* Do not wait on DEVICE 0x30 semaphore, it seems like it is something to do with queue command synchronization.
- This also fixes cellGcmSetFlipWithWaitLabel which is built specifically to enable accurate RSX flipping time, its waiting command is confirmed to be placed **AFTER** DEVICE 0x30 waiting.
* Fix default vsync state to be enabled. (and set it to enabled in cellGcmSetVBlankFrequency as well)
* Add experimental "Infinite" frame limiter mode.
* Fix spurious enabling of second vblank.
* BufferUtils: use naive function pointer on Apple arm64
Use naive function pointer on Apple arm64 because ASLR breaks asmjit.
See BufferUtils.cpp comment for explanation on why this happens and how
to fix if you want to use asmjit.
* build-macos: fix source maps for Mac
Tell Qt not to strip debug symbols when we're in debug or relwithdebinfo
modes.
* LLVM PPU: fix aarch64 on macOS
Force MachO on macOS to fix LLVM being unable to patch relocations
during codegen. Adds Aarch64 NEON intrinsics for x86 intrinsics used by
PPUTranslator/Recompiler.
* virtual memory: use 16k pages on aarch64 macOS
Temporary hack to get things working by using 16k pages instead of 4k
pages in VM emulation.
* PPU/SPU: fix NEON intrinsics and compilation for arm64 macOS
Fixes some intrinsics usage and patches usages of asmjit to properly
emit absolute jmps so ASLR doesn't cause out of bounds rel jumps. Also
patches the SPU recompiler to properly work on arm64 by telling LLVM to
target arm64.
* virtual memory: fix W^X toggles on macOS aarch64
Fixes W^X on macOS aarch64 by setting all JIT mmap'd regions to default
to RW mode. For both SPU and PPU execution threads, when initialization
finishes we toggle to RX mode. This exploits Apple's per-thread setting
for RW/RX to let us be technically compliant with the OS's W^X
enforcement while not needing to actually separate the memory
allocated for code/data.
* PPU: implement aarch64 specific functions
Implements ppu_gateway for arm64 and patches LLVM initialization to use
the correct triple. Adds some fixes for macOS W^X JIT restrictions when
entering/exiting JITed code.
* PPU: Mark rpcs3 calls as non-tail
Strictly speaking, rpcs3 JIT -> C++ calls are not tail calls. If you
call a function inside e.g. an L2 syscall, it will clobber LR on arm64
and subtly break returns in emulated code. Only JIT -> JIT "calls"
should be tail.
* macOS/arm64: compatibility fixes
* vm: patch virtual memory for arm64 macOS
Tag mmap calls with MAP_JIT to allow W^X on macOS. Fix mmap calls to
existing mmap'd addresses that were tagged with MAP_JIT on macOS. Fix
memory unmapping on 16K page machines with a hack to mark "unmapped"
pages as RW.
* PPU: remove wrong comment
* PPU: fix a merge regression
* vm: remove 16k page hacks
* PPU: formatting fixes
* PPU: fix arm64 null function assembly
* ppu: clean up arch-specific instructions
- Adds AVX512 path for upload_untouched u16 with primitive restart, and
AVX2 and AVX512 paths for upload_untouched without restart
- The AVX512 paths handle the remainder in simd code with masking, which
provided a large speedup
- On my i5-1135G7 in demons souls benchmarking a scene in boletaria with
a lot of geometry on screen via perf:
SSE4_1 0.64%
AVX2 0.59%
AVX512 0.56%
AVX512 w/ remainder masking 0.51%
- Move texture object code out of the monolithic header
- All texture binds go through the shared state
- Transient texture binds use a dedicated temp image slot shared with native UI
Part of OpenGL's achilles' heel - the API does not distinguish between VRAM and SYSTEM memory at all and relies on developers wrestling with the driver's heurestic algorithm for this.
- Turns out the AMD driver really hates it if you render with a mapped index buffer.
The driver internally seems to make a copy of the consumed indices and uses that. Very slow.
I was able to isolate this after observing that glDrawArrays is not entirely shit, but glDrawElements duration scaled linearly with the number of vertices.
