linux-kernel - Re: perf_event self-monitoring overhead regression

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] [day] [month] [year] [list]

Message-ID: <alpine.DEB.2.00.1112091717390.18889@cl320.eecs.utk.edu>
Date:	Fri, 9 Dec 2011 17:22:09 -0500
From:	Vince Weaver <vweaver1@...s.utk.edu>
To:	<eranian@...il.com>
CC:	Linus Torvalds <torvalds@...ux-foundation.org>,
	Peter Zijlstra <a.p.zijlstra@...llo.nl>,
	Ingo Molnar <mingo@...e.hu>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	Paul Mackerras <paulus@...ba.org>,
	Arnaldo Carvalho de Melo <acme@...stprotocols.net>
Subject: Re: perf_event self-monitoring overhead regression

Here are the complete results, using official kernels.

   http://web.eecs.utk.edu/~vweaver1/projects/perf-events/benchmarks/rdtsc_overhead/

The overall overhead start/stop/read has been getting worse.  3.1.0 was 
bad and the current 3.2.0-rc4-tip tree is even worse.

I'm not sure if it's worth trying to track down the changes that cause 
these regressions, as they might still be considered just noise, though 
the best to worst performance degredation is ~ 15% when attempting to 
self-monitor with 3 events on a core2 machine.

Vince

--
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/