs 1
[57399.800142] BTRFS: state leak: start 1048576 end 4194303 state 33 in tree 1 refs 1
[57399.801126] BTRFS: state leak: start 67108864 end 75497471 state 1 in tree 1 refs 1
[57399.802106] BTRFS: state leak: start 33554432 end 67108863 state 33 in tree 1 refs 1
[57399.803119] BTRFS: state leak: start 1048576 end 4194303 state 33 in tree 1 refs 1
[57399.804153] BTRFS: state leak: start 67108864 end 75497471 state 1 in tree 1 refs 1
[57399.805196] BTRFS: state leak: start 33554432 end 67108863 state 33 in tree 1 refs 1
[57399.806191] BTRFS: state leak: start 1048576 end 4194303 state 33 in tree 1 refs 1
The start and end offsets reported correspond exactly to the ranges
used by the test.
So fix that by clearing all the ranges when the test finishes.
Fixes: 1eaebb341d2b41 ("btrfs: Don't trim returned range based on input value in find_first_clear_extent_bit") Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
David Sterba [Thu, 13 Jun 2019 15:23:02 +0000 (17:23 +0200)]
btrfs: sysfs: add debugging exports
Add 'debug' directories to global sysfs and per-filesystem. This will
replace the debugfs directory. The sysfs location is simpler and builds
on top of the existing file hierarchy so there will hopefully be no more
questions about the sample debugfs file.
The directory is called 'debug' and only present under
CONFIG_BTRFS_DEBUG so this will not affect productions builds.
Josef Bacik [Thu, 20 Jun 2019 19:37:52 +0000 (15:37 -0400)]
btrfs: make caching_thread use btrfs_find_next_key
extent-tree.c has a find_next_key that just walks up the path to find
the next key, but it is used for both the caching stuff and the snapshot
delete stuff. The snapshot deletion stuff is special so it can't really
use btrfs_find_next_key, but the caching thread stuff can. We just need
to fix btrfs_find_next_key to deal with ->skip_locking and then it works
exactly the same as the private find_next_key helper.
Signed-off-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: David Sterba <dsterba@suse.com>
Josef Bacik [Thu, 20 Jun 2019 19:37:45 +0000 (15:37 -0400)]
btrfs: migrate the block group lookup code
Move these bits first as they are the easiest to move. Export two of
the helpers so they can be moved all at once.
Signed-off-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: David Sterba <dsterba@suse.com>
[ minor style updates ] Signed-off-by: David Sterba <dsterba@suse.com>
Josef Bacik [Thu, 20 Jun 2019 19:37:44 +0000 (15:37 -0400)]
btrfs: move basic block_group definitions to their own header
This is prep work for moving all of the block group cache code into its
own file.
Signed-off-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: David Sterba <dsterba@suse.com>
[ minor comment updates ] Signed-off-by: David Sterba <dsterba@suse.com>
Josef Bacik [Thu, 20 Jun 2019 19:37:43 +0000 (15:37 -0400)]
btrfs: move btrfs_add_free_space out of a header file
This is prep work for moving block_group_cache around. Having this in
the header file makes the header file include need to be in a certain
order, which is awkward, so just move it into free-space-cache.c and
then we can re-arrange later.
Signed-off-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
David Sterba [Thu, 1 Aug 2019 17:53:02 +0000 (19:53 +0200)]
btrfs: remove unused btrfs_device::flush_bio_sent
The status of flush bio is tracked as a status bit, changed in commit 1c3063b6dbfa ("btrfs: cleanup device states define
BTRFS_DEV_STATE_FLUSH_SENT"), the flush_bio_sent was forgotten.
Reviewed-by: Anand Jain <anand.jain@oracle.com> Signed-off-by: David Sterba <dsterba@suse.com>
Filipe Manana [Tue, 2 Jul 2019 14:23:07 +0000 (15:23 +0100)]
Btrfs: remove unnecessary condition in btrfs_clone() to avoid too much nesting
The bulk of the work done when cloning extents, at ioctl.c:btrfs_clone(),
is done inside an if statement that checks if the found key has the type
BTRFS_EXTENT_DATA_KEY. That if statement is redundant however, because
btrfs_search_slot() always leaves us in a leaf slot that points to a key
that is always greater then or equals to the search key, and our search
key here has that type, and because we bail out before that if statement
if the key at the given leaf slot is greater then BTRFS_EXTENT_DATA_KEY.
Therefore just remove that if statement, not only because it is useless
but mostly because it increases the nesting/indentation level in this
function which is quite big and makes things a bit awkward whenever I need
to fix something that requires changing btrfs_clone() (and it has been
like that for many years already).
Signed-off-by: Filipe Manana <fdmanana@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Nikolay Borisov [Wed, 3 Jul 2019 12:32:59 +0000 (15:32 +0300)]
btrfs: Refactor btrfs_calc_avail_data_space
Simplify the code by removing variables that don't bring any real value
as well as simplifying the checks when buidling the candidate list of
devices. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Nikolay Borisov [Thu, 23 May 2019 11:51:26 +0000 (14:51 +0300)]
btrfs: Remove unnecessary check from join_running_log_trans
join_running_log_trans checks btrfs_root::log_root outside of
btrfs_root::log_mutex to avoid contention on the mutex. Turns out this
check is not necessary because the two callers of join_running_log_trans
(both of which deal with removing entries from the tree-log during
unlink) explicitly check whether the respective inode has been logged in
the current transaction.
If it hasn't then it won't have any items in the tree-log and call path
will return before calling join_running_log_trans. If the check passes,
however, then it's guaranteed that btrfs_root::log_root is set because
the inode is logged.
Those guarantees allows us to remove the speculative as well as the
implicity and tricky memory barrier.
Reviewed-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Filipe Manana [Thu, 4 Jul 2019 15:25:00 +0000 (16:25 +0100)]
Btrfs: wake up inode cache waiters sooner to reduce waiting time
If we need to start an inode caching thread, because none currently exists
on disk, we can wake up all waiters as soon as we mark the range starting
at root's highest objectid + 1 and ending at BTRFS_LAST_FREE_OBJECTID as
free, so that they don't need to wait for the caching thread to start and
do some progress. We follow the same approach within the caching thread,
since as soon as it finds a free range and marks it as free space in the
cache, it wakes up all waiters. So improve this by adding such a wakeup
call after marking that initial range as free space.
Fixes: a47d6b70e28040 ("Btrfs: setup free ino caching in a more asynchronous way") Reviewed-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Filipe Manana [Thu, 4 Jul 2019 15:24:44 +0000 (16:24 +0100)]
Btrfs: fix inode cache waiters hanging on path allocation failure
If the caching thread fails to allocate a path, it returns without waking
up any cache waiters, leaving them hang forever. Fix this by following the
same approach as when we fail to start the caching thread: print an error
message, disable inode caching and make the wakers fallback to non-caching
mode behaviour (calling btrfs_find_free_objectid()).
Fixes: 581bb050941b4f ("Btrfs: Cache free inode numbers in memory") Reviewed-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Filipe Manana [Thu, 4 Jul 2019 15:24:32 +0000 (16:24 +0100)]
Btrfs: fix inode cache waiters hanging on failure to start caching thread
If we fail to start the inode caching thread, we print an error message
and disable the inode cache, however we never wake up any waiters, so they
hang forever waiting for the caching to finish. Fix this by waking them
up and have them fallback to a call to btrfs_find_free_objectid().
Fixes: e60efa84252c05 ("Btrfs: avoid triggering bug_on() when we fail to start inode caching task") Reviewed-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Filipe Manana [Thu, 4 Jul 2019 15:24:19 +0000 (16:24 +0100)]
Btrfs: fix inode cache block reserve leak on failure to allocate data space
If we failed to allocate the data extent(s) for the inode space cache, we
were bailing out without releasing the previously reserved metadata. This
was triggering the following warnings when unmounting a filesystem:
Several fstests test cases triggered this often, such as generic/083,
generic/102, generic/172, generic/269 and generic/300 at least, producing
stack traces like the following in dmesg/syslog:
Fix this by releasing the reserved metadata on failure to allocate data
extent(s) for the inode cache.
Fixes: 69fe2d75dd91d0 ("btrfs: make the delalloc block rsv per inode") Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Filipe Manana [Thu, 4 Jul 2019 15:24:09 +0000 (16:24 +0100)]
Btrfs: fix hang when loading existing inode cache off disk
If we are able to load an existing inode cache off disk, we set the state
of the cache to BTRFS_CACHE_FINISHED, but we don't wake up any one waiting
for the cache to be available. This means that anyone waiting for the
cache to be available, waiting on the condition that either its state is
BTRFS_CACHE_FINISHED or its available free space is greather than zero,
can hang forever.
This could be observed running fstests with MOUNT_OPTIONS="-o inode_cache",
in particular test case generic/161 triggered it very frequently for me,
producing a trace like the following:
[63795.739712] BTRFS info (device sdc): enabling inode map caching
[63795.739714] BTRFS info (device sdc): disk space caching is enabled
[63795.739716] BTRFS info (device sdc): has skinny extents
[64036.653886] INFO: task btrfs-transacti:3917 blocked for more than 120 seconds.
[64036.654079] Not tainted 5.2.0-rc4-btrfs-next-50 #1
[64036.654143] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[64036.654232] btrfs-transacti D 0 3917 2 0x80004000
[64036.654239] Call Trace:
[64036.654258] ? __schedule+0x3ae/0x7b0
[64036.654271] schedule+0x3a/0xb0
[64036.654325] btrfs_commit_transaction+0x978/0xae0 [btrfs]
[64036.654339] ? remove_wait_queue+0x60/0x60
[64036.654395] transaction_kthread+0x146/0x180 [btrfs]
[64036.654450] ? btrfs_cleanup_transaction+0x620/0x620 [btrfs]
[64036.654456] kthread+0x103/0x140
[64036.654464] ? kthread_create_worker_on_cpu+0x70/0x70
[64036.654476] ret_from_fork+0x3a/0x50
[64036.654504] INFO: task xfs_io:3919 blocked for more than 120 seconds.
[64036.654568] Not tainted 5.2.0-rc4-btrfs-next-50 #1
[64036.654617] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[64036.654685] xfs_io D 0 3919 3633 0x00000000
[64036.654691] Call Trace:
[64036.654703] ? __schedule+0x3ae/0x7b0
[64036.654716] schedule+0x3a/0xb0
[64036.654756] btrfs_find_free_ino+0xa9/0x120 [btrfs]
[64036.654764] ? remove_wait_queue+0x60/0x60
[64036.654809] btrfs_create+0x72/0x1f0 [btrfs]
[64036.654822] lookup_open+0x6bc/0x790
[64036.654849] path_openat+0x3bc/0xc00
[64036.654854] ? __lock_acquire+0x331/0x1cb0
[64036.654869] do_filp_open+0x99/0x110
[64036.654884] ? __alloc_fd+0xee/0x200
[64036.654895] ? do_raw_spin_unlock+0x49/0xc0
[64036.654909] ? do_sys_open+0x132/0x220
[64036.654913] do_sys_open+0x132/0x220
[64036.654926] do_syscall_64+0x60/0x1d0
[64036.654933] entry_SYSCALL_64_after_hwframe+0x49/0xbe
Fix this by adding a wake_up() call right after setting the cache state to
BTRFS_CACHE_FINISHED, at start_caching(), when we are able to load the
cache from disk.
Fixes: 82d5902d9c681b ("Btrfs: Support reading/writing on disk free ino cache") Reviewed-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
This patch will introduce ROOT_ITEM check, which includes:
- Key->objectid and key->offset check
Currently only some easy check, e.g. 0 as rootid is invalid.
- Item size check
Root item size is fixed.
- Generation checks
Generation, generation_v2 and last_snapshot should not be greater than
super generation + 1
- Level and alignment check
Level should be in [0, 7], and bytenr must be aligned to sector size.
- Flags check
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=203261 Reported-by: Jungyeon Yoon <jungyeon.yoon@gmail.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
[CAUSE]
This situation is caused by several factors:
- Fuzzed image
The extent tree of this fs missed one backref for extent tree root.
So we can allocated space from that slot.
- MIXED_BG feature
Super block has MIXED_BG flag.
- No mixed block groups exists
All block groups are just regular ones.
This makes data space_info->block_groups[] contains metadata block
groups. And when we reserve space for data, we can use space in
metadata block group.
Then we hit the following file operations:
- fallocate
We need to allocate data extents.
find_free_extent() choose to use the metadata block to allocate space
from, and choose the space of extent tree root, since its backref is
missing.
This generate one delayed ref head with is_data = 1.
- extent tree update
We need to update extent tree at run_delayed_ref time.
This generate one delayed ref head with is_data = 0, for the same
bytenr of old extent tree root.
Then we trigger the BUG_ON().
[FIX]
The quick fix here is to check block_group->flags before using it.
The problem can only happen for MIXED_GROUPS fs. Regular filesystems
won't have space_info with DATA|METADATA flag, and no way to hit the
bug.
btrfs: delayed-inode: Kill the BUG_ON() in btrfs_delete_delayed_dir_index()
There is one report of fuzzed image which leads to BUG_ON() in
btrfs_delete_delayed_dir_index().
