linux-kernel - Re: [PATCH 4/4] perf vendor events amd: Add Zen 4 metrics

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: <6675c596-eeb4-3347-ec85-2bd67b970973@amd.com>
Date:   Wed, 7 Dec 2022 12:28:02 +0530
From:   Sandipan Das <sandipan.das@....com>
To:     Ian Rogers <irogers@...gle.com>
Cc:     linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org,
        acme@...nel.org, peterz@...radead.org, mingo@...hat.com,
        mark.rutland@....com, alexander.shishkin@...ux.intel.com,
        jolsa@...nel.org, namhyung@...nel.org, eranian@...gle.com,
        jhladky@...hat.com, ravi.bangoria@....com, ananth.narayan@....com
Subject: Re: [PATCH 4/4] perf vendor events amd: Add Zen 4 metrics

On 12/7/2022 11:35 AM, Ian Rogers wrote:
> On Tue, Dec 6, 2022 at 9:32 PM Sandipan Das <sandipan.das@....com> wrote:
>>
>> Add metrics taken from Section 2.1.15.2 "Performance Measurement" in
>> the Processor Programming Reference (PPR) for AMD Family 19h Model 11h
>> Revision B1 processors.
>>
>> The recommended metrics are sourced from Table 27 "Guidance for Common
>> Performance Statistics with Complex Event Selects".
>>
>> The pipeline utilization metrics are sourced from Table 28 "Guidance
>> for Pipeline Utilization Analysis Statistics". These are new to Zen 4
>> processors and useful for finding performance bottlenecks by analyzing
>> activity at different stages of the pipeline. Metric groups have been
>> added for Level 1 and Level 2 analysis.
>>
>> Signed-off-by: Sandipan Das <sandipan.das@....com>
>> ---
>>  .../pmu-events/arch/x86/amdzen4/pipeline.json |  98 +++++
>>  .../arch/x86/amdzen4/recommended.json         | 334 ++++++++++++++++++
>>  2 files changed, 432 insertions(+)
>>  create mode 100644 tools/perf/pmu-events/arch/x86/amdzen4/pipeline.json
>>  create mode 100644 tools/perf/pmu-events/arch/x86/amdzen4/recommended.json
>>
<snip>
>> +  {
>> +    "MetricName": "dram_read_data_bytes_for_local_processor",
> 
> nit: Is "bytes" redundant in the name here? It may even be confusing
> given the units.
> 

Agreed. I can replace "bytes" with "mbytes" or "megabytes" for these bandwidth metrics.

