[libc-commits] [PATCH] D134002: [libc][math] Improve tanhf performance.

[Tue Ly via Phabricator via libc-commits]

lntue updated this revision to Diff 460775.
lntue added a comment.

Change to use 32 sub-intervals for middle part to reduce the degree of
polynomial approximation for lower part to 5 from 7.  Also update `sinhf`,
`coshf`, `exp2f` to use the same range reduction.  This slightly reduce their
latencies and reciprocal throughputs.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D134002/new/

https://reviews.llvm.org/D134002

Files:
  libc/docs/math.rst
  libc/src/math/generic/exp2f.cpp
  libc/src/math/generic/explogxf.cpp
  libc/src/math/generic/explogxf.h
  libc/src/math/generic/tanhf.cpp
  libc/test/src/math/explogxf_test.cpp

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D134002.460775.patch
Type: text/x-patch
Size: 23486 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/libc-commits/attachments/20220916/929abe4d/attachment-0001.bin>