libretro/ppsspp
1
0
Fork 0
mirror of https://github.com/libretro/ppsspp.git synced 2024-12-14 04:28:42 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
15,088 Commits 19 Branches 54 Tags 247 MiB
4db6b7f3e2
Commit Graph

122 Commits

Author SHA1 Message Date
Unknown W. Brackets
ab7dd0df25 x86jit: Add an option to enable/disable vpfu simd. 2014-11-17 20:37:27 -08:00
Unknown W. Brackets
d0a2ced2f9 x86jit: Flip cc in stl* to avoid reg loads. 
Unfortunately, this zero thing is now concerning me...
 2014-11-09 08:15:39 -08:00
Unknown W. Brackets
b53f13480a x86jit: Centralize continuing logic. 2014-10-12 19:01:04 -07:00
Unknown W. Brackets
d98adf27d6 x86jit: Add proxy blocks for continuing. 2014-10-12 17:15:31 -07:00
Unknown W. Brackets
2f598e8f38 jit: Statically jump for fixed branches. 
This handles both loops (first step is known) and static branches (some
code uses them instead of jumps, and we disassemble that to "b".)

Not likely to be a big improvement, but might help if the branch predictor
was wrong.

This is as opposed to continuing, which would build a larger jit block.
 2014-10-12 12:51:47 -07:00
Unknown W. Brackets
928e2adfc9 jit: Avoid applying/restoring the rounding mode. 
If the game never sets it, we can skip around syscalls, interpreter,
replacements, etc.
 2014-10-12 12:51:45 -07:00
Unknown W. Brackets
8d0dca71fe jit: Rename the rounding mode funcs to clarify. 
They apply/restore the value, set/clear is confusing.
 2014-10-12 11:35:20 -07:00
Unknown W. Brackets
4459b8f483 jit: Actually jit vmtfc/vmfvc. 
Sicne we have them and they are easy.
 2014-09-01 23:13:39 -07:00
Unknown W. Brackets
925557ed47 x86jit: Maintain the rounding mode always. 
This should be less often than doing it per block that uses fpu, unless
the game doesn't use fpu much at all.
 2014-08-22 09:53:00 -07:00
Unknown W. Brackets
dc91dc1ce8 x86jit: Support fpu rounding modes for mul, etc. 
Fixes Gods Eater Burst loading PSP savedata, but can no longer load old
savedata.
 2014-08-21 23:59:55 -07:00
Henrik Rydgard
903ddbc513 x86 JIT: Fix bug where NOR would not get computed correctly in corner case 
(CompTriArith can end up not actually mapping rd to a register when taking
a shortcut)

May fix the JIT issue mentioned by CPkmn and located by daniel229 as an aside in #6638
 2014-07-27 21:41:41 +02:00
Unknown W. Brackets
09b9d2ad81 Keep track of ranges that have emuhack ops. 
So that we can invalidate them smarter.
 2014-07-05 16:25:16 -07:00
Unknown W. Brackets
9efbc2694b Add an invalidate all method to the jit. 2014-06-19 01:13:06 -07:00
Unknown W. Brackets
6ccae8f5a7 x86jit: Use a faster safemem fallback. 
Really helps performance in games that use uncached addresses a lot,
without really impacting performance of most games which don't.

Of course, fastmem is faster.
 2014-05-06 08:05:12 -07:00
Unknown W. Brackets
d65af7353b Avoid some ChunkFile includes. 2014-03-15 11:32:57 -07:00
Henrik Rydgard
c80510fb3b MemMap should not be included in MIPS.h. 2014-03-15 10:45:39 +01:00
Unknown W. Brackets
2347498667 x86jit: Use templates to avoid some void * casts. 
Makes it a bit cleaner and potentially safer.
 2014-01-18 09:57:13 -08:00
Henrik Rydgård
00c32ddadb Mostly get rid of including "Globals.h" 2013-12-30 10:17:11 +01:00
Henrik Rydgard
215a269b34 Optimize dl_write_matrix just because. not expecting a big speedup... 2013-12-21 12:39:34 +01:00
Unknown W. Brackets
438361d0bc Clean up code pointer naming for the jit. 
Now it properly identifies thunk code which is actually a decent percent
when fastmem is off at least.
 2013-12-18 23:57:39 -08:00
Henrik Rydgard
1e300447e1 Fix some replace-related bugs. Add "jal" replace inlining, not activated. 2013-12-18 16:27:23 +01:00
Henrik Rydgard
2eab4aa1bf Play around with function replacement. Turned off by default of course. 2013-12-17 23:40:27 +01:00
Henrik Rydgard
2140892074 Initial preparations for ability to replace game functions with custom implementations. 
Also auto-saves hashmap additions and reapplies the hashmap on function
rename so that if you rename a function that exists in several copies
they will all be labelled.

