[PATCH] D29162: AMDGPU: Add a test checking alignments of emitted globals/allocas

[Matt Arsenault via Phabricator via cfe-commits]

arsenm closed this revision.
arsenm added a comment.

r294278


https://reviews.llvm.org/D29162



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D29162: AMDGPU: Add a test checking alignments of emitted globals/allocas

[Yaxun Liu via Phabricator via cfe-commits]

yaxunl accepted this revision.
yaxunl added a comment.

LGTM. Thanks!


https://reviews.llvm.org/D29162



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D29162: AMDGPU: Add a test checking alignments of emitted globals/allocas

[Konstantin Zhuravlyov via Phabricator via cfe-commits]

kzhuravl accepted this revision.
kzhuravl added a comment.
This revision is now accepted and ready to land.

LGTM.


https://reviews.llvm.org/D29162



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D29162: AMDGPU: Add a test checking alignments of emitted globals/allocas

[Matt Arsenault via Phabricator via cfe-commits]

arsenm added a comment.

ping


https://reviews.llvm.org/D29162



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D29162: AMDGPU: Add a test checking alignments of emitted globals/allocas

[Matt Arsenault via Phabricator via cfe-commits]

arsenm created this revision.
Herald added a reviewer: tstellarAMD.
Herald added subscribers: tony-tye, nhaehnle, wdng, kzhuravl.

Make sure the spec required type alignments are used in preparation
for a possible change which may break this.


https://reviews.llvm.org/D29162

Files:
  test/CodeGenOpenCL/amdgpu-alignment.cl

Index: test/CodeGenOpenCL/amdgpu-alignment.cl
===
--- /dev/null
+++ test/CodeGenOpenCL/amdgpu-alignment.cl
@@ -0,0 +1,524 @@
+// REQUIRES: amdgpu-registered-target
+// RUN: %clang_cc1 -triple amdgcn-unknown-unknown -S -disable-llvm-passes -emit-llvm -o - %s | FileCheck %s
+
+#pragma OPENCL EXTENSION cl_khr_fp64 : enable
+#pragma OPENCL EXTENSION cl_khr_fp16 : enable
+
+typedef char __attribute__((ext_vector_type(2))) char2;
+typedef char __attribute__((ext_vector_type(3))) char3;
+typedef char __attribute__((ext_vector_type(4))) char4;
+typedef char __attribute__((ext_vector_type(8))) char8;
+typedef char __attribute__((ext_vector_type(16))) char16;
+
+typedef short __attribute__((ext_vector_type(2))) short2;
+typedef short __attribute__((ext_vector_type(3))) short3;
+typedef short __attribute__((ext_vector_type(4))) short4;
+typedef short __attribute__((ext_vector_type(8))) short8;
+typedef short __attribute__((ext_vector_type(16))) short16;
+
+typedef int __attribute__((ext_vector_type(2))) int2;
+typedef int __attribute__((ext_vector_type(3))) int3;
+typedef int __attribute__((ext_vector_type(4))) int4;
+typedef int __attribute__((ext_vector_type(8))) int8;
+typedef int __attribute__((ext_vector_type(16))) int16;
+
+typedef long __attribute__((ext_vector_type(2))) long2;
+typedef long __attribute__((ext_vector_type(3))) long3;
+typedef long __attribute__((ext_vector_type(4))) long4;
+typedef long __attribute__((ext_vector_type(8))) long8;
+typedef long __attribute__((ext_vector_type(16))) long16;
+
+typedef half __attribute__((ext_vector_type(2))) half2;
+typedef half __attribute__((ext_vector_type(3))) half3;
+typedef half __attribute__((ext_vector_type(4))) half4;
+typedef half __attribute__((ext_vector_type(8))) half8;
+typedef half __attribute__((ext_vector_type(16))) half16;
+
+typedef float __attribute__((ext_vector_type(2))) float2;
+typedef float __attribute__((ext_vector_type(3))) float3;
+typedef float __attribute__((ext_vector_type(4))) float4;
+typedef float __attribute__((ext_vector_type(8))) float8;
+typedef float __attribute__((ext_vector_type(16))) float16;
+
+typedef double __attribute__((ext_vector_type(2))) double2;
+typedef double __attribute__((ext_vector_type(3))) double3;
+typedef double __attribute__((ext_vector_type(4))) double4;
+typedef double __attribute__((ext_vector_type(8))) double8;
+typedef double __attribute__((ext_vector_type(16))) double16;
+
+// CHECK: @local_memory_alignment_global.lds_i8 = internal addrspace(3) global [4 x i8] undef, align 1
+// CHECK: @local_memory_alignment_global.lds_v2i8 = internal addrspace(3) global [4 x <2 x i8>] undef, align 2
+// CHECK: @local_memory_alignment_global.lds_v3i8 = internal addrspace(3) global [4 x <3 x i8>] undef, align 4
+// CHECK: @local_memory_alignment_global.lds_v4i8 = internal addrspace(3) global [4 x <4 x i8>] undef, align 4
+// CHECK: @local_memory_alignment_global.lds_v8i8 = internal addrspace(3) global [4 x <8 x i8>] undef, align 8
+// CHECK: @local_memory_alignment_global.lds_v16i8 = internal addrspace(3) global [4 x <16 x i8>] undef, align 16
+// CHECK: @local_memory_alignment_global.lds_i16 = internal addrspace(3) global [4 x i16] undef, align 2
+// CHECK: @local_memory_alignment_global.lds_v2i16 = internal addrspace(3) global [4 x <2 x i16>] undef, align 4
+// CHECK: @local_memory_alignment_global.lds_v3i16 = internal addrspace(3) global [4 x <3 x i16>] undef, align 8
+// CHECK: @local_memory_alignment_global.lds_v4i16 = internal addrspace(3) global [4 x <4 x i16>] undef, align 8
+// CHECK: @local_memory_alignment_global.lds_v8i16 = internal addrspace(3) global [4 x <8 x i16>] undef, align 16
+// CHECK: @local_memory_alignment_global.lds_v16i16 = internal addrspace(3) global [4 x <16 x i16>] undef, align 32
+// CHECK: @local_memory_alignment_global.lds_i32 = internal addrspace(3) global [4 x i32] undef, align 4
+// CHECK: @local_memory_alignment_global.lds_v2i32 = internal addrspace(3) global [4 x <2 x i32>] undef, align 8
+// CHECK: @local_memory_alignment_global.lds_v3i32 = internal addrspace(3) global [4 x <3 x i32>] undef, align 16
+// CHECK: @local_memory_alignment_global.lds_v4i32 = internal addrspace(3) global [4 x <4 x i32>] undef, align 16
+// CHECK: @local_memory_alignment_global.lds_v8i32 = internal addrspace(3) global [4 x <8 x i32>] undef, align 32
+// CHECK: @local_memory_alignment_global.lds_v16i32 = internal addrspace(3) global [4 x <16 x i32>] undef, align 64
+// CHECK: @local_memory_alignment_global.lds_i64 = internal addrspace(3) global [4 x i64] undef, align 8
+// CHECK: