| 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
| ||
|
Message-ID: <3937e057-6b47-77fe-9440-ade079ee2cfe@amd.com>
Date: Wed, 28 Feb 2024 10:13:08 +0530
From: K Prateek Nayak <kprateek.nayak@....com>
To: John Stultz <jstultz@...gle.com>, LKML <linux-kernel@...r.kernel.org>
Cc: Joel Fernandes <joelaf@...gle.com>, Qais Yousef <qyousef@...gle.com>,
Ingo Molnar <mingo@...hat.com>, Peter Zijlstra <peterz@...radead.org>,
Juri Lelli <juri.lelli@...hat.com>,
Vincent Guittot <vincent.guittot@...aro.org>,
Dietmar Eggemann <dietmar.eggemann@....com>,
Valentin Schneider <vschneid@...hat.com>,
Steven Rostedt <rostedt@...dmis.org>, Ben Segall <bsegall@...gle.com>,
Zimuzo Ezeozue <zezeozue@...gle.com>, Youssef Esmat
<youssefesmat@...gle.com>, Mel Gorman <mgorman@...e.de>,
Daniel Bristot de Oliveira <bristot@...hat.com>,
Will Deacon <will@...nel.org>, Waiman Long <longman@...hat.com>,
Boqun Feng <boqun.feng@...il.com>, "Paul E. McKenney" <paulmck@...nel.org>,
Metin Kaya <Metin.Kaya@....com>, Xuewen Yan <xuewen.yan94@...il.com>,
Thomas Gleixner <tglx@...utronix.de>, kernel-team@...roid.com
Subject: Re: [RESEND][PATCH v8 0/7] Preparatory changes for Proxy Execution v8
Hello John,
Happy to report that I did not see any regressions with the series
as expected. Full results below.
On 2/24/2024 5:41 AM, John Stultz wrote:
> After sending out v7 of Proxy Execution, I got feedback that the
> patch series was getting a bit unwieldy to review, and Qais
> suggested I break out just the cleanups/preparatory components
> of the patch series and submit them on their own in the hope we
> can start to merge the less complex bits and discussion can focus
> on the more complicated portions afterwards.
>
> So for the v8 of this series, I only submitted those earlier
> cleanup/preparatory changes:
> https://lore.kernel.org/lkml/20240210002328.4126422-1-jstultz@google.com/
>
> After sending this out a few weeks back, I’ve not heard much, so
> I wanted to resend this again.
>
> (I did correct one detail here, which was that I had accidentally
> lost the author credit to one of the patches, and I’ve fixed that
> in this submission).
>
> As before, If you are interested, the full v8 series, it can be
> found here:
> https://github.com/johnstultz-work/linux-dev/commits/proxy-exec-v8-6.8-rc3
> https://github.com/johnstultz-work/linux-dev.git proxy-exec-v8-6.8-rc3
>
> However, I’ve been focusing pretty intensely on the series to
> shake out some issues with the more complicated later patches in
> the series (not in what I’m submitting here), and have resolved
> a number of problems I uncovered in doing wider testing (along
> with lots of review feedback from Metin), so v9 and all of its
> improvements will hopefully be ready to send out soon.
>
> If you want a preview, my current WIP tree (careful, as I rebase
> it frequently) is here:
> https://github.com/johnstultz-work/linux-dev/commits/proxy-exec-WIP
> https://github.com/johnstultz-work/linux-dev.git proxy-exec-WIP
>
> Review and feedback would be greatly appreciated!
o System Details
- 3rd Generation EPYC System
- 2 x 64C/128T
- NPS1 mode
o Kernels
tip: tip:sched/core at commit 8cec3dd9e593 ("sched/core:
Simplify code by removing duplicate #ifdefs")
proxy-setup: tip + this series
o Results
==================================================================
Test : hackbench
Units : Normalized time in seconds
Interpretation: Lower is better
Statistic : AMean
==================================================================
Case: tip[pct imp](CV) proxy-setup[pct imp](CV)
1-groups 1.00 [ -0.00]( 2.08) 1.01 [ -0.53]( 2.45)
2-groups 1.00 [ -0.00]( 0.89) 1.03 [ -3.32]( 1.48)
4-groups 1.00 [ -0.00]( 0.81) 1.02 [ -2.26]( 1.22)
8-groups 1.00 [ -0.00]( 0.78) 1.00 [ -0.29]( 0.97)
16-groups 1.00 [ -0.00]( 1.60) 1.00 [ -0.27]( 1.86)
==================================================================
Test : tbench
Units : Normalized throughput
Interpretation: Higher is better
Statistic : AMean
