[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Libunwind-devel] unw_backtrace(): Performance slowdown for ARM?

From: Milian Wolff
Subject: Re: [Libunwind-devel] unw_backtrace(): Performance slowdown for ARM?
Date: Wed, 17 Jan 2018 10:07:57 +0100

On Tuesday, January 16, 2018 4:33:12 AM CET Luke Diamand wrote:
> Hi!
> I could be wrong about this, but I'm finding a roughly 10x slowdown on
> ARM between 1.2.1 and current HEAD (e1ca8748).
> I have a small test program here:
> On my Raspberry Pi 2, this is getting ~180us to do a self-backtrace in
> 1.2.1 and around 1.8ms in HEAD.
> I tried running perf, and that seems to be saying that it is opening
> and closing files a lot.

How did you run perf? For starters, can you use `perf stat` on the two 
versions and your benchmark so that we can see whether it's CPU bound or not?

Then, please record a profile with

perf record --call-graph dwarf 

and then inspect that with "perf report" on your ARM machine or hotspot [1] on 
your devel machine, assuming you have a sysroot with the debug symbols 
available there.

This should make it much easier to find out where the performance is going to.


> I also had a go at git bisecting, but there's a separate bug which was
> introduced and then fixed which caused it to segfault, which overlaps
> with this problem, along with some other performance regression fixes.
> 90d0d15c4ec3bb812dcb44c081757ebf6756cc6b didn't crash, but was slow.
> 2a5d1a629634b3225fccb1ed8a0a765bccc77f2e crashed.
> v1.2.1 is fast.
> I'm configuring with:
> CC=arm-linux-gnueabihf-gcc CFLAGS="-g" configure --disable-debug
> --host=arm-linux \
>     --enable-shared --disable-static --with-sysroot=/home/lgd/arm
> --enable-maintainer-mode \
>     --disable-debug-frame --disable-block-signals
> Any ideas as to what's going on?
> Thanks
> Luke
> _______________________________________________
> Libunwind-devel mailing list
> address@hidden

Milian Wolff

reply via email to

[Prev in Thread] Current Thread [Next in Thread]