Linux Perf Users
 help / color / mirror / Atom feed
From: "Mi, Dapeng" <dapeng1.mi@linux.intel.com>
To: "Falcon, Thomas" <thomas.falcon@intel.com>,
	"alexander.shishkin@linux.intel.com"
	<alexander.shishkin@linux.intel.com>,
	"peterz@infradead.org" <peterz@infradead.org>,
	"acme@kernel.org" <acme@kernel.org>,
	"mingo@redhat.com" <mingo@redhat.com>,
	"Hunter, Adrian" <adrian.hunter@intel.com>,
	"namhyung@kernel.org" <namhyung@kernel.org>,
	"Rogers, Ian" <irogers@google.com>
Cc: "Chen, Zide" <zide.chen@intel.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"linux-perf-users@vger.kernel.org"
	<linux-perf-users@vger.kernel.org>,
	"Mi, Dapeng1" <dapeng1.mi@intel.com>,
	"Hao, Xudong" <xudong.hao@intel.com>
Subject: Re: [Patch v3] perf tests: Add auto counter reload (ACR) sampling test
Date: Fri, 22 May 2026 09:10:28 +0800	[thread overview]
Message-ID: <ff569303-c30b-43b7-a3d2-2eb949893f30@linux.intel.com> (raw)
In-Reply-To: <29ad63162ca81262778d4f732c9bf0bcba23a838.camel@intel.com>


On 5/22/2026 2:01 AM, Falcon, Thomas wrote:
> On Tue, 2026-04-21 at 12:55 +0800, Dapeng Mi wrote:
>> Add auto counter reload sampling test to verify that the intended
>> event
>> records can be captured and the self-reloaded events won't generate
>> any
>> records.
>>
>> Additionally, enhance ratio_to_prev test to verify ACR sampling as
>> well
>> if ACR is supported.
>>
>> Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
>> ---
>>
>> V3: Move ratio-to-prev ACR sampling test into ratio_to_prev test.
>> V2:
>> https://lore.kernel.org/all/20260420025501.2133495-1-dapeng1.mi@linux.intel.com/
>> v1:
>> https://lore.kernel.org/all/20260413010920.546501-1-dapeng1.mi@linux.intel.com/
>>
>>  tools/perf/tests/shell/record.sh | 92
>> +++++++++++++++++++++++++++++++-
>>  1 file changed, 91 insertions(+), 1 deletion(-)
>>
>> diff --git a/tools/perf/tests/shell/record.sh
>> b/tools/perf/tests/shell/record.sh
>> index 7cb81cf3444a..091d8b0a998b 100755
>> --- a/tools/perf/tests/shell/record.sh
>> +++ b/tools/perf/tests/shell/record.sh
>> @@ -402,7 +402,52 @@ test_callgraph() {
>>    echo "Callgraph test [Success]"
>>  }
>>  
>> +test_acr_sampling() {
>> + 
>> events="{instructions/period=40000,acr_mask=0x2/u,cycles/period=20000
>> ,acr_mask=0x3/u}"
>> + 
>> pebs_events="{instructions/period=40000,acr_mask=0x2/pu,cycles/period
>> =20000,acr_mask=0x3/u}"
>> +  echo "Auto counter reload (ACR) sampling test"
>> +  if ! perf record -o "${perfdata}" -e "${events}" ${testprog} 2>
>> /dev/null
>> +  then
>> +    echo "Auto counter reload sampling [Skipped not supported]"
>> +    return
>> +  fi
>> +  if ! perf script -i "${perfdata}" -F event | grep -q
>> "instructions"
>> +  then
>> +    echo "Auto counter reload sampling [Failed missing instructions
>> event]"
>> +    err=1
>> +    return
>> +  fi
>> +  if perf script -i "${perfdata}" -F event | grep -q "cycles"
>> +  then
>> +    echo "Auto counter reload sampling [Failed cycles event
>> shouldn't be sampled]"
>> +    err=1
>> +    return
>> +  fi
>> +  if ! perf record -o "${perfdata}" -e "${pebs_events}" ${testprog}
>> 2> /dev/null
>> +  then
>> +    echo "Auto counter reload PEBS sampling [Skipped not supported]"
>> +    echo "Auto counter reload sampling [Success]"
> Should this "Success" message be here?

Yes, If we didn't add the "Success" message here, then no "Success" message
is printed if PEBS is not supported. Thanks.


