diff options
Diffstat (limited to 'llvm/test/CodeGen/NVPTX/bf16x2-instructions.ll')
| -rw-r--r-- | llvm/test/CodeGen/NVPTX/bf16x2-instructions.ll | 14 |
1 files changed, 7 insertions, 7 deletions
diff --git a/llvm/test/CodeGen/NVPTX/bf16x2-instructions.ll b/llvm/test/CodeGen/NVPTX/bf16x2-instructions.ll index bd4c7775354a..3c6fb4b7517b 100644 --- a/llvm/test/CodeGen/NVPTX/bf16x2-instructions.ll +++ b/llvm/test/CodeGen/NVPTX/bf16x2-instructions.ll @@ -1,8 +1,8 @@ ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5 ; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_80 -mattr=+ptx71 | FileCheck --check-prefixes=CHECK,SM80 %s ; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_90 -mattr=+ptx78 | FileCheck --check-prefixes=CHECK,SM90 %s -; RUN: %if ptxas-11.8 %{ llc < %s -mtriple=nvptx64 -mcpu=sm_80 -mattr=+ptx71 | %ptxas-verify -arch=sm_80 %} -; RUN: %if ptxas-11.8 %{ llc < %s -mtriple=nvptx64 -mcpu=sm_90 -mattr=+ptx78 | %ptxas-verify -arch=sm_90 %} +; RUN: %if ptxas-sm_80 && ptxas-isa-7.1 %{ llc < %s -mtriple=nvptx64 -mcpu=sm_80 -mattr=+ptx71 | %ptxas-verify -arch=sm_80 %} +; RUN: %if ptxas-sm_90 && ptxas-isa-7.8 %{ llc < %s -mtriple=nvptx64 -mcpu=sm_90 -mattr=+ptx78 | %ptxas-verify -arch=sm_90 %} target datalayout = "e-m:o-i64:64-i128:128-n32:64-S128" @@ -711,11 +711,11 @@ define <2 x bfloat> @test_copysign(<2 x bfloat> %a, <2 x bfloat> %b) #0 { ; CHECK-NEXT: .reg .b32 %r<6>; ; CHECK-EMPTY: ; CHECK-NEXT: // %bb.0: -; CHECK-NEXT: ld.param.b32 %r1, [test_copysign_param_0]; -; CHECK-NEXT: ld.param.b32 %r2, [test_copysign_param_1]; -; CHECK-NEXT: and.b32 %r3, %r2, -2147450880; -; CHECK-NEXT: and.b32 %r4, %r1, 2147450879; -; CHECK-NEXT: or.b32 %r5, %r4, %r3; +; CHECK-NEXT: ld.param.b32 %r1, [test_copysign_param_1]; +; CHECK-NEXT: and.b32 %r2, %r1, -2147450880; +; CHECK-NEXT: ld.param.b32 %r3, [test_copysign_param_0]; +; CHECK-NEXT: and.b32 %r4, %r3, 2147450879; +; CHECK-NEXT: or.b32 %r5, %r4, %r2; ; CHECK-NEXT: st.param.b32 [func_retval0], %r5; ; CHECK-NEXT: ret; %r = call <2 x bfloat> @llvm.copysign.f16(<2 x bfloat> %a, <2 x bfloat> %b) |