Cool VL Viewer forum

edit

edit

And the interesting lines are still not there... I need the lines just before the crash happens (it occurred at 2016-06-18T19:45:43Z, from the first things you posted, so please provide every line logged at that time and before that time and covering at the very least several seconds of log) !

edit

edit

edit

edit

Thank you !

Alas... This time, the crash dump shows that you did not crash during the texture decoding but instead during a mesh LOD decoding (something that was seen before, but that seemed to have been fully solved like TheADX confirmed in the other forum thread)...

The more I look at your logs and crash dumps, the more I get pushed towards the conclusion that it is not a viewer bug (mind you, the Cool VL Viewer code, in these specific parts where you crash, is no different from LL's or other TPVs code: only the timing in the threads execution may differ, which may trigger a bug or not, but not a bug in the viewer code, as far as I can see...).

Are you positively, absolutely, utterly sure you didn't caught a virus (if possible, try and install another antivirus, just for the sake of checking your computer memory and disk, then uninstall it (or your old antivirus) once done, since it's not a good thing to have multiple antivirus programs running together) ?
I'd recommend ClamWin for such a check

Did you check about CPU microcode updates for your CPU (see this article about "This CPU has a known and already fixed errata which causes VMs to crash. This bug has already been fixed with a microcode update, provided by AMD and has not yet been implemented into the BIOS of the MB") ?
Did you check about your motherboard's BIOS updates (which may also contain fixes, either related to the CPU itself, or to the chipset) ?
Also, and just as one example, is this Windows hot-fix installed on your system ?
Searching on the web for "AMD FX 6300 microcode" may give you pointers to more such fixes and issues; yes, all CPUs have bugs too, not only the first-Pentium div bug, but many more (you will find them as "CPU errata" documents on the web, here is one for your CPU), and microcode hot-fixes is today's reply to such bugs...

edit

I hope you did check seriously (i.e. by installing and scanning your memory and disk with a second anti-virus).

"The" hotfix ?... The one I pointed you to was just one out of many such hotfixes, one I found with a single search on the web, at the top of the first page of the many results... It's not even a microcode hotfix, but just an optimization for threaded applications running under Windows with your specific CPU brand and model (it's not a microcode hotfix, but a fix to Windows itself)...
You'll have to check for all the published hotfixes, including the ones that really count: the microcode hotfixes..

You will also have to check for your motherboard BIOS updates, like I already explained...

This is irrelevant (they are compiled with Visual Studio 2013, while the Cool VL Viewer is (for now) compiled with VS2010 Express... Different compilers produce different code out of the same sources, and some compilers may trigger CPU bugs when others (newer) compilers have been modified to work around the said bugs. The Cool VL Viewer will be compiled with VS2013 too, in the future (I'm waiting for the 64 bits move from LL).

Again, totally irrelevant here...

No, it's not "the way it is coded", the best proof being that you crashed in code that is exactly the same in my viewer and in LL's viewer, or in other TPVs. Please, let knowledgeable people diagnose the bugs: when you will have over 35 years of experience in programing, then perhaps I will accept your own diagnostic over mine...

The crashes they encountered have all been fixed. You are the only one, so far, still crashing in various places, where there is strictly no bug whatsoever. I'm myself using my viewer for hours long (I'm not speaking about a couple of hours, here, but 10 or more hours) sessions and don't crash with it. I am, also, unable to reproduce any of the crashes you are experiencing, which is a clear sign that something is fishy with your setup.

Yes, and the crash dump shows that you crashed in yet a third (different place)... Problem: there's no reason for a crash there (no risk of pointer dereferencing, bad/dead objects, race condition or anything: you crashed during a plain integer arithmetic calculation !). Again, the code part where you crashed is not specific to the Cool VL Viewer, and this time, it's not even in a thread, but in the main loop... I cannot explain this but to deduce that the problem is with your specific setup/system.

One explanation (beside a CPU bug) could be your anti-virus or another running program (a virus, typically) interfering with the viewer: try disabling your antivirus and see if the crashes stop...