Tom Stellard 2ef5ec6b2b Fix implementation of sqrt v2 ...

Passing values less than 0 to the llvm.sqrt() intrinsic results in
undefined behavior, so we need to check the input and return NaN if
is is less than 0.

v2:
  - Fix build failures.

llvm-svn: 241906

2015-07-10 13:37:07 +00:00

..

async

…

atomic

…

cl_khr_global_int32_base_atomics

…

cl_khr_global_int32_extended_atomics

…

cl_khr_local_int32_base_atomics

…

cl_khr_local_int32_extended_atomics

…

common

Move mix from math to common

2015-03-03 21:25:08 +00:00

geometric

geometric: Limit fast_{distance,length} functions to single precision

2015-05-09 22:31:01 +00:00

integer

…

LLVM3.6

Allow compilation depending to the LLVM version

2015-04-29 15:37:06 +00:00

math

Fix implementation of sqrt v2

2015-07-10 13:37:07 +00:00

relational

Fix bitselect for float/double types v2

2015-03-05 15:31:05 +00:00

shared

…

workitem

…

clcmacro.h

math: Add ldexp implementation

2015-05-06 20:53:32 +00:00

gen_convert.py

…

SOURCES

Fix implementation of sqrt v2

2015-07-10 13:37:07 +00:00

SOURCES_LLVM3.6

Allow compilation depending to the LLVM version

2015-04-29 15:37:06 +00:00

subnormal_config.cl

configure: Add --enable-runtime-subnormal option

2015-04-20 18:49:50 +00:00

subnormal_disable.ll

configure: Add --enable-runtime-subnormal option

2015-04-20 18:49:50 +00:00

subnormal_helper_func.ll

configure: Add --enable-runtime-subnormal option

2015-04-20 18:49:50 +00:00

subnormal_use_default.ll

configure: Add --enable-runtime-subnormal option

2015-04-20 18:49:50 +00:00