diff options
| author | Linus Torvalds <torvalds@linux-foundation.org> | 2022-05-24 10:18:23 -0700 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2022-05-24 10:18:23 -0700 |
| commit | 2319be135672f6e45aa937bceaae6c2668c7867c (patch) | |
| tree | 5d2210243fac4345bbb204064f9602fe9477b785 /kernel/locking/qspinlock.c | |
| parent | 143a6252e1b8ab424b4b293512a97cca7295c182 (diff) | |
| parent | 8491d1bdf5de152f27fc941e2dcdc4e66c950542 (diff) | |
| download | linux-2319be135672f6e45aa937bceaae6c2668c7867c.tar.gz linux-2319be135672f6e45aa937bceaae6c2668c7867c.tar.bz2 linux-2319be135672f6e45aa937bceaae6c2668c7867c.zip | |
Merge tag 'locking-core-2022-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull locking updates from Ingo Molnar:
- rwsem cleanups & optimizations/fixes:
- Conditionally wake waiters in reader/writer slowpaths
- Always try to wake waiters in out_nolock path
- Add try_cmpxchg64() implementation, with arch optimizations - and use
it to micro-optimize sched_clock_{local,remote}()
- Various force-inlining fixes to address objdump instrumentation-check
warnings
- Add lock contention tracepoints:
lock:contention_begin
lock:contention_end
- Misc smaller fixes & cleanups
* tag 'locking-core-2022-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/clock: Use try_cmpxchg64 in sched_clock_{local,remote}
locking/atomic/x86: Introduce arch_try_cmpxchg64
locking/atomic: Add generic try_cmpxchg64 support
futex: Remove a PREEMPT_RT_FULL reference.
locking/qrwlock: Change "queue rwlock" to "queued rwlock"
lockdep: Delete local_irq_enable_in_hardirq()
locking/mutex: Make contention tracepoints more consistent wrt adaptive spinning
locking: Apply contention tracepoints in the slow path
locking: Add lock contention tracepoints
locking/rwsem: Always try to wake waiters in out_nolock path
locking/rwsem: Conditionally wake waiters in reader/writer slowpaths
locking/rwsem: No need to check for handoff bit if wait queue empty
lockdep: Fix -Wunused-parameter for _THIS_IP_
x86/mm: Force-inline __phys_addr_nodebug()
x86/kvm/svm: Force-inline GHCB accessors
task_stack, x86/cea: Force-inline stack helpers
Diffstat (limited to 'kernel/locking/qspinlock.c')
| -rw-r--r-- | kernel/locking/qspinlock.c | 5 |
1 files changed, 5 insertions, 0 deletions
diff --git a/kernel/locking/qspinlock.c b/kernel/locking/qspinlock.c index cbff6ba53d56..65a9a10caa6f 100644 --- a/kernel/locking/qspinlock.c +++ b/kernel/locking/qspinlock.c @@ -22,6 +22,7 @@ #include <linux/prefetch.h> #include <asm/byteorder.h> #include <asm/qspinlock.h> +#include <trace/events/lock.h> /* * Include queued spinlock statistics code @@ -401,6 +402,8 @@ pv_queue: idx = node->count++; tail = encode_tail(smp_processor_id(), idx); + trace_contention_begin(lock, LCB_F_SPIN); + /* * 4 nodes are allocated based on the assumption that there will * not be nested NMIs taking spinlocks. That may not be true in @@ -554,6 +557,8 @@ locked: pv_kick_node(lock, next); release: + trace_contention_end(lock, 0); + /* * release the node */ |