I hesitate to post this problem to a board I've just now joined, but suffice it to say that I'm very interested in getting GZDoom to work on my system:
AMD Phenom II X4 965
3.42GHz, 1.75 GB of RAM
(256MB Shared Video)
Windows XP Service Pack 2
Gigabyte MA785GMT-UD2H Motherboard w/
ATI Radeon HD 4200 integrated graphics
ATI Catalyst Control Center 10.3 Installed via clean uninstall/reinstall from scratch
- - - - - - - - - - -
As a primer I'll explain that its been a few years for me and I'm just now catching up to all the changes in the Doom landscape. I was using ZDoomGL before its demise, and in upgrading now to the latest release of ZDoom (2.4.1) I have been reminded just why I wanted the GL support as software-only rendering of course looks awful in comparison.
Put simply, I can't get GZDoom to run at all. It just begins to initialize and then I get an "Access Violation" error message. I went through a lengthy uninstall/clean reinstall process to get version 10.3, which was apparently a successful fix for most, but unfortunately I am still getting the same error screen of death...albeit with a cute little dead marine in the corner.
GZDoom Crash.jpg (60.47 KiB) Viewed 2532 times
I have been successful in running GZDoom using the software renderer only, but clearly that's sort of pointless...and that's basically all she wrote about my relationship with GZDoom thus far...I can tell its an amazing port and I'd love to use it. I know right now is probably not the greatest time to get in depth support considering recent events but I figured I may as well let someone know I'm having an issue rather than move on as though GZDoom didn't exist. If anyone has any thoughts please let me know if there may be a solution, however my fear is that this is most likely NOT a simple problem.
Thanks and btw it's nice to see that Doom is still alive and well.
As you may have read in these and other forums, there are issues with running GZDooM on a computer that uses ATI graphics chip-sets. However, to my knowledge, even though GZDooM may not run at an optimal pace with ATI cards, there's no reason for it not to run at all. As a matter of fact, I have a MOBILITY RADEON 9000 with an earlier generation processor than you, and 2 GB RAM, and I can generally get GZDooM to run fine on all but the most demanding wads.
I'm not a systems guru, so I'll let others analyze your crash report and help you identify the source of the problem. Good luck.
I forgot to specify my software revisions; ZDoom 2.4.1 and GZDoom 1.4.8. ZDoom runs fine, as does the software-only option in GZDoom. Getting full GZDoom to load in ANY configuration that involves utilizing OpenGL is the problem...as in I've never even seen the title page.
On a side note, today did bring a bit more illumination to the problem (or at least assisted in narrowing down the possible suspects).
I have two systems in my house running on nearly identical Gigabyte boards, and I believe the integrated ATI graphics chipsets are identical. The other system was recently formatted and reinstalled with Catalyst 10.3, and has XP SP3 installed (as compared to my SP2 system). After installing the recent versions (ZDoom 2.4.1 and GZDoom 1.4., it has developed the same issue. So I think that clears up the possibility of the problem being related to the driver installation or to my Windows version, in case anyone was wondering.
I'm aware of the compatibility issues between ATI and GZDoom from reading some other posts, but I'm just hoping my experience will contribute to the discussion or possibly result in a fix or workaround. In the meantime I'm actually still running ZDoomGL...(from like a decade ago) and testing wads in software-only GZDoom/Zdoom. Thanks again for the reply.
At the moment, all I can say is that you could try downgrading your driver to Catalyst 9.11. If I'm not mistaken, GZDoom broke when the OpenGL drivers were rewritten to add support for OpenGL 3.2, and that was apparently in 9.12.
This one's a bit weird. It crashes while executing a function for the self-modifying code of the software renderer. The address it tries to write to looks perfectly fine. This whole thing looks like something revoked the write permissions for the self-modifying code segment. If this is done by ATI's driver something really fucked up is happening here.
You know Graf, I think you're right about that last part. LOL. I have now rolled back to an even earlier version than 9.11, because - get ready for this - the INSTALLER for 9.11 is crashing before it completes the install.
I'm now successfully running GZDoom, however, using Catalyst version 8.6 (I think - anyway a very old version that was on my Giganyte install disc and carries the distinction of not crashing). I'm away from home right now so I can't remember the exact version number, but it's definitely from '08.
Thx for the reply Graf - I'm loving the port btw despite this little ATI debacle.
The driver I used from my Gigabyte install disc (which worked), was a proprietary driver version 8.641. That roughly correlates to Catalyst Software Suite 9.8. It wasn't until after I upgraded to Catalyst 10.3 that I installed GZDoom for the first time, and had the problem of it crashing with the "Access Violation" error that I posted above. Per your suggestions, I then downgraded from Catalyst Suite 10.3 (or display driver 8.712) back down to Catalyst 9.8 (driver 8.641) and that fixed the problem.*
Then I left town and in the meantime ATI released Catalyst 10.4 (or driver 8.723), which I read about in a separate bug posting from another new user. After reading of the new SVN build I decided to be bold and I upgraded to 10.4....as expected this made GZDoom crash the same as it had under 10.3. But using the r786.7 SVN build its now working again.
So the lovely GZDoom is now running on my system with the newest driver from ATI - 8.723, Catalyst 10.4. Thanks again Graf for the new build and your attention to this - clearly the real cause of the "access violation" is still shrouded in mystery but nonetheless I'm just happy to have everything working now.
-TT
*The only reason I didn't try Catalyst 9.11 is that it wouldn't install on my system, and for that matter neither did 9.10...(after some researching version numbers it occurred to me that perhaps some of these older drivers weren't installing properly because they weren't specifically designed for my motherboard-integrated chipset...at least that's the theory).
, it has developed the same issue. So I think that clears up the possibility of the problem being related to the driver installation or to my Windows version, in case anyone was wondering.
