mbox

[0/4] benchtests: Add atan2f, expm1f, log1pf and tanf to bench-math

Message ID 20211029204431.61718-1-skpgkp2@gmail.com
Headers

Message

Sunil Pandey Oct. 29, 2021, 8:44 p.m. UTC 
  Motivation for this patch is to prepare for upcoming libmvec new
functions.  Float and double version of libmvec functions stays
together.

I understand double version of input files are not correct as Joseph
mentioned in this thread.

https://sourceware.org/pipermail/libc-alpha/2021-October/132294.html

Float version of input file needed for libmvec and I will rerun
conversion program once double input file get fixed for these
functions.

Float input files of these function generated from existing double
input files using following scaling formula:

f = d * (FLT_MAX/DBL_MAX)

Where d is input(double) and f is output(float).  If scaled float value
is duplicate in new input file, nextafterf() function used to find next
float value, ensuring no duplicates.


Sunil K Pandey (4):
  benchtests: Add atan2f function to bench-math
  benchtests: Add expm1f function to bench-math
  benchtests: Add log1pf function to bench-math
  benchtests: Add tanf function to bench-math

 benchtests/Makefile      |    4 +
 benchtests/atan2f-inputs | 1005 +++++++++++
 benchtests/expm1f-inputs |  798 +++++++++
 benchtests/log1pf-inputs | 1004 +++++++++++
 benchtests/tanf-inputs   | 3449 ++++++++++++++++++++++++++++++++++++++
 5 files changed, 6260 insertions(+)
 create mode 100644 benchtests/atan2f-inputs
 create mode 100644 benchtests/expm1f-inputs
 create mode 100644 benchtests/log1pf-inputs
 create mode 100644 benchtests/tanf-inputs