[PowerPC] Allow using initial-exec TLS with PIC

Using initial-exec TLS variables is a reasonable performance optimisation for system libraries. Use the correct PIC mechanism to get hold of the GOT to avoid text relocations. Differential Revision: https://reviews.llvm.org/D61026 llvm-svn: 359146
llvm · Apr 24, 2019 · 8372b46 · 8372b46
1 parent 2f49676
commit 8372b46
Show file tree

Hide file tree

Showing 3 changed files with 21 additions and 6 deletions.
diff --git a/llvm/lib/Target/PowerPC/PPCISelLowering.cpp b/llvm/lib/Target/PowerPC/PPCISelLowering.cpp
@@ -2757,7 +2757,8 @@ SDValue PPCTargetLowering::LowerGlobalTLSAddress(SDValue Op,
   const Module *M = DAG.getMachineFunction().getFunction().getParent();
   PICLevel::Level picLevel = M->getPICLevel();
 
-  TLSModel::Model Model = getTargetMachine().getTLSModel(GV);
+  const TargetMachine &TM = getTargetMachine();
+  TLSModel::Model Model = TM.getTLSModel(GV);
 
   if (Model == TLSModel::LocalExec) {
     SDValue TGAHi = DAG.getTargetGlobalAddress(GV, dl, PtrVT, 0,
@@ -2781,8 +2782,14 @@ SDValue PPCTargetLowering::LowerGlobalTLSAddress(SDValue Op,
       SDValue GOTReg = DAG.getRegister(PPC::X2, MVT::i64);
       GOTPtr = DAG.getNode(PPCISD::ADDIS_GOT_TPREL_HA, dl,
                            PtrVT, GOTReg, TGA);
-    } else
-      GOTPtr = DAG.getNode(PPCISD::PPC32_GOT, dl, PtrVT);
+    } else {
+      if (!TM.isPositionIndependent())
+        GOTPtr = DAG.getNode(PPCISD::PPC32_GOT, dl, PtrVT);
+      else if (picLevel == PICLevel::SmallPIC)
+        GOTPtr = DAG.getNode(PPCISD::GlobalBaseReg, dl, PtrVT);
+      else
+        GOTPtr = DAG.getNode(PPCISD::PPC32_PICGOT, dl, PtrVT);
+    }
     SDValue TPOffset = DAG.getNode(PPCISD::LD_GOT_TPREL_L, dl,
                                    PtrVT, TGA, GOTPtr);
     return DAG.getNode(PPCISD::ADD_TLS, dl, PtrVT, TPOffset, TGATLS);

diff --git a/llvm/lib/Target/PowerPC/PPCISelLowering.h b/llvm/lib/Target/PowerPC/PPCISelLowering.h
@@ -264,11 +264,11 @@ namespace llvm {
       CR6UNSET,
 
       /// GPRC = address of _GLOBAL_OFFSET_TABLE_. Used by initial-exec TLS
-      /// on PPC32.
+      /// for non-position independent code on PPC32.
       PPC32_GOT,
 
       /// GPRC = address of _GLOBAL_OFFSET_TABLE_. Used by general dynamic and
-      /// local dynamic TLS on PPC32.
+      /// local dynamic TLS and position indendepent code on PPC32.
       PPC32_PICGOT,
 
       /// G8RC = ADDIS_GOT_TPREL_HA %x2, Symbol - Used by the initial-exec

diff --git a/llvm/test/CodeGen/PowerPC/tls.ll b/llvm/test/CodeGen/PowerPC/tls.ll
@@ -1,6 +1,7 @@
 ; RUN: llc -relocation-model=static -verify-machineinstrs -O0 < %s -mcpu=ppc64 | FileCheck -check-prefix=OPT0 %s
 ; RUN: llc -relocation-model=static -verify-machineinstrs -O1 < %s -mcpu=ppc64 | FileCheck -check-prefix=OPT1 %s
 ; RUN: llc -verify-machineinstrs -O0 < %s -mtriple=ppc32-- -mcpu=ppc | FileCheck -check-prefix=OPT0-PPC32 %s
+; RUN: llc -relocation-model=pic -verify-machineinstrs -O0 < %s -mtriple=ppc32-- -mcpu=ppc | FileCheck -check-prefix=OPT0-PPC32-PIC %s
 
 target triple = "powerpc64-unknown-linux-gnu"
 
@@ -24,7 +25,7 @@ entry:
 ; Test correct assembly code generation for thread-local storage
 ; using the initial-exec model.
 
-@a2 = external thread_local global i32
+@a2 = external thread_local(initialexec) global i32
 
 define signext i32 @main2() nounwind {
 entry:
@@ -44,3 +45,10 @@ entry:
 ;OPT0-PPC32:       addis [[REG1]], [[REG1]], _GLOBAL_OFFSET_TABLE_@ha
 ;OPT0-PPC32:       lwz [[REG2:[0-9]+]], a2@got@tprel@l([[REG1]])
 ;OPT0-PPC32:       add 3, [[REG2]], a2@tls
+
+;OPT0-PPC32-PIC-LABEL:  main2:
+;OPT0-PPC32-PIC:        .long _GLOBAL_OFFSET_TABLE_-{{.*}}
+;OPT0-PPC32-PIC-NOT:    li {{[0-9]+}}, _GLOBAL_OFFSET_TABLE_@l
+;OPT0-PPC32-PIC-NOT:    addis {{[0-9]+}}, {{[0-9+]}}, _GLOBAL_OFFSET_TABLE_@ha
+;OPT0-PPC32-PIC-NOT:    bl __tls_get_addr(a2@tlsgd)@PLT
+;OPT0-PPC32-PIC:        lwz {{[0-9]+}}, a2@got@tprel@l({{[0-9]+}})