- Sandipan

>> +    "BriefDescription": "DRAM read data bytes for local processor.",
>> +    "MetricExpr": "local_processor_read_data_beats_cs0 + local_processor_read_data_beats_cs1 + local_processor_read_data_beats_cs2 + local_processor_read_data_beats_cs3 + local_processor_read_data_beats_cs4 + local_processor_read_data_beats_cs5 + local_processor_read_data_beats_cs6 + local_processor_read_data_beats_cs7 + local_processor_read_data_beats_cs8 + local_processor_read_data_beats_cs9 + local_processor_read_data_beats_cs10 + local_processor_read_data_beats_cs11",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "dram_write_data_bytes_for_local_processor",
>> +    "BriefDescription": "DRAM write data bytes for local processor.",
>> +    "MetricExpr": "local_processor_write_data_beats_cs0 + local_processor_write_data_beats_cs1 + local_processor_write_data_beats_cs2 + local_processor_write_data_beats_cs3 + local_processor_write_data_beats_cs4 + local_processor_write_data_beats_cs5 + local_processor_write_data_beats_cs6 + local_processor_write_data_beats_cs7 + local_processor_write_data_beats_cs8 + local_processor_write_data_beats_cs9 + local_processor_write_data_beats_cs10 + local_processor_write_data_beats_cs11",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "dram_read_data_bytes_for_remote_processor",
>> +    "BriefDescription": "DRAM read data bytes for remote processor.",
>> +    "MetricExpr": "remote_processor_read_data_beats_cs0 + remote_processor_read_data_beats_cs1 + remote_processor_read_data_beats_cs2 + remote_processor_read_data_beats_cs3 + remote_processor_read_data_beats_cs4 + remote_processor_read_data_beats_cs5 + remote_processor_read_data_beats_cs6 + remote_processor_read_data_beats_cs7 + remote_processor_read_data_beats_cs8 + remote_processor_read_data_beats_cs9 + remote_processor_read_data_beats_cs10 + remote_processor_read_data_beats_cs11",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "dram_write_data_bytes_for_remote_processor",
>> +    "BriefDescription": "DRAM write data bytes for remote processor.",
>> +    "MetricExpr": "remote_processor_write_data_beats_cs0 + remote_processor_write_data_beats_cs1 + remote_processor_write_data_beats_cs2 + remote_processor_write_data_beats_cs3 + remote_processor_write_data_beats_cs4 + remote_processor_write_data_beats_cs5 + remote_processor_write_data_beats_cs6 + remote_processor_write_data_beats_cs7 + remote_processor_write_data_beats_cs8 + remote_processor_write_data_beats_cs9 + remote_processor_write_data_beats_cs10 + remote_processor_write_data_beats_cs11",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "local_socket_upstream_dma_read_data_bytes",
>> +    "BriefDescription": "Local socket upstream DMA read data bytes.",
>> +    "MetricExpr": "local_socket_upstream_read_beats_iom0 + local_socket_upstream_read_beats_iom1 + local_socket_upstream_read_beats_iom2 + local_socket_upstream_read_beats_iom3",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "local_socket_upstream_dma_write_data_bytes",
>> +    "BriefDescription": "Local socket upstream DMA write data bytes.",
>> +    "MetricExpr": "local_socket_upstream_write_beats_iom0 + local_socket_upstream_write_beats_iom1 + local_socket_upstream_write_beats_iom2 + local_socket_upstream_write_beats_iom3",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "remote_socket_upstream_dma_read_data_bytes",
>> +    "BriefDescription": "Remote socket upstream DMA read data bytes.",
>> +    "MetricExpr": "remote_socket_upstream_read_beats_iom0 + remote_socket_upstream_read_beats_iom1 + remote_socket_upstream_read_beats_iom2 + remote_socket_upstream_read_beats_iom3",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "remote_socket_upstream_dma_write_data_bytes",
>> +    "BriefDescription": "Remote socket upstream DMA write data bytes.",
>> +    "MetricExpr": "remote_socket_upstream_write_beats_iom0 + remote_socket_upstream_write_beats_iom1 + remote_socket_upstream_write_beats_iom2 + remote_socket_upstream_write_beats_iom3",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "local_socket_inbound_data_bytes_to_cpu",
>> +    "BriefDescription": "Local socket inbound data bytes to the CPU (e.g. read data).",
