Break false dependencies before partial register updates.
Two new TargetInstrInfo hooks lets the target tell ExecutionDepsFix
about instructions with partial register updates causing false unwanted
dependencies.
The ExecutionDepsFix pass will break the false dependencies if the
updated register was written in the previoius N instructions.
The small loop added to sse-domains.ll runs twice as fast with
dependency-breaking instructions inserted.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144602 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/test/CodeGen/X86/sse-domains.ll b/test/CodeGen/X86/sse-domains.ll
index d26d322..3b66f4f 100644
--- a/test/CodeGen/X86/sse-domains.ll
+++ b/test/CodeGen/X86/sse-domains.ll
@@ -43,3 +43,44 @@
while.end:
ret void
}
+
+; CHECK: f2
+;
+; This loop contains two cvtsi2ss instructions that update the same xmm
+; register. Verify that the execution dependency fix pass breaks those
+; dependencies by inserting xorps instructions.
+;
+; If the register allocator chooses different registers for the two cvtsi2ss
+; instructions, they are still dependent on themselves.
+; CHECK: xorps [[XMM1:%xmm[0-9]+]]
+; CHECK: , [[XMM1]]
+; CHECK: cvtsi2ss %{{.*}}, [[XMM1]]
+; CHECK: xorps [[XMM2:%xmm[0-9]+]]
+; CHECK: , [[XMM2]]
+; CHECK: cvtsi2ss %{{.*}}, [[XMM2]]
+;
+define float @f2(i32 %m) nounwind uwtable readnone ssp {
+entry:
+ %tobool3 = icmp eq i32 %m, 0
+ br i1 %tobool3, label %for.end, label %for.body
+
+for.body: ; preds = %entry, %for.body
+ %m.addr.07 = phi i32 [ %dec, %for.body ], [ %m, %entry ]
+ %s1.06 = phi float [ %add, %for.body ], [ 0.000000e+00, %entry ]
+ %s2.05 = phi float [ %add2, %for.body ], [ 0.000000e+00, %entry ]
+ %n.04 = phi i32 [ %inc, %for.body ], [ 1, %entry ]
+ %conv = sitofp i32 %n.04 to float
+ %add = fadd float %s1.06, %conv
+ %conv1 = sitofp i32 %m.addr.07 to float
+ %add2 = fadd float %s2.05, %conv1
+ %inc = add nsw i32 %n.04, 1
+ %dec = add nsw i32 %m.addr.07, -1
+ %tobool = icmp eq i32 %dec, 0
+ br i1 %tobool, label %for.end, label %for.body
+
+for.end: ; preds = %for.body, %entry
+ %s1.0.lcssa = phi float [ 0.000000e+00, %entry ], [ %add, %for.body ]
+ %s2.0.lcssa = phi float [ 0.000000e+00, %entry ], [ %add2, %for.body ]
+ %sub = fsub float %s1.0.lcssa, %s2.0.lcssa
+ ret float %sub
+}