include/linux/ring_buffer.h | 4 +-- kernel/trace/ring_buffer.c | 63 ++++++------------------------------- kernel/trace/trace.c | 14 +++------ kernel/trace/trace_kdb.c | 8 ++--- 4 files changed, 18 insertions(+), 71 deletions(-)
From: Steven Rostedt <rostedt@goodmis.org>
When the ring buffer was first introduced, reading the non-consuming
"trace" file required disabling the writing of the ring buffer. To make
sure the writing was fully disabled before iterating the buffer with a
non-consuming read, it would set the disable flag of the buffer and then
call an RCU synchronization to make sure all the buffers were
synchronized.
The function ring_buffer_read_start() originally would initialize the
iterator and call an RCU synchronization, but this was for each individual
per CPU buffer where this would get called many times on a machine with
many CPUs before the trace file could be read. The commit 72c9ddfd4c5bf
("ring-buffer: Make non-consuming read less expensive with lots of cpus.")
separated ring_buffer_read_start into ring_buffer_read_prepare(),
ring_buffer_read_sync() and then ring_buffer_read_start() to allow each of
the per CPU buffers to be prepared, call the read_buffer_read_sync() once,
and then the ring_buffer_read_start() for each of the CPUs which made
things much faster.
The commit 1039221cc278 ("ring-buffer: Do not disable recording when there
is an iterator") removed the requirement of disabling the recording of the
ring buffer in order to iterate it, but it did not remove the
synchronization that was happening that was required to wait for all the
buffers to have no more writers. It's now OK for the buffers to have
writers and no synchronization is needed.
Remove the synchronization and put back the interface for the ring buffer
iterator back before commit 72c9ddfd4c5bf was applied.
Reported-by: David Howells <dhowells@redhat.com>
Fixes: 1039221cc278 ("ring-buffer: Do not disable recording when there is an iterator")
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
include/linux/ring_buffer.h | 4 +--
kernel/trace/ring_buffer.c | 63 ++++++-------------------------------
kernel/trace/trace.c | 14 +++------
kernel/trace/trace_kdb.c | 8 ++---
4 files changed, 18 insertions(+), 71 deletions(-)
diff --git a/include/linux/ring_buffer.h b/include/linux/ring_buffer.h
index cd7f0ae26615..bc90c3c7b5fd 100644
--- a/include/linux/ring_buffer.h
+++ b/include/linux/ring_buffer.h
@@ -152,9 +152,7 @@ ring_buffer_consume(struct trace_buffer *buffer, int cpu, u64 *ts,
unsigned long *lost_events);
struct ring_buffer_iter *
-ring_buffer_read_prepare(struct trace_buffer *buffer, int cpu, gfp_t flags);
-void ring_buffer_read_prepare_sync(void);
-void ring_buffer_read_start(struct ring_buffer_iter *iter);
+ring_buffer_read_start(struct trace_buffer *buffer, int cpu, gfp_t flags);
void ring_buffer_read_finish(struct ring_buffer_iter *iter);
struct ring_buffer_event *
diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
index 00fc38d70e86..24bb5287c415 100644
--- a/kernel/trace/ring_buffer.c
+++ b/kernel/trace/ring_buffer.c
@@ -5846,24 +5846,20 @@ ring_buffer_consume(struct trace_buffer *buffer, int cpu, u64 *ts,
EXPORT_SYMBOL_GPL(ring_buffer_consume);
/**
- * ring_buffer_read_prepare - Prepare for a non consuming read of the buffer
+ * ring_buffer_read_start - start a non consuming read of the buffer
* @buffer: The ring buffer to read from
* @cpu: The cpu buffer to iterate over
* @flags: gfp flags to use for memory allocation
*
- * This performs the initial preparations necessary to iterate
- * through the buffer. Memory is allocated, buffer resizing
- * is disabled, and the iterator pointer is returned to the caller.
- *
- * After a sequence of ring_buffer_read_prepare calls, the user is
- * expected to make at least one call to ring_buffer_read_prepare_sync.
- * Afterwards, ring_buffer_read_start is invoked to get things going
- * for real.
+ * This creates an iterator to allow non-consuming iteration through
+ * the buffer. If the buffer is disabled for writing, it will produce
+ * the same information each time, but if the buffer is still writing
+ * then the first hit of a write will cause the iteration to stop.
*
- * This overall must be paired with ring_buffer_read_finish.
+ * Must be paired with ring_buffer_read_finish.
*/
struct ring_buffer_iter *
-ring_buffer_read_prepare(struct trace_buffer *buffer, int cpu, gfp_t flags)
+ring_buffer_read_start(struct trace_buffer *buffer, int cpu, gfp_t flags)
{
struct ring_buffer_per_cpu *cpu_buffer;
struct ring_buffer_iter *iter;
@@ -5889,51 +5885,12 @@ ring_buffer_read_prepare(struct trace_buffer *buffer, int cpu, gfp_t flags)
atomic_inc(&cpu_buffer->resize_disabled);
- return iter;
-}
-EXPORT_SYMBOL_GPL(ring_buffer_read_prepare);
-
-/**
- * ring_buffer_read_prepare_sync - Synchronize a set of prepare calls
- *
- * All previously invoked ring_buffer_read_prepare calls to prepare
- * iterators will be synchronized. Afterwards, read_buffer_read_start
- * calls on those iterators are allowed.
- */
-void
-ring_buffer_read_prepare_sync(void)
-{
- synchronize_rcu();
-}
-EXPORT_SYMBOL_GPL(ring_buffer_read_prepare_sync);
-
-/**
- * ring_buffer_read_start - start a non consuming read of the buffer
- * @iter: The iterator returned by ring_buffer_read_prepare
- *
- * This finalizes the startup of an iteration through the buffer.
- * The iterator comes from a call to ring_buffer_read_prepare and
- * an intervening ring_buffer_read_prepare_sync must have been
- * performed.
- *
- * Must be paired with ring_buffer_read_finish.
- */
-void
-ring_buffer_read_start(struct ring_buffer_iter *iter)
-{
- struct ring_buffer_per_cpu *cpu_buffer;
- unsigned long flags;
-
- if (!iter)
- return;
-
- cpu_buffer = iter->cpu_buffer;
-
- raw_spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
+ guard(raw_spinlock_irqsave)(&cpu_buffer->reader_lock);
arch_spin_lock(&cpu_buffer->lock);
rb_iter_reset(iter);
arch_spin_unlock(&cpu_buffer->lock);
- raw_spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
+
+ return iter;
}
EXPORT_SYMBOL_GPL(ring_buffer_read_start);
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 95ae7c4e5835..7996f26c3f46 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -4735,21 +4735,15 @@ __tracing_open(struct inode *inode, struct file *file, bool snapshot)
if (iter->cpu_file == RING_BUFFER_ALL_CPUS) {
for_each_tracing_cpu(cpu) {
iter->buffer_iter[cpu] =
- ring_buffer_read_prepare(iter->array_buffer->buffer,
- cpu, GFP_KERNEL);
- }
- ring_buffer_read_prepare_sync();
- for_each_tracing_cpu(cpu) {
- ring_buffer_read_start(iter->buffer_iter[cpu]);
+ ring_buffer_read_start(iter->array_buffer->buffer,
+ cpu, GFP_KERNEL);
tracing_iter_reset(iter, cpu);
}
} else {
cpu = iter->cpu_file;
iter->buffer_iter[cpu] =
- ring_buffer_read_prepare(iter->array_buffer->buffer,
- cpu, GFP_KERNEL);
- ring_buffer_read_prepare_sync();
- ring_buffer_read_start(iter->buffer_iter[cpu]);
+ ring_buffer_read_start(iter->array_buffer->buffer,
+ cpu, GFP_KERNEL);
tracing_iter_reset(iter, cpu);
}
diff --git a/kernel/trace/trace_kdb.c b/kernel/trace/trace_kdb.c
index d7b135de958a..896ff78b8349 100644
--- a/kernel/trace/trace_kdb.c
+++ b/kernel/trace/trace_kdb.c
@@ -43,17 +43,15 @@ static void ftrace_dump_buf(int skip_entries, long cpu_file)
if (cpu_file == RING_BUFFER_ALL_CPUS) {
for_each_tracing_cpu(cpu) {
iter.buffer_iter[cpu] =
- ring_buffer_read_prepare(iter.array_buffer->buffer,
- cpu, GFP_ATOMIC);
- ring_buffer_read_start(iter.buffer_iter[cpu]);
+ ring_buffer_read_start(iter.array_buffer->buffer,
+ cpu, GFP_ATOMIC);
tracing_iter_reset(&iter, cpu);
}
} else {
iter.cpu_file = cpu_file;
iter.buffer_iter[cpu_file] =
- ring_buffer_read_prepare(iter.array_buffer->buffer,
+ ring_buffer_read_start(iter.array_buffer->buffer,
cpu_file, GFP_ATOMIC);
- ring_buffer_read_start(iter.buffer_iter[cpu_file]);
tracing_iter_reset(&iter, cpu_file);
}
--
2.47.2
Hello,
kernel test robot noticed a 46.6% improvement of stress-ng.mmapfiles.ops_per_sec on:
commit: e554d9f934596e6e1c81f99d4617d3c3061129bb ("[PATCH] ring-buffer: Remove ring_buffer_read_prepare_sync()")
url: https://github.com/intel-lab-lkp/linux/commits/Steven-Rostedt/ring-buffer-Remove-ring_buffer_read_prepare_sync/20250701-060627
base: https://git.kernel.org/cgit/linux/kernel/git/trace/linux-trace for-next
patch link: https://lore.kernel.org/all/20250630180440.3eabb514@batman.local.home/
patch subject: [PATCH] ring-buffer: Remove ring_buffer_read_prepare_sync()
testcase: stress-ng
config: x86_64-rhel-9.4
compiler: gcc-12
test machine: 192 threads 2 sockets Intel(R) Xeon(R) 6740E CPU @ 2.4GHz (Sierra Forest) with 256G memory
parameters:
nr_threads: 100%
testtime: 60s
test: mmapfiles
cpufreq_governor: performance
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20250715/202507150815.f6cdbee0-lkp@intel.com
=========================================================================================
compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
gcc-12/performance/x86_64-rhel-9.4/100%/debian-12-x86_64-20240206.cgz/lkp-srf-2sp2/mmapfiles/stress-ng/60s
commit:
9cca84d73f ("Merge tools/for-next")
e554d9f934 ("ring-buffer: Remove ring_buffer_read_prepare_sync()")
9cca84d73fae0f7c e554d9f934596e6e1c81f99d461
---------------- ---------------------------
%stddev %change %stddev
\ | \
1.191e+10 -30.8% 8.242e+09 ± 5% cpuidle..time
557160 +182.9% 1576368 ± 12% cpuidle..usage
517840 ± 8% +499.0% 3101638 ± 7% numa-numastat.node0.local_node
592074 ± 10% +446.4% 3235092 ± 6% numa-numastat.node0.numa_hit
487358 ± 8% +500.3% 2925629 ± 4% numa-numastat.node1.local_node
610968 ± 9% +389.4% 2990283 ± 5% numa-numastat.node1.numa_hit
98.45 -29.1% 69.78 ± 5% vmstat.cpu.id
2.41 ± 16% +2346.5% 59.04 ± 11% vmstat.procs.r
3491 +162.4% 9160 ± 12% vmstat.system.cs
14593 +813.0% 133235 ± 22% vmstat.system.in
6.50 ± 53% +38700.0% 2522 ± 44% perf-c2c.DRAM.local
40.17 ± 21% +11012.2% 4463 ± 59% perf-c2c.DRAM.remote
83.50 ± 26% +12360.8% 10404 ± 84% perf-c2c.HITM.local
25.50 ± 22% +9441.2% 2433 ± 81% perf-c2c.HITM.remote
109.00 ± 23% +11677.8% 12837 ± 81% perf-c2c.HITM.total
143881 +61.6% 232515 meminfo.KReclaimable
10550821 ± 2% +17.2% 12364230 ± 4% meminfo.Memused
44747 ± 2% +2922.7% 1352558 ± 34% meminfo.PageTables
143881 +61.6% 232515 meminfo.SReclaimable
233317 +40.9% 328842 ± 13% meminfo.Shmem
11374925 ± 2% +17.1% 13323634 ± 5% meminfo.max_used_kB
85738 ± 39% +103.8% 174738 ± 3% numa-meminfo.node0.KReclaimable
5806365 ± 25% +35.4% 7863059 ± 2% numa-meminfo.node0.MemUsed
25460 ± 38% +2636.2% 696665 ± 37% numa-meminfo.node0.PageTables
85738 ± 39% +103.8% 174738 ± 3% numa-meminfo.node0.SReclaimable
19286 ± 52% +3358.4% 667006 ± 32% numa-meminfo.node1.PageTables
99582 ± 88% +136.2% 235198 ± 51% numa-meminfo.node1.Shmem
98.49 -29.7 68.82 ± 5% mpstat.cpu.all.idle%
0.02 ± 5% +0.1 0.16 ± 11% mpstat.cpu.all.irq%
0.05 ± 2% +0.1 0.14 ± 11% mpstat.cpu.all.soft%
1.27 +29.3 30.53 ± 12% mpstat.cpu.all.sys%
0.17 +0.2 0.35 ± 11% mpstat.cpu.all.usr%
2.00 +410.0% 10.20 ± 33% mpstat.max_utilization.seconds
56.84 ± 3% +75.9% 100.00 mpstat.max_utilization_pct
6362 ± 38% +2597.0% 171604 ± 38% numa-vmstat.node0.nr_page_table_pages
21433 ± 39% +103.7% 43659 ± 3% numa-vmstat.node0.nr_slab_reclaimable
592326 ± 10% +447.6% 3243498 ± 6% numa-vmstat.node0.numa_hit
518092 ± 8% +500.3% 3110044 ± 7% numa-vmstat.node0.numa_local
4823 ± 52% +3265.9% 162349 ± 31% numa-vmstat.node1.nr_page_table_pages
24897 ± 88% +134.9% 58481 ± 52% numa-vmstat.node1.nr_shmem
610128 ± 9% +391.6% 2999506 ± 5% numa-vmstat.node1.numa_hit
486517 ± 8% +503.2% 2934852 ± 4% numa-vmstat.node1.numa_local
173139 ± 7% -28.5% 123716 ± 5% stress-ng.mmapfiles.file_munmap_per_sec
3255621 +189.4% 9420391 ± 17% stress-ng.mmapfiles.file_pages_mmap'd_per_sec
4406848 ± 7% +177.4% 12226005 ± 17% stress-ng.mmapfiles.file_pages_munmap'd_per_sec
6750720 +45.8% 9839893 ± 4% stress-ng.mmapfiles.ops
111233 +46.6% 163043 ± 4% stress-ng.mmapfiles.ops_per_sec
25.45 +185.7% 72.72 ± 17% stress-ng.mmapfiles.pages_per_mapping
2385 ± 11% +1274.5% 32790 ± 3% stress-ng.time.involuntary_context_switches
126772 ± 7% +39.7% 177103 ± 7% stress-ng.time.minor_page_faults
233.83 +2464.3% 5996 ± 12% stress-ng.time.percent_of_cpu_this_job_got
139.02 ± 2% +2502.9% 3618 ± 12% stress-ng.time.system_time
4.73 ± 2% +325.0% 20.11 ± 7% stress-ng.time.user_time
23391 +393.7% 115477 ± 27% stress-ng.time.voluntary_context_switches
945786 +2.6% 970204 proc-vmstat.nr_file_pages
68051 +11.7% 76009 ± 6% proc-vmstat.nr_mapped
11189 ± 2% +2912.5% 337076 ± 34% proc-vmstat.nr_page_table_pages
58330 +40.5% 81973 ± 13% proc-vmstat.nr_shmem
35968 +61.4% 58056 proc-vmstat.nr_slab_reclaimable
923417 +4.1% 961412 proc-vmstat.nr_slab_unreclaimable
6994 ± 2% +282.2% 26735 ± 88% proc-vmstat.numa_hint_faults
1205445 +417.1% 6233925 ± 6% proc-vmstat.numa_hit
1007600 +499.0% 6035817 ± 6% proc-vmstat.numa_local
500.50 ± 46% +3114.1% 16086 ± 82% proc-vmstat.numa_pages_migrated
8249 +920.6% 84190 ± 65% proc-vmstat.numa_pte_updates
1867778 ± 3% +476.0% 10758231 ± 4% proc-vmstat.pgalloc_normal
480874 +15.1% 553383 ± 6% proc-vmstat.pgfault
1816801 ± 3% +417.9% 9409800 ± 8% proc-vmstat.pgfree
500.50 ± 46% +3114.1% 16086 ± 82% proc-vmstat.pgmigrate_success
18020 +39.3% 25094 ± 7% proc-vmstat.pgreuse
1.50 ± 5% -21.0% 1.19 ± 2% perf-stat.i.MPKI
9.549e+08 ± 3% +526.1% 5.979e+09 ± 5% perf-stat.i.branch-instructions
2.68 -2.1 0.58 ± 6% perf-stat.i.branch-miss-rate%
18077507 +39.1% 25144949 ± 3% perf-stat.i.branch-misses
20.48 ± 3% +15.9 36.39 ± 4% perf-stat.i.cache-miss-rate%
5467752 ± 11% +471.9% 31267720 ± 6% perf-stat.i.cache-misses
17535422 ± 4% +384.2% 84912182 ± 7% perf-stat.i.cache-references
3219 +184.2% 9149 ± 12% perf-stat.i.context-switches
1.75 +268.0% 6.43 ± 14% perf-stat.i.cpi
7.74e+09 ± 4% +2371.9% 1.913e+11 ± 11% perf-stat.i.cpu-cycles
278.42 +12.2% 312.44 ± 2% perf-stat.i.cpu-migrations
2859 ± 8% +147.1% 7066 ± 13% perf-stat.i.cycles-between-cache-misses
4.615e+09 ± 3% +524.5% 2.882e+10 ± 5% perf-stat.i.instructions
0.71 -55.8% 0.31 ± 18% perf-stat.i.ipc
5250 ± 3% +20.7% 6340 ± 7% perf-stat.i.minor-faults
5250 ± 3% +20.7% 6340 ± 7% perf-stat.i.page-faults
1.89 ± 3% -1.5 0.42 ± 2% perf-stat.overall.branch-miss-rate%
31.05 ± 6% +5.8 36.86 ± 3% perf-stat.overall.cache-miss-rate%
1.68 ± 2% +295.8% 6.65 ± 7% perf-stat.overall.cpi
1431 ± 7% +329.5% 6147 ± 9% perf-stat.overall.cycles-between-cache-misses
0.60 ± 2% -74.6% 0.15 ± 7% perf-stat.overall.ipc
9.436e+08 ± 3% +523.6% 5.884e+09 ± 5% perf-stat.ps.branch-instructions
17808666 +37.8% 24537286 ± 3% perf-stat.ps.branch-misses
5399377 ± 11% +468.8% 30710020 ± 6% perf-stat.ps.cache-misses
17331532 ± 4% +381.3% 83414797 ± 7% perf-stat.ps.cache-references
3165 +183.8% 8984 ± 12% perf-stat.ps.context-switches
7.661e+09 ± 4% +2367.5% 1.89e+11 ± 11% perf-stat.ps.cpu-cycles
273.88 +12.1% 306.96 ± 2% perf-stat.ps.cpu-migrations
4.561e+09 ± 3% +521.7% 2.836e+10 ± 5% perf-stat.ps.instructions
5162 ± 3% +20.5% 6221 ± 7% perf-stat.ps.minor-faults
5162 ± 3% +20.5% 6221 ± 7% perf-stat.ps.page-faults
2.827e+11 ± 4% +518.2% 1.748e+12 ± 6% perf-stat.total.instructions
0.03 ± 22% +114.4% 0.07 ± 48% sched_debug.cfs_rq:/.h_nr_queued.avg
0.16 ± 8% +46.6% 0.24 ± 14% sched_debug.cfs_rq:/.h_nr_queued.stddev
0.03 ± 22% +114.4% 0.07 ± 48% sched_debug.cfs_rq:/.h_nr_runnable.avg
0.16 ± 8% +46.6% 0.24 ± 14% sched_debug.cfs_rq:/.h_nr_runnable.stddev
28.55 ± 22% +105.6% 58.68 ± 30% sched_debug.cfs_rq:/.load_avg.avg
129.70 ± 12% +68.8% 219.00 ± 21% sched_debug.cfs_rq:/.load_avg.stddev
0.03 ± 23% +110.4% 0.07 ± 49% sched_debug.cfs_rq:/.nr_queued.avg
0.17 ± 10% +44.6% 0.24 ± 16% sched_debug.cfs_rq:/.nr_queued.stddev
19.92 ± 31% +131.3% 46.08 ± 38% sched_debug.cfs_rq:/.removed.load_avg.avg
96.83 ± 15% +102.7% 196.23 ± 25% sched_debug.cfs_rq:/.removed.load_avg.stddev
8.64 ± 31% +113.6% 18.44 ± 35% sched_debug.cfs_rq:/.removed.runnable_avg.avg
263.25 ± 2% +80.2% 474.40 ± 20% sched_debug.cfs_rq:/.removed.runnable_avg.max
43.62 ± 16% +87.4% 81.72 ± 22% sched_debug.cfs_rq:/.removed.runnable_avg.stddev
8.64 ± 31% +113.6% 18.44 ± 35% sched_debug.cfs_rq:/.removed.util_avg.avg
263.25 ± 2% +80.1% 474.20 ± 20% sched_debug.cfs_rq:/.removed.util_avg.max
43.62 ± 16% +87.3% 81.72 ± 22% sched_debug.cfs_rq:/.removed.util_avg.stddev
73.73 ± 5% +107.0% 152.62 ± 11% sched_debug.cfs_rq:/.runnable_avg.avg
132.86 ± 2% +75.9% 233.71 ± 18% sched_debug.cfs_rq:/.runnable_avg.stddev
73.61 ± 5% +107.0% 152.40 ± 11% sched_debug.cfs_rq:/.util_avg.avg
132.55 ± 2% +76.2% 233.59 ± 18% sched_debug.cfs_rq:/.util_avg.stddev
6.04 ± 37% +243.7% 20.77 ± 85% sched_debug.cfs_rq:/.util_est.avg
372.25 ± 17% +89.1% 704.00 ± 20% sched_debug.cfs_rq:/.util_est.max
41.23 ± 23% +122.7% 91.83 ± 38% sched_debug.cfs_rq:/.util_est.stddev
66163 ± 20% -93.2% 4511 ± 59% sched_debug.cpu.avg_idle.min
124590 ± 7% +63.4% 203555 ± 20% sched_debug.cpu.avg_idle.stddev
239770 -10.1% 215480 ± 5% sched_debug.cpu.clock.avg
239792 -10.1% 215503 ± 5% sched_debug.cpu.clock.max
239740 -10.1% 215453 ± 5% sched_debug.cpu.clock.min
239688 -10.1% 215399 ± 5% sched_debug.cpu.clock_task.avg
239761 -10.1% 215478 ± 5% sched_debug.cpu.clock_task.max
233073 -10.4% 208806 ± 5% sched_debug.cpu.clock_task.min
102.79 ± 24% +146.7% 253.62 ± 72% sched_debug.cpu.curr->pid.avg
615.65 ± 8% +39.6% 859.42 ± 28% sched_debug.cpu.curr->pid.stddev
0.03 ± 27% +128.4% 0.06 ± 54% sched_debug.cpu.nr_running.avg
0.14 ± 11% +57.0% 0.22 ± 17% sched_debug.cpu.nr_running.stddev
0.50 -83.0% 0.09 ±181% sched_debug.cpu.nr_uninterruptible.avg
239751 -10.1% 215460 ± 5% sched_debug.cpu_clk
239489 -10.1% 215201 ± 5% sched_debug.ktime
239978 -10.1% 215654 ± 5% sched_debug.sched_clk
0.00 ±223% +23036.0% 0.19 ± 54% perf-sched.sch_delay.avg.ms.__cond_resched.__kmalloc_cache_noprof.perf_event_mmap_event.perf_event_mmap.__mmap_region
0.01 +106.7% 0.02 ± 48% perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
0.00 ±223% +6920.0% 0.05 ±112% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
0.00 ±145% +14564.0% 0.24 ±126% perf-sched.sch_delay.avg.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
0.02 ± 12% -43.3% 0.01 ± 29% perf-sched.sch_delay.avg.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
0.01 ± 8% +22.3% 0.01 ± 4% perf-sched.sch_delay.avg.ms.anon_pipe_read.fifo_pipe_read.vfs_read.ksys_read
0.00 ±104% +1168.6% 0.03 ± 56% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown]
0.02 ± 47% -75.5% 0.00 ± 93% perf-sched.sch_delay.avg.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
0.02 ± 9% -33.6% 0.01 ± 19% perf-sched.sch_delay.avg.ms.schedule_hrtimeout_range_clock.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
2.30 ±219% -99.9% 0.00 ±200% perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.__tracing_open
2.35 ±222% -100.0% 0.00 perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_open
0.01 ± 11% -100.0% 0.00 perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_open_pipe
0.02 ± 45% -90.5% 0.00 ±200% perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release
0.01 ± 3% +501.8% 0.07 ± 54% perf-sched.sch_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.__wait_rcu_gp
0.01 +68.0% 0.01 ± 40% perf-sched.sch_delay.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
0.01 ± 34% +252.6% 0.04 ± 76% perf-sched.sch_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
0.03 ± 41% -58.0% 0.01 ± 8% perf-sched.sch_delay.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
0.00 ±223% +1392.5% 0.04 ± 85% perf-sched.sch_delay.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.pte_alloc_one
0.00 ±223% +1.6e+05% 1.36 ± 71% perf-sched.sch_delay.max.ms.__cond_resched.__kmalloc_cache_noprof.perf_event_mmap_event.perf_event_mmap.__mmap_region
0.04 ± 31% +12614.8% 5.36 ±136% perf-sched.sch_delay.max.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
0.00 ±223% +23720.0% 0.16 ±142% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
0.00 ±145% +43904.0% 0.73 ±110% perf-sched.sch_delay.max.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
0.04 ± 11% -46.3% 0.02 ± 34% perf-sched.sch_delay.max.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
0.04 ± 15% +38.7% 0.06 ± 5% perf-sched.sch_delay.max.ms.devkmsg_read.vfs_read.ksys_read.do_syscall_64
0.03 ± 23% -37.5% 0.02 ± 26% perf-sched.sch_delay.max.ms.do_nanosleep.hrtimer_nanosleep.common_nsleep.__x64_sys_clock_nanosleep
0.00 ±104% +8077.1% 0.19 ± 75% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown]
0.01 ± 79% +25841.2% 2.20 ±159% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
2.85 ±114% -99.9% 0.00 ± 93% perf-sched.sch_delay.max.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
0.04 ± 7% -33.2% 0.03 ± 29% perf-sched.sch_delay.max.ms.schedule_hrtimeout_range_clock.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
162.34 ±220% -100.0% 0.00 ±200% perf-sched.sch_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.__tracing_open
151.86 ±223% -100.0% 0.00 perf-sched.sch_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_open
0.03 ± 12% -83.7% 0.00 ±200% perf-sched.sch_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_release
0.04 ± 34% +2146.1% 0.81 ± 88% perf-sched.sch_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_check_open_get_tr
0.02 ± 12% -100.0% 0.00 perf-sched.sch_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_open_pipe
1.86 ±218% -99.9% 0.00 ±200% perf-sched.sch_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release
0.03 ± 5% +1199.8% 0.41 ± 51% perf-sched.sch_delay.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.__wait_rcu_gp
0.02 ± 7% +17495.5% 3.87 ±122% perf-sched.sch_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
0.02 ± 22% +3550.0% 0.58 ± 79% perf-sched.sch_delay.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
318.12 -64.9% 111.82 ± 44% perf-sched.total_wait_and_delay.average.ms
7352 +251.9% 25870 ± 78% perf-sched.total_wait_and_delay.count.ms
318.06 -64.8% 111.81 ± 44% perf-sched.total_wait_time.average.ms
3.93 -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
556.93 ± 6% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.__tracing_open
598.50 ± 7% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_open
605.96 ± 6% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_release
648.70 ± 3% -97.8% 14.58 ±127% perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_check_open_get_tr
729.67 ± 7% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_open_pipe
242.46 ± 5% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release
747.54 ± 5% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_generic_tr
621.23 ± 23% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_pipe
228.85 ± 8% -96.1% 8.96 ±200% perf-sched.wait_and_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.__wait_rcu_gp
3.26 +28.1% 4.17 ± 2% perf-sched.wait_and_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
0.00 ± 7% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.wait_for_partner.fifo_open.do_dentry_open.vfs_open
922.74 ± 2% -87.3% 116.81 ± 12% perf-sched.wait_and_delay.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
417.33 ± 2% -100.0% 0.00 perf-sched.wait_and_delay.count.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
69.17 ± 8% -100.0% 0.00 perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.__tracing_open
65.00 ± 7% -100.0% 0.00 perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_open
83.00 ± 9% -100.0% 0.00 perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_release
537.83 ± 2% -96.2% 20.60 ±145% perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_check_open_get_tr
12.50 ± 27% -100.0% 0.00 perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_open_pipe
572.00 ± 2% -100.0% 0.00 perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release
59.00 ± 13% -100.0% 0.00 perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_generic_tr
17.00 ± 10% -100.0% 0.00 perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_pipe
463.33 -99.5% 2.20 ±200% perf-sched.wait_and_delay.count.schedule_timeout.__wait_for_common.wait_for_completion_state.__wait_rcu_gp
1031 +16.0% 1196 ± 2% perf-sched.wait_and_delay.count.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
36.00 -100.0% 0.00 perf-sched.wait_and_delay.count.wait_for_partner.fifo_open.do_dentry_open.vfs_open
611.50 ± 3% +758.1% 5247 ± 7% perf-sched.wait_and_delay.count.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
1670 ± 27% -52.7% 789.82 ± 50% perf-sched.wait_and_delay.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
5.68 ± 26% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
1030 ± 2% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.__tracing_open
1031 ± 3% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_open
1034 ± 2% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_release
1046 ± 3% -94.2% 60.44 ±123% perf-sched.wait_and_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_check_open_get_tr
1013 ± 3% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_open_pipe
1031 ± 3% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release
1027 ± 2% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_generic_tr
1006 ± 2% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_pipe
1044 ± 3% -98.5% 15.36 ±200% perf-sched.wait_and_delay.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.__wait_rcu_gp
0.01 ± 29% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.wait_for_partner.fifo_open.do_dentry_open.vfs_open
4529 ± 6% -15.7% 3817 ± 10% perf-sched.wait_and_delay.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
0.29 ± 6% +16.7% 0.33 ± 5% perf-sched.wait_time.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 ±223% +28060.0% 0.70 ±154% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown]
3.91 -79.2% 0.81 ±117% perf-sched.wait_time.avg.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
554.63 ± 5% -98.1% 10.38 ±200% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.__tracing_open
596.15 ± 6% -100.0% 0.00 perf-sched.wait_time.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_open
605.94 ± 6% -96.0% 23.97 ±200% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_release
648.69 ± 3% -91.9% 52.84 ± 70% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_check_open_get_tr
729.66 ± 7% -100.0% 0.00 perf-sched.wait_time.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_open_pipe
242.44 ± 5% -99.0% 2.44 ±200% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release
747.53 ± 5% -97.3% 19.82 ± 89% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_generic_tr
621.22 ± 23% -95.6% 27.30 ±129% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_pipe
228.84 ± 8% -84.4% 35.69 ± 29% perf-sched.wait_time.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.__wait_rcu_gp
0.68 ± 9% +62.9% 1.11 ± 6% perf-sched.wait_time.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
3.24 +28.2% 4.16 ± 2% perf-sched.wait_time.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
922.72 ± 2% -87.3% 116.80 ± 12% perf-sched.wait_time.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
1670 ± 27% -29.8% 1172 ± 31% perf-sched.wait_time.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
0.01 ±223% +1.5e+05% 7.73 ±167% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown]
5.00 -83.7% 0.81 ±117% perf-sched.wait_time.max.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
1030 ± 2% -99.0% 10.38 ±200% perf-sched.wait_time.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.__tracing_open
1031 ± 3% -100.0% 0.00 perf-sched.wait_time.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_open
1034 ± 2% -97.7% 23.97 ±200% perf-sched.wait_time.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_buffers_release
1046 ± 3% -86.4% 141.95 ± 38% perf-sched.wait_time.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_check_open_get_tr
1013 ± 3% -100.0% 0.00 perf-sched.wait_time.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_open_pipe
1031 ± 3% -99.5% 4.88 ±200% perf-sched.wait_time.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release
1027 ± 2% -97.3% 27.26 ±104% perf-sched.wait_time.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_generic_tr
1006 ± 2% -95.7% 43.61 ±149% perf-sched.wait_time.max.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.tracing_release_pipe
1044 ± 3% -87.7% 128.89 ± 47% perf-sched.wait_time.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.__wait_rcu_gp
5.06 ± 2% +55.8% 7.88 ± 17% perf-sched.wait_time.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
4529 ± 6% -15.7% 3817 ± 10% perf-sched.wait_time.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
48.04 ± 7% -46.7 1.38 ± 71% perf-profile.calltrace.cycles-pp.__cmd_record
46.88 ± 8% -45.7 1.20 ± 80% perf-profile.calltrace.cycles-pp.record__finish_output.__cmd_record
46.87 ± 8% -45.7 1.20 ± 80% perf-profile.calltrace.cycles-pp.perf_session__process_events.record__finish_output.__cmd_record
46.74 ± 8% -45.5 1.20 ± 80% perf-profile.calltrace.cycles-pp.reader__read_event.perf_session__process_events.record__finish_output.__cmd_record
40.46 ± 8% -39.7 0.76 ±104% perf-profile.calltrace.cycles-pp.process_simple.reader__read_event.perf_session__process_events.record__finish_output.__cmd_record
38.04 ± 8% -37.3 0.73 ±105% perf-profile.calltrace.cycles-pp.ordered_events__queue.process_simple.reader__read_event.perf_session__process_events.record__finish_output
37.64 ± 8% -36.9 0.73 ±105% perf-profile.calltrace.cycles-pp.queue_event.ordered_events__queue.process_simple.reader__read_event.perf_session__process_events
21.58 ± 8% -21.3 0.24 ±124% perf-profile.calltrace.cycles-pp.perf_session__process_user_event.reader__read_event.perf_session__process_events.record__finish_output.__cmd_record
21.56 ± 8% -21.3 0.24 ±124% perf-profile.calltrace.cycles-pp.__ordered_events__flush.perf_session__process_user_event.reader__read_event.perf_session__process_events.record__finish_output
20.36 ± 8% -20.2 0.13 ±200% perf-profile.calltrace.cycles-pp.perf_session__deliver_event.__ordered_events__flush.perf_session__process_user_event.reader__read_event.perf_session__process_events
19.36 ± 8% -19.2 0.15 ±200% perf-profile.calltrace.cycles-pp.handle_internal_command.main
19.36 ± 8% -19.2 0.15 ±200% perf-profile.calltrace.cycles-pp.main
19.36 ± 8% -19.2 0.15 ±200% perf-profile.calltrace.cycles-pp.run_builtin.handle_internal_command.main
18.32 ± 8% -18.2 0.13 ±200% perf-profile.calltrace.cycles-pp.cmd_record.perf_c2c__record.run_builtin.handle_internal_command.main
18.32 ± 8% -18.2 0.13 ±200% perf-profile.calltrace.cycles-pp.perf_c2c__record.run_builtin.handle_internal_command.main
18.31 ± 8% -18.2 0.13 ±200% perf-profile.calltrace.cycles-pp.__cmd_record.cmd_record.perf_c2c__record.run_builtin.handle_internal_command
15.63 ± 9% -15.6 0.00 perf-profile.calltrace.cycles-pp.record__finish_output.__cmd_record.cmd_record.perf_c2c__record.run_builtin
15.63 ± 9% -15.6 0.00 perf-profile.calltrace.cycles-pp.perf_session__process_events.record__finish_output.__cmd_record.cmd_record.perf_c2c__record
15.58 ± 9% -15.6 0.00 perf-profile.calltrace.cycles-pp.reader__read_event.perf_session__process_events.record__finish_output.__cmd_record.cmd_record
14.08 ± 18% -13.5 0.62 ± 96% perf-profile.calltrace.cycles-pp.common_startup_64
13.01 ± 8% -13.0 0.00 perf-profile.calltrace.cycles-pp.machine__process_mmap2_event.perf_session__deliver_event.__ordered_events__flush.perf_session__process_user_event.reader__read_event
11.85 ± 21% -11.3 0.51 ± 98% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.common_startup_64
11.85 ± 21% -11.3 0.51 ± 98% perf-profile.calltrace.cycles-pp.start_secondary.common_startup_64
11.85 ± 21% -11.3 0.51 ± 98% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.common_startup_64
11.03 ± 16% -10.6 0.46 ± 97% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry
9.69 ± 21% -9.2 0.49 ± 98% perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.common_startup_64
9.16 ± 16% -8.8 0.34 ±136% perf-profile.calltrace.cycles-pp.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle
9.16 ± 16% -8.8 0.34 ±136% perf-profile.calltrace.cycles-pp.acpi_idle_do_entry.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call
9.16 ± 16% -8.8 0.34 ±136% perf-profile.calltrace.cycles-pp.acpi_safe_halt.acpi_idle_do_entry.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter
8.82 ± 20% -8.4 0.46 ± 97% perf-profile.calltrace.cycles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary
7.17 ± 15% -7.0 0.18 ±200% perf-profile.calltrace.cycles-pp.pv_native_safe_halt.acpi_safe_halt.acpi_idle_do_entry.acpi_idle_enter.cpuidle_enter_state
6.86 ± 21% -6.7 0.19 ±200% perf-profile.calltrace.cycles-pp.asm_sysvec_apic_timer_interrupt.pv_native_safe_halt.acpi_safe_halt.acpi_idle_do_entry.acpi_idle_enter
6.41 ± 9% -6.4 0.00 perf-profile.calltrace.cycles-pp.evlist__parse_sample.perf_session__deliver_event.__ordered_events__flush.perf_session__process_user_event.reader__read_event
6.00 ± 10% -6.0 0.00 perf-profile.calltrace.cycles-pp.map__new.machine__process_mmap2_event.perf_session__deliver_event.__ordered_events__flush.perf_session__process_user_event
5.92 ± 5% -5.7 0.27 ±122% perf-profile.calltrace.cycles-pp.kthread.ret_from_fork.ret_from_fork_asm
5.92 ± 5% -5.7 0.27 ±122% perf-profile.calltrace.cycles-pp.ret_from_fork.ret_from_fork_asm
5.92 ± 5% -5.7 0.27 ±122% perf-profile.calltrace.cycles-pp.ret_from_fork_asm
0.53 ± 47% +0.7 1.18 ± 44% perf-profile.calltrace.cycles-pp.__close
1.14 ± 11% +0.8 1.89 ± 41% perf-profile.calltrace.cycles-pp.do_dentry_open.vfs_open.do_open.path_openat.do_filp_open
1.15 ± 12% +0.8 1.95 ± 41% perf-profile.calltrace.cycles-pp.vfs_open.do_open.path_openat.do_filp_open.do_sys_openat2
0.00 +1.3 1.25 ± 45% perf-profile.calltrace.cycles-pp.lockref_put_return.dput.step_into.link_path_walk.path_openat
0.00 +1.3 1.26 ± 49% perf-profile.calltrace.cycles-pp.lookup_fast.open_last_lookups.path_openat.do_filp_open.do_sys_openat2
0.00 +1.3 1.34 ± 51% perf-profile.calltrace.cycles-pp.open_last_lookups.path_openat.do_filp_open.do_sys_openat2.__x64_sys_openat
0.00 +1.7 1.65 ± 51% perf-profile.calltrace.cycles-pp.vma_interval_tree_insert.__mmap_new_vma.__mmap_region.do_mmap.vm_mmap_pgoff
0.00 +1.7 1.68 ± 41% perf-profile.calltrace.cycles-pp.down_read.kernfs_iop_permission.inode_permission.may_open.do_open
0.00 +1.9 1.89 ± 53% perf-profile.calltrace.cycles-pp.iterate_dir.__x64_sys_getdents64.do_syscall_64.entry_SYSCALL_64_after_hwframe.getdents64
0.00 +1.9 1.92 ± 53% perf-profile.calltrace.cycles-pp.__x64_sys_getdents64.do_syscall_64.entry_SYSCALL_64_after_hwframe.getdents64
0.00 +1.9 1.93 ± 53% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.getdents64
0.00 +1.9 1.94 ± 52% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.getdents64
0.00 +2.0 1.98 ± 52% perf-profile.calltrace.cycles-pp.getdents64
0.00 +2.1 2.07 ± 42% perf-profile.calltrace.cycles-pp.down_read.kernfs_iop_getattr.vfs_getattr_nosec.vfs_fstat.__do_sys_newfstatat
0.00 +2.5 2.49 ± 46% perf-profile.calltrace.cycles-pp.kernfs_iop_permission.inode_permission.may_open.do_open.path_openat
0.00 +2.7 2.67 ± 45% perf-profile.calltrace.cycles-pp.inode_permission.may_open.do_open.path_openat.do_filp_open
0.00 +2.8 2.82 ± 46% perf-profile.calltrace.cycles-pp.down_read.kernfs_dop_revalidate.lookup_fast.walk_component.link_path_walk
0.00 +2.9 2.85 ± 45% perf-profile.calltrace.cycles-pp.may_open.do_open.path_openat.do_filp_open.do_sys_openat2
0.00 +2.9 2.86 ± 37% perf-profile.calltrace.cycles-pp.kernfs_iop_getattr.vfs_getattr_nosec.vfs_fstat.__do_sys_newfstatat.do_syscall_64
0.00 +3.1 3.10 ± 36% perf-profile.calltrace.cycles-pp.vfs_getattr_nosec.vfs_fstat.__do_sys_newfstatat.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 +3.1 3.14 ± 36% perf-profile.calltrace.cycles-pp.vfs_fstat.__do_sys_newfstatat.do_syscall_64.entry_SYSCALL_64_after_hwframe.fstatat64
