[polly] r302132 - Adjust test case to not trigger the SCEV optimization committed in r302096

[Tobias Grosser via llvm-commits]

Author: grosser
Date: Thu May  4 03:56:54 2017
New Revision: 302132

URL: http://llvm.org/viewvc/llvm-project?rev=302132&view=rev
Log:
Adjust test case to not trigger the SCEV optimization committed in r302096

This makes sure we still test the case that a PHI-NODE cannot be analyzed by
scalar evolution and consequently must be code generated explicitly.  As
Michael's optimization triggers only on a very specific "add %iv, %step"
pattern, just changing 'add' to 'mul' adds back test coverage.

Modified:
    polly/trunk/test/Isl/CodeGen/scop_expander_insert_point.ll

Modified: polly/trunk/test/Isl/CodeGen/scop_expander_insert_point.ll
URL: http://llvm.org/viewvc/llvm-project/polly/trunk/test/Isl/CodeGen/scop_expander_insert_point.ll?rev=302132&r1=302131&r2=302132&view=diff
==============================================================================
--- polly/trunk/test/Isl/CodeGen/scop_expander_insert_point.ll (original)
+++ polly/trunk/test/Isl/CodeGen/scop_expander_insert_point.ll Thu May  4 03:56:54 2017
@@ -29,7 +29,7 @@ for.body17:
   %outvalue.141 = phi i64 [ undef, %for.body17.lr.ph ], [ %add19, %for.body17 ]
   %inptr.040 = phi i8* [ %add.ptr, %for.body17.lr.ph ], [ undef, %for.body17 ]
   %1 = load i8, i8* %inptr.040
-  %add19 = add nsw i64 0, %outvalue.141
+  %add19 = mul nsw i64 0, %outvalue.141
   br i1 false, label %for.body17, label %for.end22
 
 for.end22:                                        ; preds = %for.body17