System crash

[fear]

Just tried it, same behaviour. Fine on 487, fine on 488 without mtgs, with mtgs on, the computer hangs, holding down ctrl-alt-del will make things resume after about twenty seconds.

[Air]

Ok, I'm hot on the trail of what I suspect is the reason for this crash. I'll post a test build for you to try shortly.

[Air]

Attached is a devbuild which might solve your problem. Let me know how it fairs, either as: works, the same, or breaks differently. Good luck.

[fear]

Sorry, it exhibits identical behaviour as Pcsx2 snapshot 528 - if mgts is off it works fine, if it is on it hangs at "MTGS Thread started > opening GS plugin", and hard locks the system. If I then hold down ctrl alt del for a good 20 seconds, it will moves things forwards for a second, the pcsx2 game window opens, but then it freezes again with nothing in the pcsx2 game window, and the system becomes totally unresponsive.

[Air]

Unfortunately no one else has reported having a problem anything like this, and I honestly can't find anything in the code that's wrong. Even my fix above wasn't really a fix of anything relevant. I just cleaned up the plugin init procedures so that they don't double-init plugins quite as often (which was something that got added in Pg488). In fact, Pg r488 has no MTGS-specific changes to GS init code, so I had to assume that the general init procedures that apply to both GS and MTGS affected only MTGS for some reason.

There's nothing more I can do at this time, except hope that some future code change will randomly fix it; just as something seems to have randomly broken it. That said, thousands of people use MTGS without any problems, and the init code is the simplest portion of MTGS -- the part most likely to either work or not work for everyone and any game/bios/etc. So it's a stumper.

[fear]

No worries. Thanks for looking into it.

[Air]

fear:

Download this tool: http://technet.microsoft.com/en-us/sysin...96647.aspx

Enable the DEBUG runtimes in your DX control panel (they'll be very slow, but can give useful info or assertions). Run DbgView. Run Pcsx2. Initiate MTGS and die.

a) debug runtimes may change your behavior. They do parameter validation and other things. Report if it throws an assertion instead of crashing, and what the assertion window says, along with the log from DbgView.

b) using DbgView, you can recover the logs/errors outputted by Pcsx2 after you've rebooted. Just re-run DbgView and it'll have everything right where it was when windows froze.

Post the log here.
We'll see what we can figure out from it.

[fear]

Ok, I've attached the log file. Same behaviour as normal, I had to get DbgView to write to it on the fly, when I restarted DbgView the first time it was blank. ^_^

[Air]

Try disabling Spybot's SDHelper. Chances are it won't fix the problem, but that's the first and most obvious difference between my log dump and yours. Plus, it looks like the first freeze happens on SDHelper's watch, judging by the timestamps.

[fear]

Ok, I've disabled SD helper, attached are the log files. The second one seems to go a little further; in that run-through I held down ctrl alt del after it locks up, which moves it forward a bit to a blank pcsx2 video window before locking up for good.