I did some changes on padhook (IGR) exit routine... Anyone interested on testing ?

[l_oliveira]

Backup /ee_core/src/padhook.c and place this file on it's place.

This changes the way the IOP is stopped and reset during the end of the IGR exit routine, when IGR thread is woken up.

IOP Reset code was SHAMELESSLY copied from (inspired by ? ) uLE...

Please check if the changes break anything (or improve the success rate of) in IGR exit...

[RandQalan]

I call thief but will test

Just a joke

Well first bug broke in game reset path it now resets to OSD

Not what is inputted for reset

Update
Well what ever you did might break reset but works now with Okami and resets better on most game will keep testing

[mgh]

Sadly this doesn't fix Tales of the Abyss...the game still don't like IGR

[l_oliveira]

Well first bug broke in game reset path it now resets to OSD

Well, that's kind of intentional, because MINI OPL only uses exit to OSD, since it's GUI-LESS and is booted from OSD, obviously it has to reset to OSD again. It isn't a big issue for normal OPL users who use FMCB as boot loader but if you think it really help and makes IGR better, I'll try to figure out how to make my change function with the path function again ...


About IGR not working on certain games, I just modified the code on what it does when it's exiting OPL, not the actual code that detect the buttons being pushed or the code that hooks into the game so if the problem is on that part (Aka: IGR does not respond to the key combo or make the game glitch while it's being played) won't change anything. Just changed what happens after the key combo is detected.

[RandQalan]

It is not a big deal for me either just letting you know

Yea I figured same key layout for reset and power down
and games like star ocean still need it off

Don't know why Okami works better now with your but sure is nice to have IGR back for it

[l_oliveira]

Don't know why Okami works better now with your but sure is nice to have IGR back for it

It has to do with the memory being a mess when OPL tries to exit. Because it has to prepare the PS2 at least a bit (EE side cleanup works properly as it's done by the kernel when it uses the kernel elf loader call ) and IOP side was not very clean, there was a chance for a hang up on exit. I think it would crash on a yellow screen (debug color on) most times.

[RandQalan]

So another words you changes more like cleaning the code to work fluidly less frag in memory less likely for hang or lag in game from searching available memory to use

[l_oliveira]

So another words you changes more like cleaning the code to work fluidly less frag in memory less likely for hang or lag in game from searching available memory to use

Not the code, the IOP memory that needed cleaning....

Quite like how originally HD LOADER and OPL would not work on the 10K PS2 because the kernel were not removing what was installed on RAM before wiping it out for the new ELF that was just loaded, crashing the system on a restart (the PS2 restarts like 4 to 6 times when loading a game with OPL, for example) ...

Restart = Remove old hooks, remove loaded modules used by previous program, clean RAM, re-init co-processors and the GS then load target elf and jump on it for execution. So, restart is not the same thing as reset but is quite close lol.

What IGR does is restart and then jump to OSD (or set path elf) instead of doing a full system reset.

[RandQalan]

So you took IOP reset from uLE and coded it for OPL

[l_oliveira]

So you took IOP reset from uLE and coded it for OPL

I just copied like 5 lines of the uLE code that does IOP memory clean up before launching an elf.