diff options
author | sheaf <sam.derbyshire@gmail.com> | 2023-04-08 13:42:58 +0200 |
---|---|---|
committer | Marge Bot <ben+marge-bot@smart-cactus.org> | 2023-05-11 11:55:22 -0400 |
commit | 87eebf98cb485f7c9175330051736e147ade9848 (patch) | |
tree | ffa226b3fefa8b0a03e1798fa4f55affbddf654b /libraries | |
parent | 630b1fea1e41a1e00860a30742b6ab8ade8a0de0 (diff) | |
download | haskell-87eebf98cb485f7c9175330051736e147ade9848.tar.gz |
Add fused multiply-add instructions
This patch adds eight new primops that fuse a multiplication and an
addition or subtraction:
- `{fmadd,fmsub,fnmadd,fnmsub}{Float,Double}#`
fmadd x y z is x * y + z, computed with a single rounding step.
This patch implements code generation for these primops in the following
backends:
- X86, AArch64 and PowerPC NCG,
- LLVM
- C
WASM uses the C implementation. The primops are unsupported in the
JavaScript backend.
The following constant folding rules are also provided:
- compute a * b + c when a, b, c are all literals,
- x * y + 0 ==> x * y,
- ±1 * y + z ==> z ± y and x * ±1 + z ==> z ± x.
NB: the constant folding rules incorrectly handle signed zero.
This is a known limitation with GHC's floating-point constant folding
rules (#21227), which we hope to resolve in the future.
Diffstat (limited to 'libraries')
-rw-r--r-- | libraries/ghc-prim/changelog.md | 18 |
1 files changed, 18 insertions, 0 deletions
diff --git a/libraries/ghc-prim/changelog.md b/libraries/ghc-prim/changelog.md index 1cf411c029..39e5face03 100644 --- a/libraries/ghc-prim/changelog.md +++ b/libraries/ghc-prim/changelog.md @@ -23,6 +23,24 @@ - `copyAddrToAddrNonOverlapping#` - `setAddrRange#` +- New primops for fused multiply-add operations. These primops combine a + multiplication and an addition, compiling to a single instruction when + the `-mfma` flag is enabled and the architecture supports it. + + The new primops are `fmaddFloat#, fmsubFloat#, fnmaddFloat#, fnmsubFloat# :: Float# -> Float# -> Float# -> Float#` + and `fmaddDouble#, fmsubDouble#, fnmaddDouble#, fnmsubDouble# :: Double# -> Double# -> Double# -> Double#`. + + These implement the following operations, while performing one single + rounding at the end, leading to a more accurate result: + + - `fmaddFloat# x y z`, `fmaddDouble# x y z` compute `x * y + z`. + - `fmsubFloat# x y z`, `fmsubDouble# x y z` compute `x * y - z`. + - `fnmaddFloat# x y z`, `fnmaddDouble# x y z` compute `- x * y + z`. + - `fnmsubFloat# x y z`, `fnmsubDouble# x y z` compute `- x * y - z`. + + Warning: on unsupported architectures, the software emulation provided by + the fallback to the C standard library is not guaranteed to be IEEE-compliant. + ## 0.10.0 - Shipped with GHC 9.6.1 |