From 749691ca2ad79eda60a66527aac0aab0d6ed0a0f Mon Sep 17 00:00:00 2001
From: Nick Lewycky <nick@wasmer.io>
Date: Fri, 4 Oct 2019 11:50:11 -0700
Subject: [PATCH] Add a comment explaining why we don't use the intrinsics for
 these.

---
 lib/llvm-backend/src/code.rs | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)

diff --git a/lib/llvm-backend/src/code.rs b/lib/llvm-backend/src/code.rs
index a1c69c08262..13dcd2bd9b3 100644
--- a/lib/llvm-backend/src/code.rs
+++ b/lib/llvm-backend/src/code.rs
@@ -2692,6 +2692,9 @@ impl FunctionCodeGenerator<CodegenError> for LLVMFunctionCodeGenerator {
                 state.push1(bits);
             }
             Operator::F32Min => {
+                // This implements the same logic as LLVM's @llvm.minimum
+                // intrinsic would, but x86 lowering of that intrinsics
+                // encounters a fatal error in LLVM 8 and LLVM 9.
                 let (v1, v2) = state.pop2()?;
                 let v1 = canonicalize_nans(builder, intrinsics, v1);
                 let v2 = canonicalize_nans(builder, intrinsics, v2);
@@ -2735,6 +2738,9 @@ impl FunctionCodeGenerator<CodegenError> for LLVMFunctionCodeGenerator {
                 state.push1(res);
             }
             Operator::F64Min => {
+                // This implements the same logic as LLVM's @llvm.minimum
+                // intrinsic would, but x86 lowering of that intrinsics
+                // encounters a fatal error in LLVM 8 and LLVM 9.
                 let (v1, v2) = state.pop2()?;
                 let v1 = canonicalize_nans(builder, intrinsics, v1);
                 let v2 = canonicalize_nans(builder, intrinsics, v2);
@@ -2778,6 +2784,9 @@ impl FunctionCodeGenerator<CodegenError> for LLVMFunctionCodeGenerator {
                 state.push1(res);
             }
             Operator::F32x4Min => {
+                // This implements the same logic as LLVM's @llvm.minimum
+                // intrinsic would, but x86 lowering of that intrinsics
+                // encounters a fatal error in LLVM 8 and LLVM 9.
                 let (v1, v2) = state.pop2()?;
                 let v1 = builder.build_bitcast(v1, intrinsics.f32x4_ty, "");
                 let v2 = builder.build_bitcast(v2, intrinsics.f32x4_ty, "");
@@ -2830,6 +2839,9 @@ impl FunctionCodeGenerator<CodegenError> for LLVMFunctionCodeGenerator {
                 state.push1(res);
             }
             Operator::F64x2Min => {
+                // This implements the same logic as LLVM's @llvm.minimum
+                // intrinsic would, but x86 lowering of that intrinsics
+                // encounters a fatal error in LLVM 8 and LLVM 9.
                 let (v1, v2) = state.pop2()?;
                 let v1 = builder.build_bitcast(v1, intrinsics.f64x2_ty, "");
                 let v2 = builder.build_bitcast(v2, intrinsics.f64x2_ty, "");
@@ -2882,6 +2894,9 @@ impl FunctionCodeGenerator<CodegenError> for LLVMFunctionCodeGenerator {
                 state.push1(res);
             }
             Operator::F32Max => {
+                // This implements the same logic as LLVM's @llvm.maximum
+                // intrinsic would, but x86 lowering of that intrinsics
+                // encounters a fatal error in LLVM 8 and LLVM 9.
                 let (v1, v2) = state.pop2()?;
                 let v1 = canonicalize_nans(builder, intrinsics, v1);
                 let v2 = canonicalize_nans(builder, intrinsics, v2);
@@ -2924,6 +2939,9 @@ impl FunctionCodeGenerator<CodegenError> for LLVMFunctionCodeGenerator {
                 state.push1(res);
             }
             Operator::F64Max => {
+                // This implements the same logic as LLVM's @llvm.maximum
+                // intrinsic would, but x86 lowering of that intrinsics
+                // encounters a fatal error in LLVM 8 and LLVM 9.
                 let (v1, v2) = state.pop2()?;
                 let v1 = canonicalize_nans(builder, intrinsics, v1);
                 let v2 = canonicalize_nans(builder, intrinsics, v2);
@@ -2966,6 +2984,9 @@ impl FunctionCodeGenerator<CodegenError> for LLVMFunctionCodeGenerator {
                 state.push1(res);
             }
             Operator::F32x4Max => {
+                // This implements the same logic as LLVM's @llvm.maximum
+                // intrinsic would, but x86 lowering of that intrinsics
+                // encounters a fatal error in LLVM 8 and LLVM 9.
                 let (v1, v2) = state.pop2()?;
                 let v1 = builder.build_bitcast(v1, intrinsics.f32x4_ty, "");
                 let v2 = builder.build_bitcast(v2, intrinsics.f32x4_ty, "");
@@ -3018,6 +3039,9 @@ impl FunctionCodeGenerator<CodegenError> for LLVMFunctionCodeGenerator {
                 state.push1(res);
             }
             Operator::F64x2Max => {
+                // This implements the same logic as LLVM's @llvm.maximum
+                // intrinsic would, but x86 lowering of that intrinsics
+                // encounters a fatal error in LLVM 8 and LLVM 9.
                 let (v1, v2) = state.pop2()?;
                 let v1 = builder.build_bitcast(v1, intrinsics.f64x2_ty, "");
                 let v2 = builder.build_bitcast(v2, intrinsics.f64x2_ty, "");