[llvm] r330950 - [x86] add tests to show potential opt-out of ftrunc optimization; NFC
Sanjay Patel via llvm-commits
llvm-commits at lists.llvm.org
Thu Apr 26 08:36:15 PDT 2018
Author: spatel
Date: Thu Apr 26 08:36:15 2018
New Revision: 330950
URL: http://llvm.org/viewvc/llvm-project?rev=330950&view=rev
Log:
[x86] add tests to show potential opt-out of ftrunc optimization; NFC
This is another preliminary step for disabling this transform as
discussed in the post-commit thread for:
rL330437
I'm using one of the names suggested there for the attribute, but
we can fix that up as needed once the clang side of this is sorted
out.
Modified:
llvm/trunk/test/CodeGen/X86/ftrunc.ll
Modified: llvm/trunk/test/CodeGen/X86/ftrunc.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/ftrunc.ll?rev=330950&r1=330949&r2=330950&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/ftrunc.ll (original)
+++ llvm/trunk/test/CodeGen/X86/ftrunc.ll Thu Apr 26 08:36:15 2018
@@ -356,3 +356,54 @@ define <4 x double> @trunc_signed_v4f64(
ret <4 x double> %r
}
+; FIXME: The attribute name is subject to change, but the fold may be
+; guarded to allow existing code to continue working based on its
+; assumptions of float->int overflow.
+
+define float @trunc_unsigned_f32_disable_via_attr(float %x) #1 {
+; SSE2-LABEL: trunc_unsigned_f32_disable_via_attr:
+; SSE2: # %bb.0:
+; SSE2-NEXT: cvttss2si %xmm0, %rax
+; SSE2-NEXT: movl %eax, %eax
+; SSE2-NEXT: xorps %xmm0, %xmm0
+; SSE2-NEXT: cvtsi2ssq %rax, %xmm0
+; SSE2-NEXT: retq
+;
+; SSE41-LABEL: trunc_unsigned_f32_disable_via_attr:
+; SSE41: # %bb.0:
+; SSE41-NEXT: roundss $11, %xmm0, %xmm0
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_unsigned_f32_disable_via_attr:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vroundss $11, %xmm0, %xmm0, %xmm0
+; AVX1-NEXT: retq
+ %i = fptoui float %x to i32
+ %r = uitofp i32 %i to float
+ ret float %r
+}
+
+define double @trunc_signed_f64_disable_via_attr(double %x) #1 {
+; SSE2-LABEL: trunc_signed_f64_disable_via_attr:
+; SSE2: # %bb.0:
+; SSE2-NEXT: cvttsd2si %xmm0, %rax
+; SSE2-NEXT: xorps %xmm0, %xmm0
+; SSE2-NEXT: cvtsi2sdq %rax, %xmm0
+; SSE2-NEXT: retq
+;
+; SSE41-LABEL: trunc_signed_f64_disable_via_attr:
+; SSE41: # %bb.0:
+; SSE41-NEXT: roundsd $11, %xmm0, %xmm0
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_signed_f64_disable_via_attr:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vroundsd $11, %xmm0, %xmm0, %xmm0
+; AVX1-NEXT: retq
+ %i = fptosi double %x to i64
+ %r = sitofp i64 %i to double
+ ret double %r
+}
+
+attributes #1 = { nounwind "fp-cast-overflow-workaround"="true" }
+
More information about the llvm-commits
mailing list