>
> Thanks,
> Tom
>
>> +    return
>> +  fi
>> +  if ! perf script -i "${perfdata}" -F event | grep -q
>> "instructions"
>> +  then
>> +    echo "Auto counter reload PEBS sampling [Failed missing
>> instructions event]"
>> +    err=1
>> +    return
>> +  fi
>> +  if perf script -i "${perfdata}" -F event | grep -q "cycles"
>> +  then
>> +    echo "Auto counter reload PEBS sampling [Failed cycles event
>> shouldn't be sampled]"
>> +    err=1
>> +    return
>> +  fi
>> +  echo "Auto counter reload sampling [Success]"
>> +}
>> +
>>  test_ratio_to_prev() {
>> +  ratio_events="{instructions:u,cycles/period=20000,ratio-to-
>> prev=0.5/u}"
>> + 
>> p_core_ratio_events="{cpu_core/instructions/u,cpu_core/cycles,period=
>> 20000,ratio-to-prev=0.5/u}"
>> + 
>> e_core_ratio_events="{cpu_atom/instructions/u,cpu_atom/cycles,period=
>> 20000,ratio-to-prev=0.5/u}"
>>    echo "ratio-to-prev test"
>>    if ! perf record -o /dev/null -e "{instructions,
>> cycles/period=100000,ratio-to-prev=0.5/}" \
>>       true 2> /dev/null
>> @@ -438,7 +483,51 @@ test_ratio_to_prev() {
>>      err=1
>>      return
>>    fi
>> -  echo "Basic ratio-to-prev record test [Success]"
>> +
>> +  ratio_err=0
>> +  do_ratio_to_prev_test() {
>> +    local events=$1
>> +    local cpu=$2
>> +    if ! perf record -o "${perfdata}" -e "${events}" ${testprog} 2>
>> /dev/null
>> +    then
>> +      echo "Auto counter reload ${cpu} ratio-to-prev sampling
>> [Failed sampling]"
>> +      ratio_err=1
>> +      return
>> +    fi
>> +    if ! perf script -i "${perfdata}" -F event | grep -q
>> "instructions"
>> +    then
>> +      echo "Auto counter reload ${cpu} ratio-to-prev sampling
>> [Failed missing instructions event]"
>> +      ratio_err=1
>> +      return
>> +    fi
>> +    if perf script -i "${perfdata}" -F event | grep -q "cycles"
>> +    then
>> +      echo "Auto counter reload ${cpu} ratio-to-prev sampling
>> [Failed cycles event shouldn't be sampled]"
>> +      ratio_err=1
>> +      return
>> +    fi
>> +  }
>> +
>> +  # ratio-to-prev sampling would fallback to normal sampling if ACR
>> sampling
>> +  # is not supported, so acr_mask must be checked before testing
>> ratio-to-prev sampling.
>> +  if [ -e /sys/bus/event_source/devices/cpu/format/acr_mask ]
>> +  then
>> +    do_ratio_to_prev_test "${ratio_events}" ""
>> +  fi
>> +  if [ -e /sys/bus/event_source/devices/cpu_core/format/acr_mask ]
>> +  then
>> +    do_ratio_to_prev_test "${p_core_ratio_events}" "P-core"
>> +  fi
>> +  if [ -e /sys/bus/event_source/devices/cpu_atom/format/acr_mask ]
>> +  then
>> +    do_ratio_to_prev_test "${e_core_ratio_events}" "E-core"
>> +  fi
>> +  if test "$ratio_err" -eq 0
>> +  then
>> +    echo "Basic ratio-to-prev record test [Success]"
>> +  else
>> +    err="${ratio_err}"
>> +  fi
>>  }
>>  
>>  # raise the limit of file descriptors to minimum
>> @@ -457,6 +546,7 @@ test_leader_sampling
>>  test_topdown_leader_sampling
>>  test_precise_max
>>  test_callgraph
>> +test_acr_sampling
>>  test_ratio_to_prev
>>  
>>  # restore the default value

      reply	other threads:[~2026-05-22  1:10 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-21  4:55 [Patch v3] perf tests: Add auto counter reload (ACR) sampling test Dapeng Mi
2026-05-21  0:52 ` Mi, Dapeng
2026-05-21 18:01 ` Falcon, Thomas
2026-05-22  1:10   ` Mi, Dapeng [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ff569303-c30b-43b7-a3d2-2eb949893f30@linux.intel.com \
    --to=dapeng1.mi@linux.intel.com \
    --cc=acme@kernel.org \
    --cc=adrian.hunter@intel.com \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=dapeng1.mi@intel.com \
    --cc=irogers@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=namhyung@kernel.org \
    --cc=peterz@infradead.org \
    --cc=thomas.falcon@intel.com \
    --cc=xudong.hao@intel.com \
    --cc=zide.chen@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox