linux-kernel - Re: Oprofile : need to adjust PC by 16 bytes

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]

Date:	Mon, 17 Nov 2008 16:24:34 +0100
From:	Andi Kleen <andi@...stfloor.org>
To:	eranian@...il.com
Cc:	Andi Kleen <andi@...stfloor.org>,
	Eric Dumazet <dada1@...mosbay.com>,
	Mikael Pettersson <mikpe@...uu.se>,
	Robert Richter <robert.richter@....com>,
	oprofile-list@...ts.sf.net, Ingo Molnar <mingo@...e.hu>,
	Jiri Kosina <jkosina@...e.cz>, Jiri Benc <jbenc@...e.cz>,
	Vilem Marsik <vmarsik@...e.cz>,
	Pekka Enberg <penberg@...helsinki.fi>,
	linux-kernel@...r.kernel.org
Subject: Re: Oprofile : need to adjust PC by 16 bytes

On Mon, Nov 17, 2008 at 04:02:19PM +0100, stephane eranian wrote:

Thanks for explaining it better than I could.

Just a minor correction.

> But it can be very large when overflow happens
> during a kernel critical section
> where interrupts are off.

Actually oprofile uses the NMI to avoid that particular problem.
It still has limits, e.g. no profiling of SMM sections.

> There is nothing SW can do about all of this.

Yes profiling on a out of order machine is hard.

> Furthermore, PEBS can record samples while
> in kernel critical sections. 

Again with oprofile that works too without PEBS.

-Andi
--
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/