Commit graph

4737 commits

Author SHA1 Message Date
Nekotekina 72574b11ff SPU: use reservation spinlocks on writes (non-TSX)
This should decrease contention by avoiding global lock
2018-05-21 21:56:14 +03:00
Nekotekina 182259e4a5 cellSaveData fix 2018-05-21 21:47:04 +03:00
kd-11 c9669818eb Facepalm
- overlays: Do not free self handle!!!!
2018-05-21 15:55:25 +03:00
Nekotekina 6fd402bcf8 Transactions: drop RDTSC usage
Use simple increment (minor optimization)
2018-05-21 00:18:37 +03:00
Nekotekina fb0f5f0c8b Fix vm::reservation_notifier 2018-05-21 00:18:37 +03:00
Nekotekina 7a6db3dd7e spu_getll_tx: diagnose loop count 2018-05-21 00:18:37 +03:00
Nekotekina 295f782b30 spu_putlluc_tx: diagnose loop count 2018-05-21 00:18:37 +03:00
Nekotekina a33f297315 Replace notifier::lock_shared() with try_lock_shared()
Also add notify_one(), try_lock() and unlock()
Move some code in cond.cpp
2018-05-21 00:18:37 +03:00
Nekotekina 8d5bbfb850 Update SPU dump: print current MFC cmd 2018-05-21 00:18:37 +03:00
Nekotekina 33a1c743a4 Transactions: move loops inside
Rewrite loops in assembler (minor optimization)
2018-05-21 00:18:37 +03:00
kd-11 f6f45b8699
Native UI refactored (#4623)
Refactor and improve native overlays
2018-05-20 23:05:00 +03:00
Megamouse 68fff54a8b cellSaveData: don't return after funcStat was successful in savedata_op
also add some missing constants
2018-05-20 23:16:11 +04:00
scribam bf2f00a163 log: add Vulkan header version 2018-05-19 10:59:10 +03:00
Nekotekina 67391322f4 Transactions: touch memory only after a failure
Minor optimization
2018-05-17 22:19:26 +03:00
Nekotekina 39088e5005 SPU: Allow waiting on LR+TM events (fixup) 2018-05-17 21:21:15 +03:00
Nekotekina 7afda26c69 spu_putllc_tx: preload a little more (WIN32) 2018-05-17 18:58:08 +03:00
Nekotekina 58e4d49bb5 Fixup (endless loop) 2018-05-17 18:33:44 +03:00
scribam 70e2873e69 Compilation fix for GCC 8 (at least) 2018-05-17 00:53:31 +04:00
Nekotekina 367f039523 Build transactions at runtime
Drop _xbegin family intrinsics due to bad codegen
Implemented `notifier` class, replacing vm::notify
Minor optimization: detach transactions from global mutex on TSX path
Minor optimization: don't acquire vm::passive_lock on PPU on TSX path
2018-05-16 17:31:58 +03:00
Nekotekina fd525ae1cf Implement build_function_asm
Uses ASMJIT to build function at startup
2018-05-16 15:48:35 +03:00
scribam fd8aae5725 emu: remove wait_engine files
These files are empty and do not serve any purpose currently
2018-05-16 03:56:06 +04:00
scribam 196f985283 hle: improve cellSync2 error checks and logging 2018-05-15 22:59:34 +04:00
sftt 27004eb934 Avoid illegal available_user_memory in sys_memory (#4399)
jarveson says "the comment and minus can be completely deleted/reverted" in 17cf24d0ed
2018-05-15 17:40:45 +01:00
scribam 04ad49de4d typos 2018-05-14 21:14:39 +04:00
isJuhn fc23243d01 Fix setParam in cellHddGameCheck 2018-05-14 16:56:47 +01:00
Nekotekina 3c70645f0b Update SPU cache (v2)
Improve SPU analyser: filter unreachable fragments
More strict NOP/LNOP analysis
Fill block predecessors info
ASMJIT: fix assertion and improve indirect branch
2018-05-13 20:40:23 +03:00
kd-11 4836a03a7d rsx: Fix build 2018-05-13 14:44:14 +03:00
kd-11 9d1f4a2538 vk: Include RADV POLARIS and RADV VEGA in the primitive restart
blacklist
2018-05-13 14:44:14 +03:00
kd-11 bff6060bd6 rsx: Improve puller state management
- Properly identify puller spin primitives
- Add a small wake delay after exiting a spin delay. Fixes desynchronization
  It seems real hw has a small delay between cell edits to commandbuffer memory at the GET address and the changes becoming visible to the DMA puller
