PPSSPP Git (2026/10/14) is compiled. PPSSPP is a fast and portable PSP emulator for Android, Windows, Mac, and Linux, written in C++.

PPSSPP Git Changelog:
* Merge pull request #22557 from hrydgard/cpu-equiv-5
* VFPU: vcmov with imm3 = 7 tests a CC bit that is always clear
* ARM JITs: Write vmfvc's 0 for a control register past the defined ones
* test.py: minmax_zero and minmax_tie pass now
* VFPU: vmin and vmax order zeros and denormals as equal, on every core
* Merge pull request #22555 from hrydgard/cpu-equiv-3
* MIPSTables: Include and for the dispatch generator
* test.py: shift_encodings and allegrex_encodings pass now
* Decode the unused bits of shifts and Allegrex bit ops as the PSP does
* Merge pull request #22554 from hrydgard/cpu-equiv-2
* Unit tests: Run the IR JIT on macOS too
* ARM JIT: Keep the guest's flush-to-zero in round, ceil and floor
* ARM JIT: round.w.s of NaN, and the guest's rounding mode after conversions
* Docs: The JITs' rounding mode gaps are fixed
* Unit tests: Add CPURoundingMode
* Headless: Support --appendconfig
* IR interpreter: Force the host's mode back in RestoreRoundingMode
* JITs: Notice a rounding mode that didn't come from a compiled ctc1
* JIT: Switch rounding modes around interpreted ctc1 and syscalls only
* Save the host's FPU mode on apply, so restoring it is just a write
* ARM32: Flush denormal inputs to float-to-int conversions under FS
* IR interpreter: Use the NEON Vec4 paths only on ARM64
* ARM32: Apply the guest's rounding mode and flush-to-zero in the interpreters
* JIT: Run interpreter fallbacks in the guest's rounding mode
* x86: Set DAZ along with FTZ for the guest's flush-to-zero mode
* IR interpreter: Apply the guest's rounding mode in the dispatcher
* Merge pull request #22553 from hrydgard/gpu-fallback-fixes
* frametests: Add 13 dumps from ppsspp-framedumps
* Compat: Turn on CorrectCullAfterClip for Nova
* Compat: Turn on CorrectCullAfterClip for GachiTora
* Blend in the shader where stencil to alpha needs dual source blending we lack
* Don't use BlueToAlpha with framebuffer fetch
* Keep the stencil when resizing a framebuffer
* Rebind the render target after writing stencil to another framebuffer
* Discard against one-sided depth ranges in the fragment fallback
* GLES: Declare u_minZmaxZ highp in both shaders
* GLES: Use highp for the texcoord and the math in CLUT8 depal
* GLES: Don't skip uniform updates while a program's locations aren't known
* GLES: Apply the texture LOD bias in the shader
* GL: Use 0..1 clip space depth where clip control is available
* GL: Set GL_UNPACK_ALIGNMENT to 1, texture uploads are tightly packed
* GL: Actually enable depth clamp on GLES
* Headless: --disable-gpu-features=dual-source-blend
* Headless: --disable-gpu-features=framebuffer-fetch
* frametests.py --adb: Name the device's SoC, and its GPU only where SurfaceFlinger reports one
* Merge pull request #22551 from hrydgard/cpu-equiv-1
* Merge remote-tracking branch 'origin/master' into cpu-equiv-1
* Merge pull request #22552 from hrydgard/ge-call-depth
* GE: Nest CALLs only two deep, like the hardware
* LoongArch IR JIT: Check Vec4Scale/Vec4Dot overlap before mapping
* x86 vertex decoder JIT: Fix the paths for constants out of RIP range
* JIT: Leave vidt.s and vidt.t to the interpreter
* MIPSAnalyst: Only writes to the CC register clobber the VFPU condition
* CPUEquivalence: Dump every core's end state, not only the interpreter's
* Docs: What the PSP does where the CPU cores disagree
* test.py: Add the CPU equivalence hardware tests to tests_next
* Docs: CPU core equivalence findings, and what needs a hardware test
* CPUEquivalence: Tell the JIT when a case sets fcr31
* x86 JIT: Leave 1 in fpcond after c.cond.s, not the compare's mask
* MIPSAnalyst: Only ctc1/cfc1 of control register 31 touch the FPU condition
* x86-64 IR JIT: Flush RCX and RDX before calls outside Windows
* ARM64 JIT: Don't write $zero in mfhi and mflo
* IR native backends: Fix ll into $zero, and sc losing its result
* IR: Don't assume a matrix that isn't a Vec4 is a transposed one
* x86 JIT: Gate more VFPU ops on prefixes like the IR, fix sc zero and vmmul crashes
* ARM64 JIT: Fix asserts and rounding differences found by CPUEquivalence
* Unit tests: Add CPUEquivalence, a differential fuzzer for the CPU cores
* Merge pull request #22550 from hrydgard/arm32-disasm-test
* ARM disassembler: Rewrite, cover everything the emitter can emit, add a test
* ARM emitter: Fix encodings found by disassembling every instruction
* Merge pull request #22548 from hrydgard/arm64-disasm-test
* ARM64 disassembler: Cover everything the emitter can emit, add a test
* ARM64 emitter: Fix encodings found by disassembling every instruction
* Merge pull request #22547 from hrydgard/jackass-postprocess
* - Fixes #13676
* Merge pull request #22545 from hrydgard/gpu-fallback-fixes
* Keep near plane clipping in the fragment depth clamp fallback
* Make the fragment depth clamp fallback match hardware depth clamp
* Bounds-check PRIM draws of any size, not only up to 2500 vertices
* Cull and clip check Bezier and spline patches like PRIM
* Android CMake build fix: Add includes libc++ no longer brings in
* Merge pull request #22544 from hrydgard/libretro-core-fixes
* libretro: let a pending boot finish before unload and reset
* libretro: update the buttons in one call
* libretro: check the state size inside MeasureAndSavePtr
* libretro: create the GL presets once
* libretro: fix OOB, UAF, data races and control flow in core glue
* libretro: include for the cheat file buffer
* Merge pull request #22543 from hrydgard/android-headless-gpu
* Headless: Add --hardware-transform and --disable-gpu-features
* frametests.py --adb: Apply compat settings on both sides by default
* Headless: Vulkan and OpenGL ES on Android, and frametests.py --adb
* Android.mk: Don't link the adrenotools hook libraries into ppsspp_common
* Merge pull request #22542 from hrydgard/faster-build
* Add includes that were only arriving transitively
* CMake: Keep glslang's HLSL frontend on Windows
* Drop unused Config.h and SmallDataConvert.h includes
* CMake: Build less, and parse less of libc++
* Update SPIRV-Cross to vulkan-sdk-1.4.363.0

Download: PPSSPP Git (2026/10/14) x64
Download: PPSSPP Git (2026/10/14) Android
Source: Here