Michal Hocko
7e7844226f
lockdep: allow to disable reclaim lockup detection ...
The current implementation of the reclaim lockup detection can lead to
false positives and those even happen and usually lead to tweak the code
to silence the lockdep by using GFP_NOFS even though the context can use
__GFP_FS just fine.
See
http://lkml.kernel.org/r/20160512080321.GA18496@dastard
as an example.
=================================
[ INFO: inconsistent lock state ]
4.5.0-rc2+ #4 Tainted: G O
---------------------------------
inconsistent {RECLAIM_FS-ON-R} -> {IN-RECLAIM_FS-W} usage.
kswapd0/543 [HC0[0]:SC0[0]:HE1:SE1] takes:
(&xfs_nondir_ilock_class){++++-+}, at: xfs_ilock+0x177/0x200 [xfs]
{RECLAIM_FS-ON-R} state was registered at:
mark_held_locks+0x79/0xa0
lockdep_trace_alloc+0xb3/0x100
kmem_cache_alloc+0x33/0x230
kmem_zone_alloc+0x81/0x120 [xfs]
xfs_refcountbt_init_cursor+0x3e/0xa0 [xfs]
__xfs_refcount_find_shared+0x75/0x580 [xfs]
xfs_refcount_find_shared+0x84/0xb0 [xfs]
xfs_getbmap+0x608/0x8c0 [xfs]
xfs_vn_fiemap+0xab/0xc0 [xfs]
do_vfs_ioctl+0x498/0x670
SyS_ioctl+0x79/0x90
entry_SYSCALL_64_fastpath+0x12/0x6f
CPU0
----
lock(&xfs_nondir_ilock_class);
<Interrupt>
lock(&xfs_nondir_ilock_class);
*** DEADLOCK ***
3 locks held by kswapd0/543:
stack backtrace:
CPU: 0 PID: 543 Comm: kswapd0 Tainted: G O 4.5.0-rc2+ #4
Call Trace:
lock_acquire+0xd8/0x1e0
down_write_nested+0x5e/0xc0
xfs_ilock+0x177/0x200 [xfs]
xfs_reflink_cancel_cow_range+0x150/0x300 [xfs]
xfs_fs_evict_inode+0xdc/0x1e0 [xfs]
evict+0xc5/0x190
dispose_list+0x39/0x60
prune_icache_sb+0x4b/0x60
super_cache_scan+0x14f/0x1a0
shrink_slab.part.63.constprop.79+0x1e9/0x4e0
shrink_zone+0x15e/0x170
kswapd+0x4f1/0xa80
kthread+0xf2/0x110
ret_from_fork+0x3f/0x70
To quote Dave:
"Ignoring whether reflink should be doing anything or not, that's a
"xfs_refcountbt_init_cursor() gets called both outside and inside
transactions" lockdep false positive case. The problem here is lockdep
has seen this allocation from within a transaction, hence a GFP_NOFS
allocation, and now it's seeing it in a GFP_KERNEL context. Also note
that we have an active reference to this inode.
So, because the reclaim annotations overload the interrupt level
detections and it's seen the inode ilock been taken in reclaim
("interrupt") context, this triggers a reclaim context warning where
it thinks it is unsafe to do this allocation in GFP_KERNEL context
holding the inode ilock..."
This sounds like a fundamental problem of the reclaim lock detection.
It is really impossible to annotate such a special usecase IMHO unless
the reclaim lockup detection is reworked completely. Until then it is
much better to provide a way to add "I know what I am doing flag" and
mark problematic places. This would prevent from abusing GFP_NOFS flag
which has a runtime effect even on configurations which have lockdep
disabled.
Introduce __GFP_NOLOCKDEP flag which tells the lockdep gfp tracking to
skip the current allocation request.
While we are at it also make sure that the radix tree doesn't
accidentaly override tags stored in the upper part of the gfp_mask.
