[PATCH v4 0/9] fpu: Add conversion routines for OCP FP{4,8}

[Richard Henderson]

Hiya,

This is the form of what I was talking about wrt spliting up the
patches and introduced FloatFmtExpMaxKind, as I suggested during
review of v3.  Since I've got no test case for these formats, I
haven't been able to do more.  Please test.

r~

Max Chou (8):
  fpu/softfloat: Refactor IEEE format NaN classification to share code
  fpu/softfloat: Refactor floatx80 format NaN classification to share
    code
  target/i386: Fix pseudo-NaN handling in FPATAN/FYL2XP1/FYL2X helpers
  fpu: Add overflow_raises_invalid to FloatFmt
  fpu: Add saturate parameter to parts_uncanon
  fpu: Add conversion routines for OCP FP8 E5M2
  fpu: Add conversion routines for OCP FP8 E4M3
  fpu: Add conversion routines for OCP FP4 E2M1

Richard Henderson (1):
  fpu: Introduce FloatFmtExpMaxKind

 include/fpu/softfloat-types.h  |   7 +
 include/fpu/softfloat.h        |  18 ++-
 fpu/softfloat.c                | 211 ++++++++++++++++++++++++----
 target/i386/tcg/fpu_helper.c   |  30 ++--
 fpu/softfloat-parts.c.inc      | 152 ++++++++++++++------
 fpu/softfloat-specialize.c.inc | 245 +++++++++++++--------------------
 6 files changed, 429 insertions(+), 234 deletions(-)

-- 
2.43.0