AArch64: Allow offsets to be folded into addresses with ELF.

This is a code size win in code that takes offseted addresses
frequently, such as C++ constructors that typically need to compute
an offseted address of a vtable. It reduces the size of Chromium for
Android's .text section by 46KB, or 56KB with ThinLTO (which exposes
more opportunities to use a direct access rather than a GOT access).

Because the addend range is limited in COFF and Mach-O, this is
enabled for ELF only.

Differential Revision: https://reviews.llvm.org/D45199

llvm-svn: 329611

2 files changed, 24 insertions, 17 deletions

diff --git a/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp b/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
index eee59f1e719..d44eee051aa 100644
--- a/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
+++ b/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
@@ -743,14 +743,16 @@ bool AArch64DAGToDAGISel::SelectAddrModeIndexed(SDValue N, unsigned Size,
     if (!GAN)
       return true;
 
-    const GlobalValue *GV = GAN->getGlobal();
-    unsigned Alignment = GV->getAlignment();
-    Type *Ty = GV->getValueType();
-    if (Alignment == 0 && Ty->isSized())
-      Alignment = DL.getABITypeAlignment(Ty);
-
-    if (Alignment >= Size)
-      return true;
+    if (GAN->getOffset() % Size == 0) {
+      const GlobalValue *GV = GAN->getGlobal();
+      unsigned Alignment = GV->getAlignment();
+      Type *Ty = GV->getValueType();
+      if (Alignment == 0 && Ty->isSized())
+        Alignment = DL.getABITypeAlignment(Ty);
+
+      if (Alignment >= Size)
+        return true;
+    }
   }
 
   if (CurDAG->isBaseWithConstantOffset(N)) {
diff --git a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
index 28edcc2e47b..546da27cc41 100644
--- a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
+++ b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
@@ -3675,7 +3675,8 @@ AArch64TargetLowering::LowerReturn(SDValue Chain, CallingConv::ID CallConv,
 SDValue AArch64TargetLowering::getTargetNode(GlobalAddressSDNode *N, EVT Ty,
                                              SelectionDAG &DAG,
                                              unsigned Flag) const {
-  return DAG.getTargetGlobalAddress(N->getGlobal(), SDLoc(N), Ty, 0, Flag);
+  return DAG.getTargetGlobalAddress(N->getGlobal(), SDLoc(N), Ty,
+                                    N->getOffset(), Flag);
 }
 
 SDValue AArch64TargetLowering::getTargetNode(JumpTableSDNode *N, EVT Ty,
@@ -3749,9 +3750,9 @@ SDValue AArch64TargetLowering::LowerGlobalAddress(SDValue Op,
                                       : AArch64II::MO_NO_FLAG);
   unsigned char OpFlags =
       Subtarget->ClassifyGlobalReference(GV, getTargetMachine());
-
-  assert(cast<GlobalAddressSDNode>(Op)->getOffset() == 0 &&
-         "unexpected offset in global node");
+  if (OpFlags != AArch64II::MO_NO_FLAG)
+    assert(cast<GlobalAddressSDNode>(Op)->getOffset() == 0 &&
+           "unexpected offset in global node");
 
   // This also catches the large code model case for Darwin.
   if ((OpFlags & AArch64II::MO_GOT) != 0) {
@@ -4989,11 +4990,15 @@ SDValue AArch64TargetLowering::LowerShiftLeftParts(SDValue Op,
 
 bool AArch64TargetLowering::isOffsetFoldingLegal(
     const GlobalAddressSDNode *GA) const {
-  DEBUG(dbgs() << "Skipping offset folding global address: ");
-  DEBUG(GA->dump());
-  DEBUG(dbgs() << "AArch64 doesn't support folding offsets into global "
-        "addresses\n");
-  return false;
+  // FIXME: Only ELF can represent the full range of possible addends here, as
+  // the format stores the addend in a 64-bit field. With Mach-O the equivalent
+  // field is 24 bits, and with COFF it is 21 bits. To make this work with the
+  // other object formats we will need to arrange to prevent the addend from
+  // going out of bounds.
+  if (!getTargetMachine().getTargetTriple().isOSBinFormatELF())
+    return false;
+  return Subtarget->ClassifyGlobalReference(
+             GA->getGlobal(), getTargetMachine()) == AArch64II::MO_NO_FLAG;
 }
 
 bool AArch64TargetLowering::isFPImmLegal(const APFloat &Imm, EVT VT) const {