Commit Graph
636 Commits
Author SHA1 Message Date
Ryan Houdek 6683b194ff ARMv7 register cache optimizations.
Enable support for not loading a destination register on FPR cache.
Dump registers if they won't be used later in the block. Stolen from Fiora.
2014-11-16 09:29:22 +00:00
Ryan Houdek 181f16c5f0 Reimplements fastmem for ARMv7 floating point loadstores.
This implements a new system for fastmem backpatching on ARMv7 that is less of a mindfsck to deal with.
This also implements stfs under the default loadstore path as well, not sure why it was by itself in the first place.

I'll be moving the rest of the loadstore methods over to this new way in a few days.
2014-11-15 21:17:50 +00:00
Ryan Houdek b0becf7af8 Enables the ARMv7 FPR cache.
The problem instructions with the FPR are disabled in the previous commit. This can now be reenabled for fairly large performance gains.
2014-11-14 15:14:10 +00:00
Ryan Houdek 69c3e6516c Disables NEON optimized instructions.
These are causing issues in games. In particular you get pink on the screen in Animal Crossing.
Disable until fully investigated.

This also disables fastmem on floating point loadstore instructions which are horribly broken and won't actually backpatch when an invalid read/write
is encountered.
2014-11-14 15:13:13 +00:00
Stevoisiak b25e1a2eb4 Various formatting and consistency fixes 2014-11-13 22:42:18 -05:00
skidau b1f8974db8 Merge pull request #1527 from FioraAeterna/mftbfix
JIT: revert accuracy improvement to mftb
2014-11-13 12:11:13 +11:00
Fiora 4b105ed0e4 JIT: revert accuracy improvement to mftb
Fixes a few games (e.g. Karaoke Revolution Party) for reasons explained in the
comments.
2014-11-10 20:31:07 -08:00
Fiora 6603f98d04 JIT: add 64-bit write support to FIFO functions
Also fix 64-bit values passed to CallAC and otherwise correct immediate
handling in FIFO writes.
Fixes 007 Nightfire.
2014-11-09 21:24:30 -08:00
skidau c36e7b9c23 Merge pull request #1494 from lioncash/statics
PPCCache: Make PLRU lookup tables static
2014-11-07 12:33:19 +11:00
Lioncash a5d304eb16 Merge pull request #1495 from lioncash/unused
Interpreter: Remove dead patches() function
2014-11-06 20:31:01 -05:00
Fiora b8d88a41e0 JIT: remove accidentally left-in debug code 2014-11-05 17:44:13 -08:00
Lioncash 606efbce10 PPCCache: Make PLRU lookup tables static 2014-11-05 19:35:30 -05:00
Rohit Nirmal 1beb047959 PowerPC: Remove unused variable. 2014-11-05 11:47:44 -05:00
Lioncash a105a9a557 Interpreter: Remove dead patches() function 2014-11-04 20:44:57 -05:00
skidau 0515ab852e Merge pull request #1230 from FioraAeterna/constaddr
JIT: improve handling of stores with a known address
2014-11-05 12:40:38 +11:00
Fiora b81686b582 JIT: fix register preloading
Partially broken by typoes in the bitset patch.
2014-11-04 04:50:05 -08:00
Fiora f8880c0284 JIT: fix typo in optimization patch
Whoops... made us flush everything on every branch.
2014-11-04 02:04:30 -08:00
comex 9cba787871 Merge pull request #1408 from randomstuff/perf
Profiling: measure time on non-Windows/POSIX using clock_gettime
2014-11-03 22:36:32 -05:00
Lioncash 30f97723db Core: Fix potentially uninitialized variable warnings 2014-11-03 22:21:10 -05:00
comex 42d41a456e Merge pull request #1489 from FioraAeterna/revertopt
JIT: revert cmpXX optimization
2014-11-03 21:07:11 -05:00
Fiora 768273f59b JIT: revert cmpXX optimization
It seems like this wasn't correct in 100% of cases.
2014-11-03 17:50:20 -08:00
skidau 027791685a Merge pull request #1483 from comex/on-demand-exi-interrupts
Make EXI use CoreTiming events like everything else instead of having its own special check.
2014-11-04 12:31:12 +11:00
Fiora ce71c3cd4e JIT: fix valid_block marking
This caused invalidations that only affected the last portion of a JIT block
to fail, breaking Wii64's block linking. It might affect a bunch of other
games too; I haven't tested.
2014-11-03 16:23:44 -08:00
Fiora fc63c7ecae JIT: genericize immediate address handling, support in float stores too 2014-11-03 01:31:39 -08:00
comex 9f683f353b Make EXI use CoreTiming events like everything else instead of having its own slow special check.
