70a270da5f
Add initial support for half precision builtins
...
v2: fix fmax implementation
use consistent checks for __CLC_FP_SIZE
add missing TODOs
fix whitespace in definitions.h
v3: undef ZERO in modf.inc
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
reviewer: Jeroen Ketema <j.ketema@xs4all.nl>
Reviewed-by: Aaron Watry <awatry@gmail.com>
Tested-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 332677
2018-05-17 22:55:30 +00:00
633d749da7
amdgpu: Use right builtn for rsq
...
The r600 path has never actually worked sinced double is not implemented
there.
llvm-svn: 276009
2016-07-19 19:02:01 +00:00
a48e15c6cb
Split sources for amdgcn and r600
...
Most files remain in a common amdgpu directory.
Also switches barriers to to use convergent,
and use llvm.amdgcn.s.barrier.
This now requires 3.9/trunk to build amdgcn.
llvm-svn: 260777
2016-02-13 01:01:59 +00:00
Jan Vesely