From: Aleksandar Markovic <aleksandar.qemu.devel@gmail.com>
To: Ahmed Karaman <ahmedkhaledkaraman@gmail.com>
Cc: "ldoktor@redhat.com" <ldoktor@redhat.com>,
"ehabkost@redhat.com" <ehabkost@redhat.com>,
"alex.bennee@linaro.org" <alex.bennee@linaro.org>,
"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
"crosa@redhat.com" <crosa@redhat.com>,
"rth@twiddle.net" <rth@twiddle.net>
Subject: Re: [PATCH v2 1/3] scripts/performance: Add topN_perf.py script
Date: Sat, 20 Jun 2020 11:59:40 +0200 [thread overview]
Message-ID: <CAHiYmc79=PaKFeF0og5BW_QctDTS-cRudr3ZXjTab7D+_snb6w@mail.gmail.com> (raw)
In-Reply-To: <20200619153632.1365-2-ahmedkhaledkaraman@gmail.com>
[-- Attachment #1: Type: text/plain, Size: 8806 bytes --]
петак, 19. јун 2020., Ahmed Karaman <ahmedkhaledkaraman@gmail.com> је
написао/ла:
> Python script that prints the top N most executed functions in QEMU
> using perf.
>
> Example Usage:
Don't use capitalization when not appropriate. This is better:
Example of usage:
> topN_perf.py -n 20 -- /path/to/qemu program -program -flags
>
> If '-n' is not specified, the default is 25.
>
>
The command line above is hardly example of usage. Furthermore, it is
unclear of what you meant with "program -program". Other things are unclear
too (what about qemu options?) The command line, as is now, is somewhere
between a syntax description and an example of usage, but is neither a
syntax description nor example of usage.
From the script, it looks that there is "-h" too. Not mentioned in commit
message ar all, and it shouldbe.
I would suggest that you rework this section inthis way, for the sake of
clarity:
Syntax: (
topN_perf.py [-h] [-n <number of displayed top functions > -- <qemu
executable> [<qemu executable options>] <target executable> [<target
execurable options>]
-h - .........explain here
-n - .........explain here
Example of usage:
topN_perf.py -n 20 -- qemu-arm coulomb_double-arm
Example of output:
..... and here you continue with:
No. Percentage Name Caller
> ---- ---------- ------------------------- -------------------------
> 1 16.25% float64_mul qemu-x86_64
> 2 12.01% float64_sub qemu-x86_64
> 3 11.99% float64_add qemu-x86_64
> 4 5.69% helper_mulsd qemu-x86_64
> 5 4.68% helper_addsd qemu-x86_64
> 6 4.43% helper_lookup_tb_ptr qemu-x86_64
> 7 4.28% helper_subsd qemu-x86_64
> 8 2.71% f64_compare qemu-x86_64
> 9 2.71% helper_ucomisd qemu-x86_64
> 10 1.04% helper_pand_xmm qemu-x86_64
> 11 0.71% float64_div qemu-x86_64
> 12 0.63% helper_pxor_xmm qemu-x86_64
> 13 0.50% 0x00007f7b7004ef95 [JIT] tid 491
> 14 0.50% 0x00007f7b70044e83 [JIT] tid 491
> 15 0.36% helper_por_xmm qemu-x86_64
> 16 0.32% helper_cc_compute_all qemu-x86_64
> 17 0.30% 0x00007f7b700433f0 [JIT] tid 491
> 18 0.30% float64_compare_quiet qemu-x86_64
> 19 0.27% soft_f64_addsub qemu-x86_64
> 20 0.26% round_to_int qemu-x86_64
>
> Signed-off-by: Ahmed Karaman <ahmedkhaledkaraman@gmail.com>
> ---
> scripts/performance/topN_perf.py | 115 +++++++++++++++++++++++++++++++
> 1 file changed, 115 insertions(+)
> create mode 100755 scripts/performance/topN_perf.py
>
> diff --git a/scripts/performance/topN_perf.py b/scripts/performance/topN_
> perf.py
> new file mode 100755
> index 0000000000..53fa503d8a
> --- /dev/null
> +++ b/scripts/performance/topN_perf.py
> @@ -0,0 +1,115 @@
> +#!/usr/bin/env python3
> +
> +# Print the top N most executed functions in QEMU using perf.
> +# Example Usage:
> +# topN_perf.py -n 20 -- /path/to/qemu program -program -flags
> +#
The similat comments from commit message section apply here too.
> +# If '-n' is not specified, the default is 25.
> +#
> +# This file is a part of the project "TCG Continuous Benchmarking".
> +#
> +# Copyright (C) 2020 Ahmed Karaman <ahmedkhaledkaraman@gmail.com>
> +# Copyright (C) 2020 Aleksandar Markovic <aleksandar.qemu.devel@gmail.
> com>
> +#
> +# This program is free software: you can redistribute it and/or modify
> +# it under the terms of the GNU General Public License as published by
> +# the Free Software Foundation, either version 2 of the License, or
> +# (at your option) any later version.
> +#
> +# This program is distributed in the hope that it will be useful,
> +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> +# GNU General Public License for more details.
> +#
> +# You should have received a copy of the GNU General Public License
> +# along with this program. If not, see <https://www.gnu.org/licenses/>.
> +
> +import argparse
> +import os
> +import subprocess
> +import sys
> +
> +
> +# Parse the command line arguments
> +parser = argparse.ArgumentParser(usage='topN_perf.py [-h] [-n
> TOP_FUNCTIONS] --'
> + ' /path/to/qemu program -[flags
> PROGRAM_FLAGS]')
> +
Same.
+parser.add_argument('-n', dest='top', type=int, default=25,
> + help='Specify the number of top functions to print.')
> +
> +parser.add_argument('command', type=str, nargs='+',
> help=argparse.SUPPRESS)
> +
> +args = parser.parse_args()
> +
> +# Extract the needed variables from the args
> +command = args.command
> +top = args.top
> +
> +# Insure that perf is installed
> +check_perf = subprocess.run(["which", "perf"], stdout=subprocess.DEVNULL)
> +if check_perf.returncode:
> + sys.exit("Please install perf before running the script!")
> +
OK. This is good.
> +# Insure user has previllage to run perf
> +check_previlage = subprocess.run(["perf", "stat", "ls", "/"],
> + stdout=subprocess.DEVNULL,
> stderr=subprocess.PIPE)
> +if check_previlage.returncode:
> + sys.exit(check_previlage.stderr.decode("utf-8") +
> + "\nOr alternatively, you can run the script with sudo
> privileges!")
I would avoid mixing stderr message and your message practically in one
sentence. Better:
error: <text from stderr>
You must run the script with sudo pivilages, or, alternatively, set kernel
Xxx option to....
> +
> +# Run perf record
> +perf_record = subprocess.run((["perf", "record"] + command),
> + stdout=subprocess.DEVNULL,
> stderr=subprocess.PIPE)
> +if perf_record.returncode:
> + os.unlink('perf.data')
> + sys.exit(perf_record.stderr.decode("utf-8"))
> +
> +# Save perf report output to tmp.perf.data
> +with open("tmp.perf.data", "w") as output:
> + perf_report = subprocess.run(
> + ["perf", "report", "--stdio"], stdout=output,
> stderr=subprocess.PIPE)
> + if perf_report.returncode:
> + os.unlink('perf.data')
> + output.close()
> + os.unlink('tmp.perf.data')
> + sys.exit(perf_report.stderr.decode("utf-8"))
> +
I am really confused by using both perf.data and tmp.perf.data names. Why?
They are in entirely different format. "tmp.perf.data should be
perf.top-function.list, let's say, if I am not mistaken about its meaning
and usage.
+# Read the reported data to functions[]
> +functions = []
> +with open("tmp.perf.data", "r") as data:
> + # Only read lines that are not comments (comments start with #)
> + # Only read lines that are not empty
> + functions = [line for line in data.readlines() if line and line[0]
> + != '#' and line[0] != "\n"]
> +
> +# Limit the number of top functions to "top"
> +number_of_top_functions = top if len(functions) > top else len(functions)
> +
> +# Store the data of the top functions in top_functions[]
> +top_functions = functions[:number_of_top_functions]
> +
> +# Print information headers
# Print table header
> +print('{:>4} {:>10} {:<30} {}\n{} {} {} {}'.format('No.',
> + 'Percentage',
> + 'Name',
> + 'Caller',
> + '-' * 4,
> + '-' * 10,
> + '-' * 30,
> + '-' * 25))
> +
> +
> +# Print top N functions
> +for (index, function) in enumerate(top_functions, start=1):
> + function_data = function.split()
> + function_percentage = function_data[0]
> + function_name = function_data[-1]
> + function_caller = ' '.join(function_data[2:-2])
> + print('{:>4} {:>10} {:<30} {}'.format(index,
> + function_percentage,
> + function_name,
> + function_caller))
> +
> +# Remove intermediate files
> +os.unlink('perf.data')
> +os.unlink('tmp.perf.data')
> --
Thanks,
Aleksandar.
> 2.17.1
>
>
[-- Attachment #2: Type: text/html, Size: 12410 bytes --]
next prev parent reply other threads:[~2020-06-20 10:01 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-19 15:36 [PATCH v2 0/3] Add Scripts for Finding Top 25 Executed Functions Ahmed Karaman
2020-06-19 15:36 ` [PATCH v2 1/3] scripts/performance: Add topN_perf.py script Ahmed Karaman
2020-06-20 9:59 ` Aleksandar Markovic [this message]
2020-06-20 10:01 ` Aleksandar Markovic
2020-06-19 15:36 ` [PATCH v2 2/3] scripts/performance: Add topN_callgrind.py script Ahmed Karaman
2020-06-20 10:03 ` Aleksandar Markovic
2020-06-19 15:36 ` [PATCH v2 3/3] MAINTAINERS: Add 'Performance Tools and Tests' subsection Ahmed Karaman
2020-06-19 16:04 ` [PATCH v2 0/3] Add Scripts for Finding Top 25 Executed Functions no-reply
2020-06-20 9:14 ` Aleksandar Markovic
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='CAHiYmc79=PaKFeF0og5BW_QctDTS-cRudr3ZXjTab7D+_snb6w@mail.gmail.com' \
--to=aleksandar.qemu.devel@gmail.com \
--cc=ahmedkhaledkaraman@gmail.com \
--cc=alex.bennee@linaro.org \
--cc=crosa@redhat.com \
--cc=ehabkost@redhat.com \
--cc=ldoktor@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=rth@twiddle.net \
/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;
as well as URLs for NNTP newsgroup(s).