Link: http://lkml.kernel.org/r/20170306131408.9828-3-mhocko@kernel.org
Signed-off-by: Michal Hocko <mhocko@suse.com>
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <clm@fb.com>
Cc: David Sterba <dsterba@suse.cz>
Cc: Jan Kara <jack@suse.cz>
Cc: Brian Foster <bfoster@redhat.com>
Cc: Darrick J. Wong <darrick.wong@oracle.com>
Cc: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-05-03 15:52:09 -07:00
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial
2016-03-17 21:38:27 -07:00
lib/fonts/Kconfig: keep non-Sparc fonts listed together
2017-02-27 18:43:46 -08:00
lib/lz4: remove back-compat wrappers
2017-02-24 17:46:57 -08:00
mpi: Fix NULL ptr dereference in mpi_powm() [ver #3 ]
2016-11-25 12:57:50 +11:00
lib/raid6: Add AVX2 optimized xor_syndrome functions
2016-11-07 15:08:20 -08:00
zlib_deflate/deftree: remove bi_reverse()
2015-09-10 13:29:01 -07:00
Merge branch 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
2016-05-17 09:33:39 -07:00
assoc_array: don't call compare_object() on a node
2016-04-06 14:06:48 +01:00
locking/atomic: Implement atomic{,64,_long}_fetch_{add,sub,and,andnot,or,xor}{,_relaxed,_acquire,_release}()
2016-06-16 10:48:32 +02:00
lib: add module support to atomic64 tests
2017-02-24 17:46:57 -08:00
kernel-api.rst: fix some complex tags at lib/bitmap.c
2017-04-02 14:29:33 -06:00
treewide: Remove old email address
2015-11-23 09:44:58 +01:00
debug: Add _ONCE() logic to report_bug()
2017-03-30 09:37:20 +02:00
random: replace non-blocking pool with a Chacha20-based CRNG
2016-07-03 00:57:23 -04:00
ipv4: Update parameters for csum_tcpudp_magic to their original types
2016-03-13 23:55:13 -04:00
boot/param: Move next_arg() function to lib/cmdline.c for later reuse
2017-04-18 10:37:13 +02:00
cpumask: Export cpumask_any_but()
2016-02-29 09:35:20 +01:00
lib: add module support to crc32 tests
2017-02-24 17:46:57 -08:00
lib: add module support to crc32 tests
2017-02-24 17:46:57 -08:00
sched/headers: Prepare for new header dependencies before moving code to <linux/sched/task_stack.h>
2017-03-02 08:42:36 +01:00
lib/decompressors: use real out buf size for gunzip with kernel
2015-09-10 13:29:01 -07:00
lib/decompressors: use real out buf size for gunzip with kernel
2015-09-10 13:29:01 -07:00
lib/decompress_unlz4: change module to work with new LZ4 module version
2017-02-24 17:46:57 -08:00
lib/decompress_unlzma: Do a NULL check for pointer
2015-09-10 13:29:01 -07:00
lib/decompressors: use real out buf size for gunzip with kernel
2015-09-10 13:29:01 -07:00
lib/decompressors: use real out buf size for gunzip with kernel
2015-09-10 13:29:01 -07:00
devres: use to_pci_dev()
2016-02-07 23:17:59 -08:00
KEYS: Differentiate uses of rcu_dereference_key() and user_key_payload()
2017-03-02 10:09:00 +11:00
__div64_32(): make it overridable at compile time
2015-11-16 14:42:12 -05:00
lib/dma-debug.c: make locking work for RT
2017-05-03 15:52:07 -07:00
lib/dma-noop: Clarify a comment
2017-01-24 12:23:35 -05:00
lib/dma-virt: Add dma_virt_ops
2017-01-24 12:23:35 -05:00
sched/headers: Prepare for new header dependencies before moving code to <linux/sched/debug.h>
2017-03-02 08:42:34 +01:00
dynamic_debug: add jump label support
2016-08-04 08:50:07 -04:00
lib/cpio: Make find_cpio_data()'s offset arg optional
2016-06-08 11:04:19 +02:00
Replace <asm/uaccess.h> with <linux/uaccess.h> globally
2016-12-24 11:46:01 -08:00
fault-inject: fix inverted interval/probability values in printk
2015-10-23 17:55:10 +09:00
lib/find_bit.c: micro-optimise find_next_*_bit
2017-02-24 17:46:57 -08:00
lib+mm: fix few spelling mistakes