  Simulated with a short busy_wait, large values will improve sync but degrade performance
2018-05-13 14:44:14 +03:00
kd-11 1aa44ede31 gl: Improve AMD multidraw workaround
- Reimplements the AMD workaround using an identity buffer to avoid the performance hit of doing multiple glDrawArrays for every single compiled set
- Reimplements first/count allocation using a scratch buffer to reduce allocation overhead when large number of draw calls is used
2018-05-13 14:44:14 +03:00
kd-11 eccb57d4b8 vk: AMD primitive restart bug workaround
- Emulate primitive restart with degenerate triangles
2018-05-13 14:44:14 +03:00
kd-11 b7979d3f57 rsx/vk: Improvements and minor optimizations
- Improve dirty state tracking affecting program state
- vk: Refactor out transform constants upload into a separate channel to avoid if possible
  transform data uploads are quite expensive
2018-05-13 14:44:14 +03:00
kd-11 440a31ef18 rsx: Optimizations for program management 2018-05-13 14:44:14 +03:00
kd-11 a52ea7f870 rsx: Improve fragment and vertex program usage
- Introduces a gpu program analyser step to examine shader contents before attempting compilation or cache search
  - Avoids detecting shader as being different because of unused textures having state changes
  - Adds better program size detection for vertex programs
- Improved vertex program decompiler
  - Properly support CAL type instructions
  - Support jumping over instructions marked with a termination marker with BRA/CAL class opcodes
  - Fix SRC checks and abort
  - Fix CC register initialization
  - NOTE: Even unused SRC registers have to be valid (usually referencing in.POS)
2018-05-13 14:44:14 +03:00
Jake 75b40931fc rsx: initial capture/replay functionality (#4510)
* rsx: initial capture/replay functionality
2018-05-13 12:18:05 +03:00
Nekotekina 7cb4672344 Travis: remove LLVM for now 2018-05-13 00:03:08 +03:00
Nekotekina 4c7afb6952 Opt-in set DAZ and FTZ 2018-05-13 00:03:08 +03:00
Nekotekina a77896c1b6 Update LLVM commit, use small code model 2018-05-13 00:03:08 +03:00
Nekotekina be5c18cc85 SPU Re: more precise jt generation
Improve analyser, set v1
Fix branch indirect conditional
2018-05-12 23:50:49 +03:00
Nekotekina 737db90058 Set DAZ and FTZ 2018-05-09 23:38:26 +03:00
Nekotekina 5d15d64ec8 Memory mirror support
Implemented utils::memory_release (not used)
Implemented utils::shm class (handler for shared memory)
Improved sys_mmapper syscalls
Rewritten ppu_patch function
Implemented vm::get_super_ptr (ignores memory protection)
Minimal allocation alignment increased to 0x10000
2018-05-09 23:35:34 +03:00
Nekotekina fe4c3c4d84 Implement SPU recompiler cache
Shared between ASMJIT/LLVM recompilers, compiled at startup
2018-05-09 23:35:18 +03:00
Nekotekina f5ee6fb113 Convert newlines in game title 2018-05-09 22:19:55 +03:00
Nekotekina acfe22e5bc SPU: improve jumptable detection 2018-05-09 22:19:55 +03:00
Nekotekina 8f91917e8c SPU ASMJIT: simplify patchpoints
Remove SPU thread reference from spu_recompiler_base
Disable support for far jumps in pathpoints (they were rare and unsafe)
2018-05-09 22:19:55 +03:00
Nekotekina 1ca51a023c SPU LLVM Recompiler (preview) 2018-05-09 22:19:55 +03:00
Nekotekina a975ecdc4f Use LLVM 6 2018-05-08 13:05:29 +03:00
Nekotekina 8b704588d0 Update cpu_translator 2018-05-08 13:05:29 +03:00
Nekotekina db229dffb0 PPU LLVM: minor update 2018-05-08 13:05:29 +03:00
Nekotekina 16f5167aed SPU Analyser improved
This restores some functionality removed in SPU ASMJIT v2.0
Also implements new experimental features
2018-05-08 13:05:29 +03:00