Rewrite the darwin tlv support to use a chain and return to copying

the output to the correct register. Fixes a hidden problem uncovered
by the last patch where we'd try to DAG combine our MVT::Other node
oddly.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121358 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/X86/X86ISelLowering.cpp b/lib/Target/X86/X86ISelLowering.cpp
index 8998b5bea9294ffd2360649d8ca8638eb25d2e85..f38cad7d37c4400d9123072ad9895eb85a97b42c 100644 (file)
--- a/lib/Target/X86/X86ISelLowering.cpp
+++ b/lib/Target/X86/X86ISelLowering.cpp
@@ -6181,7 +6181,7 @@ X86TargetLowering::LowerGlobalTLSAddress(SDValue Op, SelectionDAG &DAG) const {
       OpFlag = X86II::MO_TLVP;
     DebugLoc DL = Op.getDebugLoc();
     SDValue Result = DAG.getTargetGlobalAddress(GA->getGlobal(), DL,
-                                                getPointerTy(),
+                                                GA->getValueType(0),
                                                 GA->getOffset(), OpFlag);
     SDValue Offset = DAG.getNode(WrapperKind, DL, getPointerTy(), Result);
 
@@ -6194,8 +6194,10 @@ X86TargetLowering::LowerGlobalTLSAddress(SDValue Op, SelectionDAG &DAG) const {
 
     // Lowering the machine isd will make sure everything is in the right
     // location.
-    SDValue Args[] = { Offset };
-    SDValue RetVal = DAG.getNode(X86ISD::TLSCALL, DL, MVT::Other, Args, 1);
+    SDValue Chain = DAG.getEntryNode();
+    SDVTList NodeTys = DAG.getVTList(MVT::Other, MVT::Flag);
+    SDValue Args[] = { Chain, Offset };
+    Chain = DAG.getNode(X86ISD::TLSCALL, DL, NodeTys, Args, 2);
 
     // TLSCALL will be codegen'ed as call. Inform MFI that function has calls.
     MachineFrameInfo *MFI = DAG.getMachineFunction().getFrameInfo();
@@ -6203,7 +6205,8 @@ X86TargetLowering::LowerGlobalTLSAddress(SDValue Op, SelectionDAG &DAG) const {
     
     // And our return value (tls address) is in the standard call return value
     // location.
-    return RetVal;
+    unsigned Reg = Subtarget->is64Bit() ? X86::RAX : X86::EAX;
+    return DAG.getCopyFromReg(Chain, DL, Reg, getPointerTy());
   }
 
   assert(false &&

diff --git a/lib/Target/X86/X86InstrInfo.td b/lib/Target/X86/X86InstrInfo.td
index d6e7f580320be4618eedcc9f9fa89238e2e6a644..8b5a8cf44e147ced012b5bc4b14cf08767479f48 100644 (file)
--- a/lib/Target/X86/X86InstrInfo.td
+++ b/lib/Target/X86/X86InstrInfo.td
@@ -84,7 +84,7 @@ def SDTX86Wrapper : SDTypeProfile<1, 1, [SDTCisSameAs<0, 1>, SDTCisPtrTy<0>]>;
 
 def SDT_X86TLSADDR : SDTypeProfile<0, 1, [SDTCisInt<0>]>;
 
-def SDT_X86TLSCALL : SDTypeProfile<0, 1, [SDTCisPtrTy<0>]>;
+def SDT_X86TLSCALL : SDTypeProfile<0, 1, [SDTCisInt<0>]>;
 
 def SDT_X86EHRET : SDTypeProfile<0, 1, [SDTCisInt<0>]>;
 
@@ -212,7 +212,7 @@ def X86WinAlloca : SDNode<"X86ISD::WIN_ALLOCA", SDTX86Void,
                           [SDNPHasChain, SDNPInFlag, SDNPOutFlag]>;
 
 def X86TLSCall : SDNode<"X86ISD::TLSCALL", SDT_X86TLSCALL,
-                        []>;
+                        [SDNPHasChain, SDNPOptInFlag, SDNPOutFlag]>;
 
 //===----------------------------------------------------------------------===//
 // X86 Operand Definitions.

diff --git a/test/CodeGen/X86/tlv-1.ll b/test/CodeGen/X86/tlv-1.ll
index de694d8d471f8a0926cf5bedc17d2af58fe04cee..42940f147ed8361b897877182d2e61187bf1d280 100644 (file)
--- a/test/CodeGen/X86/tlv-1.ll
+++ b/test/CodeGen/X86/tlv-1.ll
@@ -1,5 +1,21 @@
 ; RUN: llc < %s -mtriple x86_64-apple-darwin | FileCheck %s
 
+%struct.A = type { [48 x i8], i32, i32, i32 }
+
+@c = external thread_local global %struct.A, align 4
+
+define void @main() nounwind ssp {
+entry:
+  call void @llvm.memset.p0i8.i64(i8* getelementptr inbounds (%struct.A* @c, i32 0, i32 0, i32 0), i8 0, i64 60, i32 1, i1 false)
+  unreachable  
+  ; CHECK: movq    _c@TLVP(%rip), %rdi
+  ; CHECK-NEXT: callq   *(%rdi)
+  ; CHECK-NEXT: movl    $0, 56(%rax)
+  ; CHECK-NEXT: movq    $0, 48(%rax)
+}
+
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
+
 @a = thread_local global i32 0                    ; <i32*> [#uses=0]
 @b = thread_local global i32 0                    ; <i32*> [#uses=0]