4 files changed, 159 insertions, 0 deletions
diff --git a/test/CodeGen/AMDGPU/llvm.dbg.value.ll b/test/CodeGen/AMDGPU/llvm.dbg.value.ll
new file mode 100644
index 0000000000000..d001bcb4db176
--- /dev/null
+++ b/test/CodeGen/AMDGPU/llvm.dbg.value.ll
@@ -0,0 +1,37 @@
+; RUN: llc -O0 -march=amdgcn -mtriple=amdgcn-unknown-amdhsa -verify-machineinstrs < %s | FileCheck %s
+
+; CHECK-LABEL: {{^}}test_debug_value:
+; CHECK: s_load_dwordx2
+; CHECK: DEBUG_VALUE: test_debug_value:globalptr_arg <- SGPR0_SGPR1
+; CHECK: buffer_store_dword
+; CHECK: s_endpgm
+define void @test_debug_value(i32 addrspace(1)* nocapture %globalptr_arg) #0 {
+entry:
+  tail call void @llvm.dbg.value(metadata i32 addrspace(1)* %globalptr_arg, i64 0, metadata !10, metadata !13), !dbg !14
+  store i32 123, i32 addrspace(1)* %globalptr_arg, align 4
+  ret void
+}
+
+declare void @llvm.dbg.value(metadata, i64, metadata, metadata) #1
+
+attributes #0 = { nounwind "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
+attributes #1 = { nounwind readnone }
+
+!llvm.dbg.cu = !{!0}
+!llvm.module.flags = !{!11, !12}
+
+!0 = distinct !DICompileUnit(language: DW_LANG_C99, file: !1, producer: "clang version 3.8.0 (trunk 244715) (llvm/trunk 244718)", isOptimized: true, runtimeVersion: 0, emissionKind: 1, enums: !2, subprograms: !3)
+!1 = !DIFile(filename: "/tmp/test_debug_value.cl", directory: "/Users/matt/src/llvm/build_debug")
+!2 = !{}
+!3 = !{!4}
+!4 = !DISubprogram(name: "test_debug_value", scope: !1, file: !1, line: 1, type: !5, isLocal: false, isDefinition: true, scopeLine: 2, flags: DIFlagPrototyped, isOptimized: true, function: void (i32 addrspace(1)*)* @test_debug_value, variables: !9)
+!5 = !DISubroutineType(types: !6)
+!6 = !{null, !7}
+!7 = !DIDerivedType(tag: DW_TAG_pointer_type, baseType: !8, size: 64, align: 32)
+!8 = !DIBasicType(name: "int", size: 32, align: 32, encoding: DW_ATE_signed)
+!9 = !{!10}
+!10 = !DILocalVariable(tag: DW_TAG_arg_variable, name: "globalptr_arg", arg: 1, scope: !4, file: !1, line: 1, type: !7)
+!11 = !{i32 2, !"Dwarf Version", i32 4}
+!12 = !{i32 2, !"Debug Info Version", i32 3}
+!13 = !DIExpression()
+!14 = !DILocation(line: 1, column: 42, scope: !4)
diff --git a/test/CodeGen/AMDGPU/promote-alloca-bitcast-function.ll b/test/CodeGen/AMDGPU/promote-alloca-bitcast-function.ll
new file mode 100644
index 0000000000000..10739df083790
--- /dev/null
+++ b/test/CodeGen/AMDGPU/promote-alloca-bitcast-function.ll
@@ -0,0 +1,22 @@
+; RUN: not llc -march=amdgcn < %s 2>&1 | FileCheck %s
+
+; Make sure that AMDGPUPromoteAlloca doesn't crash if the called
+; function is a constantexpr cast of a function.
+
+declare void @foo(float*) #0
+declare void @foo.varargs(...) #0
+
+; CHECK: error: unsupported call to function foo in crash_call_constexpr_cast
+define void @crash_call_constexpr_cast() #0 {
+  %alloca = alloca i32
+  call void bitcast (void (float*)* @foo to void (i32*)*)(i32* %alloca) #0
+  ret void
+}
+
+define void @crash_call_constexpr_cast_varargs() #0 {
+  %alloca = alloca i32
+  call void bitcast (void (...)* @foo.varargs to void (i32*)*)(i32* %alloca) #0
+  ret void
+}
+
+attributes #0 = { nounwind }
diff --git a/test/CodeGen/AMDGPU/promote-alloca-stored-pointer-value.ll b/test/CodeGen/AMDGPU/promote-alloca-stored-pointer-value.ll
new file mode 100644
index 0000000000000..2ee98cc3d2d25
--- /dev/null
+++ b/test/CodeGen/AMDGPU/promote-alloca-stored-pointer-value.ll
@@ -0,0 +1,52 @@
+; RUN: llc -march=amdgcn < %s | FileCheck -check-prefix=GCN %s
+
+; Pointer value is stored in a candidate for LDS usage.