==================================================================
Clients: tip[pct imp](CV) proxy-setup[pct imp](CV)
1 1.00 [ 0.00]( 0.71) 1.00 [ 0.31]( 0.37)
2 1.00 [ 0.00]( 0.25) 0.99 [ -0.56]( 0.31)
4 1.00 [ 0.00]( 0.85) 0.98 [ -2.35]( 0.69)
8 1.00 [ 0.00]( 1.00) 0.99 [ -0.99]( 0.12)
16 1.00 [ 0.00]( 1.25) 0.99 [ -0.78]( 1.35)
32 1.00 [ 0.00]( 0.35) 1.00 [ 0.12]( 2.23)
64 1.00 [ 0.00]( 0.71) 0.99 [ -0.97]( 0.55)
128 1.00 [ 0.00]( 0.46) 0.96 [ -4.38]( 0.47)
256 1.00 [ 0.00]( 0.24) 0.99 [ -1.32]( 0.95)
512 1.00 [ 0.00]( 0.30) 0.98 [ -1.52]( 0.10)
1024 1.00 [ 0.00]( 0.40) 0.98 [ -1.59]( 0.23)
==================================================================
Test : stream-10
Units : Normalized Bandwidth, MB/s
Interpretation: Higher is better
Statistic : HMean
==================================================================
Test: tip[pct imp](CV) proxy-setup[pct imp](CV)
Copy 1.00 [ 0.00]( 9.73) 1.04 [ 4.18]( 3.12)
Scale 1.00 [ 0.00]( 5.57) 0.99 [ -1.35]( 5.74)
Add 1.00 [ 0.00]( 5.43) 0.99 [ -1.29]( 5.93)
Triad 1.00 [ 0.00]( 5.50) 0.97 [ -3.47]( 7.81)
==================================================================
Test : stream-100
Units : Normalized Bandwidth, MB/s
Interpretation: Higher is better
Statistic : HMean
==================================================================
Test: tip[pct imp](CV) proxy-setup[pct imp](CV)
Copy 1.00 [ 0.00]( 3.26) 1.01 [ 0.83]( 2.69)
Scale 1.00 [ 0.00]( 1.26) 1.00 [ -0.32]( 4.52)
Add 1.00 [ 0.00]( 1.47) 1.01 [ 0.63]( 0.96)
Triad 1.00 [ 0.00]( 1.77) 1.02 [ 1.81]( 1.00)
==================================================================
Test : netperf
Units : Normalized Througput
Interpretation: Higher is better
Statistic : AMean
==================================================================
Clients: tip[pct imp](CV) proxy-setup[pct imp](CV)
1-clients 1.00 [ 0.00]( 0.22) 0.99 [ -0.53]( 0.26)
2-clients 1.00 [ 0.00]( 0.57) 1.00 [ -0.44]( 0.41)
4-clients 1.00 [ 0.00]( 0.43) 1.00 [ -0.48]( 0.39)
8-clients 1.00 [ 0.00]( 0.27) 1.00 [ -0.31]( 0.42)
16-clients 1.00 [ 0.00]( 0.46) 1.00 [ -0.11]( 0.42)
32-clients 1.00 [ 0.00]( 0.95) 1.00 [ -0.41]( 0.56)
64-clients 1.00 [ 0.00]( 1.79) 1.00 [ -0.15]( 1.65)
128-clients 1.00 [ 0.00]( 0.89) 1.00 [ -0.43]( 0.80)
256-clients 1.00 [ 0.00]( 3.88) 1.00 [ -0.37]( 4.74)
512-clients 1.00 [ 0.00](35.06) 1.01 [ 1.05](50.84)
==================================================================
Test : schbench
Units : Normalized 99th percentile latency in us
Interpretation: Lower is better
Statistic : Median
==================================================================
#workers: tip[pct imp](CV) proxy-setup[pct imp](CV)
1 1.00 [ -0.00](27.28) 1.31 [-31.25]( 2.38)
2 1.00 [ -0.00]( 3.85) 1.00 [ -0.00]( 8.85)
4 1.00 [ -0.00](14.00) 1.11 [-10.53](11.18)
8 1.00 [ -0.00]( 4.68) 1.08 [ -8.33]( 9.93)
16 1.00 [ -0.00]( 4.08) 0.92 [ 8.06]( 3.70)
32 1.00 [ -0.00]( 6.68) 0.95 [ 5.10]( 2.22)
64 1.00 [ -0.00]( 1.79) 0.99 [ 1.02]( 3.18)
128 1.00 [ -0.00]( 6.30) 1.02 [ -2.48]( 7.37)
256 1.00 [ -0.00](43.39) 1.00 [ -0.00](37.06)
512 1.00 [ -0.00]( 2.26) 0.98 [ 1.88]( 6.96)
Note: schbench is known to have high run to run variance for
16-workers and below.
==================================================================
Test : Unixbench
Units : Normalized scores
Interpretation: Lower is better
Statistic : Various (Mentioned)
==================================================================
Metric Variant tip proxy-setup
Hmean unixbench-dhry2reg-1 0.00% -0.60%
Hmean unixbench-dhry2reg-512 0.00% -0.01%
Amean unixbench-syscall-1 0.00% -0.41%
Amean unixbench-syscall-512 0.00% 0.13%
Hmean unixbench-pipe-1 0.00% 1.02%
Hmean unixbench-pipe-512 0.00% 0.53%
Hmean unixbench-spawn-1 0.00% -2.68%
Hmean unixbench-spawn-512 0.00% 3.24%
Hmean unixbench-execl-1 0.00% 0.61%
Hmean unixbench-execl-512 0.00% 1.97%
--
Tested-by: K Prateek Nayak <kprateek.nayak@....com>
>
> Thanks so much!
> -john
>
> [..snip..]
>
--
Thanks and Regards,
Prateek
Powered by blists - more mailing lists