public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Alexander Antonov <alexander.antonov@linux.intel.com>
To: "liuqi (BA)" <liuqi115@huawei.com>
Cc: acme@kernel.org, linux-kernel@vger.kernel.org, jolsa@redhat.com,
	ak@linux.intel.com, alexander.shishkin@linux.intel.com,
	mark.rutland@arm.com, namhyung@kernel.org, irogers@google.com,
	mingo@redhat.com, peterz@infradead.org
Subject: Re: [PATCH v4 4/5] perf stat: Enable iostat mode for x86 platforms
Date: Wed, 10 Mar 2021 19:19:19 +0300	[thread overview]
Message-ID: <65bccfcb-2e43-714f-f586-09aa5015c4fa@linux.intel.com> (raw)
In-Reply-To: <d2f23f76-3557-f6f1-dd0b-2491bb734b00@huawei.com>


On 3/9/2021 10:51 AM, liuqi (BA) wrote:
> Hi Alexander,
>
> On 2021/2/3 21:58, Alexander Antonov wrote:
>> This functionality is based on recently introduced sysfs attributes
>> for Intel® Xeon® Scalable processor family (code name Skylake-SP):
>> Commit bb42b3d39781 ("perf/x86/intel/uncore: Expose an Uncore unit to
>> IIO PMON mapping")
>>
>> Mode is intended to provide four I/O performance metrics in MB per each
>> PCIe root port:
>>   - Inbound Read: I/O devices below root port read from the host memory
>>   - Inbound Write: I/O devices below root port write to the host memory
>>   - Outbound Read: CPU reads from I/O devices below root port
>>   - Outbound Write: CPU writes to I/O devices below root port
>>
>> Each metric requiries only one uncore event which increments at every 4B
>> transfer in corresponding direction. The formulas to compute metrics
>> are generic:
>>      #EventCount * 4B / (1024 * 1024)
>>
>> Signed-off-by: Alexander Antonov<alexander.antonov@linux.intel.com>
>> ---
>>   tools/perf/Documentation/perf-iostat.txt |  88 ++++++
>>   tools/perf/Makefile.perf                 |   5 +-
>>   tools/perf/arch/x86/util/Build           |   1 +
>>   tools/perf/arch/x86/util/iostat.c        | 345 +++++++++++++++++++++++
>>   tools/perf/command-list.txt              |   1 +
>>   tools/perf/perf-iostat.sh                |  12 +
>>   6 files changed, 451 insertions(+), 1 deletion(-)
>>   create mode 100644 tools/perf/Documentation/perf-iostat.txt
>>   create mode 100644 tools/perf/perf-iostat.sh
>>
>> diff --git a/tools/perf/Documentation/perf-iostat.txt 
>> b/tools/perf/Documentation/perf-iostat.txt
>> new file mode 100644
>> index 000000000000..165176944031
>> --- /dev/null
>> +++ b/tools/perf/Documentation/perf-iostat.txt
>> @@ -0,0 +1,88 @@
>> +perf-iostat(1)
>> +===============
>> +
>> +NAME
>> +----
>> +perf-iostat - Show I/O performance metrics
>> +
>> +SYNOPSIS
>> +--------
>> +[verse]
>> +'perf iostat' list
>> +'perf iostat' <ports> -- <command> [<options>]
>> +
>> +DESCRIPTION
>> +-----------
>> +Mode is intended to provide four I/O performance metrics per each 
>> PCIe root port:
>> +
>> +- Inbound Read   - I/O devices below root port read from the host 
>> memory, in MB
>> +
>> +- Inbound Write  - I/O devices below root port write to the host 
>> memory, in MB
>> +
>> +- Outbound Read  - CPU reads from I/O devices below root port, in MB
>> +
>> +- Outbound Write - CPU writes to I/O devices below root port, in MB
>> +
>> +OPTIONS
>> +-------
>> +<command>...::
>> +    Any command you can specify in a shell.
>> +
>> +list::
>> +    List all PCIe root ports.
>
> I noticed that "iostat" commond and cmd_iostat() callback function is 
> not registered in cmd_struct in perf.c. So I think "perf iostat list" 
> perhaps can not work properly.
>
> I also test this patchset on x86 platform, and here is the log:
>
> root@ubuntu:/home/lq# ./perf iostat list
> perf: 'iostat' is not a perf-command. See 'perf --help'.
> root@ubuntu:/home/lq# ./perf stat --iostat
> ^C
>  Performance counter stats for 'system wide':
>
>    port             Inbound Read(MB)    Inbound Write(MB) Outbound 
> Read(MB)   Outbound Write(MB)
> 0000:00                    0 0                    0                  0
> 0000:80                    0 0                    0                  0
> 0000:17                    0 0                    0                  0
> 0000:85                    0 0                    0                  0
> 0000:3a                    0 0                    0                  0
> 0000:ae                    0 0                    0                  0
> 0000:5d                    0 0                    0                  0
> 0000:d7                    0 0                    0                  0
>
>        0.611303832 seconds time elapsed
>
>
> root@ubuntu:/home/lq# ./perf stat --iostat=0000:17
> ^C
>  Performance counter stats for 'system wide':
>
>    port             Inbound Read(MB)    Inbound Write(MB) Outbound 
> Read(MB)   Outbound Write(MB)
> 0000:17                    0 0                    0                  0
>
>        0.521317572 seconds time elapsed
>
> So how does following perf iostat list work, did I miss something?
>
> Thanks,
> Qi
>
Hello,

The 'iostat' mode uses aliases mechanism in perf same as 'perf archive' and
in this case you don't need to add function callback into cmd_struct.
For example, the command 'perf iostat list' will be converted to
'perf stat --iostat=list'.

After building the perf tool you should have two shell scripts in tools/perf
directory and one of them is executable, for example:
# make -C tools/perf
# ls -l tools/perf/perf-iostat*
-rwxr-xr-x 1 root root 290 Mar 10 18:17 perf-iostat
-rw-r--r-- 1 root root 290 Feb  3 15:14 perf-iostat.sh

It should be possible to run 'perf iostat' from build directory:
# cd tools/perf
# ./perf iostat list
S0-uncore_iio_0<0000:00>
S1-uncore_iio_0<0000:80>
S0-uncore_iio_1<0000:17>
S1-uncore_iio_1<0000:85>
S0-uncore_iio_2<0000:3a>
S1-uncore_iio_2<0000:ae>
S0-uncore_iio_3<0000:5d>
S1-uncore_iio_3<0000:d7>

Also you can copy 'perf-iostat' to ~/libexec/perf-core/ or just run 
'make install'

# make install
# cp perf /usr/bin/
# ls -lh ~/libexec/perf-core/
total 24K
-rwxr-xr-x 1 root root 1.4K Mar 10 18:17 perf-archive
-rwxr-xr-x 1 root root  290 Mar 10 18:17 perf-iostat
-rwxr-xr-x 1 root root 6.3K Mar 10 18:17 perf-with-kcore
drwxr-xr-x 4 root root 4.0K Nov  5  2019 scripts
drwxr-xr-x 4 root root 4.0K Mar 10 18:17 tests
# perf iostat 0000:17 -I 1000 --interval-count 2
#           time    port             Inbound Read(MB)    Inbound 
Write(MB)    Outbound Read(MB)   Outbound Write(MB)
      1.000220341 0000:17                    0 0                    
0                    0
      2.000569569 0000:17                    0 0                    
0                    0

Actually, Arnaldo has explained before how does aliases mechanism work.

I hope it will solve your issue. Otherwise, please email.

Thank you,
Alexander
>> +
>> +<ports>::
>> +    Select the root ports for monitoring. Comma-separated list is 
>> supported.
>> +
>> +EXAMPLES
>> +--------
>> +
>> +1. List all PCIe root ports (example for 2-S platform):
>> +
>> +   $ perf iostat list
>> +   S0-uncore_iio_0<0000:00>
>> +   S1-uncore_iio_0<0000:80>
>> +   S0-uncore_iio_1<0000:17>
>> +   S1-uncore_iio_1<0000:85>
>> +   S0-uncore_iio_2<0000:3a>
>> +   S1-uncore_iio_2<0000:ae>
>> +   S0-uncore_iio_3<0000:5d>
>> +   S1-uncore_iio_3<0000:d7>
>

  reply	other threads:[~2021-03-10 16:20 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-02-03 13:58 [PATCH v4 0/5] perf stat: Introduce iostat mode to provide I/O performance metrics Alexander Antonov
2021-02-03 13:58 ` [PATCH v4 1/5] perf stat: Add AGGR_PCIE_PORT mode Alexander Antonov
2021-02-04 12:07   ` Namhyung Kim
2021-02-08 11:30     ` Alexander Antonov
2021-02-10 10:02       ` Namhyung Kim
2021-02-03 13:58 ` [PATCH v4 2/5] perf stat: Basic support for iostat in perf Alexander Antonov
2021-02-04 12:22   ` Namhyung Kim
2021-02-08 11:55     ` Alexander Antonov
2021-02-10 10:07       ` Namhyung Kim
2021-02-03 13:58 ` [PATCH v4 3/5] perf stat: Helper functions for PCIe root ports list in iostat mode Alexander Antonov
2021-02-04 12:32   ` Namhyung Kim
2021-02-08 11:58     ` Alexander Antonov
2021-02-03 13:58 ` [PATCH v4 4/5] perf stat: Enable iostat mode for x86 platforms Alexander Antonov
2021-03-09  7:51   ` liuqi (BA)
2021-03-10 16:19     ` Alexander Antonov [this message]
2021-03-11  7:02       ` liuqi (BA)
2021-02-03 13:58 ` [PATCH v4 5/5] perf: Update .gitignore file Alexander Antonov
2021-02-03 20:55 ` [PATCH v4 0/5] perf stat: Introduce iostat mode to provide I/O performance metrics Arnaldo Carvalho de Melo
2021-02-04 12:35   ` Namhyung Kim

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=65bccfcb-2e43-714f-f586-09aa5015c4fa@linux.intel.com \
    --to=alexander.antonov@linux.intel.com \
    --cc=acme@kernel.org \
    --cc=ak@linux.intel.com \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=irogers@google.com \
    --cc=jolsa@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=liuqi115@huawei.com \
    --cc=mark.rutland@arm.com \
    --cc=mingo@redhat.com \
    --cc=namhyung@kernel.org \
    --cc=peterz@infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox