Re: Segfault when running "make check"

[Mike Miller]

On Thu, Jan 31, 2013 at 5:54 AM, Julien Bect wrote:
> Julien Bect wrote
>>
>> Mike Miller-2 wrote
>>> $ ./run-octave -qf --eval 'o = javaObject ("java.lang.Object"); test
>>> dblquad'
>>> Segmentation fault
>> Here is the result of valgrind on this command :
>> valgrind.out.jbect.20130131
>> <http://octave.1599824.n4.nabble.com/file/n4649305/valgrind.out.jbect.20130131>
>
> I have reverted from openjdk-7 (7u9-2.3.4-0ubuntu1.12.10.1) to openjdk-6
> (6b24-1.11.5-0ubuntu1~12.10.1) and rebuilt Octave.
>
> "make check" run fine now.

I get similar valgrind results, with the last error in the report:

==27366== Can't extend stack to 0xfe857600 during signal delivery for thread 1:
==27366==   too small or bad protection modes

seems to confirm what I ran into working on bug #37891, that Java uses
up a lot of the process stack space allocation just being resident and
some algorithms that worked fine before are failing when they can't
allocate another stack frame.

On Thu, Jan 31, 2013 at 7:34 AM, Michael Goffioul wrote:
> Does your build have jit enabled? If yes, can you try without it?

No JIT and no GUI. I actually did two clean builds in identically
configured chroots, one i386 and one amd64, both with --disable-jit
and --disable-gui. The amd64 build has no problems, the i386 build
segfaults on make check or with the minimal test case above. I can
also do a --disable-java build to be exhaustive, I expect the error to
go away.

-- 
mike