Unknown W. Brackets
|
a779645152
|
Wrap sceKernelGetThreadExitStatus().
So it's usable internally too.
|
2013-05-26 10:17:59 -07:00 |
|
Unknown W. Brackets
|
2c4254d0c8
|
Disallow suspend/resume on current thread.
Might be some extra subtle things with callbacks/interrupts here...
|
2013-05-26 02:25:53 -07:00 |
|
Unknown W. Brackets
|
e6795b1a41
|
Initial attempt at a suspend/resume thread impl.
|
2013-05-26 01:01:29 -07:00 |
|
Unknown W. Brackets
|
690e52bc34
|
Make sure we keep the suspend state around.
I assume it sticks around in this case.
|
2013-05-25 23:57:04 -07:00 |
|
Unknown W. Brackets
|
084ad5aa53
|
Wrap the sleep/wakeup HLE funcs.
|
2013-05-25 20:43:19 -07:00 |
|
Unknown W. Brackets
|
8adf1e9ede
|
Properly return success in sceKernelWakeupThread().
Fixes #1596, fixes #1796, possibly others. I can't believe how long
I stared at this code and debugged it without noticing such a simple
thing...
|
2013-05-25 20:28:22 -07:00 |
|
Henrik Rydgard
|
4b15fe3c5e
|
Minor cleanup
|
2013-05-19 18:45:58 +02:00 |
|
raven02
|
c378af0878
|
Fix Dragon Ball Z Tenkaichi Tag Team
|
2013-05-19 19:56:18 +08:00 |
|
Unknown W. Brackets
|
ead70f56d9
|
Add a safety check to prevent leaks.
|
2013-05-18 20:21:29 -07:00 |
|
Unknown W. Brackets
|
0fb5877d75
|
Implement sceKernelExtendThreadStack().
Fixes #1833. Implemented using fake nids instead of callbacks since
it's simpler to get them (and waits and recursion) right.
|
2013-05-18 20:21:28 -07:00 |
|
Unknown W. Brackets
|
f91135cedf
|
Prevent some threading related NULL derefs.
|
2013-05-18 10:19:34 -07:00 |
|
Henrik Rydgård
|
c13a101901
|
Merge pull request #1765 from unknownbrackets/perf
Eat cycles in sceKernelReferThreadStatus()
|
2013-05-11 14:31:42 -07:00 |
|
Unknown W. Brackets
|
16fe2c9691
|
Eat cycles in sceKernelReferThreadStatus().
Improves .hack//Link (demo)'s perf significantly, doesn't hurt anyone
else.
|
2013-05-11 12:25:55 -07:00 |
|
Unknown W. Brackets
|
5318e97a2e
|
Threads starting better threads should stay at top.
This wasn't correct, I think I must've added it because something else
should've been rescheduling.
|
2013-05-11 11:28:37 -07:00 |
|
Henrik Rydgard
|
5fc8de091d
|
Make some logging VERBOSE
|
2013-05-09 11:18:38 +02:00 |
|
Unknown W. Brackets
|
bf826ed9c1
|
Correct delay for sceKernelDelayThread(0).
It's not 200 for 0, more like 100ish.
|
2013-05-04 23:42:10 -07:00 |
|
Unknown W. Brackets
|
f1f2cf0603
|
Replace some hleEatCycles() with actual PSP timing.
Measured several runs at 8 per run with test overhead check.
|
2013-05-04 23:42:10 -07:00 |
|
nachume
|
2181b76c42
|
ISO that replaced EBOOT.BIN is playable.
implement LoadExecForUer_362A956B() and sceKernelRegisterExitCallback.
Add some ERROR define on kernel.h. I was referring to the jpcsp.
|
2013-05-04 20:34:33 +09:00 |
|
Unknown W. Brackets
|
b18235d2e0
|
Sneak in a quick logging fix.
|
2013-04-28 11:27:05 -07:00 |
|
Unknown W. Brackets
|
acdd14f0d4
|
Fix initial thread stack, arg space and alignment.
Anything that used > 256 bytes of args was probably broken...
|
2013-04-28 00:34:49 -07:00 |
|
Unknown W. Brackets
|
d131588a6d
|
Validate sceKernelStartThread() args better.
|
2013-04-28 00:34:49 -07:00 |
|
Unknown W. Brackets
|
df62cafe3d
|
Handle a few errors on thread create better.
Especially out of stack space.
|
2013-04-28 00:34:48 -07:00 |
|
Unknown W. Brackets
|
fb758de3b9
|
sceKernelWakeupThread() should reschedule on wake.
Otherwise it will have different timing than the PSP.
|
2013-04-28 00:34:48 -07:00 |
|
Unknown W. Brackets
|
c8958a253d
|
Cleanup some unused / duplicate code in threading.
|
2013-04-28 00:34:48 -07:00 |
|
Unknown W. Brackets
|
e30dedd46f
|
Allocate thread stack on create, not on start.
Also, don't free and reallocate on thread reset.
|
2013-04-28 00:34:47 -07:00 |
|