Re: application crash - looking for suggestion

2008-10-14 Thread Bernhard Reutner-Fischer
On Tue, Oct 14, 2008 at 12:42:31PM +0800, Ming-Ching Tiew wrote:

My app crashed in uClibc after running for about a week, 
and this is the kernel dmesg log :-

6my_app[5656]: segfault at fea0 ip b743a437 sp 0818ff04 
error 6 in libuClibc-0.9.28.so[b740e000+37000]

Is there any suggestions for giving additional clues to the
crash ( perhaps the function calls stack ) ? 

Run your app (perhaps via a screen session) in gdb and show us
'bt f'.

Compile uClibc and your app with something like -g3 -ggdb3 (refer to
the documentation of your compiler).

Since I have the sources to my_app and uclibc, I don't 
mind recompile everything as long as it does not impact
performance to badly.

As long as you don't build with optimizations disabled the speed should
be about the same.

Would 'sstrip' impact the capability to debug a library and
an app ?

yes, strip -d (and sstrip for that matter) will remove debug
information.

You didn't say which version of uClibc you use on which arch, nor if
your app uses threads etc.
___
uClibc mailing list
uClibc@uclibc.org
http://busybox.net/cgi-bin/mailman/listinfo/uclibc


Re: application crash - looking for suggestion

2008-10-14 Thread Ming-Ching Tiew
Bernhard Reutner-Fischer wrote:
 On Tue, Oct 14, 2008 at 12:42:31PM +0800, Ming-Ching Tiew wrote:
 
 My app crashed in uClibc after running for about a week,
 and this is the kernel dmesg log :-
 
 6my_app[5656]: segfault at fea0 ip b743a437 sp 0818ff04
 error 6 in libuClibc-0.9.28.so[b740e000+37000]
 
 Is there any suggestions for giving additional clues to the
 crash ( perhaps the function calls stack ) ?
 

Is there anyway I could make sense from the whole bunch of
numbers show above ?

 
 You didn't say which version of uClibc you use on which arch, nor if
 your app uses threads etc.

Yeah I know it's old :- uclibc-0.9.28

:-(

Yes my app uses threads extensively.


___
uClibc mailing list
uClibc@uclibc.org
http://busybox.net/cgi-bin/mailman/listinfo/uclibc


Re: How goes the -rc2 todo list?

2008-10-14 Thread Bernhard Reutner-Fischer
On Sun, Oct 12, 2008 at 10:31:29PM +0200, Bernhard Reutner-Fischer wrote:
On Sun, Oct 12, 2008 at 08:25:17AM +0200, Carmelo Amoroso wrote:

 There is a depency problem in locales, i'll fix this during the weekend
 and then announce rc2.

 I'll test Bernhard's locale patch on Manday.
 I don't have a sh4 box at home.

$ svn diff uClibc.check/ | diffstat -up0
 uClibc.check/Makefile.in|5 
 uClibc.check/extra/locale/Makefile.in   |  118 +---
 uClibc.check/extra/locale/gen_wctype.c  |   94 ++--
 uClibc.check/extra/locale/gen_collate.c |  234 
 
 uClibc.check/extra/locale/gen_locale.c  |  226 --
 uClibc.check/extra/Configs/Config.in|2 
 uClibc.check/Makerules  |   13 +
 uClibc.check/Rules.mak  |2 
 8 files changed, 403 insertions(+), 291 deletions(-)

A couple of you saw dependency problems in locale, i'm asking anybody to
provide feedback if it works for you now.

TIA,

PS: As you can see this will add a V=2 mode to be verbose in
locale-generation, as verbose as it previously was per default.
This change is responsible for the vast majority of the patch,
unfortunately. I didn't even diff the generated data, so if you can save
away extra/locale before applying the patch, please check if the
generated data is still the same afterwards (and let me know if it is
not ;)

I've now installed this as r23677. Thanks to HcE for giving it cursory
testing.
___
uClibc mailing list
uClibc@uclibc.org
http://busybox.net/cgi-bin/mailman/listinfo/uclibc


Re: application crash - looking for suggestion

2008-10-14 Thread Bernhard Reutner-Fischer
On Tue, Oct 14, 2008 at 04:32:49PM +0800, Ming-Ching Tiew wrote:
Bernhard Reutner-Fischer wrote:
 On Tue, Oct 14, 2008 at 12:42:31PM +0800, Ming-Ching Tiew wrote:
 
 My app crashed in uClibc after running for about a week,
 and this is the kernel dmesg log :-
 
 6my_app[5656]: segfault at fea0 ip b743a437 sp 0818ff04
 error 6 in libuClibc-0.9.28.so[b740e000+37000]
 
 Is there any suggestions for giving additional clues to the
 crash ( perhaps the function calls stack ) ?
 

Is there anyway I could make sense from the whole bunch of
numbers show above ?

Sure, look at your app and the library near those spots.

 
 You didn't say which version of uClibc you use on which arch, nor if
 your app uses threads etc.

Yeah I know it's old :- uclibc-0.9.28

:-(

Yes my app uses threads extensively.

So this is likely fixed on trunk.
___
uClibc mailing list
uClibc@uclibc.org
http://busybox.net/cgi-bin/mailman/listinfo/uclibc