Re: [PATCH v5 00/26] perf tool: AlderLake hybrid support series 1

[Jin, Yao]

Hi Arnaldo,

On 4/27/2021 4:41 AM, Arnaldo Carvalho de Melo wrote:
> Em Fri, Apr 23, 2021 at 01:35:15PM +0800, Jin Yao escreveu:
> > AlderLake uses a hybrid architecture utilizing Golden Cove cores
> > (core cpu) and Gracemont cores (atom cpu). Each cpu has dedicated
> > event list. Some events are available on core cpu, some events
> > are available on atom cpu and some events can be available on both.
> >
> > Kernel exports new pmus "cpu_core" and "cpu_atom" through sysfs:
> > /sys/devices/cpu_core
> > /sys/devices/cpu_atom
>
> [acme@five perf]$ b4 am -t -s -l --cc-trailers 20210423053541.12521-1-yao.jin@xxxxxxxxxxxxxxx
> Looking up https://lore.kernel.org/r/20210423053541.12521-1-yao.jin%40linux.intel.com
> Grabbing thread from lore.kernel.org/lkml
> Analyzing 29 messages in the thread
> ---
> Writing ./v5_20210423_yao_jin_perf_tool_alderlake_hybrid_support_series_1.mbx
>    [PATCH v5 01/26] tools headers uapi: Update tools's copy of linux/perf_event.h
>      + Signed-off-by: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
>      + Link: https://lore.kernel.org/r/20210423053541.12521-2-yao.jin@xxxxxxxxxxxxxxx
>      + Cc: peterz@xxxxxxxxxxxxx
>      + Cc: kan.liang@xxxxxxxxx
>      + Cc: yao.jin@xxxxxxxxx
>      + Cc: acme@xxxxxxxxxx
>      + Cc: jolsa@xxxxxxxxxx
>      + Cc: ak@xxxxxxxxxxxxxxx
>      + Cc: alexander.shishkin@xxxxxxxxxxxxxxx
>      + Cc: mingo@xxxxxxxxxx
>      + Cc: Linux-kernel@xxxxxxxxxxxxxxx
> <SNIP>
> ---
> Total patches: 26
> ---
> Cover: ./v5_20210423_yao_jin_perf_tool_alderlake_hybrid_support_series_1.cover
>   Link: https://lore.kernel.org/r/20210423053541.12521-1-yao.jin@xxxxxxxxxxxxxxx
>   Base: not found
>         git am ./v5_20210423_yao_jin_perf_tool_alderlake_hybrid_support_series_1.mbx
> [acme@five perf]$        git am ./v5_20210423_yao_jin_perf_tool_alderlake_hybrid_support_series_1.mbx
> Applying: tools headers uapi: Update tools's copy of linux/perf_event.h
> Applying: perf jevents: Support unit value "cpu_core" and "cpu_atom"
> Applying: perf pmu: Simplify arguments of __perf_pmu__new_alias
> Applying: perf pmu: Save pmu name
> Applying: perf pmu: Save detected hybrid pmus to a global pmu list
> Applying: perf pmu: Add hybrid helper functions
> Applying: perf stat: Uniquify hybrid event name
> error: patch failed: tools/perf/builtin-stat.c:68
> error: tools/perf/builtin-stat.c: patch does not apply
> error: patch failed: tools/perf/util/stat-display.c:17
> error: tools/perf/util/stat-display.c: patch does not apply
> Patch failed at 0007 perf stat: Uniquify hybrid event name
> hint: Use 'git am --show-current-patch=diff' to see the failed patch
> When you have resolved this problem, run "git am --continue".
> If you prefer to skip this patch, run "git am --skip" instead.
> To restore the original branch and stop patching, run "git am --abort".
> [acme@five perf]$
>
> [acme@five perf]$ git am --show-current-patch=diff | patch -p1
> patching file tools/perf/builtin-stat.c
> Hunk #1 FAILED at 68.
> Hunk #2 succeeded at 2402 (offset 24 lines).
> 1 out of 2 hunks FAILED -- saving rejects to file tools/perf/builtin-stat.c.rej
> patching file tools/perf/util/evsel.h
> Hunk #1 succeeded at 116 (offset 1 line).
> patching file tools/perf/util/parse-events.c
> patching file tools/perf/util/stat-display.c
> Hunk #1 FAILED at 17.
> Hunk #2 succeeded at 538 (offset 6 lines).
> Hunk #3 succeeded at 553 (offset 6 lines).
> 1 out of 3 hunks FAILED -- saving rejects to file tools/perf/util/stat-display.c.rej
> [acme@five perf]$ vim tools/perf/builtin-stat.c.rej
> [acme@five perf]$ cat tools/perf/builtin-stat.c.rej
> --- tools/perf/builtin-stat.c
> +++ tools/perf/builtin-stat.c
> @@ -68,6 +68,7 @@
>   #include "util/affinity.h"
>   #include "util/pfm.h"
>   #include "util/bpf_counter.h"
> +#include "util/pmu-hybrid.h"
>   #include "asm/bug.h"
>
>   #include <linux/time64.h>
> [acme@five perf]$ cat tools/perf/util/stat-display.c.rej
> --- tools/perf/util/stat-display.c
> +++ tools/perf/util/stat-display.c
> @@ -17,6 +17,7 @@
>   #include "cgroup.h"
>   #include <api/fs/fs.h>
>   #include "util.h"
> +#include "pmu-hybrid.h"
>
>   #define CNTR_NOT_SUPPORTED	"<not supported>"
>   #define CNTR_NOT_COUNTED	"<not counted>"
> [acme@five perf]$
>
>
> Its clashing with some BPF changes by Song that are still under review
> but I have in my tmp.perf/core branch so that I can test build it while
> I wait for Jiri to say if Song addressed all his comments.
>
> So after you address the new round of comments for v5 you can please
> rebase on tmp.perf/core or, at that point, perf/core.
>
> - Arnaldo

The v6 is ready now which is based on perf/core. It only has a minor update (adding new test case 
for cache events with pmu prefix). There is no other updates in v6.

I'm OK to rebase v6 on tmp.perf/core first and then post the new series.

Thanks
Jin Yao