Re: [PATCH 1/2] perf arm-spe: Add arm_spe_record to synthesized

From: Ali Saidi <alisaidi@amazon.com>
To: <german.gomez@arm.com>
Cc: <acme@kernel.org>, <alexander.shishkin@linux.intel.com>,
	<alisaidi@amazon.com>, <andrew.kilroy@arm.com>,
	<benh@kernel.crashing.org>, <james.clark@arm.com>,
	<john.garry@huawei.com>, <jolsa@redhat.com>, <leo.yan@linaro.org>,
	<linux-arm-kernel@lists.infradead.org>,
	<linux-kernel@vger.kernel.org>,
	<linux-perf-users@vger.kernel.org>, <mark.rutland@arm.com>,
	<mathieu.poirier@linaro.org>, <mingo@redhat.com>,
	<namhyung@kernel.org>, <peterz@infradead.org>, <will@kernel.org>
Subject: Re: [PATCH 1/2] perf arm-spe: Add arm_spe_record to synthesized
Date: Wed, 26 Jan 2022 15:58:20 +0000	[thread overview]
Message-ID: <20220126155820.21929-1-alisaidi@amazon.com> (raw)
In-Reply-To: <5ca990a5-eedd-0edb-26d3-b5e16c36ac34@arm.com>

Hi German,
>Hi Ali,
>
>On 25/01/2022 19:20, Ali Saidi wrote:
>> Providing the arm_spe_record as raw data to the synthesized SPE samples
>> allows perf scripts to read and separately process the data in ways
>> existing perf tools don't support and mirrors functionality available
>> for PEBS.
>> Signed-off-by: Ali Saidi <alisaidi@amazon.com>
>> ---
>>  tools/perf/util/arm-spe.c | 6 ++++++
>>  1 file changed, 6 insertions(+)
>>
>> diff --git a/tools/perf/util/arm-spe.c b/tools/perf/util/arm-spe.c
>> index d2b64e3f588b..a7499cde6fc0 100644
>> --- a/tools/perf/util/arm-spe.c
>> +++ b/tools/perf/util/arm-spe.c
>> @@ -336,6 +336,8 @@ static int arm_spe__synth_mem_sample(struct arm_spe_queue *speq,
>>  	sample.phys_addr = record->phys_addr;
>>  	sample.data_src = data_src;
>>  	sample.weight = record->latency;
>> +	sample.raw_size = sizeof(*record);
>> +	sample.raw_data = record;
>
>Have you tried this with perf-inject? I think it would need the PERF_SAMPLE_RAW bit in the sample_type,

Yes I've tried the following and it worked as expected with the original
perf.data or the perf.data.jitted after perf-inject. 

perf record -e arm_spe_0/jitter=1/ -k 1 java ...
perf  inject -f --jit -i perf.data -o perf.data.jitted
perf script -i perf.data -s t1.py --itrace=i1i

>
>Although I quickly looked over the perf inject code and it looks like it's expecting some type of padding:
>
>  // synthetic-events.c
>  if (type & PERF_SAMPLE_RAW) {
>    result += sizeof(u32);
>    result += sample->raw_size;
>  }
>
>I'm seeing some comments in utils/event.h related to this on the intel events.

Yes i noticed this too,but looking at how the raw data is added to the same
other places like intel-pt.c:1703 the perf_synth__raw*() functions are used to
strip away the 4 bytes bytes before the data is added to the sample. The other
places i can find the padding used is in builtin-script.c but given we have the
--dump-raw-trace option it's not clear to me that it's needed to wrap the
arm_spe_event in another struct with padding like perf_synth_intel_ptwrite?

Thanks,
Ali