Index: include/llvm/Object/RelocVisitor.h
===================================================================
--- include/llvm/Object/RelocVisitor.h
+++ include/llvm/Object/RelocVisitor.h
@@ -143,12 +143,12 @@
         switch (RelocType) {
         case llvm::ELF::R_AMDGPU_ABS32_LO:
         case llvm::ELF::R_AMDGPU_ABS32_HI:
+        case llvm::ELF::R_AMDGPU_REL32:
           return visitELF_AMDGPU_32(R, Value);
         default:
           HasError = true;
           return RelocToApply();
         }
-
       default:
         HasError = true;
         return RelocToApply();
Index: include/llvm/Support/ELFRelocs/AMDGPU.def
===================================================================
--- include/llvm/Support/ELFRelocs/AMDGPU.def
+++ include/llvm/Support/ELFRelocs/AMDGPU.def
@@ -5,3 +5,4 @@
 ELF_RELOC(R_AMDGPU_NONE,          0)
 ELF_RELOC(R_AMDGPU_ABS32_LO,      1)
 ELF_RELOC(R_AMDGPU_ABS32_HI,      2)
+ELF_RELOC(R_AMDGPU_REL32,         4)
Index: lib/Target/AMDGPU/MCTargetDesc/AMDGPUELFObjectWriter.cpp
===================================================================
--- lib/Target/AMDGPU/MCTargetDesc/AMDGPUELFObjectWriter.cpp
+++ lib/Target/AMDGPU/MCTargetDesc/AMDGPUELFObjectWriter.cpp
@@ -47,12 +47,13 @@
       return ELF::R_AMDGPU_ABS32_LO;
     if (Target.getSymA()->getSymbol().getName() == "SCRATCH_RSRC_DWORD1")
       return ELF::R_AMDGPU_ABS32_HI;
+  case FK_PCRel_4:
+    return ELF::R_AMDGPU_REL32;
   }
 
   llvm_unreachable("unhandled relocation type");
 }
 
-
 MCObjectWriter *llvm::createAMDGPUELFObjectWriter(bool Is64Bit,
                                                   bool HasRelocationAddend,
                                                   raw_pwrite_stream &OS) {
Index: lib/Target/AMDGPU/SIISelLowering.cpp
===================================================================
--- lib/Target/AMDGPU/SIISelLowering.cpp
+++ lib/Target/AMDGPU/SIISelLowering.cpp
@@ -1421,7 +1421,8 @@
                                              SelectionDAG &DAG) const {
   GlobalAddressSDNode *GSD = cast<GlobalAddressSDNode>(Op);
 
-  if (GSD->getAddressSpace() != AMDGPUAS::CONSTANT_ADDRESS)
+  if (GSD->getAddressSpace() != AMDGPUAS::CONSTANT_ADDRESS &&
+      GSD->getAddressSpace() != AMDGPUAS::GLOBAL_ADDRESS)
     return AMDGPUTargetLowering::LowerGlobalAddress(MFI, Op, DAG);
 
   SDLoc DL(GSD);
Index: test/CodeGen/AMDGPU/global-zero-initializer.ll
===================================================================
--- test/CodeGen/AMDGPU/global-zero-initializer.ll
+++ test/CodeGen/AMDGPU/global-zero-initializer.ll
@@ -1,12 +1,17 @@
-; RUN: not llc -march=amdgcn -mcpu=SI < %s 2>&1 | FileCheck %s
-; RUN: not llc -march=amdgcn -mcpu=tonga < %s 2>&1 | FileCheck %s
+; RUN: llc -march=amdgcn -mcpu=SI < %s 2>&1 | FileCheck %s
+; RUN: llc -march=amdgcn -mcpu=tonga < %s 2>&1 | FileCheck %s
 
-; CHECK: in function load_init_global_global{{.*}}: unsupported initializer for address space
-
-@lds = addrspace(1) global [256 x i32] zeroinitializer
+; CHECK: {{^}}load_init_global_global:
+; CHECK: s_getpc_b64 s{{\[}}[[PC_LO:[0-9]+]]:[[PC_HI:[0-9]+]]{{\]}}
+; CHECK: s_add_u32 s[[ADDR_LO:[0-9]+]], s[[PC_LO]], global+4
+; CHECK: s_addc_u32 s5, s[[PC_HI]], 0
+; CHECK: buffer_load_dword v{{[0-9]+}}, off, s{{\[}}[[ADDR_LO]]:7], 0 offset:40
+; CHECK: global:
+; CHECK: .zero 1024
+@global = addrspace(1) global [256 x i32] zeroinitializer
 
 define void @load_init_global_global(i32 addrspace(1)* %out, i1 %p) {
- %gep = getelementptr [256 x i32], [256 x i32] addrspace(1)* @lds, i32 0, i32 10
+ %gep = getelementptr [256 x i32], [256 x i32] addrspace(1)* @global, i32 0, i32 10
   %ld = load i32, i32 addrspace(1)* %gep
   store i32 %ld, i32 addrspace(1)* %out
   ret void
Index: test/CodeGen/AMDGPU/wqm.ll
===================================================================
--- test/CodeGen/AMDGPU/wqm.ll
+++ test/CodeGen/AMDGPU/wqm.ll
@@ -40,6 +40,7 @@
 ;CHECK: s_and_b64 exec, exec, [[ORIG]]
 ;CHECK: store
 ;CHECK-NOT: exec
+;CHECK: .size test3
 define amdgpu_ps <4 x float> @test3(<8 x i32> inreg %rsrc, <4 x i32> inreg %sampler, float addrspace(1)* inreg %ptr, <4 x i32> %c) {
 main_body:
   %tex = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> %c, <8 x i32> %rsrc, <4 x i32> %sampler, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)