2
0
mirror of https://github.com/edk2-porting/linux-next.git synced 2024-12-24 21:24:00 +08:00
linux-next/include/asm-generic/bitops
Peter Zijlstra f75d48644c bitops: Do not default to __clear_bit() for __clear_bit_unlock()
__clear_bit_unlock() is a special little snowflake. While it carries the
non-atomic '__' prefix, it is specifically documented to pair with
test_and_set_bit() and therefore should be 'somewhat' atomic.

Therefore the generic implementation of __clear_bit_unlock() cannot use
the fully non-atomic __clear_bit() as a default.

If an arch is able to do better; is must provide an implementation of
__clear_bit_unlock() itself.

Specifically, this came up as a result of hackbench livelock'ing in
slab_lock() on ARC with SMP + SLUB + !LLSC.

The issue was incorrect pairing of atomic ops.

 slab_lock() -> bit_spin_lock() -> test_and_set_bit()
 slab_unlock() -> __bit_spin_unlock() -> __clear_bit()

The non serializing __clear_bit() was getting "lost"

 80543b8e:	ld_s       r2,[r13,0] <--- (A) Finds PG_locked is set
 80543b90:	or         r3,r2,1    <--- (B) other core unlocks right here
 80543b94:	st_s       r3,[r13,0] <--- (C) sets PG_locked (overwrites unlock)

Fixes ARC STAR 9000817404 (and probably more).

Reported-by: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Tested-by: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Christoph Lameter <cl@linux.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Helge Deller <deller@gmx.de>
Cc: James E.J. Bottomley <jejb@parisc-linux.org>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Noam Camus <noamc@ezchip.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: http://lkml.kernel.org/r/20160309114054.GJ6356@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-03-21 10:50:48 +01:00
..
__ffs.h x86, generic: mark complex bitops.h inlines as __always_inline 2009-01-13 18:56:30 +01:00
__fls.h x86, generic: mark complex bitops.h inlines as __always_inline 2009-01-13 18:56:30 +01:00
arch_hweight.h arch, hweight: Fix compilation errors 2010-05-04 10:25:27 -07:00
atomic.h arch: Mass conversion of smp_mb__*() 2014-04-18 14:20:48 +02:00
builtin-__ffs.h generic: Implement generic ffs/fls using __builtin_* functions 2012-09-14 17:15:41 +01:00
builtin-__fls.h generic: Implement generic ffs/fls using __builtin_* functions 2012-09-14 17:15:41 +01:00
builtin-ffs.h generic: Implement generic ffs/fls using __builtin_* functions 2012-09-14 17:15:41 +01:00
builtin-fls.h generic: Implement generic ffs/fls using __builtin_* functions 2012-09-14 17:15:41 +01:00
const_hweight.h bitops: Fix signedness of compile-time hweight implementations 2014-02-09 13:39:26 +01:00
ext2-atomic-setbit.h asm-generic: add another generic ext2 atomic bitops 2011-07-26 16:49:46 -07:00
ext2-atomic.h asm-generic: add another generic ext2 atomic bitops 2011-07-26 16:49:46 -07:00
ffs.h [PATCH] bitops: generic ffs() 2006-03-26 08:57:11 -08:00
ffz.h [PATCH] bitops: generic ffz() 2006-03-26 08:57:10 -08:00
find.h bitops/find: clarify and extend documentation 2013-11-13 12:09:22 +09:00
fls64.h x86, generic: mark complex bitops.h inlines as __always_inline 2009-01-13 18:56:30 +01:00
fls.h x86, generic: mark complex bitops.h inlines as __always_inline 2009-01-13 18:56:30 +01:00
hweight.h bitops: Optimize hweight() by making use of compile-time evaluation 2010-04-06 15:52:11 -07:00
le.h bitops: introduce generic {clear,set}_bit_le() 2012-10-06 03:04:55 +09:00
lock.h bitops: Do not default to __clear_bit() for __clear_bit_unlock() 2016-03-21 10:50:48 +01:00
non-atomic.h define first set of BIT* macros 2007-10-19 11:53:42 -07:00
sched.h sched: simplify sched_find_first_bit() 2007-07-09 18:52:00 +02:00