My ezQuake crashes often when I'm either playing or spectating. I usually get a BSOD, but sometimes it just freezes either before a BSOD or at the BSOD. The BSOD gives a huge laundry list of files/drivers that cause the BSOD. It's rarely the same more than three times in a row. I crash at a random time, anywhere from 5 seconds into ezQuake running to 2 hours.

My system specs:
WinXP Home
Intel Core 2 Duo
Intel® Desktop Board DG33FB
nVidia GeForce 8800 GTS

Some troubleshooting I've done:
-Run Windows Memory Test overnight with 0 errors
-Run Memtest 86+ overnight with 0 errors
-Removed sound card
-Eliminated all unessential processes
-Removed Firewall
-Removed Virus Protection
-Updated video drivers
-Downgraded video drivers
-Used Driver Cleaner Pro to delete old drivers and reinstalled drivers
-Installed new motherboard drivers (sound, usb, chipset, etc)
-Installed new mouse drivers
-Turned off all video card processes such as anti-aliasing, anisotropic filtering, texture filtering, etc.

ezQuake Versions that have crashed:
1.8.2
1.8.3
1.9

I'm sure there are some other things I have tried, but nothing has worked. I believe that there is a conflict between my video card drivers and ezQuake. I am able to run FuhQuake without the program crashing at all.

Any suggestions would be greatly appreciated. I far prefer the look and feel of ezQuake to FuhQuake and would like to be able to use it without crashes.

According to what you just said, it certainly looks like a driver problem (I had such problem myself with certain NVIDIA drivers, but it crashed on desktop randomly, not during games, huge list of shit displayed).

Since your system is also freezing, there are limited amount of options left. Basically:

- GFX test: Run RTHDRIBL for 30 minutes in window (default settings) and log/check temperatures with GPU-Z

- CPU test: Run one instance of Prime95 torture test using max heat, power consumption torture test for each cpu cores you have. On dual core systems, use two instances of Prime95, on quad core systems, four. Log temperatures with Core Temp. For "some" stability, make sure Prime95s are run at least an hour. You should NOT be getting a single warning or error during this time.

- Check your PSU, it's voltages and amperage. What manufacturer, what model?

(you can optionally run Prime95 using 1 "max heat/power consumption" and 1 "blend" torture tests (3 + 1 blend on quad core), this way it tests your memory prefetching stability under heavy CPU stress. Again, at least 60mins).

If these tests run flawlessly, then it's the NVIDIA drivers that cause the problem.

Check for shared IRQs (device manager -> view by type -> IRQ, check if e.g. video card and network/sound share one IRQ number) It shouldn't be a problem but on some configurations yes it causes troubles.

These tests all run flawlessly.

My PSU is a Coolmax CX-400B 400 Watt power supply.

So, if we know it's the NVIDIA drivers, do you have any suggestions as to how to solve the problem? Why would the drivers have no conflicts with FuhQuake but conflict with ezQuake, isn't ezQuake just a new and improved version of FuhQuake?


I've removed my sound card, so there's nothing left to conflict. My video card is using IRQ Port 16. The only other device on the device manager that uses port 16 is the Intel(R) G33/G31/P35 Express Chipset PCI Express Root Port - 29C1, but I'd guess that this is necessary as it appears to be the driver for the PCI Express port.

By the way, thanks for the help guys.

It is and that you are having problems with only one of those is very rare.

I remember someone talking about multithreaded opengl/texturing or something like that in graphic card options or so (maybe someone could give better description), it was necessary to disable it, maybe try that if there such setting left that you haven't tried.

BTW disregard my "shared irq" advise, you have new enough motherboard where shared irq doesn't matter.

These are the 3d Settings I currently have:

Anisotropic filtering - Off
Antialiasing - Gamma Correction - Off
Antialiasing - Mode - Off
Antialiasing - Transparency - Off
Conformant texture clamp - Off
Extension Limit - Off
Force mipmaps - None
Maximum pre-rendered frames - 0
Texture filtering - Anisotropic sample optimization - Off
Texture filtering - Negative LOD bias - Allow
Texture filtering - Quality - High performance
Texture filtering - Trilinear optimization - Off
Theaded optimization - Off
Triple buffering - Off
Vertical sync - Force off

On a side note, I'm testing older versions of ezQuake to see if I can get them to run without crashing. I'm currently working with ezQuake 1754.

I have found a version of ezQuake that does not crash (1514). I have tried ezQuake 1754, 1.8.2, 1.8.3, and 1.9 and all of them have crashed. That makes me think that some addition to the build 1754 is conflicting with my video drivers. Should I just use 1514 and ignore all future updates? Or do you think there's a possibility that in the future the conflict will be resolved?

Thanks,
Traxler

After over 5 hours of using ezQuake 1517 without crashing, it finally crashed. This makes me wonder whether fuhQuake would have crashed if I played longer. I never made it over 2 hours without crashing before. I'm now at a complete loss as to what to do to solve this problem.

You didn't tell me what the temps were with CPU and GPU torture tests. Also I'm wondering if that 400W psu is really enough for 8800GTS card, I can't find a single review of that model.

The difference between ezQuakes is the ability to run higher fps than for example Fuhquake (without beta patch). Also some of the assembler code was removed (right?) and it makes your CPU to work more since the new ezQuakes are less optimized. In both cases the temperature of GPU and CPU will rise, so does the required wattage. In this case new PSU could mean all the difference in the world.

However if the temperatures are ok (check them and report them here, have RTHDRIBL run at least 5-10 mins and prime95 the same time, should be somewhat accurate) I would still go for the NVIDIA driver problem. You said you have 8800GTS but that doesn't quite mean anything, so you need to be more specific, is it GTS-320, GTS-640 or the newer GTS-512 model.

The max temps/speed of the GPU test with 1680x944 resolution for 45 minutes was:

GPU Core Clock 576.0 MHz
GPU Memory Clock 850.0 MHz
GPU Temperature 92.0
GPU Temperature 82.3
PCB Temperature 71.0
Fan Speed 85% 2586 RPM

The card is an EVGA e-GeForcee 8800 GTS
Memory Type GDDR3
Bus Width 320 Bit
Memory Size 320 MB
Bandwidth 68.9 GB/s
GPU Clock 567 MHz
Memory 850 MHz
Shader 1350 MHz

The max temps for each the processors on the CPU test were 62.

This weekend I hope to have time to remove my video card and install the drivers for my onboard video. If ezQuake 1.9 can run without crashing with onboard video then it must be a video driver issue. If it still crashes then it may be a motherboard issue as I did buy an open box motherboard when I put my computer together.

I had similar problems with my NV cards - similar temps but crashing after a while (although with all games). Using NTune to force fans to 100% fixed it for me.

No problem with cpu temps, however GFX card temps are a bit too high and could cause momentary freezing but bsods? If I still had to guess something I'd still say NVIDIA drivers or the temps, a slight change to be PSU related.

Your system seems otherwise stable.

I removed my video card and tried to use just the onboard video on my motherboard. I still crashed. The problem must not be tied to my video card. Oh well....

I finally found out what the problem was. My I/O Controller Hub was overheating (over 100 degrees). Redirecting some fans has eliminated my crashing problem.