Henrik RydgårdandClaude Opus 5.5 e33f38afa6 ARM64 IR JIT: Look up the block for a jump to a register at the jump ...

Each jr and jalr now does the dispatcher's lookup itself, so each has its
own indirect branch, which the CPU predicts much better than the one in
the dispatcher that every one of them shared. Out of downcount or with no
block yet, it goes to the dispatcher as before.

Cycles over 10 emulated seconds: -2.3% in ATV, -1.4% Tekken 6 and Ridge
Racer, -1.2% Daxter, -1.1% Killzone, and -0.3% to -0.8% in the rest.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>

2026-10-15 11:02:17 +02:00

..

ARM

JIT: Add --jit-profile, a sampling profiler for the CPU cores

2026-10-15 11:01:52 +02:00

ARM64

ARM64 IR JIT: Look up the block for a jump to a register at the jump

2026-10-15 11:02:17 +02:00

IR

IR: Compile on past conditional branches

2026-10-15 11:02:17 +02:00

JitCommon

IR: Compile on past conditional branches

2026-10-15 11:02:17 +02:00

LoongArch64

JIT: Add --jit-profile, a sampling profiler for the CPU cores

2026-10-15 11:01:52 +02:00

RiscV

JIT: Add --jit-profile, a sampling profiler for the CPU cores

2026-10-15 11:01:52 +02:00

x86

JIT: Add --jit-profile, a sampling profiler for the CPU cores

2026-10-15 11:01:52 +02:00

Interpreter.cpp

bltzal and friends test rs from before they link

2026-10-15 09:55:35 +02:00

Interpreter.h

Rename the replacement pseudo-op, drop the leftover emuhack defines

2026-10-11 10:33:30 +02:00

InterpreterDispatch.cpp

Decode the unused bits of shifts and Allegrex bit ops as the PSP does

2026-10-14 10:56:48 +02:00

InterpreterDispatch.h

Plumb an explicit MIPSState *mips through the interpreter's Int_* handlers

2026-08-12 14:02:19 +02:00

InterpreterVFPU.cpp

VFPU: vfim converts its half float like vh2f

2026-10-15 09:55:35 +02:00

InterpreterVFPU.h

Plumb an explicit MIPSState *mips through the interpreter's Int_* handlers

2026-08-12 14:02:19 +02:00

MIPS.cpp

JIT: Add --jit-profile, a sampling profiler for the CPU cores

2026-10-15 11:01:52 +02:00

MIPS.h

IR: Compute runs of VFPU dot products four at a time

2026-10-15 11:02:17 +02:00

MIPSAnalyst.cpp

MIPSAnalyst: Only writes to the CC register clobber the VFPU condition

2026-10-14 00:10:00 +02:00

MIPSAnalyst.h

Savestate: Rehash analyzed functions before writing replacements on load

2026-09-28 09:34:06 -06:00

MIPSAsm.cpp

Make instruction cache invalidation (for us, jit cache invalidation) clearer

2026-08-16 16:59:52 +02:00

MIPSAsm.h

Fix the API for MIPSAssembleOpcode to remove a global

2025-11-03 16:08:07 +01:00

MIPSCodeUtils.cpp

Keep JIT block entries and replacement hooks out of PSP memory

2026-10-11 10:33:30 +02:00

MIPSCodeUtils.h

Decode the unused bits of shifts and Allegrex bit ops as the PSP does

2026-10-14 10:56:48 +02:00

MIPSDebugInterface.cpp

Debugger: Show replacement hooks in the disassembly again

2026-10-11 10:33:30 +02:00

MIPSDebugInterface.h

Win32 debugger: stop cutting off register values in CtrlRegisterList

2026-09-12 13:46:12 -06:00

MIPSDis.cpp

Decode the unused bits of shifts and Allegrex bit ops as the PSP does

2026-10-14 10:56:48 +02:00

MIPSDis.h

Rename the replacement pseudo-op, drop the leftover emuhack defines

2026-10-11 10:33:30 +02:00

MIPSDisVFPU.cpp

VFPU: vrnds takes its seed from the vs field

2026-10-15 00:09:25 +02:00

MIPSDisVFPU.h

Improve debugger assemble/disassemble compatiblity

2026-07-07 20:30:51 -04:00

MIPSStackWalk.cpp

Keep JIT block entries and replacement hooks out of PSP memory

2026-10-11 10:33:30 +02:00

MIPSStackWalk.h

Show a MIPS stack trace on crash screen (#17211)

2023-03-31 10:08:12 +02:00

MIPSTables.cpp

MIPSTables: Include <algorithm> and <string_view> for the dispatch generator

2026-10-14 11:05:42 +02:00

MIPSTables.h

Plumb an explicit MIPSState *mips through the interpreter's Int_* handlers

2026-08-12 14:02:19 +02:00

MIPSTracer.cpp

MIPSTracer: don't index trace_info with stale block indices

2026-09-19 10:02:54 -06:00

MIPSTracer.h

MIPSTracer: don't index trace_info with stale block indices

2026-09-19 10:02:54 -06:00

MIPSVFPUUtils.cpp

IR: Compute runs of VFPU dot products four at a time

2026-10-15 11:02:17 +02:00

MIPSVFPUUtils.h

IR: Compute runs of VFPU dot products four at a time

2026-10-15 11:02:17 +02:00