Files
FEX-Emu--FEX/FEXCore
Ryan Houdek 76bd81af15 FEXCore: Defer setting x87 softflow rounding mode until use
Currently FEX will always jump out of the JIT any time FCW was getting
written to, ensuring that the softfloat state is setup to rounding at
the time of FCW getting written.
This has the unintended side-effect that even in "x87 reduced precision"
mode we were jumping out of the JIT.
This hit a real world use case of an installer reloading FCW after every
x87 operation and generating a block with 2297 instructions.

Instead when jumping out of the JIT for handling x87 operations, load
FCW and pass it as the first argument of the handler. Setting the
softfloat state at that point.

This helps the installer's hottest block by cutting it down to 1477
instructions. 64.3% of the original size. The code block is still
burning 90% of the CPU time of the installer but the performance is
significantly better while it is doing its decompression.

In order to optimize this installer's block of code more then we will
likely need to optimize out x87 stack usage.
2023-09-12 05:21:06 -07:00
..
2023-08-17 16:32:16 -04:00
2023-09-11 16:41:30 -07:00
2023-08-17 16:32:16 -04:00
2023-08-17 16:32:16 -04:00

FEXCore - Fast x86 Core emulation library

This is the core emulation library that is used for the FEX emulator project. This project aims to provide a fast and functional x86-64 emulation library that can meet and surpass other x86-64 emulation libraries.

Goals

  • Be as fast as possible, beating and exceeding current options for x86-64 emulation
    • 25% - 50% lower performance than native code would be desired target
    • Use an IR to efficiently translate x86-64 to our host architecture
    • Support a tiered recompiler to allow for fast runtime performance
    • Support offline compilation and offline tooling for inspection and performance analysis
    • Support threaded emulation. Including emulating x86-64's strong memory model on weak memory model architectures
  • Support a significant portion of the x86-64 instruction space.
    • Including MMX, SSE, SSE2, SSE3, SSSE3, and SSE4*
  • Support fallback routines for uncommonly used x86-64 instructions
    • Including x87 and 3DNow!
  • Only support userspace emulation.
    • All x86-64 instructions run as if they are under CPL-3(userland) security layer
  • Minimal Linux Syscall emulation for testing purposes
  • Portable library implementation in order to support easy integration in to applications

Target Host Architecture

The target host architecture for this library is AArch64. Specifically the ARMv8.1 version or newer. The CPU IR is designed with AArch64 in mind but should allow for other architectures as well. x86-64 host support is available for ease of development, but is not a priority.

Not desired

  • Kernel space emulation
  • CPL0-2 emulation
  • Real Mode, Protected Mode, Virtual-8086 Mode, System Management Mode
  • IRQs
  • SVM
  • "Cycle Accurate" emulation