Bug 17374

The file is too big to be joined here. So I share it through Owncloud :

http://popcircus.hd.free.fr/owncloud/index.php/s/kVSW5JfZPYq6Ul2

Below the crash message.

#
# A fatal error has been detected by the Java Runtime Environment:
#
#  SIGSEGV (0xb) at pc=0x8553c514, pid=2748, tid=2233465664
#
# JRE version: OpenJDK Runtime Environment (8.0_65-b17) (build 1.8.0_65-b17)
# Java VM: OpenJDK Server VM (25.65-b01 mixed mode linux-x86 )
# Problematic frame:
# C  [libGL.so.1+0x40514]
#
# Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again

Created attachment 7297 [details]
Crash with version 4.6 of MGA5 i586

Created attachment 7298 [details]
Crash with version 5.1 of MGA5 i586 updates_testing

Unfortunately version 5.1 submitted to updates_testing does not fix the problem.

Both 4.6 and 5.1 in testing crash here on mga5-64 so probably it is not the arch, but other things that is the problem.

CC: (none) => fri
Hardware: i586 => All

Created attachment 7704 [details]
Crash log tar.gz compressed,

Sweethome3d 5.1-1.mga5 from updates testing, on 64 bit mga5, i7 CPU, 8GBRAM
Crashes after less than ten seconds generating a simple video at lowest resolution of pretty simple room.

Created attachment 7705 [details]
error log .tgz from version 5.2 on mga6 cauldron 64 bit

5.2 too crashes while generating video on cauldron 64 bit on a different machine.

It seems to be more a problem with java (openjdk) or with the driver video fglrx than sweethome3d.

I don't know how to debug that correctly.

Version: Cauldron => 5

Ah, yes.
Both machines it was crashing on have ATI graphics (and use proprietary drivers)

Now i have installed it on two very different machines, one old Thinkpad T61 with Intel Duo CPU, Nvidia Quadro graphics, and one AMD hexacore with Nvidia 760 graphics.  Both work perfectly with sweethome3d 5.1-1.mga5 from updates testing, on 64 bit mga5.  (did not bother testing 4.6)

IMO you can push 5.1 as there seem not be a regression, but in this case same bug (whatever and wherever it is...)

Version: 5 => Cauldron

In fact the first bug report was on Intel 915 graphics, and it did not crash on MGA5_64 with Intel 945 graphics.

I feel the java bug is the good way to go, not the ATI graphics, maybe a too small stack? (I never programmed in java)

I also feel an 5.2 update could be pushed, just to use the latest version.

I support both your conclusions :)

I think we should get in contact with sweethome3d developers and see if they recognize the problem.

(In reply to José Jorge from comment #10)
> In fact the first bug report was on Intel 915 graphics, and it did not crash
> on MGA5_64 with Intel 945 graphics.
> 
> I feel the java bug is the good way to go, not the ATI graphics, maybe a too
> small stack? (I never programmed in java)
> 
> I also feel an 5.2 update could be pushed, just to use the latest version.

any reason this is not done yet? for mga5? I could use an updated version. gonna start building a house soon.

CC: (none) => alien

(In reply to AL13N from comment #12)
> (In reply to José Jorge from comment #10)
> > I also feel an 5.2 update could be pushed, just to use the latest version.
> 
> any reason this is not done yet? for mga5? I could use an updated version.
> gonna start building a house soon.

Time miss. Feel free to submit the latest update to updates_testing. I never got a reply from upstream...

Status: NEW => ASSIGNED

Looks like this bug was really only for Mageia 5, as fglrx is gone as of Mageia 6.  Closing.

Status: ASSIGNED => RESOLVED
Resolution: (none) => OLD
Whiteboard: MGA5TOO => (none)
Version: Cauldron => 5