0.00 +3.2 3.23 ± 76% perf-profile.calltrace.cycles-pp.zap_present_ptes.zap_pte_range.zap_pmd_range.unmap_page_range.unmap_vmas
0.00 +3.2 3.24 ± 39% perf-profile.calltrace.cycles-pp.kernfs_dop_revalidate.lookup_fast.walk_component.link_path_walk.path_openat
0.00 +3.3 3.32 ± 36% perf-profile.calltrace.cycles-pp.__do_sys_newfstatat.do_syscall_64.entry_SYSCALL_64_after_hwframe.fstatat64
0.00 +3.4 3.38 ± 35% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.fstatat64
0.00 +3.4 3.39 ± 35% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.fstatat64
0.00 +3.5 3.55 ± 35% perf-profile.calltrace.cycles-pp.fstatat64
1.36 ± 8% +3.7 5.09 ± 42% perf-profile.calltrace.cycles-pp.do_open.path_openat.do_filp_open.do_sys_openat2.__x64_sys_openat
0.00 +3.9 3.89 ± 58% perf-profile.calltrace.cycles-pp.dput.terminate_walk.path_openat.do_filp_open.do_sys_openat2
0.00 +3.9 3.91 ± 57% perf-profile.calltrace.cycles-pp.terminate_walk.path_openat.do_filp_open.do_sys_openat2.__x64_sys_openat
0.00 +3.9 3.92 ± 58% perf-profile.calltrace.cycles-pp.lockref_get_not_dead.__legitimize_path.try_to_unlazy.link_path_walk.path_openat
0.00 +4.0 3.95 ± 58% perf-profile.calltrace.cycles-pp.__legitimize_path.try_to_unlazy.link_path_walk.path_openat.do_filp_open
0.00 +4.0 3.96 ± 57% perf-profile.calltrace.cycles-pp.try_to_unlazy.link_path_walk.path_openat.do_filp_open.do_sys_openat2
0.00 +4.1 4.12 ± 77% perf-profile.calltrace.cycles-pp.zap_pte_range.zap_pmd_range.unmap_page_range.unmap_vmas.vms_clear_ptes
0.00 +4.1 4.13 ± 77% perf-profile.calltrace.cycles-pp.zap_pmd_range.unmap_page_range.unmap_vmas.vms_clear_ptes.vms_complete_munmap_vmas
0.00 +4.1 4.14 ± 77% perf-profile.calltrace.cycles-pp.unmap_page_range.unmap_vmas.vms_clear_ptes.vms_complete_munmap_vmas.do_vmi_align_munmap
0.00 +4.3 4.32 ± 77% perf-profile.calltrace.cycles-pp.unmap_vmas.vms_clear_ptes.vms_complete_munmap_vmas.do_vmi_align_munmap.do_vmi_munmap
0.00 +5.0 4.95 ± 46% perf-profile.calltrace.cycles-pp.__mmap_new_vma.__mmap_region.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff
0.00 +5.0 5.04 ± 77% perf-profile.calltrace.cycles-pp.vms_clear_ptes.vms_complete_munmap_vmas.do_vmi_align_munmap.do_vmi_munmap.__vm_munmap
0.00 +5.2 5.23 ± 77% perf-profile.calltrace.cycles-pp.vms_complete_munmap_vmas.do_vmi_align_munmap.do_vmi_munmap.__vm_munmap.__x64_sys_munmap
0.00 +5.7 5.70 ± 49% perf-profile.calltrace.cycles-pp.__mmap_region.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64
0.00 +6.2 6.16 ± 77% perf-profile.calltrace.cycles-pp.do_vmi_align_munmap.do_vmi_munmap.__vm_munmap.__x64_sys_munmap.do_syscall_64
0.00 +6.2 6.19 ± 77% perf-profile.calltrace.cycles-pp.do_vmi_munmap.__vm_munmap.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 +6.2 6.21 ± 77% perf-profile.calltrace.cycles-pp.__vm_munmap.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe.__munmap
0.00 +6.2 6.21 ± 77% perf-profile.calltrace.cycles-pp.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe.__munmap
0.00 +6.5 6.53 ± 47% perf-profile.calltrace.cycles-pp.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 +6.6 6.61 ± 48% perf-profile.calltrace.cycles-pp.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe.__mmap
0.00 +6.6 6.64 ± 76% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__munmap
0.00 +6.6 6.65 ± 76% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__munmap
0.00 +6.7 6.66 ± 47% perf-profile.calltrace.cycles-pp.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe.__mmap
0.00 +6.7 6.67 ± 76% perf-profile.calltrace.cycles-pp.__munmap
0.00 +6.7 6.73 ± 47% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__mmap
0.00 +6.7 6.73 ± 47% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__mmap
0.00 +6.9 6.89 ± 47% perf-profile.calltrace.cycles-pp.__mmap
0.00 +9.4 9.42 ± 18% perf-profile.calltrace.cycles-pp.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.__open64_nocancel
0.00 +9.4 9.42 ± 18% perf-profile.calltrace.cycles-pp.do_sys_openat2.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.__open64_nocancel
0.00 +9.5 9.46 ± 18% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__open64_nocancel
0.00 +9.5 9.46 ± 18% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__open64_nocancel
0.00 +9.5 9.49 ± 17% perf-profile.calltrace.cycles-pp.__open64_nocancel
0.00 +10.1 10.12 ± 36% perf-profile.calltrace.cycles-pp.up_read.kernfs_iop_permission.inode_permission.link_path_walk.path_openat
0.00 +12.5 12.52 ± 41% perf-profile.calltrace.cycles-pp.down_read.kernfs_iop_permission.inode_permission.link_path_walk.path_openat
0.00 +12.5 12.52 ± 71% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.dput.step_into.link_path_walk
0.00 +12.6 12.64 ± 71% perf-profile.calltrace.cycles-pp._raw_spin_lock.dput.step_into.link_path_walk.path_openat
0.00 +13.1 13.09 ± 72% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.__d_lookup.lookup_fast.walk_component
0.00 +14.1 14.06 ± 61% perf-profile.calltrace.cycles-pp.dput.step_into.link_path_walk.path_openat.do_filp_open
0.00 +14.3 14.26 ± 64% perf-profile.calltrace.cycles-pp._raw_spin_lock.__d_lookup.lookup_fast.walk_component.link_path_walk
0.00 +14.4 14.41 ± 63% perf-profile.calltrace.cycles-pp.__d_lookup.lookup_fast.walk_component.link_path_walk.path_openat
0.00 +14.8 14.80 ± 61% perf-profile.calltrace.cycles-pp.step_into.link_path_walk.path_openat.do_filp_open.do_sys_openat2
0.00 +18.2 18.21 ± 46% perf-profile.calltrace.cycles-pp.lookup_fast.walk_component.link_path_walk.path_openat.do_filp_open
0.00 +18.3 18.25 ± 46% perf-profile.calltrace.cycles-pp.walk_component.link_path_walk.path_openat.do_filp_open.do_sys_openat2
0.00 +23.8 23.84 ± 38% perf-profile.calltrace.cycles-pp.kernfs_iop_permission.inode_permission.link_path_walk.path_openat.do_filp_open
0.00 +26.6 26.62 ± 38% perf-profile.calltrace.cycles-pp.inode_permission.link_path_walk.path_openat.do_filp_open.do_sys_openat2
2.39 ± 7% +63.7 66.05 ± 15% perf-profile.calltrace.cycles-pp.do_sys_openat2.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.open64
2.39 ± 7% +63.7 66.05 ± 15% perf-profile.calltrace.cycles-pp.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.open64
2.42 ± 7% +63.8 66.23 ± 15% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.open64
2.42 ± 7% +63.8 66.25 ± 15% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.open64
0.10 ±223% +63.9 64.00 ± 19% perf-profile.calltrace.cycles-pp.link_path_walk.path_openat.do_filp_open.do_sys_openat2.__x64_sys_openat
2.48 ± 7% +64.0 66.44 ± 15% perf-profile.calltrace.cycles-pp.open64
2.30 ± 8% +72.9 75.16 ± 15% perf-profile.calltrace.cycles-pp.path_openat.do_filp_open.do_sys_openat2.__x64_sys_openat.do_syscall_64
2.30 ± 8% +72.9 75.17 ± 15% perf-profile.calltrace.cycles-pp.do_filp_open.do_sys_openat2.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe
66.48 ± 5% -64.6 1.84 ± 58% perf-profile.children.cycles-pp.__cmd_record
62.50 ± 6% -61.1 1.42 ± 74% perf-profile.children.cycles-pp.record__finish_output
62.50 ± 6% -61.1 1.42 ± 74% perf-profile.children.cycles-pp.perf_session__process_events
62.34 ± 6% -60.9 1.42 ± 74% perf-profile.children.cycles-pp.reader__read_event
40.48 ± 8% -39.4 1.07 ± 78% perf-profile.children.cycles-pp.process_simple
38.11 ± 8% -37.1 1.01 ± 78% perf-profile.children.cycles-pp.ordered_events__queue
37.76 ± 8% -36.8 1.00 ± 78% perf-profile.children.cycles-pp.queue_event
21.66 ± 8% -21.3 0.37 ± 66% perf-profile.children.cycles-pp.__ordered_events__flush
21.58 ± 8% -21.2 0.37 ± 66% perf-profile.children.cycles-pp.perf_session__process_user_event
21.37 ± 8% -21.0 0.37 ± 68% perf-profile.children.cycles-pp.perf_session__deliver_event
19.36 ± 8% -18.9 0.41 ± 51% perf-profile.children.cycles-pp.handle_internal_command
19.36 ± 8% -18.9 0.41 ± 51% perf-profile.children.cycles-pp.main
19.36 ± 8% -18.9 0.41 ± 51% perf-profile.children.cycles-pp.run_builtin
18.49 ± 8% -18.1 0.39 ± 51% perf-profile.children.cycles-pp.cmd_record
18.32 ± 8% -18.0 0.34 ± 58% perf-profile.children.cycles-pp.perf_c2c__record
14.08 ± 18% -13.3 0.78 ± 56% perf-profile.children.cycles-pp.common_startup_64
14.08 ± 18% -13.3 0.78 ± 56% perf-profile.children.cycles-pp.cpu_startup_entry
14.08 ± 18% -13.3 0.78 ± 56% perf-profile.children.cycles-pp.do_idle
13.13 ± 8% -12.9 0.20 ± 62% perf-profile.children.cycles-pp.machine__process_mmap2_event
11.85 ± 21% -11.2 0.65 ± 58% perf-profile.children.cycles-pp.start_secondary
11.91 ± 17% -11.2 0.75 ± 57% perf-profile.children.cycles-pp.cpuidle_idle_call
11.03 ± 17% -10.3 0.71 ± 57% perf-profile.children.cycles-pp.cpuidle_enter
11.03 ± 16% -10.3 0.71 ± 57% perf-profile.children.cycles-pp.cpuidle_enter_state
9.17 ± 16% -8.5 0.67 ± 58% perf-profile.children.cycles-pp.acpi_idle_enter
9.16 ± 16% -8.5 0.67 ± 59% perf-profile.children.cycles-pp.acpi_idle_do_entry
9.16 ± 16% -8.5 0.67 ± 58% perf-profile.children.cycles-pp.pv_native_safe_halt
9.16 ± 16% -8.5 0.67 ± 59% perf-profile.children.cycles-pp.acpi_safe_halt
6.48 ± 9% -6.3 0.13 ± 86% perf-profile.children.cycles-pp.evlist__parse_sample
6.10 ± 10% -6.0 0.09 ± 59% perf-profile.children.cycles-pp.map__new
5.98 ± 5% -5.6 0.40 ± 53% perf-profile.children.cycles-pp.ret_from_fork_asm
5.97 ± 5% -5.6 0.40 ± 53% perf-profile.children.cycles-pp.ret_from_fork
5.92 ± 5% -5.5 0.40 ± 53% perf-profile.children.cycles-pp.kthread
6.25 ± 19% -5.0 1.29 ± 37% perf-profile.children.cycles-pp.asm_sysvec_apic_timer_interrupt
5.05 ± 9% -4.9 0.10 ±105% perf-profile.children.cycles-pp.evlist__event2evsel
3.61 ± 11% -3.6 0.04 ± 83% perf-profile.children.cycles-pp.__dsos__find_by_longname_id
3.54 ± 13% -3.5 0.04 ± 83% perf-profile.children.cycles-pp.maps__fixup_overlap_and_insert
4.53 ± 20% -3.3 1.18 ± 35% perf-profile.children.cycles-pp.sysvec_apic_timer_interrupt
3.47 ± 14% -3.3 0.17 ± 50% perf-profile.children.cycles-pp._nohz_idle_balance
3.37 ± 12% -3.1 0.24 ± 45% perf-profile.children.cycles-pp.asm_sysvec_call_function_single
3.33 ± 5% -3.1 0.20 ± 60% perf-profile.children.cycles-pp.worker_thread
3.61 ± 10% -3.0 0.59 ± 55% perf-profile.children.cycles-pp.__irq_exit_rcu
3.70 ± 10% -3.0 0.73 ± 59% perf-profile.children.cycles-pp.handle_softirqs
2.95 ± 10% -2.9 0.04 ± 83% perf-profile.children.cycles-pp.dsos__findnew_id
3.03 ± 7% -2.9 0.13 ± 65% perf-profile.children.cycles-pp.process_one_work
2.89 ± 14% -2.9 0.04 ± 83% perf-profile.children.cycles-pp.__maps__fixup_overlap_and_insert
3.04 ± 10% -2.8 0.22 ± 46% perf-profile.children.cycles-pp.sysvec_call_function_single
2.80 ± 9% -2.7 0.07 ± 93% perf-profile.children.cycles-pp.read
2.86 ± 6% -2.6 0.23 ± 82% perf-profile.children.cycles-pp.write
2.70 ± 11% -2.6 0.07 ± 93% perf-profile.children.cycles-pp.ksys_read
2.79 ± 6% -2.6 0.21 ± 81% perf-profile.children.cycles-pp.ksys_write
2.78 ± 6% -2.6 0.20 ± 80% perf-profile.children.cycles-pp.vfs_write
2.61 ± 11% -2.5 0.07 ± 91% perf-profile.children.cycles-pp.vfs_read
2.60 ± 8% -2.5 0.05 ± 91% perf-profile.children.cycles-pp.asm_exc_page_fault
2.54 ± 11% -2.4 0.17 ± 32% perf-profile.children.cycles-pp.__schedule
2.41 ± 9% -2.4 0.04 ± 86% perf-profile.children.cycles-pp.machine__findnew_thread
2.32 ± 9% -2.3 0.04 ± 90% perf-profile.children.cycles-pp.evsel__parse_sample
2.27 ± 12% -2.2 0.06 ± 93% perf-profile.children.cycles-pp.seq_read_iter
2.24 ± 10% -2.2 0.05 ± 93% perf-profile.children.cycles-pp.exc_page_fault
2.23 ± 10% -2.2 0.05 ± 93% perf-profile.children.cycles-pp.do_user_addr_fault
2.32 ± 8% -2.1 0.20 ± 39% perf-profile.children.cycles-pp.sched_balance_rq
2.23 ± 12% -2.1 0.12 ± 66% perf-profile.children.cycles-pp.rest_init
2.23 ± 12% -2.1 0.12 ± 66% perf-profile.children.cycles-pp.start_kernel
2.23 ± 12% -2.1 0.12 ± 66% perf-profile.children.cycles-pp.x86_64_start_kernel
2.23 ± 12% -2.1 0.12 ± 66% perf-profile.children.cycles-pp.x86_64_start_reservations
2.78 ± 22% -2.1 0.73 ± 28% perf-profile.children.cycles-pp.__sysvec_apic_timer_interrupt
2.73 ± 23% -2.0 0.72 ± 28% perf-profile.children.cycles-pp.hrtimer_interrupt
2.06 ± 11% -2.0 0.05 ± 90% perf-profile.children.cycles-pp.handle_mm_fault
2.00 ± 11% -2.0 0.04 ± 93% perf-profile.children.cycles-pp.__handle_mm_fault
2.09 ± 9% -1.9 0.17 ± 42% perf-profile.children.cycles-pp.sched_balance_find_src_group
2.06 ± 13% -1.9 0.15 ± 30% perf-profile.children.cycles-pp.schedule
2.07 ± 9% -1.9 0.17 ± 42% perf-profile.children.cycles-pp.update_sd_lb_stats
1.93 ± 10% -1.8 0.16 ± 41% perf-profile.children.cycles-pp.update_sg_lb_stats
1.95 ± 6% -1.7 0.22 ± 84% perf-profile.children.cycles-pp.writen
1.91 ± 6% -1.7 0.19 ± 80% perf-profile.children.cycles-pp.shmem_file_write_iter
1.89 ± 6% -1.7 0.18 ± 78% perf-profile.children.cycles-pp.generic_perform_write
1.80 ± 10% -1.7 0.13 ± 33% perf-profile.children.cycles-pp.__pick_next_task
1.86 ± 6% -1.6 0.22 ± 84% perf-profile.children.cycles-pp.record__pushfn
1.82 ± 14% -1.6 0.17 ± 72% perf-profile.children.cycles-pp.smpboot_thread_fn
1.95 ± 6% -1.6 0.35 ± 99% perf-profile.children.cycles-pp.perf_mmap__push
1.95 ± 6% -1.6 0.36 ±100% perf-profile.children.cycles-pp.record__mmap_read_evlist
1.58 ± 16% -1.5 0.05 ± 93% perf-profile.children.cycles-pp.seq_read
1.59 ± 11% -1.5 0.12 ± 35% perf-profile.children.cycles-pp.sched_balance_newidle
1.87 ± 23% -1.3 0.53 ± 24% perf-profile.children.cycles-pp.__hrtimer_run_queues
1.37 ± 7% -1.2 0.13 ± 79% perf-profile.children.cycles-pp.copy_page_from_iter_atomic
1.69 ± 24% -1.2 0.51 ± 24% perf-profile.children.cycles-pp.tick_nohz_handler
1.17 ± 14% -1.1 0.09 ± 39% perf-profile.children.cycles-pp.vfs_fstatat
1.09 ± 5% -1.0 0.07 ± 91% perf-profile.children.cycles-pp.sched_balance_domains
1.04 ± 21% -1.0 0.05 ± 64% perf-profile.children.cycles-pp.raw_spin_rq_lock_nested
1.00 ± 21% -1.0 0.04 ± 88% perf-profile.children.cycles-pp.idle_cpu
1.03 ± 10% -0.9 0.18 ± 68% perf-profile.children.cycles-pp._raw_spin_lock_irqsave
0.80 ± 12% -0.8 0.05 ± 88% perf-profile.children.cycles-pp.sched_balance_update_blocked_averages
1.10 ± 12% -0.7 0.38 ± 39% perf-profile.children.cycles-pp.__alloc_frozen_pages_noprof
1.01 ± 11% -0.7 0.32 ± 37% perf-profile.children.cycles-pp.get_page_from_freelist
0.99 ± 10% -0.6 0.36 ± 43% perf-profile.children.cycles-pp.alloc_pages_mpol
1.04 ± 30% -0.6 0.45 ± 20% perf-profile.children.cycles-pp.update_process_times
0.72 ± 27% -0.6 0.16 ± 48% perf-profile.children.cycles-pp.ktime_get
0.62 ± 17% -0.5 0.11 ± 27% perf-profile.children.cycles-pp.pick_next_task_fair
0.55 ± 18% -0.5 0.04 ± 53% perf-profile.children.cycles-pp.update_load_avg
0.76 ± 9% -0.5 0.25 ± 40% perf-profile.children.cycles-pp.clear_page_erms
0.62 ± 28% -0.5 0.15 ± 42% perf-profile.children.cycles-pp.clockevents_program_event
0.45 ± 34% -0.2 0.22 ± 20% perf-profile.children.cycles-pp.sched_tick
0.25 ± 25% -0.2 0.04 ± 83% perf-profile.children.cycles-pp.rmqueue
0.28 ± 44% -0.2 0.09 ± 28% perf-profile.children.cycles-pp.tmigr_requires_handle_remote
0.20 ± 36% -0.2 0.04 ± 83% perf-profile.children.cycles-pp.__rmqueue_pcplist
0.24 ± 19% -0.2 0.08 ±106% perf-profile.children.cycles-pp.tracing_buffers_open
0.17 ± 50% -0.1 0.05 ± 90% perf-profile.children.cycles-pp.__memset
0.05 ± 72% +0.1 0.10 ± 11% perf-profile.children.cycles-pp.task_tick_fair
0.00 +0.1 0.06 ± 14% perf-profile.children.cycles-pp.syscall_return_via_sysret
0.00 +0.1 0.08 ± 33% perf-profile.children.cycles-pp.readdir64
0.00 +0.1 0.08 ± 40% perf-profile.children.cycles-pp.mas_ascend
0.00 +0.1 0.10 ± 57% perf-profile.children.cycles-pp.kmem_cache_free_bulk
0.00 +0.1 0.11 ± 41% perf-profile.children.cycles-pp.cp_new_stat
0.00 +0.1 0.11 ± 53% perf-profile.children.cycles-pp.simple_getattr
0.04 ± 73% +0.1 0.16 ± 76% perf-profile.children.cycles-pp.up_write
0.02 ±142% +0.1 0.14 ± 58% perf-profile.children.cycles-pp.path_init
0.00 +0.1 0.13 ± 66% perf-profile.children.cycles-pp.__rb_insert_augmented
0.00 +0.1 0.13 ± 60% perf-profile.children.cycles-pp.nd_jump_root
0.06 ± 74% +0.1 0.19 ± 43% perf-profile.children.cycles-pp.alloc_empty_file
0.00 +0.1 0.13 ± 89% perf-profile.children.cycles-pp.tlb_remove_table_rcu
0.02 ±142% +0.1 0.15 ± 45% perf-profile.children.cycles-pp.vm_area_alloc
0.00 +0.1 0.14 ± 38% perf-profile.children.cycles-pp.filldir64
0.00 +0.1 0.14 ± 54% perf-profile.children.cycles-pp.unmap_region
0.01 ±223% +0.1 0.16 ± 39% perf-profile.children.cycles-pp.mas_find
0.00 +0.1 0.15 ± 67% perf-profile.children.cycles-pp.pfn_modify_allowed
0.02 ±142% +0.1 0.17 ± 95% perf-profile.children.cycles-pp.kernfs_unlink_open_file
0.11 ± 20% +0.2 0.27 ± 42% perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack
0.02 ±144% +0.2 0.19 ± 45% perf-profile.children.cycles-pp.rb_next
0.00 +0.2 0.18 ± 33% perf-profile.children.cycles-pp.vma_migratable
0.02 ±142% +0.2 0.20 ± 32% perf-profile.children.cycles-pp.lockref_get
0.00 +0.2 0.19 ± 55% perf-profile.children.cycles-pp.generic_fillattr
0.13 ± 40% +0.2 0.32 ± 48% perf-profile.children.cycles-pp.alloc_pages_noprof
0.00 +0.2 0.19 ± 33% perf-profile.children.cycles-pp.task_numa_work
0.13 ± 22% +0.2 0.33 ± 41% perf-profile.children.cycles-pp.entry_SYSCALL_64
0.05 ±109% +0.2 0.26 ± 72% perf-profile.children.cycles-pp.mas_wr_bnode
0.00 +0.2 0.21 ± 82% perf-profile.children.cycles-pp.vma_interval_tree_remove
0.03 ±141% +0.2 0.24 ± 83% perf-profile.children.cycles-pp.kernfs_fop_release
0.00 +0.2 0.22 ± 74% perf-profile.children.cycles-pp.tracing_open_file_tr
0.00 +0.2 0.25 ± 76% perf-profile.children.cycles-pp.__kmem_cache_alloc_bulk
0.00 +0.3 0.26 ± 73% perf-profile.children.cycles-pp.kmem_cache_alloc_bulk_noprof
0.06 ± 73% +0.3 0.33 ± 54% perf-profile.children.cycles-pp.mas_preallocate
0.05 ± 79% +0.3 0.35 ± 49% perf-profile.children.cycles-pp.pte_alloc_one
0.00 +0.4 0.36 ± 48% perf-profile.children.cycles-pp.__pte_alloc
0.02 ±142% +0.4 0.43 ± 67% perf-profile.children.cycles-pp.mas_alloc_nodes
0.00 +0.4 0.44 ± 45% perf-profile.children.cycles-pp.mas_rev_awalk
0.27 ± 14% +0.5 0.76 ± 37% perf-profile.children.cycles-pp.syscall_exit_to_user_mode
0.01 ±223% +0.5 0.50 ± 52% perf-profile.children.cycles-pp.generic_permission
0.07 ± 47% +0.5 0.59 ± 34% perf-profile.children.cycles-pp.task_work_run
0.00 +0.6 0.56 ± 84% perf-profile.children.cycles-pp.vm_normal_page
0.00 +0.6 0.57 ± 44% perf-profile.children.cycles-pp.mas_empty_area_rev
0.14 ± 25% +0.6 0.74 ± 66% perf-profile.children.cycles-pp.kernfs_fop_open
0.00 +0.6 0.62 ± 78% perf-profile.children.cycles-pp.mas_store_gfp
0.02 ±142% +0.7 0.67 ± 92% perf-profile.children.cycles-pp.scan_positives
0.01 ±223% +0.7 0.66 ± 44% perf-profile.children.cycles-pp.unmapped_area_topdown
0.01 ±223% +0.7 0.67 ± 44% perf-profile.children.cycles-pp.vm_unmapped_area
0.07 ± 73% +0.7 0.73 ± 74% perf-profile.children.cycles-pp.free_pgtables
0.00 +0.7 0.67 ±120% perf-profile.children.cycles-pp.__traverse_mounts
0.01 ±223% +0.7 0.69 ± 76% perf-profile.children.cycles-pp.unlink_file_vma_batch_process
0.00 +0.7 0.69 ± 53% perf-profile.children.cycles-pp.set_nlink
0.01 ±223% +0.7 0.73 ± 43% perf-profile.children.cycles-pp.arch_get_unmapped_area_topdown
0.02 ±141% +0.7 0.75 ± 43% perf-profile.children.cycles-pp.__get_unmapped_area
0.06 ± 77% +0.8 0.82 ± 83% perf-profile.children.cycles-pp.dcache_readdir
1.20 ± 11% +0.8 2.01 ± 41% perf-profile.children.cycles-pp.do_dentry_open
0.00 +0.8 0.83 ± 50% perf-profile.children.cycles-pp.kernfs_fop_readdir
1.20 ± 11% +0.9 2.07 ± 42% perf-profile.children.cycles-pp.vfs_open
0.00 +1.0 1.02 ± 60% perf-profile.children.cycles-pp.remap_pfn_range_internal
0.00 +1.0 1.04 ± 60% perf-profile.children.cycles-pp.remap_pfn_range
0.00 +1.0 1.05 ± 60% perf-profile.children.cycles-pp.kernfs_fop_mmap
0.47 ± 8% +1.1 1.60 ± 46% perf-profile.children.cycles-pp.open_last_lookups
0.16 ± 65% +1.5 1.66 ± 51% perf-profile.children.cycles-pp.vma_interval_tree_insert
0.10 ± 36% +1.8 1.90 ± 53% perf-profile.children.cycles-pp.iterate_dir
0.10 ± 36% +1.8 1.92 ± 53% perf-profile.children.cycles-pp.__x64_sys_getdents64
0.10 ± 35% +1.9 1.99 ± 52% perf-profile.children.cycles-pp.getdents64
1.24 ± 12% +2.1 3.35 ± 35% perf-profile.children.cycles-pp.__do_sys_newfstatat
1.30 ± 12% +2.3 3.62 ± 35% perf-profile.children.cycles-pp.fstatat64
0.08 ± 51% +2.5 2.54 ± 42% perf-profile.children.cycles-pp.lockref_put_return
0.21 ± 26% +2.8 3.03 ± 40% perf-profile.children.cycles-pp.may_open
0.00 +2.9 2.86 ± 37% perf-profile.children.cycles-pp.kernfs_iop_getattr
0.05 ± 50% +3.0 3.10 ± 36% perf-profile.children.cycles-pp.vfs_getattr_nosec
0.17 ± 40% +3.1 3.28 ± 76% perf-profile.children.cycles-pp.zap_present_ptes
0.02 ±142% +3.1 3.14 ± 36% perf-profile.children.cycles-pp.vfs_fstat
1.45 ± 7% +3.7 5.20 ± 39% perf-profile.children.cycles-pp.do_open
0.23 ± 39% +3.9 4.17 ± 77% perf-profile.children.cycles-pp.zap_pte_range
0.23 ± 40% +3.9 4.18 ± 77% perf-profile.children.cycles-pp.zap_pmd_range
0.24 ± 39% +4.0 4.21 ± 76% perf-profile.children.cycles-pp.unmap_page_range
0.07 ± 61% +4.0 4.03 ± 53% perf-profile.children.cycles-pp.terminate_walk
0.10 ± 59% +4.0 4.15 ± 51% perf-profile.children.cycles-pp.try_to_unlazy
0.24 ± 30% +4.1 4.32 ± 36% perf-profile.children.cycles-pp.kernfs_dop_revalidate
0.07 ± 85% +4.1 4.19 ± 51% perf-profile.children.cycles-pp.lockref_get_not_dead
0.10 ± 66% +4.1 4.21 ± 50% perf-profile.children.cycles-pp.__legitimize_path
0.26 ± 36% +4.2 4.41 ± 76% perf-profile.children.cycles-pp.unmap_vmas
0.24 ± 25% +4.7 4.96 ± 46% perf-profile.children.cycles-pp.__mmap_new_vma
0.06 ± 84% +5.0 5.04 ± 77% perf-profile.children.cycles-pp.vms_clear_ptes
0.56 ± 19% +5.2 5.71 ± 49% perf-profile.children.cycles-pp.__mmap_region
0.04 ±110% +5.2 5.23 ± 77% perf-profile.children.cycles-pp.vms_complete_munmap_vmas
0.60 ± 18% +5.9 6.54 ± 47% perf-profile.children.cycles-pp.do_mmap
0.61 ± 19% +6.0 6.62 ± 47% perf-profile.children.cycles-pp.vm_mmap_pgoff
0.08 ± 67% +6.1 6.18 ± 77% perf-profile.children.cycles-pp.do_vmi_align_munmap
0.09 ± 52% +6.1 6.19 ± 77% perf-profile.children.cycles-pp.do_vmi_munmap
0.08 ± 68% +6.1 6.21 ± 77% perf-profile.children.cycles-pp.__vm_munmap
0.04 ±141% +6.2 6.21 ± 77% perf-profile.children.cycles-pp.__x64_sys_munmap
0.46 ± 22% +6.2 6.66 ± 47% perf-profile.children.cycles-pp.ksys_mmap_pgoff
0.04 ±151% +6.6 6.68 ± 76% perf-profile.children.cycles-pp.__munmap
0.21 ± 16% +6.7 6.92 ± 47% perf-profile.children.cycles-pp.__mmap
0.22 ± 21% +9.3 9.50 ± 17% perf-profile.children.cycles-pp.__open64_nocancel
0.09 ± 29% +12.1 12.17 ± 32% perf-profile.children.cycles-pp.up_read
0.46 ± 16% +14.6 15.04 ± 60% perf-profile.children.cycles-pp.step_into
0.41 ± 9% +14.6 15.03 ± 59% perf-profile.children.cycles-pp.__d_lookup
0.88 ± 15% +17.4 18.28 ± 46% perf-profile.children.cycles-pp.walk_component
1.19 ± 11% +18.5 19.67 ± 40% perf-profile.children.cycles-pp.lookup_fast
0.19 ± 21% +18.8 18.96 ± 58% perf-profile.children.cycles-pp.dput
0.08 ± 85% +20.2 20.29 ± 38% perf-profile.children.cycles-pp.down_read
0.12 ± 30% +26.5 26.60 ± 38% perf-profile.children.cycles-pp.kernfs_iop_permission
0.35 ± 28% +29.1 29.46 ± 38% perf-profile.children.cycles-pp.inode_permission
0.46 ± 18% +32.9 33.37 ± 70% perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath
1.51 ± 19% +33.5 35.02 ± 65% perf-profile.children.cycles-pp._raw_spin_lock
1.12 ± 13% +62.9 64.03 ± 19% perf-profile.children.cycles-pp.link_path_walk
2.48 ± 7% +64.0 66.47 ± 15% perf-profile.children.cycles-pp.open64
2.63 ± 7% +72.6 75.19 ± 15% perf-profile.children.cycles-pp.path_openat
2.63 ± 7% +72.6 75.20 ± 15% perf-profile.children.cycles-pp.do_filp_open
2.76 ± 7% +72.7 75.50 ± 15% perf-profile.children.cycles-pp.do_sys_openat2
2.76 ± 7% +72.7 75.51 ± 15% perf-profile.children.cycles-pp.__x64_sys_openat
14.50 ± 6% +81.5 95.98 ± 2% perf-profile.children.cycles-pp.do_syscall_64
14.51 ± 6% +81.5 96.03 ± 2% perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
37.26 ± 8% -36.3 0.99 ± 78% perf-profile.self.cycles-pp.queue_event
4.85 ± 9% -4.8 0.10 ±104% perf-profile.self.cycles-pp.evlist__event2evsel
3.19 ± 10% -3.2 0.04 ± 83% perf-profile.self.cycles-pp.__dsos__find_by_longname_id
2.80 ± 13% -2.8 0.04 ± 83% perf-profile.self.cycles-pp.__maps__fixup_overlap_and_insert
2.29 ± 9% -2.2 0.04 ± 90% perf-profile.self.cycles-pp.evsel__parse_sample
1.98 ± 20% -1.9 0.09 ±108% perf-profile.self.cycles-pp.pv_native_safe_halt
1.39 ± 9% -1.3 0.13 ± 43% perf-profile.self.cycles-pp.update_sg_lb_stats
1.32 ± 6% -1.2 0.13 ± 81% perf-profile.self.cycles-pp.copy_page_from_iter_atomic
0.99 ± 22% -1.0 0.04 ± 88% perf-profile.self.cycles-pp.idle_cpu
0.74 ± 10% -0.5 0.24 ± 40% perf-profile.self.cycles-pp.clear_page_erms
0.61 ± 27% -0.5 0.15 ± 45% perf-profile.self.cycles-pp.ktime_get
0.15 ± 52% -0.1 0.05 ± 90% perf-profile.self.cycles-pp.__memset
0.00 +0.1 0.06 ± 14% perf-profile.self.cycles-pp.syscall_return_via_sysret
0.00 +0.1 0.07 ± 37% perf-profile.self.cycles-pp.readdir64
0.00 +0.1 0.08 ± 42% perf-profile.self.cycles-pp.step_into
0.00 +0.1 0.08 ± 41% perf-profile.self.cycles-pp.mas_ascend
0.00 +0.1 0.10 ± 46% perf-profile.self.cycles-pp.__mmap_region
0.00 +0.1 0.10 ± 38% perf-profile.self.cycles-pp.filldir64
0.00 +0.1 0.13 ± 68% perf-profile.self.cycles-pp.__rb_insert_augmented
0.00 +0.1 0.14 ± 68% perf-profile.self.cycles-pp.pfn_modify_allowed
0.11 ± 20% +0.2 0.27 ± 41% perf-profile.self.cycles-pp.entry_SYSRETQ_unsafe_stack
0.02 ±142% +0.2 0.18 ± 33% perf-profile.self.cycles-pp.lockref_get
0.02 ±144% +0.2 0.19 ± 44% perf-profile.self.cycles-pp.rb_next
0.00 +0.2 0.17 ± 34% perf-profile.self.cycles-pp.vma_migratable
0.00 +0.2 0.18 ± 53% perf-profile.self.cycles-pp.generic_fillattr
0.00 +0.2 0.19 ± 74% perf-profile.self.cycles-pp.__kmem_cache_alloc_bulk
0.00 +0.2 0.20 ± 81% perf-profile.self.cycles-pp.vma_interval_tree_remove
0.00 +0.2 0.23 ± 53% perf-profile.self.cycles-pp.kernfs_fop_readdir
0.02 ±142% +0.2 0.27 ± 59% perf-profile.self.cycles-pp.__fput
0.06 ± 75% +0.3 0.32 ± 38% perf-profile.self.cycles-pp.do_dentry_open
0.00 +0.4 0.38 ± 45% perf-profile.self.cycles-pp.mas_rev_awalk
0.01 ±223% +0.5 0.49 ± 51% perf-profile.self.cycles-pp.generic_permission
0.00 +0.5 0.52 ± 66% perf-profile.self.cycles-pp.remap_pfn_range_internal
0.00 +0.5 0.52 ± 83% perf-profile.self.cycles-pp.vm_normal_page
0.00 +0.7 0.69 ± 52% perf-profile.self.cycles-pp.set_nlink
0.02 ±142% +0.8 0.82 ± 84% perf-profile.self.cycles-pp.zap_pte_range
0.07 ± 86% +1.0 1.07 ± 31% perf-profile.self.cycles-pp.lockref_get_not_dead
0.16 ± 65% +1.5 1.61 ± 51% perf-profile.self.cycles-pp.vma_interval_tree_insert
0.08 ± 50% +2.4 2.51 ± 42% perf-profile.self.cycles-pp.lockref_put_return
0.18 ± 21% +2.5 2.70 ± 57% perf-profile.self.cycles-pp.inode_permission
0.07 ± 74% +2.5 2.61 ± 73% perf-profile.self.cycles-pp.zap_present_ptes
0.08 ± 27% +12.0 12.10 ± 32% perf-profile.self.cycles-pp.up_read
0.08 ± 88% +19.8 19.91 ± 39% perf-profile.self.cycles-pp.down_read
0.46 ± 18% +32.7 33.19 ± 70% perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath
Disclaimer:
Results have been estimated based on internal Intel analysis and are provided
for informational purposes only. Any difference in system hardware or software
design or configuration may affect actual performance.
