mirror of
https://github.com/NVIDIA/cutlass.git
synced 2026-04-19 22:38:56 +00:00
Merge pull request #3022 from tridao/nvvm_fmin
[Cute-DSL] Add cute.arch.fmin by calling nvvm
This commit is contained in:
@@ -956,6 +956,20 @@ def fmax(
|
||||
)
|
||||
|
||||
|
||||
@dsl_user_op
|
||||
def fmin(
|
||||
a: Union[float, Float32], b: Union[float, Float32], *, loc=None, ip=None
|
||||
) -> Float32:
|
||||
return Float32(
|
||||
nvvm.fmin(
|
||||
Float32(a).ir_value(loc=loc, ip=ip),
|
||||
Float32(b).ir_value(loc=loc, ip=ip),
|
||||
loc=loc,
|
||||
ip=ip,
|
||||
)
|
||||
)
|
||||
|
||||
|
||||
@dsl_user_op
|
||||
def rcp_approx(a: Union[float, Float32], *, loc=None, ip=None):
|
||||
return Float32(
|
||||
|
||||
Reference in New Issue
Block a user