Hi.
Is there a policy of keeping trunk in usable condition?
I've found this problem with the Linux VM: http://bugs.squeak.org/view.php?id=7419
I was going to build a new VM to see if I could fix this problem and I'd like to know if the source from trunk is the best place to start.
Gulik.
I think I know this one. Are you by any chance running headless and interact using VNC? If so, it's the attempt to use the X clipboard which causes the segfault. In our images, we have the clipboard prims disabled on Unix by default because of that particular problem.
Cheers, - Andreas
Michael van der Gulik wrote:
Hi.
Is there a policy of keeping trunk in usable condition?
I've found this problem with the Linux VM: http://bugs.squeak.org/view.php?id=7419
I was going to build a new VM to see if I could fix this problem and I'd like to know if the source from trunk is the best place to start.
Gulik.
On Tue, 24 Nov 2009, Andreas Raab wrote:
I think I know this one. Are you by any chance running headless and interact using VNC? If so, it's the attempt to use the X clipboard which causes the segfault. In our images, we have the clipboard prims disabled on Unix by default because of that particular problem.
We are using -vm-display-null and didn't experience such problem.
Levente
Cheers,
- Andreas
Michael van der Gulik wrote:
Hi.
Is there a policy of keeping trunk in usable condition?
I've found this problem with the Linux VM: http://bugs.squeak.org/view.php?id=7419
I was going to build a new VM to see if I could fix this problem and I'd like to know if the source from trunk is the best place to start.
Gulik.
Michael van der Gulik wrote:
Hi.
Is there a policy of keeping trunk in usable condition?
I've found this problem with the Linux VM: http://bugs.squeak.org/view.php?id=7419
I was going to build a new VM to see if I could fix this problem and I'd like to know if the source from trunk is the best place to start.
On Tue, Nov 24, 2009 at 9:44 PM, Andreas Raab andreas.raab@gmx.de wrote:
I think I know this one. Are you by any chance running headless and interact using VNC? If so, it's the attempt to use the X clipboard which causes the segfault. In our images, we have the clipboard prims disabled on Unix by default because of that particular problem.
No, just bog-standard Squeak running bog-standard images on X11. Both a 3.10 image and the latest Pharo image do it.
Another problem is that killing the VM violently will prevent the rest of X processing events well. The XTerm that I'm returned to after killing Squeak won't accept any keyboard events. I consider this a bug in X though: by that stage, Squeak is dead and shouldn't affect the rest of the windowing system.
In other news, trunk doesn't build:
gcc -O3 -mtune=core2 -DLSB_FIRST=1 -DHAVE_CONFIG_H -I/home/mikevdg/src/squeakvm-trunk/squeak-svn/platforms/unix/bld -I/home/mikevdg/src/squeakvm-trunk/squeak-svn/platforms/unix/vm -I/home/mikevdg/src/squeakvm-trunk/squeak-svn/platforms/Cross/vm -I/home/mikevdg/src/squeakvm-trunk/squeak-svn/platforms/unix/src/vm -c /home/mikevdg/src/squeakvm-trunk/squeak-svn/platforms/unix/vm-sound-pulse/sqUnixSoundPulseAudio.c -fPIC -DPIC -DPIC -o sqUnixSoundPulseAudio.o /home/mikevdg/src/squeakvm-trunk/squeak-svn/platforms/unix/vm-sound-pulse/sqUnixSoundPulseAudio.c:51:26: error: pulse/simple.h: No such file or directory
Easily fixed, but this should pick up the lack of PulseAudio and not try to build that plugin.
Gulik.
vm-dev@lists.squeakfoundation.org