Note that actual function replacement is not activated yet.
 2013-12-17 12:27:20 +01:00
Henrik Rydgard
2d8429ac48 Assorted cleanup in the MIPS emulation 2013-12-10 13:15:16 +01:00
Unknown W. Brackets
5d2ff64252 Support for modified jit-enabled VerySleepy. 
This allows profiling the jit.  Should have zero perf impact when not
in use, since it's entirely triggered by VerySleepy.
 2013-11-30 19:20:21 -08:00
Henrik Rydgard
5bb3824dcf Implement vocp on ARM and x86. 2013-11-19 21:41:47 +01:00
Unknown W. Brackets
a334aaf6ca x86jit: Refactor and skip flushes in branch cont. 
Still not faster, but at least the code isn't as messy.
 2013-11-12 00:45:28 -08:00
Unknown W. Brackets
7e19933f64 x86jit: Try predicting branch continues. 
Still doesn't seem to work.  Something like a 4% gain in Star Ocean was
the best I saw...
 2013-11-10 22:50:23 -08:00
Unknown W. Brackets
359110f010 x86/armjit: Add jump following (off by default.) 
Inlines function calls up to a certain extent.  Allows us to get
immediates all the way to a syscall, for example, usually.

Not sure if faster.
 2013-11-10 21:59:49 -08:00
Unknown W. Brackets
aacb31bc18 armjit: Copy over (disabled) immbranch optim. 
This does a little loop unrolling.  Costs a bit more cache space, but
avoids flushing regs for longer.

Not enabled.
 2013-11-10 21:59:48 -08:00
Unknown W. Brackets
455a7e090d Compile the cache instruction to nothing. 
Was showing up in a few profiles, does nothing currently.
 2013-11-10 14:38:10 -08:00
Henrik Rydgard
0a844ce98d Delete functions for vsge and vslt, these have been rolled into VecDo3 2013-11-09 19:29:52 +01:00
Henrik Rydgard
309f904c0c Extract JitState into its own header (arm/x86) 2013-11-08 18:51:52 +01:00
Henrik Rydgard
6eb7f94065 Implement vsgn in x86/x64 and ARM jit 2013-11-07 15:29:13 +01:00
Henrik Rydgard
aa3cf34fc1 Jit: Fix valgrind warnings. 
The first time PrefixStart was entered with startDefaultPrefix = true, it would
call EatPrefix, which checks the so far entirely uninitialized prefixXFlags.
 2013-10-16 22:33:48 +02:00
Henrik Rydgard
20174d9410 Delete the lookup table version of vh2f 2013-09-28 22:15:29 +02:00
Henrik Rydgard
7ca6d73857 Two approaches to vh2f (half-float to float): lookuptable and fast SSE 2013-09-28 22:08:44 +02:00
Henrik Rydgard
aa753c88b2 ARM: implement vhdp 2013-09-28 12:30:28 +02:00
Unknown W. Brackets
157b682344 Always use fastmem for sw/lw on SP. 2013-09-07 22:44:18 -07:00
Unknown W. Brackets
538a4c064c Add a note so as not to forget. 2013-09-01 01:15:08 -07:00
Unknown W. Brackets
b558189c37 Just invalidate blocks on ClearCacheAt(). 
This makes it safe to call from a jitted syscall, etc.
 2013-09-01 00:32:43 -07:00
Unknown W. Brackets
97aa1a631e Improve typesafety in the x86 regalloc. 2013-08-24 19:41:10 -07:00
Unknown W. Brackets
6c97b66806 Cap imm branch instructions, reset compiling. 
Break and other delay slot ops could've set it to false.

It's actually sometimes faster now.
 2013-08-24 17:26:24 -07:00
Unknown W. Brackets
109ad17ac6 Use a typesafe struct for opcodes. 
Also, correctly read delayslots using Read_Instruction on ARM.
 2013-08-24 15:36:24 -07:00
Unknown W. Brackets
df32c99be6 Attempt to follow branches to a max # of ops. 
Seems to make it slower also.  Maybe taking the branch would be better...
hmmph.
 2013-08-16 01:07:11 -07:00
Unknown W. Brackets
defd2b6383 Attempt at doing branches with imm args. 2013-08-16 01:05:52 -07:00
Unknown W. Brackets
6b0b5145e5 Clean up some inconsistency in jit branches. 2013-08-16 00:44:23 -07:00
Unknown W. Brackets
64c2ea86c0 Add a method to save the gpr/fpr state in jit. 2013-08-16 00:12:49 -07:00
Henrik Rydgard
4e8958f42d A small optimization, a few jit stubs, and cross/quat product on x86. 2013-08-01 00:15:08 +02:00
Henrik Rydgard
0a8f85a919 Some JIT cleanup, implement VI2F on ARM. also disabled untested impl of viim for x86. 2013-07-31 17:27:04 +02:00