This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: [gfortran,patch, RFC] Allow coredump and backtrace on runtime error

From: FX Coudert <fxcoudert at gmail dot com>
To: Daniel Franke <franke dot daniel at gmail dot com>, "fortran at gcc dot gnu dot org List" <fortran at gcc dot gnu dot org>
Date: Sun, 26 Nov 2006 14:27:39 +0100
Subject: Re: [gfortran,patch, RFC] Allow coredump and backtrace on runtime error
References: <19c433eb0611240855v68a78b1ftda3ef385932300e6@mail.gmail.com> <640ad44b0611250809nca053c3i21128fc996ecb616@mail.gmail.com>

I'm abroad, therefore I can not test this myself. How does your patch
work with OMP, i.e. multiple threads? Will each thread dump its own
backtrace? If yes, will they be separable?

From my experiments, only one thread dumps it backtrace, and it's not a full backtrace but only from the point where the thread was created:

Backtrace for this error:
  + function gee_.omp_fn.1 (0x400AAC)
    at line 15 of file u.f90
  + /utmp/coudert/gfortran/irun/lib64/libgomp.so.1 [0x2a9570d17c]
  + /lib64/tls/libpthread.so.0 [0x3ff370610a]
  + /lib64/tls/libc.so.6(__clone+0x73) [0x3ff24c5ee3]

We can work later on improving this, but I think it's not so bad a behaviour for now.

FX

References:
- [gfortran,patch, RFC] Allow coredump and backtrace on runtime error
  - From: François-Xavier Coudert

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]