Segmentation fault when using cmake on dev build.

[Jason Sheldon]

Hi All,

I've been running into a segmentation fault when I try and use cmake on the most recent deal.ii dev build.

specifically when it tries to do the following
"
-- Include /Users/jvs5482/lib/deal.ii/cmake/configure/configure_boost.cmake
Segmentation fault: 11
"

I checked out revision 32901 (May 15 2014) and was able to run cmake without error.  I'm not sure if this is the most recent working version, I just picked one from the log that was before all the recent cmake changes.

Jason

--

Jason Sheldon, M.S.
Research Assistant at the Pennsylvania State University
Department of Engineering Science and Mechanics

[Wolfgang Bangerth]

> I've been running into a segmentation fault when I try and use cmake on the
> most recent deal.ii dev build.
>
> specifically when it tries to do the following
> "
> -- Include /Users/jvs5482/lib/deal.ii/cmake/configure/configure_boost.cmake
> Segmentation fault: 11
> "
>
> I checked out revision 32901 (May 15 2014) and was able to run cmake without
> error. I'm not sure if this is the most recent working version, I just picked
> one from the log that was before all the recent cmake changes.

I saw this as well last week on a colleague's Mac laptop. Are you on Mac as well?

In general, I don't quite know how we would address a segfault in cmake. Can
you try to find which version broke it, using bisection of the revisions
between 32901 and today?

Best
W.

--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@math.tamu.edu
www: http://www.math.tamu.edu/~bangerth/

[Scott Miller]

I saw this as well last week on a colleague's Mac laptop. Are you on Mac as well?

It is happening on my Mac as well.

 

In general, I don't quite know how we would address a segfault in cmake. Can
you try to find which version broke it, using bisection of the revisions
between 32901 and today?

Maybe I will let Jason figure that out.  I did dig around a little bit and found the following

calls that lead to the seg fault:

1. cmake/configure/configure_boost.cmake : 

CONFIGURE_FEATURE(BOOST)

2. cmake/macros/macro_configure_feature.cmake : 

FEATURE_FIND_EXTERNAL(${_feature} FEATURE_${_feature}_EXTERNAL_FOUND)

3. cmake/macros/macro_configure_feature.cmake : 

FIND_PACKAGE(${_feature})

4.  cmake/macros/macro_find_package.cmake : 

_FIND_PACKAGE (${_package_name} ${ARGN})

It looks almost as if _FIND_PACKAGE is calling FIND_PACKAGE recursively until the set fault occurs.

Printing, I see that ${_package_name_uppercase} = BOOST, 

${ARGN} = 1.44COMPONENTSiostreamsserializationsystemthread

Also, I have tried using bundled as well as pointing to an external install, and I 

get the same error each time.

Best,

-Scott

[Matthias Maier]

Can you tell me the CMake versions in question for reference?


Am 28. Mai 2014, 15:05 schrieb Scott Miller <miller....@gmail.com>:

> FIND_PACKAGE(${_feature})

> _FIND_PACKAGE (${_package_name} ${ARGN})

Good catch!

The problem is that I overrode the (CMake) FIND_PACKAGE call with our
own version and I relied on a CMake internal that now seems to have
changed..

I pushed a small fix for that in r32983 - can you please test this
revision?


Best,
Matthias

[Jason Sheldon]

At r32983 I get the same Segfault.

I am using cmake version 2.8.12.2.

Working on the bisection thing now, will update when I've found the revision that caused the issue.

[Jason Sheldon]

PS:  yes, I am using a Mac as well, forgot to mention.

[Scott Miller]

Hi Matthias,

I have cmake 2.8.12.1.

I updated, but still get the same error in the same place, with the changed macro name of course.

Also, I have not updated CMake in quite sometime, and I have never had this problem before.

-Scott

[Jason Sheldon]

Hi All,

I've found the specific revision that broke things (for Mac).
The last working revision for me is r32914, while r32915 and above result in a segfault.
Hope this helps track down the problem!

Jason

[Matthias Maier]

Am 28. Mai 2014, 16:52 schrieb Jason Sheldon <jasonp...@gmail.com>:

> Hi All,
>
> I've found the specific revision that broke things (for Mac).
> The last working revision for me is r32914, while r32915 and above
> result in a segfault.
> Hope this helps track down the problem!
>
> Jason

So close, so close, yet completely wrong with my first guess...

Second try: This is a MAC. MACs don't have learned case sensitivity in
file systems yet.. Please test r32985

Best,
Matthias

[Jason Sheldon]

Hi Matthias,

That seems to have done the job.  Cmade just fine this time.

Thanks,
Jason

--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en
---
You received this message because you are subscribed to a topic in the Google Groups "deal.II User Group" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/dealii/AvcXabYxQo0/unsubscribe.
To unsubscribe from this group and all its topics, send an email to dealii+un...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Scott Miller]

Likewise!  Thanks, Matthias!

-Scott