1 ; RUN: opt %loadPolly -polly-codegen-ppcg -polly-acc-dump-code \
2 ; RUN: -disable-output < %s | \
3 ; RUN: FileCheck -check-prefix=CODE %s
5 ; RUN: opt %loadPolly -polly-codegen-ppcg \
7 ; RUN: FileCheck -check-prefix=IR %s
11 ; void foo(long A[], long n) {
12 ; for (long i = 0; i < n; i++)
17 ; CODE-NEXT: cudaCheckReturn(cudaMemcpy(dev_MemRef_A, MemRef_A, (n) * sizeof(i64), cudaMemcpyHostToDevice));
19 ; CODE-NEXT: dim3 k0_dimBlock(32);
20 ; CODE-NEXT: dim3 k0_dimGrid(n >= 1048546 ? 32768 : floord(n + 31, 32));
21 ; CODE-NEXT: kernel0 <<<k0_dimGrid, k0_dimBlock>>> (dev_MemRef_A, n);
22 ; CODE-NEXT: cudaCheckKernel();
25 ; CODE: cudaCheckReturn(cudaMemcpy(MemRef_A, dev_MemRef_A, (n) * sizeof(i64), cudaMemcpyDeviceToHost));
29 ; CODE-NEXT: for (int c0 = 0; c0 <= (n - 32 * b0 - 1) / 1048576; c0 += 1)
30 ; CODE-NEXT: if (n >= 32 * b0 + t0 + 1048576 * c0 + 1)
31 ; CODE-NEXT: Stmt_bb2(32 * b0 + t0 + 1048576 * c0);
33 ; IR: store i64 %n, i64* %polly_launch_0_param_1
34 ; IR-NEXT: [[REGA:%.+]] = getelementptr [4 x i8*], [4 x i8*]* %polly_launch_0_params, i64 0, i64 1
35 ; IR-NEXT: [[REGB:%.+]] = bitcast i64* %polly_launch_0_param_1 to i8*
36 ; IR-NEXT: store i8* [[REGB]], i8** [[REGA]]
38 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
40 define void @foo(i64* %A, i64 %n) {
44 bb1: ; preds = %bb6, %bb
45 %i.0 = phi i64 [ 0, %bb ], [ %tmp7, %bb6 ]
46 %tmp = icmp slt i64 %i.0, %n
47 br i1 %tmp, label %bb2, label %bb8
50 %tmp3 = getelementptr inbounds i64, i64* %A, i64 %i.0
51 %tmp4 = load i64, i64* %tmp3, align 8
52 %tmp5 = add nsw i64 %tmp4, 100
53 store i64 %tmp5, i64* %tmp3, align 8
57 %tmp7 = add nuw nsw i64 %i.0, 1