>> +    "MetricExpr": "local_socket_inf0_inbound_data_beats_ccm0 + local_socket_inf1_inbound_data_beats_ccm0 + local_socket_inf0_inbound_data_beats_ccm1 + local_socket_inf1_inbound_data_beats_ccm1 + local_socket_inf0_inbound_data_beats_ccm2 + local_socket_inf1_inbound_data_beats_ccm2 + local_socket_inf0_inbound_data_beats_ccm3 + local_socket_inf1_inbound_data_beats_ccm3 + local_socket_inf0_inbound_data_beats_ccm4 + local_socket_inf1_inbound_data_beats_ccm4 + local_socket_inf0_inbound_data_beats_ccm5 + local_socket_inf1_inbound_data_beats_ccm5 + local_socket_inf0_inbound_data_beats_ccm6 + local_socket_inf1_inbound_data_beats_ccm6 + local_socket_inf0_inbound_data_beats_ccm7 + local_socket_inf1_inbound_data_beats_ccm7",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "3.0517578125e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "local_socket_outbound_data_bytes_from_cpu",
>> +    "BriefDescription": "Local socket outbound data bytes from the CPU (e.g. write data).",
>> +    "MetricExpr": "local_socket_inf0_outbound_data_beats_ccm0 + local_socket_inf1_outbound_data_beats_ccm0 + local_socket_inf0_outbound_data_beats_ccm1 + local_socket_inf1_outbound_data_beats_ccm1 + local_socket_inf0_outbound_data_beats_ccm2 + local_socket_inf1_outbound_data_beats_ccm2 + local_socket_inf0_outbound_data_beats_ccm3 + local_socket_inf1_outbound_data_beats_ccm3 + local_socket_inf0_outbound_data_beats_ccm4 + local_socket_inf1_outbound_data_beats_ccm4 + local_socket_inf0_outbound_data_beats_ccm5 + local_socket_inf1_outbound_data_beats_ccm5 + local_socket_inf0_outbound_data_beats_ccm6 + local_socket_inf1_outbound_data_beats_ccm6 + local_socket_inf0_outbound_data_beats_ccm7 + local_socket_inf1_outbound_data_beats_ccm7",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "remote_socket_inbound_data_bytes_to_cpu",
>> +    "BriefDescription": "Remote socket inbound data bytes to the CPU (e.g. read data).",
>> +    "MetricExpr": "remote_socket_inf0_inbound_data_beats_ccm0 + remote_socket_inf1_inbound_data_beats_ccm0 + remote_socket_inf0_inbound_data_beats_ccm1 + remote_socket_inf1_inbound_data_beats_ccm1 + remote_socket_inf0_inbound_data_beats_ccm2 + remote_socket_inf1_inbound_data_beats_ccm2 + remote_socket_inf0_inbound_data_beats_ccm3 + remote_socket_inf1_inbound_data_beats_ccm3 + remote_socket_inf0_inbound_data_beats_ccm4 + remote_socket_inf1_inbound_data_beats_ccm4 + remote_socket_inf0_inbound_data_beats_ccm5 + remote_socket_inf1_inbound_data_beats_ccm5 + remote_socket_inf0_inbound_data_beats_ccm6 + remote_socket_inf1_inbound_data_beats_ccm6 + remote_socket_inf0_inbound_data_beats_ccm7 + remote_socket_inf1_inbound_data_beats_ccm7",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "3.0517578125e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "remote_socket_outbound_data_bytes_from_cpu",
>> +    "BriefDescription": "Remote socket outbound data bytes from the CPU (e.g. write data).",
>> +    "MetricExpr": "remote_socket_inf0_outbound_data_beats_ccm0 + remote_socket_inf1_outbound_data_beats_ccm0 + remote_socket_inf0_outbound_data_beats_ccm1 + remote_socket_inf1_outbound_data_beats_ccm1 + remote_socket_inf0_outbound_data_beats_ccm2 + remote_socket_inf1_outbound_data_beats_ccm2 + remote_socket_inf0_outbound_data_beats_ccm3 + remote_socket_inf1_outbound_data_beats_ccm3 + remote_socket_inf0_outbound_data_beats_ccm4 + remote_socket_inf1_outbound_data_beats_ccm4 + remote_socket_inf0_outbound_data_beats_ccm5 + remote_socket_inf1_outbound_data_beats_ccm5 + remote_socket_inf0_outbound_data_beats_ccm6 + remote_socket_inf1_outbound_data_beats_ccm6 + remote_socket_inf0_outbound_data_beats_ccm7 + remote_socket_inf1_outbound_data_beats_ccm7",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  },
>> +  {
>> +    "MetricName": "local_socket_outbound_data_bytes_from_all_links",
>> +    "BriefDescription": "Outbound data bytes from all links (local socket).",
>> +    "MetricExpr": "local_socket_outbound_data_beats_link0 + local_socket_outbound_data_beats_link1 + local_socket_outbound_data_beats_link2 + local_socket_outbound_data_beats_link3 + local_socket_outbound_data_beats_link4 + local_socket_outbound_data_beats_link5 + local_socket_outbound_data_beats_link6 + local_socket_outbound_data_beats_link7",
>> +    "MetricGroup": "data_fabric",
>> +    "PerPkg": "1",
>> +    "ScaleUnit": "6.103515625e-5MiB"
>> +  }
>> +]
>> --
>> 2.34.1
>>