Home | Forums | What's new | Resources | |
Yabause |
slinga - Jul 26, 2005 |
1 | 2 | 3 | ... | 8 | Next> |
CyberWarriorX | Jul 27, 2005 | |||
Hmm... looks like you're running the latest cvs version? It may very well be that for some reason or other you're compiling with the dynarec code enabled, which last I checked wasn't fully functional. What configure settings are you using, etc.? Oh, and btw, what kind of cpu are you using? Cyber Warrior X |
slinga | Jul 27, 2005 | ||||
I've tried both the old release and the new release. I should of mentioned that I'm running the CVS code (version 0.50). Both have similar crashes. I'm running on a dual Intel Xeon 64 bit machine. Here is the output from configure:
Configure looks pretty standard to me, no warnings\errors. I think the error is in the Yabause source code somewhere, atleast for the seg faults from loading the binaries. Edit: Make gives me tons of warnings, and seems to stall on c68k/c68kexec.c for a while, but it compiles. |
CyberWarriorX | Jul 27, 2005 | ||||
Yup, that's what I thought. The 64-bit part anyways. That's something we've got to fix yet. From what I understand, it has to do with the fact that long's are 8 bytes in size instead of the 4 on 32-bit platforms. So for the C-rewrite that we've been working on the last couple months(we're getting closer to completion, actually), that issue will be addressed. Cyber Warrior X |
slinga | Jul 27, 2005 | |||
Weird...as far as I know, Intel Xeons can run 32 bit code just fine. I have plenty of packages on my box that are i586 and i686 optimized and they work. Anyway to force configure to compile it as a 32 bit application? If not I'm going to have to try to wine satourne or giri giri...::shudders:: |
CyberWarriorX | Jul 27, 2005 | |||
Well, I guess if you don't mind messing around with the configure.in script and rebuilding the files using autoconf, etc. you -could- try the -m32 compile flag. Although I haven't had a chance to try it out myself since I don't have any 64-bit machines, nor can I afford one right now. However, if anyone is will to donate something like that..... Cyber Warrior X |
slinga | Jul 27, 2005 | |||
Perhaps I'm being a moron, but can you (or any other kind soul) just post up your compiled 32 bit binary here? Unless I'm missing something... |
mrkotfw | Jul 30, 2005 | ||||
if you're running bash,
Code:
|
slinga | Aug 1, 2005 | |||
Thanks for your tip Piratero, but it didn't work. I tried it and then tried configure again, and it gave an error saying my compiler couldn't create executables. Looking through the source code, I seem to have trouble with the struct: u32 and s32. I'm not sure what they are, but I'm guessing it has to do with 32 bit architecture. Can you post your binary of yabause if it's not too much trouble? Thanks in advance. |
mrkotfw | Aug 2, 2005 | |||
get my compiled version here... |
slinga | Aug 2, 2005 | |||
Didn't work, thanks anyway man. |
mrkotfw | Aug 2, 2005 | ||||
this is the compiled version of the yabause rewrite:
Code:
|
slinga | Aug 2, 2005 | ||||
I'm getting a little closer:
|
CyberWarriorX | Aug 2, 2005 | |||
Try the latest cvs of the yabauserewrite branch again. I added a new check for detecting 64-bit compiling. Oh, and if anyone has any spare 64-bit systems they don't want, I need a test system Cyber Warrior X |
slinga | Aug 2, 2005 | |||
I grabbed it from here: cvs -d cvs -z3 -d Looked like the same thing, even version number was 0.5. Same crashes. |
CyberWarriorX | Aug 3, 2005 | ||||
Instead of that line, use: cvs -z3 -d The idea is you want the yabauserewrite branch, not the main(old) branch. Cyber Warrior X |
slinga | Aug 3, 2005 | |||||||
Ok I'm trying it out now. Slightly off topic, but how long does it take you to compile Yabause? My system goes crazy (on Yabause version > .07) on the line: c68k/c68kexec.c:255: warning: cast to pointer from integer of different size, when it's compiling c68exec.o. It takes about 10-15 minutes from that point. I think it's a problem with my system, just confirming with you guys. Edit: It compiles much faster now, thanks. Same error though:
With another BIOS:
|
antime | Aug 3, 2005 | |||
Building the m68k core takes a while, as it's actually generating a lot of code. |
CyberWarriorX | Aug 27, 2005 | |||
I finally got around to going through the entire codebase and fixed any potential problem variables along the way. So if you have time slinga, mind trying it again? Cyber Warrior X |
slinga | Aug 28, 2005 | ||||
Same error message as before:
But I now see a FPS counter, I never saw that before. It runs for like 2 seconds, I see no graphics, and then it crashes. There is also some noise from my speakers. |
1 | 2 | 3 | ... | 8 | Next> |