Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
* [Buildroot] backtrace()  depth
@ 2013-09-10 19:18 Peter Kümmel
  2013-09-10 19:40 ` Samuel Martin
  0 siblings, 1 reply; 3+ messages in thread
From: Peter Kümmel @ 2013-09-10 19:18 UTC (permalink / raw)
  To: buildroot

I've a question about the depth of a backtrace.

I've configured uClibc/Buildroot to build libubacktrace-0.9.33.2.so
via the standard Buildroot menuconfigs:
Buildroot 2013.02, GCC 4.7.2 by Buildroot, uClibc-0.9.33

When I call backtrace() on a segfault, I only see the function
where the segfault is produced, and not much more (only /lib/libc.so.0,
and two more addresses). The function name could be demanded correctly,
so I assume the compiler flags are ok (-rdynamic).

The segfault is not in the main thread
(overall 4 threads, /lib/libpthread-0.9.33.2.so)
And the bt looks equal for debug and release builds.

A build with crosstool-ng and eglibc produces the same bad backtrace.

Is there a way to get a more complete backtrace?

Many thanks,
Peter

^ permalink raw reply	[flat|nested] 3+ messages in thread

* [Buildroot] backtrace() depth
  2013-09-10 19:18 [Buildroot] backtrace() depth Peter Kümmel
@ 2013-09-10 19:40 ` Samuel Martin
  2013-09-10 19:53   ` Peter Kümmel
  0 siblings, 1 reply; 3+ messages in thread
From: Samuel Martin @ 2013-09-10 19:40 UTC (permalink / raw)
  To: buildroot

Peter,


2013/9/10 Peter K?mmel <syntheticpp@gmx.net>

> I've a question about the depth of a backtrace.
>
> I've configured uClibc/Buildroot to build libubacktrace-0.9.33.2.so
> via the standard Buildroot menuconfigs:
> Buildroot 2013.02, GCC 4.7.2 by Buildroot, uClibc-0.9.33
>
> When I call backtrace() on a segfault, I only see the function
> where the segfault is produced, and not much more (only /lib/libc.so.0,
> and two more addresses). The function name could be demanded correctly,
> so I assume the compiler flags are ok (-rdynamic).
>
> The segfault is not in the main thread
> (overall 4 threads, /lib/libpthread-0.9.33.2.so)
> And the bt looks equal for debug and release builds.
>
> A build with crosstool-ng and eglibc produces the same bad backtrace.
>
> Is there a way to get a more complete backtrace?
>

Is the strip command set to none (under the "Build options" menu)?
if not, then you should certainly added the binaries you are debgging to the
BR2_STRIP_EXCLUDE_FILES or  BR2_STRIP_EXCLUDE_DIRS lists.


Regards,

-- 
Samuel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20130910/22183164/attachment.html>

^ permalink raw reply	[flat|nested] 3+ messages in thread

* [Buildroot] backtrace() depth
  2013-09-10 19:40 ` Samuel Martin
@ 2013-09-10 19:53   ` Peter Kümmel
  0 siblings, 0 replies; 3+ messages in thread
From: Peter Kümmel @ 2013-09-10 19:53 UTC (permalink / raw)
  To: buildroot

>
> Is the strip command set to none (under the "Build options" menu)?
> if not, then you should certainly added the binaries you are debgging to the
> BR2_STRIP_EXCLUDE_FILES or  BR2_STRIP_EXCLUDE_DIRS lists.

Hi Samuel,

I even selected sstrip for the compiler and rootfs, but the the application
itself is not stripped, it's not build by buildroot.
Does the bt stop at the first stripped lib it reaches?

Peter

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2013-09-10 19:53 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-09-10 19:18 [Buildroot] backtrace() depth Peter Kümmel
2013-09-10 19:40 ` Samuel Martin
2013-09-10 19:53   ` Peter Kümmel

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox