RPCSX/rpcsx
1
0
Fork 0
mirror of https://github.com/RPCSX/rpcsx.git synced 2025-12-06 07:12:14 +01:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity
13209 commits 5 branches 1 tag 150 MiB
Commit graph

168 commits

Author SHA1 Message Date
Nekotekina 349f251d14 PPU LLVM: use masked stores for STVLX/STVRX 
Drop maskmove intrinsic, not portable.
Its implicit NT hint may also hurt performance.
 2022-01-20 21:16:00 +03:00
Nekotekina 14cca55b50 PPU: refactor vector rounding instructions 
Fix: nearbyint -> roundeven
 2022-01-18 00:10:32 +03:00
Nekotekina f95395b351 PPU LLVM: improve accuracy of VSL/VSR 
Passes tests, should now be equal to interpreter.
 2022-01-15 21:13:31 +03:00
Nekotekina df24cff0b1 PPU LLVM: fix VMINFP and VMAXFP accuracy 
PPU cache needs to be cleared.
 2022-01-15 17:36:57 +03:00
Nekotekina 6dda047128 PPU LLVM: fix VNMSUBFP sign handling 
PPU cache needs to be cleared.
 2022-01-15 17:36:57 +03:00
Nekotekina e9efa73eed PPU: restore previous NJ mode handling option 
Fix the divergence between PPU Interpreter and LLVM.
 2022-01-15 17:36:57 +03:00
Nekotekina 580bd2b25e Initial Linux Aarch64 support 
* Update asmjit dependency (aarch64 branch)
* Disable USE_DISCORD_RPC by default
* Dump some JIT objects in rpcs3 cache dir
* Add SIGILL handler for all platforms
* Fix resetting zeroing denormals in thread pool
* Refactor most v128:: utils into global gv_** functions
* Refactor PPU interpreter (incomplete), remove "precise"
* - Instruction specializations with multiple accuracy flags
* - Adjust calling convention for speed
* - Removed precise/fast setting, replaced with static
* - Started refactoring interpreters for building at runtime JIT
*   (I got tired of poor compiler optimizations)
* - Expose some accuracy settings (SAT, NJ, VNAN, FPCC)
* - Add exec_bytes PPU thread variable (akin to cycle count)
* PPU LLVM: fix VCTUXS+VCTSXS instruction NaN results
* SPU interpreter: remove "precise" for now (extremely non-portable)
* - As with PPU, settings changed to static/dynamic for interpreters.
* - Precise options will be implemented later
* Fix termination after fatal error dialog
 2022-01-15 06:48:04 +03:00
Eladash a60cee6536 Update PPUTranslator::MTFSFI for its intention to be clearer 2022-01-12 03:37:39 +03:00
Nekotekina e3e39e8de3 PPU LLVM: rewrite and optimize saturation bit 
Use vector accumulator
 2021-12-03 00:14:06 +03:00
Nekotekina 209b14fbac PPU LLVM: inline remaining vector instructions 2021-12-03 00:14:06 +03:00
Nekotekina 04c9d01390 PPU LLVM: modernize most vector instructions 
Rewritten VSUM instructions:
VSUMSWS, VSUM2SWS, VSUM4SBS, VSUM4SHS, VSUM4UBS
 2021-12-03 00:14:06 +03:00
Nekotekina c9d8e59dbf PPU LLVM: allow to drop setting SAT flag (optimization, module-wide) 
Implement ppu_attr::has_mfvscr (partially, module-wide search).
If this instruction isn't found, allow to drop setting SAT flag.
It's based on presumption that only MFVSCR can retrieve SAT flag.
 2021-12-03 00:14:06 +03:00
Nekotekina 86b194014b PPU LLVM: rewrite more packing instructions 
Rewritten VPKUHUM, VPKUHUS, VPKUWUM, VPKUWUS.
Decoupled saturation test from sat pack pattern.
 2021-12-03 00:14:06 +03:00
Nekotekina e7c827f73b PPU LLVM: rewrite some packing instructions 
Rewritten VPKSHSS, VPKSHUS, VPKSWSS, VPKSWUS.
Decoupled saturation test from sat pack pattern.
 2021-12-03 00:14:06 +03:00
Nekotekina abe498f35c PPU LLVM: modernize some code with new DSL 
PPU: rewritten instructions VMHADDSHS, VMHRADDSHS
PPU: added optimized path for VPERM (ra=rb)
 2021-12-03 00:14:06 +03:00
Nekotekina 69f321a471 LLVM 13 2021-11-02 20:11:08 +03:00
Malcolm Jestadt f06c8b22e8 PPU/SPU LLVM: Emulate VPERM2B with a 256 bit wide VPERMB 
- Save 1 uop by using 256 wide VPERMB instead of VPERM2B. (Compiles down to a vinserti128 and vpermb)
 2021-10-13 17:51:54 +03:00
Nekotekina 4b8ee85995 LLVM DSL: reimplement pshufb, add 'calli' 
Implement postponed custom intrinsic replacement.
Make bitcast operator static like other ones.
 2021-09-17 10:23:43 +03:00
Nekotekina 7cf9d1380b LLVM DSL: add line number in get_const_vector automatically 2021-09-17 10:23:43 +03:00
Eladash f98595bee5 Patches/PPU: Add jump_link patch type 2021-09-10 11:46:39 +03:00
Nekotekina 06f733a7f2 Fixup No.2 for #10779 2021-09-01 16:56:38 +03:00
Eladash b40ed5bdb7
Patches/PPU: Extend and improve patching capabilities (code allocations, jumps to any address) (#10779) 
* Patches/PPU: Implement dynamic code allocation + Any-Address jump patches

Also fix deallocation path of fixed allocation patches.
 2021-09-01 13:38:17 +03:00
Eladash ddb042148d Patches/LLVM: Implement Complex Patches Support 2021-08-26 23:04:32 +03:00
Nekotekina 160b131de3 types.hpp: implement smin, smax, amin, amax 
Rewritten the following global utility constants:
`umax` returns max number, restricted to unsigned.
`smax` returns max signed number, restricted to integrals.
`smin` returns min signed number, restricted to signed.
`amin` returns smin or zero, less restricted.
`amax` returns smax or umax, less restricted.

Fix operators == and <=> for synthesized rel-ops.
 2021-05-22 12:10:57 +03:00
Megamouse a16d8ba3ea More random changes 2021-04-11 14:01:51 +03:00
Nekotekina 87af905018 Enable -Wunused-parameter 2021-03-06 18:07:08 +03:00
Nekotekina 0c034ad7de PPU LLVM: upgrade to GHC call conv 
Get rid of some global variables.
Implement ppu_escape (unused yet).
Bump PPU cache version to v4.
 2021-02-01 11:30:50 +03:00
Nekotekina c89362f6a2 PPU LLVM: don't use module name as PRX indicator 2021-02-01 11:30:50 +03:00
Nekotekina 8a029159cd PPU Analyser: compile certain functions on per-instruction basis 
PPU LLVM: optimize small blocks
 2021-02-01 11:30:50 +03:00
Nekotekina 382509d778 PPU LLVM: Implement inline __add_get_ov 2021-02-01 11:30:50 +03:00
Nekotekina f9ee8978ff PPU LLVM: improve analyser 
Compile possibly executable holes between detected functions.
Add unused "PPU LLVM Greedy Mode" option (for future updates).
Add "nounwind" attribute to compiled functions (reduces size).
 2021-02-01 11:30:50 +03:00
Nekotekina db8e6fe7a7 Enable -Wunused-variable 2021-01-12 14:34:14 +03:00
Nekotekina bd269bccaf types.hpp: remove intrinsic includes 
Replace v128 with u128 in some places.
Removed some unused files.
 2020-12-21 21:11:25 +03:00
Nekotekina fb29933d3d Add usz alias for std::size_t 2020-12-18 12:23:53 +03:00
Eladash 7eb16e13bb PRX loader: Fix libfs_155.sprx loading 
Fix relocations' segments referencing when there are "empty" (memsize=0) LOAD segments.
 2020-12-15 11:16:45 +03:00
Nekotekina e321765c54 Split BEType.h to util/v128.hpp and util/to_endian.hpp 2020-12-13 16:34:45 +03:00
Nekotekina 65c04e4ddd Remove constexpr from ppu/spu decoders. 
We don't need them at compile time (yet).
But can reduce compile time and complexity.
 2020-12-10 15:06:01 +03:00
Nekotekina 36c8654fb8 Remove HERE macro 
Some cleanup.
Add location to some functions.
 2020-12-10 12:30:22 +03:00
Nekotekina 5d934c8759 Improve narrow() and size32() with src_loc detection 2020-12-09 16:26:20 +03:00
Nekotekina e055d16b2c Replace verify() with ensure() with auto src location. 
Expression ensure(x) returns x.
Using comma operator removed.
 2020-12-09 15:43:38 +03:00
RipleyTom af8c661a64 Remove BOM markers 2020-12-06 15:30:12 +03:00
Nekotekina 1b8bf081b5 Upgrade to LLVM 11 Stable 2020-11-02 21:23:25 +03:00
Eladash 443c2b920d PPU: Handle cache line inconsistencies (PPU 128 reservations) 2020-10-16 22:51:30 +03:00
Nekotekina f2d2a6b605 JIT cleanup for PPU LLVM 
Remove MemoryManager3 as unnecessary.
Rewrite MemoryManager1 to use its own 512M reservations.
Disabled unwind info registration on all platforms.
Use 64-bit executable pointers under vm::g_exec_addr area.
Stop relying on deploying PPU LLVM objects in first 2G of address space.
Implement jit_module_manager, protect its data with mutex.
 2020-10-11 17:22:28 +03:00
Eladash f4ca6f02a1 PPU: Implement support for 128-byte reservations coherency 2020-09-28 22:34:42 +03:00
Eladash 09cddc84be SPU/PPU: Implement Atomic Cache Line Stores 2020-09-27 20:09:21 +03:00
Eladash 8cdfe5952a
SPU/PPU LLVM: Improve 0 addend FMA detection (#8709) 2020-08-13 04:13:08 +03:00
Whatcookie 4ce2ad54a8
PPU LLVM: Use VPERM2B to emulate VPERM (#8704) 
- The VPERM2B instructions are a match of VPERM's behavior, besides operating in reverse byte order
 2020-08-09 01:50:26 +01:00
Eladash 7e11855330 SPU/PPU LLVM: Fix FMA signed zeroes handling 2020-08-08 22:21:22 +01:00
Eladash 6a51c27fde PPU LLVM: Fix VMAXFP, VMINFP NaN handling 2020-08-03 15:43:00 +01:00