Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1647696 > unrolled thread

Re: [PATCH 7/7] DWARF: add the config option

Started byPeter Zijlstra <peterz@infradead.org>
First post2017-05-23 09:10 +0200
Last post2017-05-23 09:30 +0200
Articles 2 — 2 participants

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Re: [PATCH 7/7] DWARF: add the config option Peter Zijlstra <peterz@infradead.org> - 2017-05-23 09:10 +0200
    Re: [PATCH 7/7] DWARF: add the config option Ingo Molnar <mingo@kernel.org> - 2017-05-23 09:30 +0200

#1647696 — Re: [PATCH 7/7] DWARF: add the config option

FromPeter Zijlstra <peterz@infradead.org>
Date2017-05-23 09:10 +0200
SubjectRe: [PATCH 7/7] DWARF: add the config option
Message-ID<tKc78-8iP-15@gated-at.bofh.it>
On Wed, May 10, 2017 at 10:32:06AM +0200, Jiri Slaby wrote:
> > But it does hurt, in the sense that the complicated format of DWARF CFI
> > means the unwinder has to jump through a lot more hoops to read it.
> 
> Why that matters, actually? Unwinder is nothing to be performance
> oriented. And if somebody is doing a lot of unwinding during runtime,
> they can switch to in-this-case-faster FP unwinder.

perf (and ftrace) like the unwinder to be considered performance
oriented.

[toc] | [next] | [standalone]


#1647723

FromIngo Molnar <mingo@kernel.org>
Date2017-05-23 09:30 +0200
Message-ID<tKcqu-8q9-17@gated-at.bofh.it>
In reply to#1647696
* Peter Zijlstra <peterz@infradead.org> wrote:

> On Wed, May 10, 2017 at 10:32:06AM +0200, Jiri Slaby wrote:
> > > But it does hurt, in the sense that the complicated format of DWARF CFI
> > > means the unwinder has to jump through a lot more hoops to read it.
> > 
> > Why that matters, actually? Unwinder is nothing to be performance
> > oriented. And if somebody is doing a lot of unwinding during runtime,
> > they can switch to in-this-case-faster FP unwinder.
> 
> perf (and ftrace) like the unwinder to be considered performance
> oriented.

Yes, and given how critical debugging is there's a kind of useful synergy here: 
overall the perf unwinder is run about 10 orders of magnitude more often than the 
debugging unwinder, so it's a very big help in shaking out unwinder/debug-info 
bugs and increasing robustness overall.

The 'price' for using the unwinder in perf is that it has to be fast.

Thanks,

	Ingo

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web