2016-02-15 11:18:23 +01:00
lib/GCD.c: use binary GCD algorithm instead of Euclidean
2016-05-20 17:58:30 -07:00
lib/genalloc.c: start search from start of chunk
2016-10-27 18:43:43 -07:00
lib: add module support to glob tests
2017-02-24 17:46:57 -08:00
lib: add module support to glob tests
2017-02-24 17:46:57 -08:00
lib/hexdump.c: truncate output in case of overflow
2015-11-06 17:50:42 -08:00
x86/hweight: Get rid of the special calling convention
2016-06-08 15:01:02 +02:00
idr: Add missing __rcu annotations
2017-02-13 21:44:10 -05:00
lib/iomap_copy.c: add __ioread32_copy()
2016-01-20 17:09:18 -08:00
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc
2015-11-05 16:34:48 -08:00
lib/iommu-helper: skip to next segment
2016-08-02 19:35:07 -04:00
mm: convert generic code to 5-level paging
2017-03-09 11:48:47 -08:00
Merge branch 'work.uaccess' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
2017-05-01 14:41:04 -07:00
This adds a new gcc plugin named "latent_entropy". It is designed to
2016-10-15 10:03:15 -07:00
sched/headers: Prepare to move 'init_task' and 'init_thread_union' from <linux/sched.h> to <linux/sched/task.h>
2017-03-02 08:42:38 +01:00
lib/kasprintf.c: add sanity check to kvasprintf
2016-01-16 11:17:27 -08:00
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
2017-02-28 10:00:39 -08:00
A reasonably busy cycle for documentation this time around. There is a new
2017-05-02 10:21:17 -07:00
mm, kasan: switch SLUB to stackdepot, enable memory quarantine for SLUB
2016-07-28 16:07:41 -07:00
kgdb: depends on VT
2016-05-23 17:04:14 -07:00
Kconfig: lib/Kconfig.ubsan fix reference to ubsan documentation
2016-12-14 16:04:08 -08:00
klist: fix starting point removed bug in klist iterators
2016-02-07 22:18:47 -08:00
kobject: Export kobject_get_unless_zero()
2017-03-22 20:11:35 -06:00
kobject: improve function-level documentation
2016-10-28 02:42:10 -04:00
Replace <asm/uaccess.h> with <linux/uaccess.h> globally
2016-12-24 11:46:01 -08:00
Merge branch 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
2016-01-22 11:58:43 -08:00
bug: switch data corruption check to __must_check
2017-02-24 17:46:56 -08:00
lib/llist.c: fix data race in llist_del_first
2015-11-06 17:50:42 -08:00
locking/selftest: Fix output since KERN_CONT changes
2016-11-25 07:12:19 +01:00
locking/core: Remove cpu_relax_lowlatency() users
2016-11-16 10:15:10 +01:00
lru_cache: Converted lc_seq_printf_status to return void
2015-11-25 09:22:02 -07:00
Merge branch 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
2017-05-02 15:53:46 -07:00
net: Add support for CHANGEUPPER notifier error injection
2015-12-03 11:49:23 -05:00
netlink: pass extended ACK struct to parsing functions
2017-04-13 13:58:22 -04:00
sched/headers: Prepare for new header dependencies before moving code to <linux/sched/debug.h>
2017-03-02 08:42:34 +01:00
include/linux/nodemask.h: create next_node_in() helper
2016-05-19 19:12:14 -07:00
once: make helper generic for calling functions once
2015-10-08 05:26:36 -07:00
lib: Introduce priority array area manager
2017-02-03 16:35:42 -05:00
parser: add u64 number parser
2016-12-06 10:17:03 +02:00
libnvdimm for 4.3:
2015-09-08 14:35:59 -07:00
percpu-refcount: init ->confirm_switch member properly
2016-08-11 13:52:23 -04:00
percpu_counter: percpu_counter_hotcpu_callback() cleanup
2017-01-20 10:06:56 -05:00
sched/headers: Prepare to remove the <linux/gfp.h> include from <linux/sched.h>
2017-03-02 08:42:34 +01:00
sched/headers: Prepare for new header dependencies before moving code to <linux/sched/clock.h>
2017-03-02 08:42:27 +01:00
