The MainWin issue should be fixed in the latest version of VirtualGL. Please try upgrading to 2.3.2.
Can you send me the script containing the 'ls -l' command off-list? I also need to know where that script is located in the filesystem and the version of Ansys that it affects. That should be enough to document the workaround. On Nov 29, 2012, at 5:04 AM, "R. David" <da...@unistra.fr> wrote: > Hello, > > Thanks for your answer. > > Le 29 nov. 2012 à 11:13, DRC a écrit : > >> I am an independent software contractor, so the answer to the question >> "do I have a copy of such-and-such commercial software" is almost always >> "no", and often, the companies that make such applications are reluctant >> to issue evaluation copies to anyone who isn't interested in actually >> buying the app. I've even had companies try to sell me a copy of their >> application when all I needed was to use it for 2 weeks to debug a >> problem. <sigh> >> >> I don't think I've seen a copy of Ansys Workbench in 5 years, so you'll >> have to provide more details about what exactly is hanging and where. >> If this is only occurring on CentOS 6.3, then my first inclination would >> be to think that it is an interaction issue with the script that Ansys >> uses to launch Design Modeler. In other words, VirtualGL isn't tripping >> up the application itself but rather some other process that is invoked >> by the launcher. In such cases, it's necessary to edit the script and >> temporarily save and unset the LD_PRELOAD environment variable until >> right before the actual application is launched. > The scriptS are really awful and, as the whole thing works with RedHat 5 > *and* VirtiualGL, I am not totally sure this is an easy-way. > > Currently the code hangs in : > #0 0x00002ac4741e41fb in InterlockedCompareExchangePointer () > from > /opt/ansys_inc/v145/aisol/../commonfiles/MainWin/linx64/mw/lib-amd64_linux_optimized/libkernel32.so > #1 0x00002ac4dc05955e in MpHeapFree(void*, void*) () > from > /opt/ansys_inc/v145/commonfiles/MainWin/linx64/mw/lib-amd64_linux_optimized/libmsxml.so > #2 0x00002ac4dc05a529 in operator delete(void*) () > from > /opt/ansys_inc/v145/commonfiles/MainWin/linx64/mw/lib-amd64_linux_optimized/libmsxml.so > #3 0x00002ac4dc0569c2 in SlotAllocator::FreePage(SlotPage*) () > from > /opt/ansys_inc/v145/commonfiles/MainWin/linx64/mw/lib-amd64_linux_optimized/libmsxml.so > #4 0x00002ac4dc056b6c in SlotAllocator::Release() () > from > /opt/ansys_inc/v145/commonfiles/MainWin/linx64/mw/lib-amd64_linux_optimized/libmsxml.so > #5 0x00002ac4dc0e386e in Document::finalize() () > from > /opt/ansys_inc/v145/commonfiles/MainWin/linx64/mw/lib-amd64_linux_optimized/libmsxml.so > #6 0x00002ac4dc05adda in Base::freeRentalObjects(TLSDATA*, bool) () > from > /opt/ansys_inc/v145/commonfiles/MainWin/linx64/mw/lib-amd64_linux_optimized/libmsxml.so > #7 0x00002ac4dc05bafa in Base::StackExitNormal(TLSDATA*) () > > > Here, the steps to reproduce are simple : > - log-on the vnc/VirtualGL Machine, start everything as usual > - launch the ansys workbench > - double-click on the "Fluid flow (Fluent)" icon > - on the window that popped-up, double-click on "Geometry", and there wait > forever. > > I had too the "ls -l" hanging, it ran away when I disabled SELINUX on the box. > > Best regards, > > > > >> >> >> On 11/29/12 3:03 AM, R. David wrote: >>> Dear all, >>> >>> I have a strange problem with the Ansys Workbench while running under >>> Turbovnc + Virtual GL. >>> >>> I have two OS settings on the same machine : >>> - Redhat 5.0 Enterprise 64 bits >>> - Centos 6.3 64 bits >>> >>> My goal is to start the Design Modeler of Fluent. This graphic design tool >>> uses OpenGL and is started via the ansys Workbench. I use Turbovnc 1.0.2 >>> and VirtualGL 2.3 rpms. >>> >>> On the Redhat 5 OS, all runs smoothly with VirtualGL. >>> >>> On the Centos 6.3, the Design Modeler does not start and the corresponding >>> process hangs in a pointer-exchange function (as seen with gdb) when >>> launched with VirtualGL. >>> On the Centos 6.3 using display redirection with ssh, all runs smoothly. >>> Sometimes all runs smoothly on the Centos (1 out of 100 tries, >>> approximately). >>> >>> Could you have this software working on your configurations ? >> >> ------------------------------------------------------------------------------ >> Keep yourself connected to Go Parallel: >> VERIFY Test and improve your parallel project with help from experts >> and peers. http://goparallel.sourceforge.net >> _______________________________________________ >> VirtualGL-Users mailing list >> VirtualGL-Users@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/virtualgl-users > > --------------------------------------------------------- > R. David - da...@unistra.fr > Responsable du meso-centre > UdS / Direction Informatique > Tel. : 03 68 85 45 48 > --------------------------------------------------------- > > > > > ------------------------------------------------------------------------------ > Keep yourself connected to Go Parallel: > VERIFY Test and improve your parallel project with help from experts > and peers. http://goparallel.sourceforge.net > _______________________________________________ > VirtualGL-Users mailing list > VirtualGL-Users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/virtualgl-users ------------------------------------------------------------------------------ Keep yourself connected to Go Parallel: VERIFY Test and improve your parallel project with help from experts and peers. http://goparallel.sourceforge.net _______________________________________________ VirtualGL-Users mailing list VirtualGL-Users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/virtualgl-users
