lkml.org 
[lkml]   [2022]   [Jun]   [20]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH v1 1/3] docs: perf: Add description for Alibaba's T-Head PMU driver
On Fri, 17 Jun 2022 19:18:23 +0800
Shuai Xue <xueshuai@linux.alibaba.com> wrote:

> Alibaba's T-Head SoC implements uncore PMU for performance and functional
> debugging to facilitate system maintenance. Document it to provide guidance
> on how to use it.
>
> Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com>

Hi Shuia Xue,

A few quick comments inline,

Thanks,

Jonathan

> ---
> .../admin-guide/perf/alibaba_pmu.rst | 94 +++++++++++++++++++
> Documentation/admin-guide/perf/index.rst | 1 +
> 2 files changed, 95 insertions(+)
> create mode 100644 Documentation/admin-guide/perf/alibaba_pmu.rst
>
> diff --git a/Documentation/admin-guide/perf/alibaba_pmu.rst b/Documentation/admin-guide/perf/alibaba_pmu.rst
> new file mode 100644
> index 000000000000..337f4f1d4c54
> --- /dev/null
> +++ b/Documentation/admin-guide/perf/alibaba_pmu.rst
> @@ -0,0 +1,94 @@
> +=============================================================
> +Alibaba's T-Head SoC Uncore Performance Monitoring Unit (PMU)
> +=============================================================
> +
> +The Yitian 710, custom-built by Alibaba Group's chip development business,
> +T-Head, implements uncore PMU for performance and functional debugging to
> +facilitate system maintenance.
> +
> +DDR Sub-System Driveway (DRW) PMU Driver
> +=========================================
> +
> +The Yitian 710 SoC supports the most advanced DDR5/4 DRAM to provide
> +tremendous memory bandwidth for cloud computing and HPC. The Driveway is a
> +module which is a bridge between a router of mesh network and memory
> +controller. It provides various functions like secure control, address map
> +and so on.

I'd focus on the PMU aspect. No point in describing other features in this
document.

> +
> +Yitian 710 employs eight DDR5/4 channels, four on each die. Each channel is
> +independent of others to service system memory requests. And one DDR5

Each DDR5 channel...

> +channel is split into two independent sub-channels. The DDR Sub-System
> +Driveway implements separate PMUs for each sub-channel to monitor various
> +performance metrics.
> +
> +The Driveway PMU devices are named as ali_drw_<sys_base_addr> with perf.
> +For example, ali_drw_21000 and ali_drw_21080 are two PMU devices for two
> +sub-channels of the same channel in die 0. And the PMU device of die 1 is
> +prefixed with ali_drw_400XXXXX, e.g. ali_drw_40021000.
> +
> +Each sub-channel has 36 PMU counters in total, which is classified into
> +four groups:
> +
> +- Group 0: PMU Cycle Counter. This group has one pair of counters
> + pmu_cycle_cnt_low and pmu_cycle_cnt_high, that is used as the cycle count
> + based on DDRC core clock.
> +
> +- Group 1: PMU Bandwidth Counters. This group has 8 counters that are used
> + to count the total access number of either the eight bank groups in a
> + selected rank, or four ranks separately in the first 4 counters. The base
> + transfer unit is 64B.
> +
> +- Group 2: PMU Retry Counters. This group has 10 counters, that intend to
> + count the total retry number of each type of uncorrectable error.
> +
> +- Group 3: PMU Common Counters. This group has 16 counters, that are used
> + to count the common events.
> +
> +For now, the Driveway PMU driver only uses counters in group 0 and group 3.
> +
> +Example usage of counting memory data bandwidth::
> +
> + perf stat \
> + -e ali_drw_21000/perf_hif_wr/ \

What does hif stand for? Also, perf seems redundant for
a perf event.


> + -e ali_drw_21000/perf_hif_rd/ \
> + -e ali_drw_21000/perf_hif_rmw/ \
> + -e ali_drw_21000/perf_cycle/ \
> + -e ali_drw_21080/perf_hif_wr/ \
> + -e ali_drw_21080/perf_hif_rd/ \
> + -e ali_drw_21080/perf_hif_rmw/ \
> + -e ali_drw_21080/perf_cycle/ \
> + -e ali_drw_23000/perf_hif_wr/ \
> + -e ali_drw_23000/perf_hif_rd/ \
> + -e ali_drw_23000/perf_hif_rmw/ \
> + -e ali_drw_23000/perf_cycle/ \
> + -e ali_drw_23080/perf_hif_wr/ \
> + -e ali_drw_23080/perf_hif_rd/ \
> + -e ali_drw_23080/perf_hif_rmw/ \
> + -e ali_drw_23080/perf_cycle/ \
> + -e ali_drw_25000/perf_hif_wr/ \
> + -e ali_drw_25000/perf_hif_rd/ \
> + -e ali_drw_25000/perf_hif_rmw/ \
> + -e ali_drw_25000/perf_cycle/ \
> + -e ali_drw_25080/perf_hif_wr/ \
> + -e ali_drw_25080/perf_hif_rd/ \
> + -e ali_drw_25080/perf_hif_rmw/ \
> + -e ali_drw_25080/perf_cycle/ \
> + -e ali_drw_27000/perf_hif_wr/ \
> + -e ali_drw_27000/perf_hif_rd/ \
> + -e ali_drw_27000/perf_hif_rmw/ \
> + -e ali_drw_27000/perf_cycle/ \
> + -e ali_drw_27080/perf_hif_wr/ \
> + -e ali_drw_27080/perf_hif_rd/ \
> + -e ali_drw_27080/perf_hif_rmw/ \
> + -e ali_drw_27080/perf_cycle/ -- sleep 10
> +
> +The average DRAM bandwidth can be calculated as follows:
> +
> +- Read Bandwidth = perf_hif_rd * DDRC_WIDTH * DDRC_Freq / DDRC_Cycle
> +- Write Bandwidth = (perf_hif_wr + perf_hif_rmw) * DDRC_WIDTH * DDRC_Freq / DDRC_Cycle
> +
> +Here, DDRC_WIDTH = 64 bytes.
> +
> +The current driver does not support sampling. So "perf record" is
> +unsupported. Also attach to a task is unsupported as the events are all
> +uncore.
> diff --git a/Documentation/admin-guide/perf/index.rst b/Documentation/admin-guide/perf/index.rst
> index 69b23f087c05..bf466ae91c6c 100644
> --- a/Documentation/admin-guide/perf/index.rst
> +++ b/Documentation/admin-guide/perf/index.rst
> @@ -17,3 +17,4 @@ Performance monitor support
> xgene-pmu
> arm_dsu_pmu
> thunderx2-pmu
> + thead_pmu

\
 
 \ /
  Last update: 2022-06-20 13:51    [W:0.315 / U:5.652 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site