lib/prime_numbers: Suppress warn on kmalloc failure
2017-01-23 09:17:12 +01:00
lockdep: allow to disable reclaim lockup detection
2017-05-03 15:52:09 -07:00
This adds a new gcc plugin named "latent_entropy". It is designed to
2016-10-15 10:03:15 -07:00
ratelimit: extend to print suppressed messages on release
2016-08-02 19:35:06 -04:00
rbtree: use designated initializers
2017-02-24 17:46:57 -08:00
locking/refcounts: Use atomic_try_cmpxchg()
2017-03-23 08:54:41 +01:00
rhashtable: compact struct rhashtable_params
2017-05-01 16:22:40 -04:00
sbitmap: add sbitmap_get_shallow() operation
2017-04-14 14:06:52 -06:00
scatterlist: do not disable IRQs in sg_copy_buffer
2017-02-27 18:43:46 -08:00
tracing: Use seq_buf_used() in seq_buf_to_user() instead of len
2015-12-23 14:27:20 -05:00
lib: scatterlist: move SG pool code from SCSI driver to lib/sg_pool.c
2016-04-15 16:53:14 -04:00
lib/show_mem.c: teach show_mem to work with the given nodemask
2017-02-22 16:41:30 -08:00
siphash: implement HalfSipHash1-3 for hash tables
2017-01-09 13:58:57 -05:00
sched/core: Remove the tsk_cpus_allowed() wrapper
2017-03-02 08:42:24 +01:00
lib: add CONFIG_TEST_SORT to enable self-test of sort()
2017-02-24 17:46:57 -08:00
lib/stackdepot: export save/fetch stack for drivers
2016-11-11 08:12:37 -08:00
kernel-api.rst: fix a series of errors when parsing C files
2017-04-02 14:31:49 -06:00
string_helpers: add kstrdup_quotable_file
2016-04-21 10:47:26 +10:00
lib: harden strncpy_from_user
2016-10-11 15:06:30 -07:00
unsafe_[get|put]_user: change interface to use a error target label
2016-08-08 13:02:01 -07:00
swiotlb: ensure that page-sized mappings are page-aligned
2017-01-15 12:37:24 -05:00
lib/syscall: Clear return values when no stack
2017-03-24 07:43:35 +01:00
kstrto*: accept "-0" for signed conversion
2015-09-10 13:29:01 -07:00
lib/test-string_helpers.c: fix and improve string_get_size() tests
2016-02-03 08:28:43 -08:00
test_bitmap: unit tests for lib/bitmap.c
2016-02-19 22:54:09 -05:00
bpf, arm64: fix jit branch offset related to ldimm64
2017-05-02 15:04:50 -04:00
test_firmware: add test custom fallback trigger
2017-01-25 11:52:34 +01:00
lib/test_hash.c: fix warning in preprocessor symbol evaluation
2016-09-01 17:52:01 -07:00
test_hexdump: print statistics at the end
2016-01-20 17:09:18 -08:00
kasan: report only the first error by default
2017-03-31 17:13:30 -07:00
lib: fix spelling mistake: "actualy" -> "actually"
2017-02-26 11:03:38 -05:00
mm, printk: introduce new format string for flags
2016-03-15 16:55:16 -07:00
rhashtable-test: Fix max_size parameter description
2016-08-08 12:52:42 -07:00
siphash: implement HalfSipHash1-3 for hash tables
2017-01-09 13:58:57 -05:00
lib/test_sort.c: make it explicitly non-modular
2017-02-24 17:46:57 -08:00
locking/static_keys: Avoid nested functions
2016-02-09 10:27:29 +01:00
lib: remove check for AVR32 arch in test_user_copy
2017-05-01 09:36:30 +02:00
lib/uuid: add a test module
2016-05-30 15:26:57 -07:00
timerqueue: Use rb_entry_safe() instead of open-coding it
2017-01-20 08:03:42 +01:00
UBSAN: fix typo in format string
2016-08-02 17:31:41 -04:00
UBSAN: run-time undefined behavior sanity checker
2016-01-20 17:09:18 -08:00
lib/ucs2_string: Speed up ucs2_utf8size()
2016-09-09 16:08:46 +01:00
generic ...copy_..._user primitives
2017-03-28 18:22:11 -04:00
lib/uuid.c: use correct offset in uuid parser
2016-05-30 15:26:57 -07:00
kernel-api.rst: fix output of the vsnprintf() documentation
2017-04-02 14:29:19 -06:00
lib/win_minmax: windowed min or max estimator
2016-09-21 00:22:59 -04:00
7e7844226f