Microphone is probably wrong/mistimed because it doesn't take into
account cycles late, but that's not a new issue here.
2014-11-03 00:28:46 -05:00
Fiora e729fc4a28 JIT: fix dumb mistake in crclr optimization patch 2014-11-02 21:03:11 -08:00
Ryan Houdek 204598a082 Merge pull request #1350 from FioraAeterna/integeropts
Various smallish JIT optimizations
2014-11-02 20:13:20 -06:00
Gabriel Corona 641e820257 Profiling: measure time on POSIX systems using clock_gettime 2014-11-03 00:07:12 +01:00
Ryan Houdek 6e43562496 Merge pull request #1468 from Tilka/cleanup
Small cleanup
2014-11-02 11:02:35 -06:00
Ryan Houdek 0d70880d89 Merge pull request #1466 from Sonicadvance1/ARMv7-and-optimization
Optimizes ARMv7 andi{s,}_rc implementations.
2014-11-02 09:33:37 -06:00
Ryan Houdek 824bad458c Merge pull request #1454 from lioncash/interp
Interpreter: Remove a redundant macro
2014-11-02 09:33:19 -06:00
Tillmann Karras f4fed0dea0 JitAsm: remove unused code pointers 2014-11-02 02:00:47 +01:00
Ryan Houdek 86ca63658b Optimizes ARMv7 andi{s,}_rc implementations.
Cuts down from a 3 instruction max implementation down to 1 instruction if the immediate can fit in to the instruction encoding.
2014-11-01 13:06:52 +00:00
Fiora 7deaf00c44 JIT: more mftb fixes
A very subtle difference in how I calculated the timebase value seems
to have broken Karaoke Revolution; this seems to fix it. Also be a bit more
paranoid in conditions for mftb merging.
2014-11-01 03:15:25 -07:00
Lioncash 475bb40364 Interpreter: Remove a redundant macro 2014-10-31 10:55:25 -04:00
Fiora fb0960f0ee JIT: flush unused registers during branch merges
Also correct some flags in interpreter tables.
2014-10-29 00:32:59 -07:00
Fiora 1ec1a9c33a JIT: optimize crclr special case of crxor 2014-10-29 00:30:27 -07:00
Fiora 97fba41860 JIT: merge fcmpx and cror
Almost all uses of boolean condition-register ops in real code seem to be
the combination fcmpx + cror (e.g. for <= or >=). This merges the two.
2014-10-29 00:30:27 -07:00
Fiora a666bb6bf6 JIT: optimize mulhwu 2014-10-29 00:30:26 -07:00
Fiora 5b5e462200 JIT: reorder blr comparisons
This should allow macro-op fusion in blr instructions.
2014-10-29 00:30:26 -07:00
Fiora 7388c62439 JIT: use BLR optimization to avoid anding LR with 0xFFFFFFFC
Should save roughly one instruction per blr.
2014-10-29 00:30:26 -07:00
Fiora cec2cb9d38 JIT: micro-optimize cmpXX a bit more for lower latency/code size 2014-10-29 00:30:25 -07:00
comex 089e32ba7d Merge pull request #1307 from comex/bitset
Higher level bitset wrapper
2014-10-28 23:39:35 -04:00
skidau 1630b0c684 Merge pull request #1291 from skidau/debugger-step-out
Dolphin debugger enhancements
2014-10-28 12:53:22 +11:00
skidau daf977e84e Disabled optimizations, block merging and instruction skipping code while the debugger is stepping. 2014-10-27 23:38:45 +11:00
skidau 4570dd7eeb Fixed a crash that would occur if a new watch were added by entering a watch name.
Code style updates.
2014-10-26 23:23:45 +11:00
skidau 2a3a8520a2 Removed the EXRAM_MASK from the safe address check as it was causing invalid accesses on Wii games in DEBUGFAST builds 2014-10-26 14:56:03 +11:00
skidau 290e1bed37 Disable block linking while debugger stepping or if there are breakpoints 2014-10-26 14:56:02 +11:00
skidau 613cae613a Added a RAM Watch window to the debugger
Conflicts:
	Source/Core/Core/HW/Memmap.cpp
	Source/Core/Core/HW/Memmap.h
	Source/Core/DolphinWX/Debugger/CodeWindow.h
2014-10-26 14:56:02 +11:00
skidau df37649b9f Changed the step over routine to a single stepping version that steps until a blr is encountered.
Cleared out all temporary breakpoints on each step to prevent phantom breakpoints from stopping the debugger.
2014-10-26 14:56:02 +11:00