Although that fuzzed image can already be addressed by enhanced
extent-tree error handler, it's still better to hunt down more BUG_ON().
This patch will hunt down two BUG_ON()s in
btrfs_delete_delayed_dir_index():
- One for error from btrfs_delayed_item_reserve_metadata()
Instead of BUG_ON(), we output an error message and free the item.
And return the error.
All callers of this function handles the error by aborting current
trasaction.
- One for possible EEXIST from __btrfs_add_delayed_deletion_item()
That function can return -EEXIST.
We already have a good enough error message for that, only need to
clean up the reserved metadata space and allocated item.
To help above cleanup, also modifiy __btrfs_remove_delayed_item() called
in btrfs_release_delayed_item(), to skip unassociated item.
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=203253 Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
Test case btrfs/156 fails since commit 302167c50b32 ("btrfs: don't end
the transaction for delayed refs in throttle") with ENOSPC.
[CAUSE]
The ENOSPC is reported from btrfs_can_relocate().
This function will check:
- If this block group is empty, we can relocate
- If we can enough free space, we can relocate
Above checks are valid but the following check is vague due to its
implementation:
- If and only if we can allocated a new block group to contain all the
used space, we can relocate
This design itself is OK, but the way to determine if we can allocate a
new block group is problematic.
btrfs_can_relocate() uses find_free_dev_extent() to find free space on a
device.
However find_free_dev_extent() only searches commit root and excludes
dev extents allocated in current trans, this makes it unable to use dev
extent just freed in current transaction.
So for the following example, btrfs_can_relocate() will report ENOSPC:
The example block group layout:
1M 129M 257M 385M 513M 550M
|///////|///////////|//////////| | |
// = Used bg, consider all bg is 100% used for easy calculation.
And all block groups are SINGLE, on-disk bytenr is the same as the
logical bytenr.
1) Bg in [129M, 257M) get relocated to [385M, 513M), transid=100
1M 129M 257M 385M 513M 550M
|///////| |//////////|/////////|
In transid 100, bg in [129M, 257M) get relocated to [385M, 513M)
However transid 100 is not committed yet, so in dev commit tree, we
still have the old dev extents layout:
1M 129M 257M 385M 513M 550M
|///////|///////////|//////////| | |
2) Try to relocate bg [257M, 385M)
We goes into btrfs_can_relocate(), no free space in current bgs, so we
check if we can find large enough free dev extents.
The first slot is [385M, 513M), but that is already used by new bg at
[385M, 513M), so we continue search.
The remaining slot is [512M, 550M), smaller than the bg's length 128M.
So btrfs_can_relocate report ENOSPC.
However this is over killed, in fact if we just skip btrfs_can_relocate()
check, and go into regular relocation routine, at extent reservation time,
if we can't find free extent, then we fallback to commit transaction,
which will free up the dev extents and allow new block group to be created.
[FIX]
The fix here is to remove btrfs_can_relocate() completely.
If we hit the false ENOSPC case just like btrfs/156, extent allocator
will push harder by committing transaction and we will have space for
new block group, avoiding the false ENOSPC.
If we really ran out of space, we will hit ENOSPC at
relocate_block_group(), and btrfs will just reports the ENOSPC error as
usual.
Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
btrfs: extent-tree: Add comment for inc_block_group_ro()
inc_block_group_ro() is only designed to mark one block group read-only,
it doesn't really care if other block groups have enough free space to
contain the used space in the block group.
However due to the close connection between this function and
relocation, sometimes we can be confused and think this function is
responsible for balance space reservation, which is not true.
Add some comment to make the functionality clear.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
btrfs: volumes: Add comment for find_free_dev_extent_start()
Since commit 6df9a95e6339 ("Btrfs: make the chunk allocator completely
tree lockless") we search commit root of device tree to avoid deadlock.
This introduced a safety feature, find_free_dev_extent_start() won't
use dev extents which just get freed in current transaction.
This safety feature makes sure we won't allocate new block group using
just freed dev extents to break CoW.
However, this feature also makes find_free_dev_extent_start() not
reliable reporting free device space. Just add such comment to make
later viewer careful about this behavior.
This behavior makes one caller, btrfs_can_relocate() unreliable
determining the device free space.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
David Sterba [Wed, 27 Mar 2019 15:19:55 +0000 (16:19 +0100)]
btrfs: assert extent map tree lock in add_extent_mapping
As add_extent_mapping is called from several functions, let's add the
lock annotation. The tree is going to be modified so it must be the
exclusive lock.
Reviewed-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Jia-Ju Bai [Sat, 27 Jul 2019 08:51:13 +0000 (16:51 +0800)]
btrfs: Add an assertion to warn incorrect case in insert_inline_extent()
In insert_inline_extent(), the case that checks compressed_size > 0
and compressed_pages = NULL cannot occur, otherwise a null-pointer
dereference may occur on line 215:
cpage = compressed_pages[i];
To catch this incorrect case, an assertion is added.
Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Nikolay Borisov [Wed, 17 Jul 2019 13:18:17 +0000 (16:18 +0300)]
btrfs: Remove leftover of in-band dedupe
It's unlikely in-band dedupe is going to land so just remove any
leftovers - dedupe.h header as well as the 'dedupe' parameter to
btrfs_set_extent_delalloc.
Signed-off-by: Nikolay Borisov <nborisov@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Nikolay Borisov [Wed, 17 Jul 2019 13:18:16 +0000 (16:18 +0300)]
btrfs: Remove delalloc_end argument from extent_clear_unlock_delalloc
It was added in ba8b04c1d4ad ("btrfs: extend btrfs_set_extent_delalloc
and its friends to support in-band dedupe and subpage size patchset") as
a preparatory patch for in-band and subapge block size patchsets.
However neither of those are likely to be merged anytime soon and the
code has diverged significantly from the last public post of either
of those patchsets.
It's unlikely either of the patchests are going to use those preparatory
steps so just remove the variables. Since cow_file_range also took
delalloc_end to pass it to extent_clear_unlock_delalloc remove the
parameter from that function as well.
Signed-off-by: Nikolay Borisov <nborisov@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Nikolay Borisov [Wed, 17 Jul 2019 11:41:45 +0000 (14:41 +0300)]
btrfs: Move free_pages_out label in inline extent handling branch in compress_file_range
This label is only executed if compress_file_range fails to create an
inline extent. So move its code in the semantically related inline
extent handling branch. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Nikolay Borisov [Wed, 17 Jul 2019 11:41:44 +0000 (14:41 +0300)]
btrfs: Return number of compressed extents directly in compress_file_range
compress_file_range returns a void, yet uses a function parameter as a
return value. Make that more idiomatic by simply returning the number
of compressed extents directly. Also track such extents in more aptly
named variables. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Nikolay Borisov [Thu, 25 Jul 2019 08:27:28 +0000 (11:27 +0300)]
btrfs: Remove unused locking functions
Those were split out of btrfs_clear_lock_blocking_rw by aa12c02778a9 ("btrfs: split btrfs_clear_lock_blocking_rw to read and write helpers")
however at that time this function was unused due to commit 523983401644 ("Btrfs: kill btrfs_clear_path_blocking"). Put the final
nail in the coffin of those 2 functions.
Signed-off-by: Nikolay Borisov <nborisov@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
btrfs: reduce stack usage for btrfsic_process_written_block
btrfsic_process_written_block() cals btrfsic_process_metablock(),
which has a fairly large stack usage due to the btrfsic_stack_frame
variable. It also calls btrfsic_test_for_metadata(), which now
needs several hundreds of bytes for its SHASH_DESC_ON_STACK().
In some configurations, we end up with both functions on the
same stack, and gcc warns about the excessive stack usage that
might cause the available stack space to run out:
fs/btrfs/check-integrity.c:1743:13: error: stack frame size of 1152 bytes in function 'btrfsic_process_written_block' [-Werror,-Wframe-larger-than=]
Marking both child functions as noinline_for_stack helps because
this guarantees that the large variables are not on the same
stack frame.
Fixes: d5178578bcd4 ("btrfs: directly call into crypto framework for checksumming") Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: David Sterba <dsterba@suse.com>
fs/btrfs/volumes.c: In function __btrfs_map_block:
fs/btrfs/volumes.c:6023:6: warning:
variable offset set but not used [-Wunused-but-set-variable]
It is not used any more since commit 343abd1c0ca9 ("btrfs: Use
btrfs_get_io_geometry appropriately")
Reported-by: Hulk Robot <hulkci@huawei.com> Signed-off-by: YueHaibing <yuehaibing@huawei.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Filipe Manana [Fri, 5 Jul 2019 10:09:50 +0000 (11:09 +0100)]
Btrfs: fix ENOSPC errors, leading to transaction aborts, when cloning extents
When cloning extents (or deduplicating) we create a transaction with a
space reservation that considers we will drop or update a single file
extent item of the destination inode (that we modify a single leaf). That
is fine for the vast majority of scenarios, however it might happen that
we need to drop many file extent items, and adjust at most two file extent
items, in the destination root, which can span multiple leafs. This will
lead to either the call to btrfs_drop_extents() to fail with ENOSPC or
the subsequent calls to btrfs_insert_empty_item() or btrfs_update_inode()
(called through clone_finish_inode_update()) to fail with ENOSPC. Such
failure results in a transaction abort, leaving the filesystem in a
read-only mode.
In order to fix this we need to follow the same approach as the hole
punching code, where we create a local reservation with 1 unit and keep
ending and starting transactions, after balancing the btree inode,
when __btrfs_drop_extents() returns ENOSPC. So fix this by making the
extent cloning call calls the recently added btrfs_punch_hole_range()
helper, which is what does the mentioned work for hole punching, and
make sure whenever we drop extent items in a transaction, we also add a
replacing file extent item, to avoid corruption (a hole) if after ending
a transaction and before starting a new one, the old transaction gets
committed and a power failure happens before we finish cloning.
Filipe Manana [Fri, 28 Jun 2019 22:11:26 +0000 (23:11 +0100)]
Btrfs: factor out extent dropping code from hole punch handler
Move the code that is responsible for dropping extents in a range out of
btrfs_punch_hole() into a new helper function, btrfs_punch_hole_range(),
so that later it can be used by the reflinking (extent cloning and dedup)
code to fix a ENOSPC bug.
Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
Merge tag 'gpio-v5.3-5' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio
Pull GPIO fixes from Linus Walleij:
"All related to the PCA953x driver when handling chips with more than 8
ports, now that works again"
* tag 'gpio-v5.3-5' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio:
gpio: pca953x: use pca953x_read_regs instead of regmap_bulk_read
gpio: pca953x: correct type of reg_direction
Nick Desaulniers [Wed, 28 Aug 2019 22:55:23 +0000 (15:55 -0700)]
include/linux/compiler.h: fix Oops for Clang-compiled kernels
GCC unescapes escaped string section names while Clang does not. Because
__section uses the `#` stringification operator for the section name, it
doesn't need to be escaped.
This fixes an Oops observed in distro's that use systemd and not
net.core.bpf_jit_enable=1, when their kernels are compiled with Clang.
Chris Wilson reports that it breaks his CPU hotplug test scripts. In
particular, it breaks offlining and then re-onlining the boot CPU, which
we treat specially (and the BIOS does too).
The symptoms are that we can offline the CPU, but it then does not come
back online again:
smpboot: CPU 0 is now offline
smpboot: Booting Node 0 Processor 0 APIC 0x0
smpboot: do_boot_cpu failed(-1) to wakeup CPU#0
Thomas says he knows why it's broken (my personal suspicion: our magic
handling of the "cpu0_logical_apicid" thing), but for 5.3 the right fix
is to just revert it, since we've never touched the LDR bits before, and
it's not worth the risk to do anything else at this stage.
[ Hotpluging of the boot CPU is special anyway, and should be off by
default. See the "BOOTPARAM_HOTPLUG_CPU0" config option and the
cpu0_hotplug kernel parameter.
In general you should not do it, and it has various known limitations
(hibernate and suspend require the boot CPU, for example).
But it should work, even if the boot CPU is special and needs careful
treatment - Linus ]
Merge tag 'char-misc-5.3-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc
Pull Documentation updates from Greg KH:
"A few small patches for the documenation file that came in through the
char-misc tree in -rc7 for your tree.
They fix the mistake in the .rst format that kept the table of
companies from showing up in the html output, and most importantly,
add people's names to the list showing support for our process"
* tag 'char-misc-5.3-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
Documentation/process: Add Qualcomm process ambassador for hardware security issues
Documentation/process/embargoed-hardware-issues: Microsoft ambassador
Documentation/process: Add Google contact for embargoed hardware issues
Documentation/process: Volunteer as the ambassador for Xen
Merge tag 'libnvdimm-fix-5.3-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm
Pull libnvdimm fix from Dan Williams:
"Restore support for 1GB alignment namespaces, truncate the end of
misaligned namespaces"
* tag 'libnvdimm-fix-5.3-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:
libnvdimm/pfn: Fix namespace creation on misaligned addresses
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input
Pull input fix from Dmitry Torokhov:
"A tiny update from Benjamin removing a mistakenly added Elan PNP ID so
that the device is again handled by hid-multitouch"
Merge tag 'configfs-for-5.3' of git://git.infradead.org/users/hch/configfs
Pull configfs fixes from Christoph Hellwig:
"Late configfs fixes from Al that fix pretty nasty removal vs attribute
access races"
* tag 'configfs-for-5.3' of git://git.infradead.org/users/hch/configfs:
configfs: provide exclusion between IO and removals
configfs: new object reprsenting tree fragments
configfs_register_group() shouldn't be (and isn't) called in rmdirable parts
configfs: stash the data we need into configfs_buffer at open time
Merge tag 'iommu-fixes-v5.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu
Pull IOMMU fixes from Joerg Roedel:
- Revert an Intel VT-d patch that caused problems for some users.
