Handle small memory compares against 0

The previous commit missed the TYP_BOOL case that the original code handled. But then the original code failed to do this thinking that "test" will be used. But that means we end up with a "movzx" as well:
       movzx    rax, byte  ptr [rsi+24]
       test     eax, eax
instead of just
       cmp      byte  ptr [rsi+24], 0

The Intel manual actually recomends against using the "cmp mem, imm" form but not if other instructions need to be added. This may warrant further investigation though.

FX diff shows a 7579 bytes improvement without any regressions.

Commit migrated from https://github.com/dotnet/coreclr/commit/9bc2e775a85667fa85a1254fa4138fb98f417543

diff --git a/src/coreclr/src/jit/lower.cpp b/src/coreclr/src/jit/lower.cpp
index 80d3bad..8d45d7f 100644 (file)
--- a/src/coreclr/src/jit/lower.cpp
+++ b/src/coreclr/src/jit/lower.cpp
@@ -1997,7 +1997,7 @@ void Lowering::LowerCompare(GenTree* cmp)
         GenTreeIntCon* op2      = cmp->gtGetOp2()->AsIntCon();
         ssize_t        op2Value = op2->IconValue();
 
-        if (op1->isMemoryOp() && varTypeIsSmallInt(op1Type))
+        if (op1->isMemoryOp() && varTypeIsSmall(op1Type))
         {
             //
             // If op1's type is small then try to narrow op2 so it has the same type as op1.
@@ -2006,7 +2006,7 @@ void Lowering::LowerCompare(GenTree* cmp)
             // (e.g "cmp ubyte, 200") we also get a smaller instruction encoding.
             //
 
-            if ((op1Type == TYP_UBYTE) && FitsIn<UINT8>(op2Value))
+            if (((op1Type == TYP_BOOL) || (op1Type == TYP_UBYTE)) && FitsIn<UINT8>(op2Value))
             {
                 cmp->gtFlags |= GTF_UNSIGNED;
                 op2->gtType = op1Type;