linux-kernel - Re: [PATCH v1 1/2] perf auxtrace: Add 'T' itrace option for timestamp trace

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <4bc9774a-ff11-b382-2904-cbceea4c538f@arm.com>
Date:   Thu, 19 Oct 2023 12:17:59 +0100
From:   James Clark <james.clark@....com>
To:     Adrian Hunter <adrian.hunter@...el.com>,
        Leo Yan <leo.yan@...aro.org>,
        Arnaldo Carvalho de Melo <acme@...nel.org>,
        Suzuki K Poulose <suzuki.poulose@....com>,
        Mike Leach <mike.leach@...aro.org>
Cc:     John Garry <john.g.garry@...cle.com>,
        Will Deacon <will@...nel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Ingo Molnar <mingo@...hat.com>,
        Mark Rutland <mark.rutland@....com>,
        Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        Jiri Olsa <jolsa@...nel.org>,
        Namhyung Kim <namhyung@...nel.org>,
        Ian Rogers <irogers@...gle.com>, coresight@...ts.linaro.org,
        linux-arm-kernel@...ts.infradead.org,
        linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v1 1/2] perf auxtrace: Add 'T' itrace option for timestamp
 trace



On 19/10/2023 11:47, Adrian Hunter wrote:
> On 14/10/23 10:45, Leo Yan wrote:
>> An AUX trace can contain timestamp, but in some situations, the hardware
>> trace module (e.g. Arm CoreSight) cannot decide the traced timestamp is
>> the same source with CPU's time, thus the decoder can not use the
>> timestamp trace for samples.
>>
>> This patch introduces 'T' itrace option. If users know the platforms
> 
> "If users know" <- how would users know?  Could the kernel
> or tools also figure it out?
> 

The problem is this was only made into a discoverable feature with
Feat_TRF in Armv8.4. So this workaround is to support devices that
already had the right kind of timestamps before that feature.

It might be possible to list every device in the driver, but maybe there
could even be some firmware that disconnects the clock source on a
device that we thought would have it.

IMO adding this option is the best way to workaround it.

>> they are working on have the same time counter with CPUs, users can
>> use this new option to tell a decoder for using timestamp trace as
>> kernel time.
>>
>> Signed-off-by: Leo Yan <leo.yan@...aro.org>
>> ---
>>  tools/perf/Documentation/itrace.txt | 1 +
>>  tools/perf/util/auxtrace.c          | 3 +++
>>  tools/perf/util/auxtrace.h          | 3 +++
>>  3 files changed, 7 insertions(+)
>>
>> diff --git a/tools/perf/Documentation/itrace.txt b/tools/perf/Documentation/itrace.txt
>> index a97f95825b14..19cc179be9a7 100644
>> --- a/tools/perf/Documentation/itrace.txt
>> +++ b/tools/perf/Documentation/itrace.txt
>> @@ -25,6 +25,7 @@
>>  		q	quicker (less detailed) decoding
>>  		A	approximate IPC
>>  		Z	prefer to ignore timestamps (so-called "timeless" decoding)
>> +		T	use the timestamp trace as kernel time
>>  
>>  	The default is all events i.e. the same as --itrace=iybxwpe,
>>  	except for perf script where it is --itrace=ce
>> diff --git a/tools/perf/util/auxtrace.c b/tools/perf/util/auxtrace.c
>> index a0368202a746..f528c4364d23 100644
>> --- a/tools/perf/util/auxtrace.c
>> +++ b/tools/perf/util/auxtrace.c
>> @@ -1638,6 +1638,9 @@ int itrace_do_parse_synth_opts(struct itrace_synth_opts *synth_opts,
>>  		case 'Z':
>>  			synth_opts->timeless_decoding = true;
>>  			break;
>> +		case 'T':
>> +			synth_opts->use_timestamp = true;
>> +			break;
>>  		case ' ':
>>  		case ',':
>>  			break;
>> diff --git a/tools/perf/util/auxtrace.h b/tools/perf/util/auxtrace.h
>> index 29eb82dff574..55702215a82d 100644
>> --- a/tools/perf/util/auxtrace.h
>> +++ b/tools/perf/util/auxtrace.h
>> @@ -99,6 +99,7 @@ enum itrace_period_type {
>>   * @remote_access: whether to synthesize remote access events
>>   * @mem: whether to synthesize memory events
>>   * @timeless_decoding: prefer "timeless" decoding i.e. ignore timestamps
>> + * @use_timestamp: use the timestamp trace as kernel time
>>   * @vm_time_correlation: perform VM Time Correlation
>>   * @vm_tm_corr_dry_run: VM Time Correlation dry-run
>>   * @vm_tm_corr_args:  VM Time Correlation implementation-specific arguments
>> @@ -146,6 +147,7 @@ struct itrace_synth_opts {
>>  	bool			remote_access;
>>  	bool			mem;
>>  	bool			timeless_decoding;
>> +	bool			use_timestamp;
>>  	bool			vm_time_correlation;
>>  	bool			vm_tm_corr_dry_run;
>>  	char			*vm_tm_corr_args;
>> @@ -678,6 +680,7 @@ bool auxtrace__evsel_is_auxtrace(struct perf_session *session,
>>  "				q:			quicker (less detailed) decoding\n" \
>>  "				A:			approximate IPC\n" \
>>  "				Z:			prefer to ignore timestamps (so-called \"timeless\" decoding)\n" \
>> +"				T:			use the timestamp trace as kernel time\n" \
>>  "				PERIOD[ns|us|ms|i|t]:   specify period to sample stream\n" \
>>  "				concatenate multiple options. Default is iybxwpe or cewp\n"
>>  
> 
>