Commit Graph
3548 Commits
Author SHA1 Message Date
Lioncash 1f0710caf7 Remove some more includes. 2013-10-19 19:59:03 -04:00
Lioncash 6adf3e37e2 Clean out unnecessary includes from the DSP JIT. 2013-10-19 19:59:00 -04:00
Lioncash af1b5fe767 Woops forgot to clean out some files in the Jit64IL. All clean now. 2013-10-19 19:58:59 -04:00
Lioncash 80b4cb6cf3 Clean out includes from regular JitCommon. 2013-10-19 19:58:59 -04:00
Lioncash 3360ec013f Clean out includes in JitILCommon. 2013-10-19 19:58:58 -04:00
Lioncash 22cdc0f56e Remove some more unnecessary includes, as well as organize a little more. 2013-10-19 19:58:57 -04:00
Lioncash 97cd42f5c3 Remove some unnecessary includes as well as simplifying exisiting ones if possible. 2013-10-19 19:58:56 -04:00
comex 33ee04626d Don't define __rdtsc if x86intrin.h already does. 2013-10-19 12:40:21 -04:00
Lioncash 81964517d6 Remove a redundant assignment in Init() within PowerPC.cpp. We already assign zero to ppcState.dtlb_last. 2013-10-17 20:38:49 -04:00
Lioncash b8a947ce0a Fix a filter in the VS project file for JitILCommon-related files. 2013-10-17 06:36:32 -04:00
comex de1773affb Basic precompiled header support for Linux/OS X. Shaves 20-30% off full rebuild time on my system. 2013-10-17 00:07:35 -04:00
Tillmann Karras e4a81f32d4 Document some UGeckoInstruction fields 2013-10-17 04:59:13 +02:00
Rachel Bryk f605ea9eeb [Netplay] Allow a spectator to leave without affecting players. 2013-10-16 21:44:49 -04:00
degasus 5a212a8fe0 jit64: enable faddsx again
It was disabled because of issue 182, but as this game depeneds on FPRF, it was just 'fixed' because of the fallback to interpreter (which implements FPRF by default).
Also enables FPRF for this game via GameIni, so that the issue is still workaround.

If there are any regressions because of this commit, please try to enable FPRF in GameIni.
2013-10-17 01:43:16 +02:00
Tillmann Karras c234dc97c0 Jit64: optimize floating-point/paired-single arith
The "else if (b != d)" branch was dead code and only works if b == d. Now
the last else block with two temporary XMM registers is not needed anymore.
2013-10-16 00:52:21 +02:00
skidau b4513313bb Increase the trampoline cache size to cater for fastmem writes.
Fixes issue 6722.
2013-10-16 09:26:08 +11:00
Lioncash f3af8ee0f0 Revert "Remove some unnecessary includes."
Turns out this explodes on Linux.

This reverts commit c4696568cc.
2013-10-14 04:05:38 -04:00
Lioncash c4696568cc Remove some unnecessary includes. 2013-10-14 04:01:57 -04:00
Ryan Houdek 2f0aca2055 [ARM] Fix the JITILArm and JITArm dispatchers from comex's recent changes. 2013-10-12 08:30:26 +00:00
Ryan Houdek 0556193442 Build fix for previous commit. 2013-10-11 21:45:44 -05:00
comex 78a4dbced8 Remove more unused stuff. 2013-10-09 21:37:29 -04:00
comex 8962b9606b One shudders to imagine what inhuman thoughts lie behind that code.
(Read_Opcode_JIT and Write_Opcode_JIT read/write from unrelated memory
areas.*  Rename the latter and refactor.)

