Consider only references to an IV within the loop when

figuring out the base of the IV.  This produces better
code in the example.  (Addresses use (IV) instead of 
(BASE,IV) - a significant improvement on low-register
machines like x86).



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@60374 91177308-0d34-0410-b5e6-96231b3b80d8

test/CodeGen/X86/2008-12-01-loop-iv-used-outside-loop.ll

@@ -0,0 +1,28 @@
+; RUN: llvm-as < %s | llc -mtriple=i386-apple-darwin | grep -v lea
+; ModuleID = '<stdin>'
+target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128"
+target triple = "i386-apple-darwin9.5"
+
+define i8* @test(i8* %Q, i32* %L) nounwind {
+entry:
+	br label %bb1
+
+bb:		; preds = %bb1, %bb1
+	%indvar.next = add i32 %P.0.rec, 1		; <i32> [#uses=1]
+	br label %bb1
+
+bb1:		; preds = %bb, %entry
+	%P.0.rec = phi i32 [ 0, %entry ], [ %indvar.next, %bb ]		; <i32> [#uses=3]
+	%P.0 = getelementptr i8* %Q, i32 %P.0.rec		; <i8*> [#uses=2]
+	%0 = load i8* %P.0, align 1		; <i8> [#uses=1]
+	switch i8 %0, label %bb3 [
+		i8 12, label %bb
+		i8 42, label %bb
+	]
+
+bb3:		; preds = %bb1
+	%P.0.sum = add i32 %P.0.rec, 2		; <i32> [#uses=1]
+	%1 = getelementptr i8* %Q, i32 %P.0.sum		; <i8*> [#uses=1]
+	store i8 4, i8* %1, align 1
+	ret i8* %P.0
+}