- Removal of a feature in the Intel VT-d driver that was never
supported in hardware. This qualifies as a fix because the code for
this feature sets reserved bits in the invalidation queue descriptor,
causing failed invalidations on real hardware.
- Two fixes for AMD IOMMU driver to fix a race condition and to add a
missing IOTLB flush when kernel is booted in kdump mode.
* tag 'iommu-fixes-v5.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:
iommu/amd: Fix race in increase_address_space()
iommu/amd: Flush old domains in kdump kernel
iommu/vt-d: Remove global page flush support
Revert "iommu/vt-d: Avoid duplicated pci dma alias consideration"
Merge tag 'drm-fixes-2019-09-06' of git://anongit.freedesktop.org/drm/drm
Pull drm fixes from Dave Airlie:
"Live from my friend's couch in Barcelona, latest round of drm fixes.
The command line parser regression fixes look a bit larger because
they come with selftests included for the bugs they fix. Otherwise a
single nouveau, single ingenic and single vmwgfx fix:
* tag 'drm-fixes-2019-09-06' of git://anongit.freedesktop.org/drm/drm:
drm/vmwgfx: Fix double free in vmw_recv_msg()
drm/nouveau/sec2/gp102: add missing MODULE_FIRMWAREs
drm/selftests: modes: Add more unit tests for the cmdline parser
drm/modes: Introduce a whitelist for the named modes
drm/modes: Fix the command line parser to take force options into account
drm/modes: Add a switch to differentiate free standing options
drm/ingenic: Hardcode panel type to DPI
Merge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost
Pull virtio fixes from Michael Tsirkin:
"virtio, vhost, and balloon bugfixes.
A couple of last minute bugfixes. And a revert of a failed attempt at
metadata access optimization - we'll try again in the next cycle"
* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:
mm/balloon_compaction: suppress allocation warnings
Revert "vhost: access vq metadata through kernel virtual address"
vhost: Remove unnecessary variable
virtio-net: lower min ring num_free for efficiency
vhost/test: fix build for vhost test
vhost/test: fix build for vhost test
Merge tag 'powerpc-5.3-5' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux
Pull powerpc fixes from Michael Ellerman:
"One fix for a boot hang on some Freescale machines when PREEMPT is
enabled.
Two CVE fixes for bugs in our handling of FP registers and
transactional memory, both of which can result in corrupted FP state,
or FP state leaking between processes.
Thanks to: Chris Packham, Christophe Leroy, Gustavo Romero, Michael
Neuling"
* tag 'powerpc-5.3-5' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/tm: Fix restoring FP/VMX facility incorrectly on interrupts
powerpc/tm: Fix FP/VMX unavailable exceptions inside a transaction
powerpc/64e: Drop stale call to smp_processor_id() which hangs SMP startup
Lee Jones [Thu, 5 Sep 2019 08:25:55 +0000 (09:25 +0100)]
soc: qcom: geni: Provide parameter error checking
When booting with ACPI, the Geni Serial Engine is not set as the I2C/SPI
parent and thus, the wrapper (parent device) is unassigned. This causes
the kernel to crash with a null dereference error.
Link: https://lore.kernel.org/r/20190905082555.15020-1-lee.jones@linaro.org Fixes: 8bc529b25354 ("soc: qcom: geni: Add support for ACPI") Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org> Reviewed-by: Stephen Boyd <sboyd@kernel.org> Signed-off-by: Lee Jones <lee.jones@linaro.org> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
After the conversion to lock-less dma-api call the
increase_address_space() function can be called without any
locking. Multiple CPUs could potentially race for increasing
the address space, leading to invalid domain->mode settings
and invalid page-tables. This has been happening in the wild
under high IO load and memory pressure.
Fix the race by locking this operation. The function is
called infrequently so that this does not introduce
a performance regression in the dma-api path again.
Reported-by: Qian Cai <cai@lca.pw> Fixes: 256e4621c21a ('iommu/amd: Make use of the generic IOVA allocator') Signed-off-by: Joerg Roedel <jroedel@suse.de>
Stuart Hayes [Thu, 5 Sep 2019 17:09:48 +0000 (12:09 -0500)]
iommu/amd: Flush old domains in kdump kernel
When devices are attached to the amd_iommu in a kdump kernel, the old device
table entries (DTEs), which were copied from the crashed kernel, will be
overwritten with a new domain number. When the new DTE is written, the IOMMU
is told to flush the DTE from its internal cache--but it is not told to flush
the translation cache entries for the old domain number.
Without this patch, AMD systems using the tg3 network driver fail when kdump
tries to save the vmcore to a network system, showing network timeouts and
(sometimes) IOMMU errors in the kernel log.
This patch will flush IOMMU translation cache entries for the old domain when
a DTE gets overwritten with a new domain number.
Signed-off-by: Stuart Hayes <stuart.w.hayes@gmail.com> Fixes: 3ac3e5ee5ed5 ('iommu/amd: Copy old trans table from old kernel') Signed-off-by: Joerg Roedel <jroedel@suse.de>
keys: Fix missing null pointer check in request_key_auth_describe()
If a request_key authentication token key gets revoked, there's a window in
which request_key_auth_describe() can see it with a NULL payload - but it
makes no check for this and something like the following oops may occur:
Fix this by checking for a NULL pointer when describing such a key.
Also make the read routine check for a NULL pointer to be on the safe side.
