linux-kernel - Re: [PATCH 4/6] trace: Add tracepoints to reschedule interrupt handler

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: <CAL26m8JvhzkweuCD1NGvbr_xDEskF3me5FxTpuX7o5DiwHCJTg@mail.gmail.com>
Date:	Thu, 21 Jul 2011 11:04:01 -0700
From:	Vaibhav Nagarnaik <vnagarnaik@...gle.com>
To:	Peter Zijlstra <a.p.zijlstra@...llo.nl>
Cc:	David Sharp <dhsharp@...gle.com>,
	Frederic Weisbecker <fweisbec@...il.com>,
	Thomas Gleixner <tglx@...utronix.de>,
	Ingo Molnar <mingo@...hat.com>,
	Steven Rostedt <rostedt@...dmis.org>,
	Michael Rubin <mrubin@...gle.com>,
	linux-kernel@...r.kernel.org, x86@...nel.org
Subject: Re: [PATCH 4/6] trace: Add tracepoints to reschedule interrupt handler

On Thu, Jul 21, 2011 at 5:31 AM, Peter Zijlstra <a.p.zijlstra@...llo.nl> wrote:
> On Fri, 2011-07-15 at 10:50 -0700, David Sharp wrote:
>> On Fri, Jul 15, 2011 at 1:48 AM, Peter Zijlstra <a.p.zijlstra@...llo.nl> wrote:
>> > On Fri, 2011-07-15 at 00:02 +0200, Frederic Weisbecker wrote:
>> >> > This is a part of overall effort to trace all the interrupts happening
>> >> > in a system to figure out what time is spent in kernel space versus user
>> >> > space.
>> >
>> > How about you enable CONFIG_IRQ_TIME_ACCOUNTING and get all that?
>> >
>>
>> It's not just about how long, it's when, and what other events the
>> interrupts are interacting with.
>
> But that's not what the Changelog said, how am I supposed to smell that?

I will update the changelog to be clearer.

>
> As to the patch in particular, I hate it, why do you need separate
> tracepoints for the resched ipi and wtf is up with that ignore argument.
>

By separate tracepoints if you mean the entry and exit tracepoints, we
use it to trace the interrupts in the system and get the timing for each
interrupt to complete. When interleaved with the other system trace, we
get how it affected other tasks. It would be hard to know that by using
CONFIG_IRQ_TIME_ACCOUNTING.

The 'ignore' argument is due to the limitation of the TRACE_EVENT macro,
which requires at least 1 argument passed to it. I could not think of a
suitable data argument for the tracepoint, so I had to use a dummy
argument. If you can suggest something that can be used instead, I will
be glad to update the patch.


Vaibhav Nagarnaik
--
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/