Add some xfailed R600 tests for 64-bit private accesses.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@200620 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/test/CodeGen/R600/indirect-private-64.ll b/test/CodeGen/R600/indirect-private-64.ll
new file mode 100644 (file)
index 0000000..a6f1e93
--- /dev/null
+++ b/test/CodeGen/R600/indirect-private-64.ll
@@ -0,0 +1,31 @@
+; REQUIRES: asserts
+; XFAIL: *
+; RUN: llc -march=r600 -mcpu=SI -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s
+
+declare void @llvm.AMDGPU.barrier.local() noduplicate nounwind
+
+; SI-LABEL: @indirect_access_f64_alloca:
+; SI: BUFFER_STORE_DWORD
+define void @f64_alloca(double addrspace(1)* noalias %out, double addrspace(1)* noalias %in, i32 %b) nounwind {
+  %val = load double addrspace(1)* %in, align 8
+  %array = alloca double, i32 16, align 8
+  %ptr = getelementptr double* %array, i32 %b
+  store double %val, double* %ptr, align 8
+  call void @llvm.AMDGPU.barrier.local() noduplicate nounwind
+  %result = load double* %ptr, align 8
+  store double %result, double addrspace(1)* %out, align 8
+  ret void
+}
+
+; SI-LABEL: @indirect_access_v2f64_alloca:
+; SI: BUFFER_STORE_DWORDX4
+define void @v2f64_alloca(<2 x double> addrspace(1)* noalias %out, <2 x double> addrspace(1)* noalias %in, i32 %b) nounwind {
+  %val = load <2 x double> addrspace(1)* %in, align 16
+  %array = alloca <2 x double>, i32 16, align 16
+  %ptr = getelementptr <2 x double>* %array, i32 %b
+  store <2 x double> %val, <2 x double>* %ptr, align 16
+  call void @llvm.AMDGPU.barrier.local() noduplicate nounwind
+  %result = load <2 x double>* %ptr, align 16
+  store <2 x double> %result, <2 x double> addrspace(1)* %out, align 16
+  ret void
+}

diff --git a/test/CodeGen/R600/insert_vector_elt_f64.ll b/test/CodeGen/R600/insert_vector_elt_f64.ll
new file mode 100644 (file)
index 0000000..e334be1
--- /dev/null
+++ b/test/CodeGen/R600/insert_vector_elt_f64.ll
@@ -0,0 +1,36 @@
+; REQUIRES: asserts
+; XFAIL: *
+; RUN: llc -march=r600 -mcpu=SI < %s | FileCheck -check-prefix=SI %s
+
+
+; SI-LABEL: @dynamic_insertelement_v2f64:
+; SI: BUFFER_STORE_DWORDX4
+define void @dynamic_insertelement_v2f64(<2 x double> addrspace(1)* %out, <2 x double> %a, i32 %b) nounwind {
+  %vecins = insertelement <2 x double> %a, double 8.0, i32 %b
+  store <2 x double> %vecins, <2 x double> addrspace(1)* %out, align 16
+  ret void
+}
+
+; SI-LABEL: @dynamic_insertelement_v2f64:
+; SI: BUFFER_STORE_DWORDX4
+define void @dynamic_insertelement_v2i64(<2 x i64> addrspace(1)* %out, <2 x i64> %a, i32 %b) nounwind {
+  %vecins = insertelement <2 x i64> %a, i64 5, i32 %b
+  store <2 x i64> %vecins, <2 x i64> addrspace(1)* %out, align 8
+  ret void
+}
+
+; SI-LABEL: @dynamic_insertelement_v4f64:
+; SI: BUFFER_STORE_DWORDX4
+define void @dynamic_insertelement_v4f64(<4 x double> addrspace(1)* %out, <4 x double> %a, i32 %b) nounwind {
+  %vecins = insertelement <4 x double> %a, double 8.0, i32 %b
+  store <4 x double> %vecins, <4 x double> addrspace(1)* %out, align 16
+  ret void
+}
+
+; SI-LABEL: @dynamic_insertelement_v8f64:
+; SI: BUFFER_STORE_DWORDX4
+define void @dynamic_insertelement_v8f64(<8 x double> addrspace(1)* %out, <8 x double> %a, i32 %b) nounwind {
+  %vecins = insertelement <8 x double> %a, double 8.0, i32 %b
+  store <8 x double> %vecins, <8 x double> addrspace(1)* %out, align 16
+  ret void
+}