linux-kernel - Re: [PATCH] Expand CPU compiler options

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1355054935.13447.YahooMailNeo@web140004.mail.bf1.yahoo.com>
Date:	Sun, 9 Dec 2012 04:08:55 -0800 (PST)
From:	John <da_audiophile@...oo.com>
To:	Borislav Petkov <bp@...en8.de>
Cc:	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] Expand CPU compiler options

> Let's see, if I'm reading the log file correctly, the average values of

> each test run differ by ~ 0.1 seconds tops.
> 
> For example, i7-3770K generic build gives on average 69.41404 while
> the more optimized version 69.33554. The diff between the two is even
> less than 0.1 second. The other two machines' diff is a bit higher. And
> from looking at your graphs, this is all eaten up by stddev so I'd say
> there aren't any improvements from using a different uarch target - just
> noise. AFAICT, at least.

While I agree that the differences as small - on the order of ms - they are not insignificant nor are they in the noise of the measurements.  
1) All the assumptions for ANOVA are met:
*Data are normally distributed as show in the normal quantile plots.

*The population variances are fairly equal (Levene and Barlett tests).

2) The ANOVA plots clearly show significance.
*Pair-wise analysis by Tukey-Kramer shows significance at the 0.05 level for all CPUs compared.  Below are the differences in median values:
core2 +87.5 ms
corei7-avx +79.7 ms
core-avx-i+257.2 ms

> In any case, these results are too marginal to warrant any code change
> since they're basically disappearing in noise.

Currently, the 'core2' optimization is included in as an option for a CPU Family.  In the case of core-avx-i, I am showing an improvement of ~3x over that.  And in the corei7-avx case, the improvement is approximately equal to it.  I don't understand why these wouldn't be considered value-added since they preform at least as good as the option already included.

Thank you for the consideration and suggestions.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/