Hi everyone,
I'm still working on the dynarec. I reached a bug where everything will work, now get this, if a fopen/fclose I used for debugging is in a certain function. If its not there, it will crash as it seems the recompiled function won't return properly in one spot. Why this is happening is beyond me at this point, and why an fopen/fclose or usleep(500) makes it work is also beyond me. I figure it's a stack issue, but I seem to be doing things right. I'll have some people look at it and see what I'm doing wrong. I'm refusing to blame the compiler. ;P
With the trick in place above, the BIOS is completely recompiled dynamically, all 2.6 million cycles of it. Though it doesn't get as far as displaying anything from a game yet. I want to find out whats causing the memory issue mentioned above first before tackling other bugs, as they are probably related.
My feelings about the second core is that at the current state of the emulator and the way the second core is accessed, it's fairly useless. psx4all wasn't written with threading in mind, though, yes the GPU and SPU could possibly be put on the second core with a few advancements. It's just isn't happening in the immediate future as it seems so far. The best bet would be putting the GPU core on the second core, but thats in Unai's hands.
I couldn't work on the dynarec the past few days since I had to complete a commercial project, a Palm OS game. Another Palm OS game is in the works for me to do, but I'm going to get some work on the dynarec done inbetween, starting now.
I've been working non-stop, sometimes to the point I can't update everyone on the status of things.
ZodTTD