[clang] [CIR] Maximum across vector (IEEE754) (PR #199779)

[Andrzej Warzyński via cfe-commits]

================
@@ -24,6 +24,40 @@
 
 #include <arm_neon.h>
 
+//===------------------------------------------------------===//
+// 2.1.1.13.5  Maximum across vector (IEEE754)
+//===------------------------------------------------------===//
+
+//ALL-LABEL: @test_vmaxnmv_f32(
+float32_t test_vmaxnmv_f32(float32x2_t a) {
+    //CIR: cir.call @vmaxnmv_f32
+    // 
+    // LLVM-SAME: <2 x float> {{.*}} [[A:%.*]])
+    // LLVM:  [[VMAXNMV_F32_I:%.*]] = call float 
@llvm.aarch64.neon.fmaxnmv.f32.v2f32(<2 x float> [[A]])
+    // LLVM: ret float [[VMAXNMV_F32_I]]
+  return vmaxnmv_f32(a);
----------------
banach-space wrote:

[nit]
```suggestion
  // CIR: cir.call @vmaxnmv_f32
  
  // LLVM-SAME: <2 x float> {{.*}} [[A:%.*]])
  // LLVM:  [[VMAXNMV_F32_I:%.*]] = call float 
@llvm.aarch64.neon.fmaxnmv.f32.v2f32(<2 x float> [[A]])
  // LLVM: ret float [[VMAXNMV_F32_I]]
  return vmaxnmv_f32(a);
```

https://github.com/llvm/llvm-project/pull/199779
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits