search for: dummy_sext

...bb: > br label %bb2 > > bb1: ; preds = %bb2 > ret void > > bb2: ; preds = %bb2, %bb > %k.01 = phi i64 [ 0, %bb ], [ %tmp15, %bb2 ] > %tmp = shl nsw i64 %k.01, 1 > %dummy_sext = sext i64 %tmp to i128 > %tmp3 = getelementptr inbounds [1024 x float], [1024 x float]* @x, > i64 0, i64 %tmp > %tmp4 = load float, float* %tmp3, align 16, !tbaa !2 > %tmp5 = getelementptr inbounds [1024 x float], [1024 x float]* @y, > i64 0, i64 %k.01 > %tmp6 = load...

I am testing vectorization on the following test case: float x[1024], y[1024]; void myloop1() { for (long int k = 0; k < 512; k++) { x[2*k] = x[2*k]+y[k]; } } Vectorization failed due to "unsafe dependent memory operation". I traced the LoopAccessAnalysis.cpp and found the reason is the NoWrapFlag for SCEVAddRecExpr is not set and consequently the

...; preds = %bb2 >>> ret void >>> >>> bb2: ; preds = %bb2, %bb >>> %k.01 = phi i64 [ 0, %bb ], [ %tmp15, %bb2 ] >>> %tmp = shl nsw i64 %k.01, 1 >>> %dummy_sext = sext i64 %tmp to i128 >>> %tmp3 = getelementptr inbounds [1024 x float], [1024 x float]* @x, >>> i64 0, i64 %tmp >>> %tmp4 = load float, float* %tmp3, align 16, !tbaa !2 >>> %tmp5 = getelementptr inbounds [1024 x float], [1024 x float]* @y, >>> i...