[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <YmfE6wIA1Dbet1nX@rric.localdomain>
Date: Tue, 26 Apr 2022 12:09:47 +0200
From: Robert Richter <rrichter@....com>
To: Ravi Bangoria <ravi.bangoria@....com>
CC: <peterz@...radead.org>, <acme@...nel.org>, <mingo@...hat.com>,
<mark.rutland@....com>, <jolsa@...nel.org>, <namhyung@...nel.org>,
<tglx@...utronix.de>, <bp@...en8.de>, <irogers@...gle.com>,
<yao.jin@...ux.intel.com>, <james.clark@....com>,
<leo.yan@...aro.org>, <kan.liang@...ux.intel.com>,
<ak@...ux.intel.com>, <eranian@...gle.com>,
<like.xu.linux@...il.com>, <x86@...nel.org>,
<linux-perf-users@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
<sandipan.das@....com>, <ananth.narayan@....com>,
<kim.phillips@....com>, <santosh.shukla@....com>
Subject: Re: [PATCH 3/6] perf/tool/amd/ibs: Warn about sampling period skew
On 25.04.22 10:13:20, Ravi Bangoria wrote:
> @@ -29,3 +31,32 @@ void arch_evsel__fixup_new_cycles(struct perf_event_attr *attr)
>
> free(env.cpuid);
> }
> +
> +void arch_evsel__warn_ambiguity(struct evsel *evsel, struct perf_event_attr *attr)
Have an 'ibs_' string in the name?
> +{
> + struct perf_env *env = evsel__env(evsel);
> + struct perf_pmu *evsel_pmu = evsel__find_pmu(evsel);
> + struct perf_pmu *ibs_fetch_pmu = perf_pmu__find("ibs_fetch");
> + struct perf_pmu *ibs_op_pmu = perf_pmu__find("ibs_op");
> + static int warned_once;
> +
> + if (warned_once || !perf_env__cpuid(env) || !env->cpuid ||
> + !strstarts(env->cpuid, "AuthenticAMD") || !evsel_pmu)
> + return;
> +
> + if (ibs_fetch_pmu && ibs_fetch_pmu->type == evsel_pmu->type) {
> + if (attr->config & (1ULL << 59)) {
> + pr_warning(
> +"WARNING: Hw internally resets sampling period when L3 Miss Filtering is enabled\n"
> +"and tagged operation does not cause L3 Miss. This causes sampling period skew.\n");
> + warned_once = 1;
> + }
> + } else if (ibs_op_pmu && ibs_op_pmu->type == evsel_pmu->type) {
> + if (attr->config & (1ULL << 16)) {
> + pr_warning(
> +"WARNING: Hw internally resets sampling period when L3 Miss Filtering is enabled\n"
> +"and tagged operation does not cause L3 Miss. This causes sampling period skew.\n");
Avoid duplicate code. You could move it to the end of the function and
early return if the config bit is not set.
> + warned_once = 1;
> + }
> + }
> +}
-Robert
Powered by blists - more mailing lists