Possible bug in ifort linux when using PETSc

[Wee_Beng_T_]

Hi,

I'm using a library called PETSc to solve my linear equations. They have a new feature which works in gfortran and ifort windows in VS2008. However, it gives segmentation fault in ifort linux. I was told by the developers that there's a bug in ifort linux.

I was wondering if there's a way to test if it's really a bug in ifort linux, and whether there's a way to fix it.

Can I send in a code or something like that?

Thanks!

[Steven_L_Intel1]

If you have a commercial or academic license, you can submit a bug report through Intel Premier Support. Even without such a license, you can submit a report here, being sure to give us everything needed to reproduce the problem. Please also identify which exact version of ifort you are using. It is very unusual for there to be a compiler bug on one OS only (though it does happen.)

[TimP]

One of the more common reasons for segfault in linux is missing a requirement to adjust stack, e.g. 'ulimit -s unlimited'. There is a checklist posted at the top of the forum for diagnosing segfaults. I

[Wee_Beng_T_]

Hi,

I've used the checklist to compile and run, but the same error happens. My school uses the academic version but I think it's easier to post here.

The code uses the PETSc library (http://www.mcs.anl.gov/petsc/). The library is also compiled using the same ifort compiler.

The code is an example given in the code.

What do I have to supply specifically?

I have the source code. Is the a.out file enough?

The version is :

Intel Fortran Intel 64 Compiler Professional for applications running on Intel 64, Version 11.1 Build 20091012 Package ID: l_cprof_p_11.1.059
Copyright (C) 1985-2009 Intel Corporation. All rights reserved.

Thank you!