[DH: Modified to not take already-held rcu lock and modified to also check
in the read routine]
Fixes: 04c567d9313e ("[PATCH] Keys: Fix race between two instantiators of a key") Reported-by: Sachin Sant <sachinp@linux.vnet.ibm.com> Signed-off-by: Hillf Danton <hdanton@sina.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Sachin Sant <sachinp@linux.vnet.ibm.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge tag 'sound-5.3-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Pull sound fixes from Takashi Iwai:
"A collection of small HD-audio fixes:
- A regression fix for Realtek codecs due to the recent
initialization procedure change
- A fix for potential endless loop at the quirk table lookup
- Quirks for Lenovo, ASUS and HP machines"
* tag 'sound-5.3-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
ALSA: hda/realtek - Fix the problem of two front mics on a ThinkCentre
ALSA: hda/realtek - Enable internal speaker & headset mic of ASUS UX431FL
ALSA: hda/realtek - Add quirk for HP Pavilion 15
ALSA: hda/realtek - Fix overridden device-specific initialization
ALSA: hda - Fix potential endless loop at applying quirks
Merge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Ingo Molnar:
"Misc fixes:
- EFI boot fix for signed kernels
- an AC flags fix related to UBSAN
- Hyper-V infinite loop fix"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/hyper-v: Fix overflow bug in fill_gva_list()
x86/uaccess: Don't leak the AC flags into __get_user() argument evaluation
x86/boot: Preserve boot_params.secure_boot from sanitizing
Merge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler fixes from Ingo Molnar:
"This fixes an ABI bug introduced this cycle, plus fixes a throttling
bug"
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/core: Fix uclamp ABI bug, clean up and robustify sched_read_attr() ABI logic and code
sched/fair: Don't assign runtime for throttled cfs_rq
Dan Carpenter [Thu, 15 Aug 2019 08:30:50 +0000 (11:30 +0300)]
drm/vmwgfx: Fix double free in vmw_recv_msg()
We recently added a kfree() after the end of the loop:
if (retries == RETRIES) {
kfree(reply);
return -EINVAL;
}
There are two problems. First the test is wrong and because retries
equals RETRIES if we succeed on the last iteration through the loop.
Second if we fail on the last iteration through the loop then the kfree
is a double free.
When you're reading this code, please note the break statement at the
end of the while loop. This patch changes the loop so that if it's not
successful then "reply" is NULL and we can test for that afterward.
Cc: <stable@vger.kernel.org> Fixes: 6b7c3b86f0b6 ("drm/vmwgfx: fix memory leak when too many retries have occurred") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Al Viro [Sat, 31 Aug 2019 07:43:43 +0000 (09:43 +0200)]
configfs: provide exclusion between IO and removals
Make sure that attribute methods are not called after the item
has been removed from the tree. To do so, we
* at the point of no return in removals, grab ->frag_sem
exclusive and mark the fragment dead.
* call the methods of attributes with ->frag_sem taken
shared and only after having verified that the fragment is still
alive.
The main benefit is for method instances - they are
guaranteed that the objects they are accessing *and* all ancestors
are still there. Another win is that we don't need to bother
with extra refcount on config_item when opening a file -
the item will be alive for as long as it stays in the tree, and
we won't touch it/attributes/any associated data after it's
been removed from the tree.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Christoph Hellwig <hch@lst.de>
sched/core: Fix uclamp ABI bug, clean up and robustify sched_read_attr() ABI logic and code
Thadeu Lima de Souza Cascardo reported that 'chrt' broke on recent kernels:
$ chrt -p $$
chrt: failed to get pid 26306's policy: Argument list too long
and he has root-caused the bug to the following commit increasing sched_attr
size and breaking sched_read_attr() into returning -EFBIG:
a509a7cd7974 ("sched/uclamp: Extend sched_setattr() to support utilization clamping")
The other, bigger bug is that the whole sched_getattr() and sched_read_attr()
logic of checking non-zero bits in new ABI components is arguably broken,
and pretty much any extension of the ABI will spuriously break the ABI.
That's way too fragile.
Instead implement the perf syscall's extensible ABI instead, which we
already implement on the sched_setattr() side:
- if user-attributes have the same size as kernel attributes then the
logic is unchanged.
- if user-attributes are larger than the kernel knows about then simply
skip the extra bits, but set attr->size to the (smaller) kernel size
so that tooling can (in principle) handle older kernel as well.
- if user-attributes are smaller than the kernel knows about then just
copy whatever user-space can accept.
Also clean up the whole logic:
- Simplify the code flow - there's no need for 'ret' for example.
- Standardize on 'kattr/uattr' and 'ksize/usize' naming to make sure we
always know which side we are dealing with.
- Why is it called 'read' when what it does is to copy to user? This
code is so far away from VFS read() semantics that the naming is
actively confusing. Name it sched_attr_copy_to_user() instead, which
mirrors other copy_to_user() functionality.
- Move the attr->size assignment from the head of sched_getattr() to the
sched_attr_copy_to_user() function. Nothing else within the kernel
should care about the size of the structure.
With these fixes the sched_getattr() syscall now nicely supports an
extensible ABI in both a forward and backward compatible fashion, and
will also fix the chrt bug.
As an added bonus the bogus -EFBIG return is removed as well, which as
Thadeu noted should have been -E2BIG to begin with.
Reported-by: Thadeu Lima de Souza Cascardo <cascardo@canonical.com> Tested-by: Dietmar Eggemann <dietmar.eggemann@arm.com> Tested-by: Thadeu Lima de Souza Cascardo <cascardo@canonical.com> Acked-by: Thadeu Lima de Souza Cascardo <cascardo@canonical.com> Cc: Arnaldo Carvalho de Melo <acme@infradead.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Patrick Bellasi <patrick.bellasi@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: a509a7cd7974 ("sched/uclamp: Extend sched_setattr() to support utilization clamping") Link: https://lkml.kernel.org/r/20190904075532.GA26751@gmail.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
Merge tag 'renesas-fixes-for-v5.3' of git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas into arm/fixes
Renesas ARM Based SoC Fixes for v5.3
* R-Car D3 (r8a77995) based Draak Board
- Correct backlight regulator name in device tree
* tag 'renesas-fixes-for-v5.3' of git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas:
arm64: dts: renesas: r8a77995: draak: Fix backlight regulator name
powerpc/tm: Fix restoring FP/VMX facility incorrectly on interrupts
When in userspace and MSR FP=0 the hardware FP state is unrelated to
the current process. This is extended for transactions where if tbegin
is run with FP=0, the hardware checkpoint FP state will also be
unrelated to the current process. Due to this, we need to ensure this
hardware checkpoint is updated with the correct state before we enable
FP for this process.
Unfortunately we get this wrong when returning to a process from a
hardware interrupt. A process that starts a transaction with FP=0 can
take an interrupt. When the kernel returns back to that process, we
change to FP=1 but with hardware checkpoint FP state not updated. If
this transaction is then rolled back, the FP registers now contain the
wrong state.
The process looks like this:
Userspace: Kernel
Start userspace
with MSR FP=0 TM=1
< -----
...
tbegin
bne
Hardware interrupt
---- >
<do_IRQ...>
....
ret_from_except
restore_math()
/* sees FP=0 */
restore_fp()
tm_active_with_fp()
/* sees FP=1 (Incorrect) */
load_fp_state()
FP = 0 -> 1
< -----
Return to userspace
with MSR TM=1 FP=1
with junk in the FP TM checkpoint
TM rollback
reads FP junk
When returning from the hardware exception, tm_active_with_fp() is
incorrectly making restore_fp() call load_fp_state() which is setting
FP=1.
The fix is to remove tm_active_with_fp().
tm_active_with_fp() is attempting to handle the case where FP state
has been changed inside a transaction. In this case the checkpointed
and transactional FP state is different and hence we must restore the
FP state (ie. we can't do lazy FP restore inside a transaction that's
used FP). It's safe to remove tm_active_with_fp() as this case is
handled by restore_tm_state(). restore_tm_state() detects if FP has
been using inside a transaction and will set load_fp and call
restore_math() to ensure the FP state (checkpoint and transaction) is
restored.
This is a data integrity problem for the current process as the FP
registers are corrupted. It's also a security problem as the FP
registers from one process may be leaked to another.
Similarly for VMX.
A simple testcase to replicate this will be posted to
tools/testing/selftests/powerpc/tm/tm-poison.c
This fixes CVE-2019-15031.
Fixes: a7771176b439 ("powerpc: Don't enable FP/Altivec if not checkpointed") Cc: stable@vger.kernel.org # 4.15+ Signed-off-by: Gustavo Romero <gromero@linux.ibm.com> Signed-off-by: Michael Neuling <mikey@neuling.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20190904045529.23002-2-gromero@linux.vnet.ibm.com
powerpc/tm: Fix FP/VMX unavailable exceptions inside a transaction
When we take an FP unavailable exception in a transaction we have to
account for the hardware FP TM checkpointed registers being
incorrect. In this case for this process we know the current and
checkpointed FP registers must be the same (since FP wasn't used
inside the transaction) hence in the thread_struct we copy the current
FP registers to the checkpointed ones.
This copy is done in tm_reclaim_thread(). We use thread->ckpt_regs.msr
to determine if FP was on when in userspace. thread->ckpt_regs.msr
represents the state of the MSR when exiting userspace. This is setup
by check_if_tm_restore_required().
Unfortunatley there is an optimisation in giveup_all() which returns
early if tsk->thread.regs->msr (via local variable `usermsr`) has
FP=VEC=VSX=SPE=0. This optimisation means that
check_if_tm_restore_required() is not called and hence
thread->ckpt_regs.msr is not updated and will contain an old value.
This can happen if due to load_fp=255 we start a userspace process
with MSR FP=1 and then we are context switched out. In this case
thread->ckpt_regs.msr will contain FP=1. If that same process is then
context switched in and load_fp overflows, MSR will have FP=0. If that
process now enters a transaction and does an FP instruction, the FP
unavailable will not update thread->ckpt_regs.msr (the bug) and MSR
FP=1 will be retained in thread->ckpt_regs.msr. tm_reclaim_thread()
will then not perform the required memcpy and the checkpointed FP regs
in the thread struct will contain the wrong values.
The code path for this happening is:
Userspace: Kernel
Start userspace
with MSR FP/VEC/VSX/SPE=0 TM=1
< -----
...
tbegin
bne
fp instruction
FP unavailable
---- >
fp_unavailable_tm()
tm_reclaim_current()
tm_reclaim_thread()
giveup_all()
return early since FP/VMX/VSX=0
/* ckpt MSR not updated (Incorrect) */
tm_reclaim()
/* thread_struct ckpt FP regs contain junk (OK) */
/* Sees ckpt MSR FP=1 (Incorrect) */
no memcpy() performed
/* thread_struct ckpt FP regs not fixed (Incorrect) */
tm_recheckpoint()
/* Put junk in hardware checkpoint FP regs */
....
< -----
Return to userspace
with MSR TM=1 FP=1
with junk in the FP TM checkpoint
TM rollback
reads FP junk
This is a data integrity problem for the current process as the FP
registers are corrupted. It's also a security problem as the FP
registers from one process may be leaked to another.
This patch moves up check_if_tm_restore_required() in giveup_all() to
ensure thread->ckpt_regs.msr is updated correctly.
A simple testcase to replicate this will be posted to
tools/testing/selftests/powerpc/tm/tm-poison.c
Similarly for VMX.
This fixes CVE-2019-15030.
Fixes: f48e91e87e67 ("powerpc/tm: Fix FP and VMX register corruption") Cc: stable@vger.kernel.org # 4.12+ Signed-off-by: Gustavo Romero <gromero@linux.vnet.ibm.com> Signed-off-by: Michael Neuling <mikey@neuling.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20190904045529.23002-1-gromero@linux.vnet.ibm.com
There is no reason to print warnings when balloon page allocation fails,
as they are expected and can be handled gracefully. Since VMware
balloon now uses balloon-compaction infrastructure, and suppressed these
warnings before, it is also beneficial to suppress these warnings to
keep the same behavior that the balloon had before.
Cc: Jason Wang <jasowang@redhat.com> Signed-off-by: Nadav Amit <namit@vmware.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: David Hildenbrand <david@redhat.com>
Revert "vhost: access vq metadata through kernel virtual address"
This reverts commit 7f466032dc ("vhost: access vq metadata through
kernel virtual address"). The commit caused a bunch of issues, and
while commit 73f628ec9e ("vhost: disable metadata prefetch
optimization") disabled the optimization it's not nice to keep lots of
dead code around.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
? jiang [Tue, 20 Aug 2019 02:51:23 +0000 (02:51 +0000)]
virtio-net: lower min ring num_free for efficiency
This change lowers ring buffer reclaim threshold from 1/2*queue to budget
for better performance. According to our test with qemu + dpdk, packet
dropping happens when the guest is not able to provide free buffer in
avail ring timely with default 1/2*queue. The value in the patch has been
tested and does show better performance.
Test setup: iperf3 to generate packets to guest (total 30mins, pps 400k, UDP)
avg packets drop before: 2842
avg packets drop after: 360(-87.3%)
Further, current code suffers from a starvation problem: the amount of
work done by try_fill_recv is not bounded by the budget parameter, thus
(with large queues) once in a while userspace gets blocked for a long
time while queue is being refilled. Trigger refills earlier to make sure
the amount of work to do is limited.
Signed-off-by: jiangkidd <jiangkidd@hotmail.com> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Tiwei Bie [Wed, 28 Aug 2019 05:37:00 +0000 (13:37 +0800)]
vhost/test: fix build for vhost test
Since vhost_exceeds_weight() was introduced, callers need to specify
the packet weight and byte weight in vhost_dev_init(). Note that, the
packet weight isn't counted in this patch to keep the original behavior
unchanged.
