1; RUN: opt -S %loadPolly -polly-pattern-matching-based-opts=false \ 2; RUN: -polly-vectorizer=stripmine -polly-opt-isl -polly-ast -analyze \ 3; RUN: < %s | FileCheck %s 4; CHECK: // 1st level tiling - Tiles 5; CHECK-NEXT: #pragma known-parallel 6; CHECK-NEXT: for (int c0 = 0; c0 <= floord(ni - 1, 32); c0 += 1) 7; CHECK-NEXT: for (int c1 = 0; c1 <= floord(nj - 1, 32); c1 += 1) 8; CHECK-NEXT: #pragma minimal dependence distance: 1 9; CHECK-NEXT: for (int c2 = 0; c2 <= floord(nk - 1, 32); c2 += 1) { 10; CHECK-NEXT: // 1st level tiling - Points 11; CHECK-NEXT: for (int c3 = 0; c3 <= min(31, ni - 32 * c0 - 1); c3 += 1) { 12; CHECK-NEXT: for (int c4 = 0; c4 <= min(7, -8 * c1 + nj / 4 - 1); c4 += 1) 13; CHECK-NEXT: #pragma minimal dependence distance: 1 14; CHECK-NEXT: for (int c5 = 0; c5 <= min(31, nk - 32 * c2 - 1); c5 += 1) { 15; CHECK-NEXT: // SIMD 16; CHECK-NEXT: #pragma simd 17; CHECK-NEXT: for (int c6 = 0; c6 <= 3; c6 += 1) 18; CHECK-NEXT: Stmt_for_body_6(32 * c0 + c3, 32 * c1 + 4 * c4 + c6, 32 * c2 + c5); 19; CHECK-NEXT: } 20; CHECK-NEXT: if (32 * c1 + 31 >= nj) 21; CHECK-NEXT: #pragma minimal dependence distance: 1 22; CHECK-NEXT: for (int c5 = 0; c5 <= min(31, nk - 32 * c2 - 1); c5 += 1) { 23; CHECK-NEXT: // SIMD 24; CHECK-NEXT: #pragma simd 25; CHECK-NEXT: for (int c6 = 0; c6 < nj % 4; c6 += 1) 26; CHECK-NEXT: Stmt_for_body_6(32 * c0 + c3, -(nj % 4) + nj + c6, 32 * c2 + c5); 27; CHECK-NEXT: } 28; CHECK-NEXT: } 29; CHECK-NEXT: } 30 31; Function Attrs: nounwind uwtable 32define void @kernel_gemm(i32 %ni, i32 %nj, i32 %nk, double %alpha, double %beta, [1024 x double]* %C, [1024 x double]* %A, [1024 x double]* %B) #0 { 33entry: 34 %cmp.27 = icmp sgt i32 %ni, 0 35 br i1 %cmp.27, label %for.cond.1.preheader.lr.ph, label %for.end.22 36 37for.cond.1.preheader.lr.ph: ; preds = %entry 38 br label %for.cond.1.preheader 39 40for.cond.1.preheader: ; preds = %for.cond.1.preheader.lr.ph, %for.inc.20 41 %indvars.iv33 = phi i64 [ 0, %for.cond.1.preheader.lr.ph ], [ %indvars.iv.next34, %for.inc.20 ] 42 %cmp2.25 = icmp sgt i32 %nj, 0 43 br i1 %cmp2.25, label %for.cond.4.preheader.lr.ph, label %for.inc.20 44 45for.cond.4.preheader.lr.ph: ; preds = %for.cond.1.preheader 46 br label %for.cond.4.preheader 47 48for.cond.4.preheader: ; preds = %for.cond.4.preheader.lr.ph, %for.inc.17 49 %indvars.iv29 = phi i64 [ 0, %for.cond.4.preheader.lr.ph ], [ %indvars.iv.next30, %for.inc.17 ] 50 %cmp5.23 = icmp sgt i32 %nk, 0 51 br i1 %cmp5.23, label %for.body.6.lr.ph, label %for.inc.17 52 53for.body.6.lr.ph: ; preds = %for.cond.4.preheader 54 br label %for.body.6 55 56for.body.6: ; preds = %for.body.6.lr.ph, %for.body.6 57 %indvars.iv = phi i64 [ 0, %for.body.6.lr.ph ], [ %indvars.iv.next, %for.body.6 ] 58 %arrayidx8 = getelementptr inbounds [1024 x double], [1024 x double]* %A, i64 %indvars.iv33, i64 %indvars.iv 59 %0 = load double, double* %arrayidx8, align 8 60 %arrayidx12 = getelementptr inbounds [1024 x double], [1024 x double]* %B, i64 %indvars.iv, i64 %indvars.iv29 61 %1 = load double, double* %arrayidx12, align 8 62 %mul = fmul double %0, %1 63 %arrayidx16 = getelementptr inbounds [1024 x double], [1024 x double]* %C, i64 %indvars.iv33, i64 %indvars.iv29 64 %2 = load double, double* %arrayidx16, align 8 65 %add = fadd double %2, %mul 66 store double %add, double* %arrayidx16, align 8 67 %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1 68 %lftr.wideiv = trunc i64 %indvars.iv.next to i32 69 %exitcond = icmp ne i32 %lftr.wideiv, %nk 70 br i1 %exitcond, label %for.body.6, label %for.cond.4.for.inc.17_crit_edge 71 72for.cond.4.for.inc.17_crit_edge: ; preds = %for.body.6 73 br label %for.inc.17 74 75for.inc.17: ; preds = %for.cond.4.for.inc.17_crit_edge, %for.cond.4.preheader 76 %indvars.iv.next30 = add nuw nsw i64 %indvars.iv29, 1 77 %lftr.wideiv31 = trunc i64 %indvars.iv.next30 to i32 78 %exitcond32 = icmp ne i32 %lftr.wideiv31, %nj 79 br i1 %exitcond32, label %for.cond.4.preheader, label %for.cond.1.for.inc.20_crit_edge 80 81for.cond.1.for.inc.20_crit_edge: ; preds = %for.inc.17 82 br label %for.inc.20 83 84for.inc.20: ; preds = %for.cond.1.for.inc.20_crit_edge, %for.cond.1.preheader 85 %indvars.iv.next34 = add nuw nsw i64 %indvars.iv33, 1 86 %lftr.wideiv35 = trunc i64 %indvars.iv.next34 to i32 87 %exitcond36 = icmp ne i32 %lftr.wideiv35, %ni 88 br i1 %exitcond36, label %for.cond.1.preheader, label %for.cond.for.end.22_crit_edge 89 90for.cond.for.end.22_crit_edge: ; preds = %for.inc.20 91 br label %for.end.22 92 93for.end.22: ; preds = %for.cond.for.end.22_crit_edge, %entry 94 ret void 95} 96