|
From: Tom H. <th...@cy...> - 2004-10-10 16:02:08
|
In message <Pin...@jd...>
Igmar Palsenberg <mai...@jd...> wrote:
> I was force to do some coding on an older system (glibc 2.2.4 based), and
> I noticed it takes considerably more time to run an app under valgrind
> then on a somewhat newer system.
I am aware of this, as it makes RedHat 7.2 systems very slow. I've
never really been able to get to the bottom of it though. The time
is all taken during startup - once things are up and running it seems
to be fine.
The best I was able to work out was that it was something to do with
the memory mapping as strace indicated that mmap was about the only
thing being done during the delay.
> I can see VG looping the CPU. To give an indication :
>
> real 1m44.480s
> user 1m31.300s
> sys 0m0.950s
Interesting that it's all user time. I was thinking it was the kernel
taking a long time to do the mmap calls.
> Ik locks up a long time at
>
> ==13832== Reading syms from /lib/libc-2.2.4.so (0xB1027000)
> <messages about redirection>
> ==13832== Reading syms from /lib/libc-2.2.4.so (0x1B915000)
>
> bit eventually does finish. I'll see if I can bump the libc version to
> 2.2.5 and see if that makes a difference. It this bug-file worthy material
> ?
It's probably worth filing a bug to get it on the record.
Tom
--
Tom Hughes (th...@cy...)
Software Engineer, Cyberscience Corporation
http://www.cyberscience.com/
|