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
On Mon, 30 Jun 2025 18:04:40 -0400
Steven Rostedt <rostedt@goodmis.org> wrote:
> From: Steven Rostedt <rostedt@goodmis.org>
>
> When the ring buffer was first introduced, reading the non-consuming
> "trace" file required disabling the writing of the ring buffer. To make
> sure the writing was fully disabled before iterating the buffer with a
> non-consuming read, it would set the disable flag of the buffer and then
> call an RCU synchronization to make sure all the buffers were
> synchronized.
>
> The function ring_buffer_read_start() originally would initialize the
> iterator and call an RCU synchronization, but this was for each individual
> per CPU buffer where this would get called many times on a machine with
> many CPUs before the trace file could be read. The commit 72c9ddfd4c5bf
> ("ring-buffer: Make non-consuming read less expensive with lots of cpus.")
> separated ring_buffer_read_start into ring_buffer_read_prepare(),
> ring_buffer_read_sync() and then ring_buffer_read_start() to allow each of
> the per CPU buffers to be prepared, call the read_buffer_read_sync() once,
> and then the ring_buffer_read_start() for each of the CPUs which made
> things much faster.
>
> The commit 1039221cc278 ("ring-buffer: Do not disable recording when there
> is an iterator") removed the requirement of disabling the recording of the
> ring buffer in order to iterate it, but it did not remove the
> synchronization that was happening that was required to wait for all the
> buffers to have no more writers. It's now OK for the buffers to have
> writers and no synchronization is needed.
>
> Remove the synchronization and put back the interface for the ring buffer
> iterator back before commit 72c9ddfd4c5bf was applied.
>
Looks good to me.
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Thank you,
> Reported-by: David Howells <dhowells@redhat.com>
> Fixes: 1039221cc278 ("ring-buffer: Do not disable recording when there is an iterator")
> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
> ---
> include/linux/ring_buffer.h | 4 +--
> kernel/trace/ring_buffer.c | 63 ++++++-------------------------------
> kernel/trace/trace.c | 14 +++------
> kernel/trace/trace_kdb.c | 8 ++---
> 4 files changed, 18 insertions(+), 71 deletions(-)
>
> diff --git a/include/linux/ring_buffer.h b/include/linux/ring_buffer.h
> index cd7f0ae26615..bc90c3c7b5fd 100644
> --- a/include/linux/ring_buffer.h
> +++ b/include/linux/ring_buffer.h
> @@ -152,9 +152,7 @@ ring_buffer_consume(struct trace_buffer *buffer, int cpu, u64 *ts,
> unsigned long *lost_events);
>
> struct ring_buffer_iter *
> -ring_buffer_read_prepare(struct trace_buffer *buffer, int cpu, gfp_t flags);
> -void ring_buffer_read_prepare_sync(void);
> -void ring_buffer_read_start(struct ring_buffer_iter *iter);
> +ring_buffer_read_start(struct trace_buffer *buffer, int cpu, gfp_t flags);
> void ring_buffer_read_finish(struct ring_buffer_iter *iter);
>
> struct ring_buffer_event *
> diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
> index 00fc38d70e86..24bb5287c415 100644
> --- a/kernel/trace/ring_buffer.c
> +++ b/kernel/trace/ring_buffer.c
> @@ -5846,24 +5846,20 @@ ring_buffer_consume(struct trace_buffer *buffer, int cpu, u64 *ts,
> EXPORT_SYMBOL_GPL(ring_buffer_consume);
>
> /**
> - * ring_buffer_read_prepare - Prepare for a non consuming read of the buffer
> + * ring_buffer_read_start - start a non consuming read of the buffer
> * @buffer: The ring buffer to read from
> * @cpu: The cpu buffer to iterate over
> * @flags: gfp flags to use for memory allocation
> *
> - * This performs the initial preparations necessary to iterate
> - * through the buffer. Memory is allocated, buffer resizing
> - * is disabled, and the iterator pointer is returned to the caller.
> - *
> - * After a sequence of ring_buffer_read_prepare calls, the user is
> - * expected to make at least one call to ring_buffer_read_prepare_sync.
> - * Afterwards, ring_buffer_read_start is invoked to get things going
> - * for real.
> + * This creates an iterator to allow non-consuming iteration through
> + * the buffer. If the buffer is disabled for writing, it will produce
> + * the same information each time, but if the buffer is still writing
> + * then the first hit of a write will cause the iteration to stop.
> *
> - * This overall must be paired with ring_buffer_read_finish.
> + * Must be paired with ring_buffer_read_finish.
> */
> struct ring_buffer_iter *
> -ring_buffer_read_prepare(struct trace_buffer *buffer, int cpu, gfp_t flags)
> +ring_buffer_read_start(struct trace_buffer *buffer, int cpu, gfp_t flags)
> {
> struct ring_buffer_per_cpu *cpu_buffer;
> struct ring_buffer_iter *iter;
> @@ -5889,51 +5885,12 @@ ring_buffer_read_prepare(struct trace_buffer *buffer, int cpu, gfp_t flags)
>
> atomic_inc(&cpu_buffer->resize_disabled);
>
> - return iter;
> -}
> -EXPORT_SYMBOL_GPL(ring_buffer_read_prepare);
> -
> -/**
> - * ring_buffer_read_prepare_sync - Synchronize a set of prepare calls
> - *
> - * All previously invoked ring_buffer_read_prepare calls to prepare
> - * iterators will be synchronized. Afterwards, read_buffer_read_start
> - * calls on those iterators are allowed.
> - */
> -void
> -ring_buffer_read_prepare_sync(void)
> -{
> - synchronize_rcu();
> -}
> -EXPORT_SYMBOL_GPL(ring_buffer_read_prepare_sync);
> -
> -/**
> - * ring_buffer_read_start - start a non consuming read of the buffer
> - * @iter: The iterator returned by ring_buffer_read_prepare
> - *
> - * This finalizes the startup of an iteration through the buffer.
> - * The iterator comes from a call to ring_buffer_read_prepare and
> - * an intervening ring_buffer_read_prepare_sync must have been
> - * performed.
> - *
> - * Must be paired with ring_buffer_read_finish.
> - */
> -void
> -ring_buffer_read_start(struct ring_buffer_iter *iter)
> -{
> - struct ring_buffer_per_cpu *cpu_buffer;
> - unsigned long flags;
> -
> - if (!iter)
> - return;
> -
> - cpu_buffer = iter->cpu_buffer;
> -
> - raw_spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
> + guard(raw_spinlock_irqsave)(&cpu_buffer->reader_lock);
> arch_spin_lock(&cpu_buffer->lock);
> rb_iter_reset(iter);
> arch_spin_unlock(&cpu_buffer->lock);
> - raw_spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
> +
> + return iter;
> }
> EXPORT_SYMBOL_GPL(ring_buffer_read_start);
>
> diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
> index 95ae7c4e5835..7996f26c3f46 100644
> --- a/kernel/trace/trace.c
> +++ b/kernel/trace/trace.c
> @@ -4735,21 +4735,15 @@ __tracing_open(struct inode *inode, struct file *file, bool snapshot)
> if (iter->cpu_file == RING_BUFFER_ALL_CPUS) {
> for_each_tracing_cpu(cpu) {
> iter->buffer_iter[cpu] =
> - ring_buffer_read_prepare(iter->array_buffer->buffer,
> - cpu, GFP_KERNEL);
> - }
> - ring_buffer_read_prepare_sync();
> - for_each_tracing_cpu(cpu) {
> - ring_buffer_read_start(iter->buffer_iter[cpu]);
> + ring_buffer_read_start(iter->array_buffer->buffer,
> + cpu, GFP_KERNEL);
> tracing_iter_reset(iter, cpu);
> }
> } else {
> cpu = iter->cpu_file;
> iter->buffer_iter[cpu] =
> - ring_buffer_read_prepare(iter->array_buffer->buffer,
> - cpu, GFP_KERNEL);
> - ring_buffer_read_prepare_sync();
> - ring_buffer_read_start(iter->buffer_iter[cpu]);
> + ring_buffer_read_start(iter->array_buffer->buffer,
> + cpu, GFP_KERNEL);
> tracing_iter_reset(iter, cpu);
> }
>
> diff --git a/kernel/trace/trace_kdb.c b/kernel/trace/trace_kdb.c
> index d7b135de958a..896ff78b8349 100644
> --- a/kernel/trace/trace_kdb.c
> +++ b/kernel/trace/trace_kdb.c
> @@ -43,17 +43,15 @@ static void ftrace_dump_buf(int skip_entries, long cpu_file)
> if (cpu_file == RING_BUFFER_ALL_CPUS) {
> for_each_tracing_cpu(cpu) {
> iter.buffer_iter[cpu] =
> - ring_buffer_read_prepare(iter.array_buffer->buffer,
> - cpu, GFP_ATOMIC);
> - ring_buffer_read_start(iter.buffer_iter[cpu]);
> + ring_buffer_read_start(iter.array_buffer->buffer,
> + cpu, GFP_ATOMIC);
> tracing_iter_reset(&iter, cpu);
> }
> } else {
> iter.cpu_file = cpu_file;
> iter.buffer_iter[cpu_file] =
> - ring_buffer_read_prepare(iter.array_buffer->buffer,
> + ring_buffer_read_start(iter.array_buffer->buffer,
> cpu_file, GFP_ATOMIC);
> - ring_buffer_read_start(iter.buffer_iter[cpu_file]);
> tracing_iter_reset(&iter, cpu_file);
> }
>
> --
> 2.47.2
>
--
Masami Hiramatsu (Google) <mhiramat@kernel.org>
Steven Rostedt <rostedt@goodmis.org> wrote:
> From: Steven Rostedt <rostedt@goodmis.org>
>
> When the ring buffer was first introduced, reading the non-consuming
> "trace" file required disabling the writing of the ring buffer. To make
> sure the writing was fully disabled before iterating the buffer with a
> non-consuming read, it would set the disable flag of the buffer and then
> call an RCU synchronization to make sure all the buffers were
> synchronized.
>
> The function ring_buffer_read_start() originally would initialize the
> iterator and call an RCU synchronization, but this was for each individual
> per CPU buffer where this would get called many times on a machine with
> many CPUs before the trace file could be read. The commit 72c9ddfd4c5bf
> ("ring-buffer: Make non-consuming read less expensive with lots of cpus.")
> separated ring_buffer_read_start into ring_buffer_read_prepare(),
> ring_buffer_read_sync() and then ring_buffer_read_start() to allow each of
> the per CPU buffers to be prepared, call the read_buffer_read_sync() once,
> and then the ring_buffer_read_start() for each of the CPUs which made
> things much faster.
>
> The commit 1039221cc278 ("ring-buffer: Do not disable recording when there
> is an iterator") removed the requirement of disabling the recording of the
> ring buffer in order to iterate it, but it did not remove the
> synchronization that was happening that was required to wait for all the
> buffers to have no more writers. It's now OK for the buffers to have
> writers and no synchronization is needed.
>
> Remove the synchronization and put back the interface for the ring buffer
> iterator back before commit 72c9ddfd4c5bf was applied.
>
> Reported-by: David Howells <dhowells@redhat.com>
> Fixes: 1039221cc278 ("ring-buffer: Do not disable recording when there is an iterator")
> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Tested-by: David Howells <dhowells@redhat.com>
© 2016 - 2025 Red Hat, Inc.