*except at the one specific exception handler where it doesn't.  I
have no idea what this is supposed to do, but it probably doesn't do
it correctly.  For now, remove the exception.
2013-10-09 21:37:29 -04:00
comex 2e32f11f03 Remove "deprecated" and unused function Read_Opcode_JIT_LC. 2013-10-09 21:37:28 -04:00
comex 5847e26d34 Minor JitCache cleanup. 2013-10-09 21:37:28 -04:00
comex 86f2e1734a Make JIT_UNLIMITED_ICACHE always on.
(1) The alternative doesn't compile.
(2) Despite "unlimited" sounding like a hack, it's actually
    significantly more correct then the alternative, which is no
    emulated icache.
(3) Easier to wrap my head around.
2013-10-09 21:37:28 -04:00
Ryan Houdek 715d5ae8a7 Merge branch 'JitArmIL'
This implements a partial JITIL based off of the JIT64IL. It's enough to run most games, albiet at a slow speed.
Implementing instructions for this IL is really simple since it basically is just enabling based on what is already in JIT64IL, and then enabling each individual IL instruction.
2013-10-09 23:16:07 +00:00
Ryan Houdek bd79603c66 [ARM-JitArmIL] Enable a bunch of instructions. 2013-10-09 23:03:39 +00:00
Ryan Houdek 4914665429 [ARM-JITArmIL] Remove the SystemRegisters file. 2013-10-09 20:59:15 +00:00
comex d5e40bfb01 Don't waste time looking up the block if enableBlocklink is off anyway. 2013-10-09 15:01:55 -04:00
Matthew Parlane 935ed814ea Missed a accept error handler.
Init instead of memset.
2013-10-09 23:35:29 +13:00
Matthew Parlane f811dbb575 Fix accept() bug, which was using the wrong isRW for error conversion.
Also fixed a debug issue where local_name is used uninitialised.
2013-10-09 22:16:04 +13:00
Ryan Houdek 932586f743 [JitArmIL] Fix Windows build? 2013-10-09 00:55:43 +00:00
Ryan Houdek 4263f062b0 [JITArmIL] Fix JITIL compiling on x86. 2013-10-08 19:16:27 -05:00
Ryan Houdek ce5dbfe7a7 [ARM-JITArmIL] Begin merging of the two IL cores in to a common source base. 2013-10-08 19:23:37 +00:00
Ryan Houdek a0f2183424 [ARM-JITArmIL] Provide the necessary instructions to allow the JitArmIL to actually run. Disable branch instructions as well for now since one is wrong somewhere. 2013-10-08 16:42:33 +00:00
Ryan Houdek 0236ba3f86 [ARM-JITARMIL] CMP optimization. Int3 IR. DownCount added. 2013-10-07 23:25:13 +00:00
Ryan Houdek a4eab75d15 [ARM] Beginning of JitArmIL JIT recompiler. This code is entirely based off of magumagu's work with the JIT64IL. This puts down the base needed for the recompiler. Starfield actually runs quite a few blocks being mysteriously dying while most other games just exit out due to an instruction not being done yet. A lot of this code can be tidied up and actually shared between the two IL recompilers but isn't yet done. 2013-10-07 14:25:07 +00:00
comex d6f0ecebb4 Fix updating the register even if an exception occurred in MMU mode. 2013-10-07 02:34:38 -04:00
comex a9908fdf09 Fix build issues on OS X i386. 2013-10-07 02:16:51 -04:00
skidau 2b08172a45 Corrected the XMM0 processing for pairedStoreQuantized in the x86 build.
Updated the PC for the fifo writes processing in the less common case.
2013-10-06 16:31:34 +11:00
Matthew Parlane 8722b9cfb5 Missing break for IOCTL_ES_DELETETITLECONTENT 2013-10-06 16:22:44 +13:00
Lioncash a7d073b0da [Core] Simplify a loop within function Run() in Interpreter.cpp. The increment variable for the for loop is never used at all, so it's sensible to replace it with a while loop. 2013-10-05 14:38:21 -04:00
skidau 2d00c3a4f8 Added code to update the PC before doing a fastmem write. Fixes games with freezing and FIFO errors since r2a339c926e43c11f7b9acc8d3af202f0be54e2b0. 2013-10-05 23:23:49 +10:00
comex 6498a77362 Fix storeSingleFloat stack alignment. 2013-10-05 07:56:57 -04:00
skidau 8324510d65 Removed the IOSVersion check (caused some Wii games like Battalion Wars 2 not to boot). 2013-10-05 13:07:40 +10:00
skidau b7d1c03c8a Added a check for the IOS version before overriding the OSGetResetCode() function.
Fixes issue 6700.
2013-10-05 12:41:59 +10:00
comex a51eb5fd19 Fix idle skipping.
It incorrectly continued to test EAX after it was changed to load
directly to the assigned register.

Also switch from a flush to ABI_PushRegistersAndAdjustStack, to avoid
needless flushing in the no-idle case.
2013-10-04 20:33:39 -04:00
comex 3679f9ba60 Don't push registers before pairedStoreQuantized, that's dumb.
And fix some stuff up.  It would probably be good to unify the stack
handling some more rather than having ABI_PushRegistersAndAdjustStack do
part of it and ABI_AlignStack the rest, causing unnecessary subtract
instructions on Linux x86 (only).
2013-10-04 15:48:06 -04:00
comex a91469ffa5 Fix stfd, which was broken in the fastmem writes commit. 2013-10-03 18:22:35 -04:00
comex 5e4665301b Finish replacing ThunkManager with ABI_PushRegistersAndAdjustStack.
As part of that, change SafeLoadToEAX to SafeLoadToReg, and have JitIL
use that, which should fix fastmem on JitIL.

This should also fix a potential stack corruption issue with x86.
2013-10-03 18:22:35 -04:00