+
+; GCN-LABEL: {{^}}stored_lds_pointer_value:
+; GCN: buffer_store_dword v
+define void @stored_lds_pointer_value(float* addrspace(1)* %ptr) #0 {
+  %tmp = alloca float
+  store float 0.0, float *%tmp
+  store float* %tmp, float* addrspace(1)* %ptr
+  ret void
+}
+
+; GCN-LABEL: {{^}}stored_lds_pointer_value_gep:
+; GCN-DAG: s_mov_b32 s{{[0-9]+}}, SCRATCH_RSRC_DWORD0
+; GCN-DAG: s_mov_b32 s{{[0-9]+}}, SCRATCH_RSRC_DWORD1
+; GCN: buffer_store_dword v
+; GCN: buffer_store_dword v
+define void @stored_lds_pointer_value_gep(float* addrspace(1)* %ptr, i32 %idx) #0 {
+bb:
+  %tmp = alloca float, i32 16
+  store float 0.0, float* %tmp
+  %tmp2 = getelementptr inbounds float, float* %tmp, i32 %idx
+  store float* %tmp2, float* addrspace(1)* %ptr
+  ret void
+}
+
+; Pointer value is stored in a candidate for vector usage
+; GCN-LABEL: {{^}}stored_vector_pointer_value:
+; GCN-DAG: s_mov_b32 s{{[0-9]+}}, SCRATCH_RSRC_DWORD0
+; GCN-DAG: s_mov_b32 s{{[0-9]+}}, SCRATCH_RSRC_DWORD1
+; GCN: buffer_store_dword
+; GCN: buffer_store_dword
+; GCN: buffer_store_dword
+; GCN: buffer_store_dword
+define void @stored_vector_pointer_value(i32* addrspace(1)* %out, i32 %index) {
+entry:
+  %tmp0 = alloca [4 x i32]
+  %x = getelementptr [4 x i32], [4 x i32]* %tmp0, i32 0, i32 0
+  %y = getelementptr [4 x i32], [4 x i32]* %tmp0, i32 0, i32 1
+  %z = getelementptr [4 x i32], [4 x i32]* %tmp0, i32 0, i32 2
+  %w = getelementptr [4 x i32], [4 x i32]* %tmp0, i32 0, i32 3
+  store i32 0, i32* %x
+  store i32 1, i32* %y
+  store i32 2, i32* %z
+  store i32 3, i32* %w
+  %tmp1 = getelementptr [4 x i32], [4 x i32]* %tmp0, i32 0, i32 %index
+  store i32* %tmp1, i32* addrspace(1)* %out
+  ret void
+}
+
+attributes #0 = { nounwind }
diff --git a/test/CodeGen/AMDGPU/trunc-store.ll b/test/CodeGen/AMDGPU/trunc-store.ll
new file mode 100644
index 0000000000000..4ba815f26690f
--- /dev/null
+++ b/test/CodeGen/AMDGPU/trunc-store.ll
@@ -0,0 +1,48 @@
+; RUN: llc -march=amdgcn -mcpu=verde -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
+; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
+
+; FUNC-LABEL: {{^}}truncstore_arg_v16i32_to_v16i8:
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+define void @truncstore_arg_v16i32_to_v16i8(<16 x i8> addrspace(1)* %out, <16 x i32> %in) {
+  %trunc = trunc <16 x i32> %in to <16 x i8>
+  store <16 x i8> %trunc, <16 x i8> addrspace(1)* %out
+  ret void
+}
+
+; FUNC-LABEL: {{^}}truncstore_arg_v16i64_to_v16i8:
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+; SI: buffer_store_byte
+define void @truncstore_arg_v16i64_to_v16i8(<16 x i8> addrspace(1)* %out, <16 x i64> %in) {
+  %trunc = trunc <16 x i64> %in to <16 x i8>
+  store <16 x i8> %trunc, <16 x i8> addrspace(1)* %out
+  ret void
+}