Fixes: e82b9b0727ff ("vhost: introduce vhost_exceeds_weight()") Cc: stable@vger.kernel.org Signed-off-by: Tiwei Bie <tiwei.bie@intel.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Jason Wang <jasowang@redhat.com>
Tiwei Bie [Wed, 28 Aug 2019 05:36:59 +0000 (13:36 +0800)]
vhost/test: fix build for vhost test
Since below commit, callers need to specify the iov_limit in
vhost_dev_init() explicitly.
Fixes: b46a0bf78ad7 ("vhost: fix OOB in get_rx_bufs()") Cc: stable@vger.kernel.org Signed-off-by: Tiwei Bie <tiwei.bie@intel.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Jason Wang <jasowang@redhat.com>
Hui Wang [Wed, 4 Sep 2019 05:53:27 +0000 (13:53 +0800)]
ALSA: hda/realtek - Fix the problem of two front mics on a ThinkCentre
This ThinkCentre machine has a new realtek codec alc222, it is not
in the support list, we add it in the realtek.c then this machine
can apply FIXUPs for the realtek codec.
And this machine has two front mics which can't be handled
by PA so far, it uses the pin 0x18 and 0x19 as the front mics, as
a result the existing FIXUP ALC294_FIXUP_LENOVO_MIC_LOCATION doesn't
work on this machine. Fortunately another FIXUP
ALC283_FIXUP_HEADSET_MIC also can change the location for one of the
two mics on this machine.
dmaengine: rcar-dmac: Fix DMACHCLR handling if iommu is mapped
The commit 20c169aceb45 ("dmaengine: rcar-dmac: clear pertinence
number of channels") forgets to clear the last channel by
DMACHCLR in rcar_dmac_init() (and doesn't need to clear the first
channel) if iommu is mapped to the device. So, this patch fixes it
by using "channels_mask" bitfield.
Note that the hardware and driver don't support more than 32 bits
in DMACHCLR register anyway, so this patch should reject more than
32 channels in rcar_dmac_parse_of().
Fixes: 20c169aceb459575 ("dmaengine: rcar-dmac: clear pertinence number of channels") Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Reviewed-by: Simon Horman <horms+renesas@verge.net.au> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Link: https://lore.kernel.org/r/1567424643-26629-1-git-send-email-yoshihiro.shimoda.uh@renesas.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
Baolin Wang [Fri, 30 Aug 2019 07:37:45 +0000 (15:37 +0800)]
dmaengine: sprd: Fix the DMA link-list configuration
For the Spreadtrum DMA link-list mode, when the DMA engine got a slave
hardware request, which will trigger the DMA engine to load the DMA
configuration from the link-list memory automatically. But before the
slave hardware request, the slave will get an incorrect residue due
to the first node used to trigger the link-list was configured as the
last source address and destination address.
Thus we should make sure the first node was configured the start source
address and destination address, which can fix this issue.
Jan Kaisrlik [Tue, 20 Aug 2019 11:42:29 +0000 (13:42 +0200)]
Revert "mmc: core: do not retry CMD6 in __mmc_switch()"
Turns out the commit 3a0681c7448b ("mmc: core: do not retry CMD6 in
__mmc_switch()") breaks initialization of a Toshiba THGBMNG5 eMMC card,
when using the meson-gx-mmc.c driver on a custom board based on Amlogic
A113D.
The CMD6 that switches the card into HS200 mode is then one that fails and
according to the below printed messages from the log:
After some analyze, it turns out that adding a delay of ~5ms inside
mmc_select_bus_width() but after mmc_compare_ext_csds() has been executed,
also fixes the problem. Adding yet some more debug code, trying to figure
out if potentially the card could be in a busy state, both by using CMD13
and ->card_busy() ops concluded that this was not the case.
Therefore, let's simply revert the commit that dropped support for retrying
of CMD6, as this also fixes the problem.
Fixes: 3a0681c7448b ("mmc: core: do not retry CMD6 in __mmc_switch()") Cc: stable@vger.kernel.org Signed-off-by: Jan Kaisrlik <ja.kaisrlik@gmail.com> Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Jacob Pan [Mon, 26 Aug 2019 15:53:29 +0000 (08:53 -0700)]
iommu/vt-d: Remove global page flush support
Global pages support is removed from VT-d spec 3.0. Since global pages G
flag only affects first-level paging structures and because DMA request
with PASID are only supported by VT-d spec. 3.0 and onward, we can
safely remove global pages support.
For kernel shared virtual address IOTLB invalidation, PASID
granularity and page selective within PASID will be used. There is
no global granularity supported. Without this fix, IOTLB invalidation
will cause invalid descriptor error in the queued invalidation (QI)
interface.
Fixes: 1c4f88b7f1f9 ("iommu/vt-d: Shared virtual address in scalable mode") Reported-by: Sanjay K Kumar <sanjay.k.kumar@intel.com> Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com> Signed-off-by: Joerg Roedel <jroedel@suse.de>
Liangyan [Mon, 26 Aug 2019 12:16:33 +0000 (20:16 +0800)]
sched/fair: Don't assign runtime for throttled cfs_rq
do_sched_cfs_period_timer() will refill cfs_b runtime and call
distribute_cfs_runtime to unthrottle cfs_rq, sometimes cfs_b->runtime
will allocate all quota to one cfs_rq incorrectly, then other cfs_rqs
attached to this cfs_b can't get runtime and will be throttled.
We find that one throttled cfs_rq has non-negative
cfs_rq->runtime_remaining and cause an unexpetced cast from s64 to u64
in snippet:
The runtime here will change to a large number and consume all
cfs_b->runtime in this cfs_b period.
According to Ben Segall, the throttled cfs_rq can have
account_cfs_rq_runtime called on it because it is throttled before
idle_balance, and the idle_balance calls update_rq_clock to add time
that is accounted to the task.
This commit prevents cfs_rq to be assgined new runtime if it has been
throttled until that distribute_cfs_runtime is called.
Signed-off-by: Liangyan <liangyan.peng@linux.alibaba.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Valentin Schneider <valentin.schneider@arm.com> Reviewed-by: Ben Segall <bsegall@google.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: shanpeic@linux.alibaba.com Cc: stable@vger.kernel.org Cc: xlpang@linux.alibaba.com Fixes: d3d9dc330236 ("sched: Throttle entities exceeding their allowed bandwidth") Link: https://lkml.kernel.org/r/20190826121633.6538-1-liangyan.peng@linux.alibaba.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
1. Has duplicated internal speakers (0x14 & 0x17) which makes the output
route become confused. So, the output volume cannot be changed by
setting.
2. Misses the headset mic pin node.
This patch disables the confusing speaker (NID 0x14) and enables the
headset mic (NID 0x19).