Closed picca closed 1 year ago
Hello, I have a similar problem i I run asy under Xvfb:
As a root: /usr/bin/Xvfb :42 -screen 0 1024x768x24 -ac +extension GLX +render -noreset
As a normal user: export DISPLAY=:42 asy -vvvv -f pdf -render 10 3S+1D.asy
The last lines of the output are:
GLSL version 4.50 No fragment shader interlock support shipout3(prefix,f,preview ? nativeformat() : format, ^ /usr/share/asymptote/three.asy: 2896.13
Best regards, Rafel Amer
This issue sounds very much like this problem, which was solved by upgrading the NVIDIA driver. Please give that a try and let us know. (I am still using X and haven't yet tried out Wayland since it doesn't support fvwm, but I think the problem is more likely due to a bug in the graphics driver.)
https://sourceforge.net/p/asymptote/discussion/409349/thread/d535c62247/
I'm running asy in a virtual server whitout any X environment, NVIDIA card and nvidia drivers. I run Fedora 37 server and the version of freeglut is freeglut-3.2.2-2.fc37.x86_64
Best regards, Rafel Amer
Regarding the debian report, try: ./configure --disable-gc make clean make CFLAGS=-g
and then see if the backtrace changes. Running under valgrind is a good way to find the first instance of a segfault, uninitialized memory, or floating point exception.
Program terminated with signal SIGFPE, Arithmetic exception.
This suggests an FPE exception in your graphics driver. You can temporarily work around this by specifying -mask on the command line. A better work around is to identify which OpenGL call is triggering the FPE. For example, the OSMesa library in Fedora 37 generates an FPE in glLinkProgram. Commit 72670f2b7e77a8d6e1b44b82f3a240f75205416b works around an FPE on that platform. Perhaps that will solve your issue too.
Now that the 2.85 release is out, can we now close this issue?
I will juste wait for the Debian release of 2.85, and report back, thanks
Frederic
Hello, I am using asy in order to build the documentation of my hkl library, from the sway environment (pure wayland).
the Debian bug report I filled against asymtote is here
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1025049
the test file is here
https://bugs.debian.org/cgi-bin/bugreport.cgi?att=1;bug=1025049;filename=B_b3_y.asy;msg=5
to my opinion the
the glrender call from the backtrace seems suspicious to mee
I would like your help in order to tackle this issue.
thanks for considering