[LLD][ELF] Support R_ARM_SBREL32 Relocation

[peter.smith]

Summary

This change adds support for the R_ARM_SBREL32 relocation. The relocation is a base relative relocation that is produced by clang/llvm when -frwpi is used. The use case for the -frwpi option is position independent data for embedded systems that do not have a GOT. With -frwpi all read-write data is accessed via an offset from a base register (usually r9), where r9 is set at run time to where the RW data has been loaded. The base of the RW data is known as the static base.

Resolving the relocation is simple it is (S + A) - B where B is the static base. Choosing the location of and obtaining the value of the static base is more complicated. For lld I have chosen to use the base of the program segment containing the RW to be the static base. This supports the most common use case where there is a single PT_LOAD segment for the data.

Addresses PR32924 https://bugs.llvm.org/show_bug.cgi?id=32924

ELF/InputSection.cpp

static uint64_t getAArch64UndefinedRelativeWeakVA(uint64_t Type, uint64_t A,
   case R_AARCH64_JUMP26:
   case R_AARCH64_TSTBR14:
     return P + 4;
   default:
     return A;
   }
 }
 
+// ARM SBREL relocations are of the form S + A - B where B is the static base
+// We defined the static base to be the base of the PT_LOAD segment containing
+// the Body.
+static uint64_t getARMStaticBase(const SymbolBody &Body) {
+  OutputSection *OS = Body.getOutputSection();
+  if (!OS || !OS->FirstInPtLoad)
+    fatal("SBREL relocation to " + Body.getName() + " without static base\n");
+  return OS->FirstInPtLoad->Addr;
+}
+
 template <class ELFT>
 static typename ELFT::uint
 getRelocTargetVA(uint32_t Type, int64_t A, typename ELFT::uint P,
                  const SymbolBody &Body, RelExpr Expr) {
   switch (Expr) {
   case R_ABS:
   case R_RELAX_GOT_PC_NOPIC:
     return Body.getVA(A);
+  case R_ARM_SBREL:
+    return Body.getVA(A) - getARMStaticBase(Body);
   case R_GOT:
   case R_RELAX_TLS_GD_TO_IE_ABS:
     return Body.getGotVA() + A;
   case R_GOTONLY_PC:
     return InX::Got->getVA() + A - P;
   case R_GOTONLY_PC_FROM_END:
     return InX::Got->getVA() + A - P + InX::Got->getSize();
   case R_GOTREL:

ELF/Relocations.h

 class InputSectionBase;
 class OutputSection;
 
 // List of target-independent relocation types. Relocations read
 // from files are converted to these types so that the main code
 // doesn't have to know about architecture-specific details.
 enum RelExpr {
   R_ABS,
+  R_ARM_SBREL,
   R_GOT,
   R_GOTONLY_PC,
   R_GOTONLY_PC_FROM_END,
   R_GOTREL,
   R_GOTREL_FROM_END,
   R_GOT_FROM_END,
   R_GOT_OFF,
   R_GOT_PAGE_PC,

ELF/Target.cpp

case R_ARM_GOTOFF32:
     return R_GOTREL;
   case R_ARM_GOT_BREL:
     // GOT(S) + A - GOT_ORG
     return R_GOT_OFF;
   case R_ARM_GOT_PREL:
   case R_ARM_TLS_IE32:
     // GOT(S) + A - P
     return R_GOT_PC;
+  case R_ARM_SBREL32:
+    return R_ARM_SBREL;
   case R_ARM_TARGET1:
     return Config->Target1Rel ? R_PC : R_ABS;
   case R_ARM_TARGET2:
     if (Config->Target2 == Target2Policy::Rel)
       return R_PC;
     if (Config->Target2 == Target2Policy::Abs)
       return R_ABS;
     return R_GOT_PC;
void ARMTargetInfo::relocateOne(uint8_t *Loc, uint32_t Type,
   case R_ARM_ABS32:
   case R_ARM_BASE_PREL:
   case R_ARM_GLOB_DAT:
   case R_ARM_GOTOFF32:
   case R_ARM_GOT_BREL:
   case R_ARM_GOT_PREL:
   case R_ARM_REL32:
   case R_ARM_RELATIVE:
+  case R_ARM_SBREL32:
   case R_ARM_TARGET1:
   case R_ARM_TARGET2:
   case R_ARM_TLS_GD32:
   case R_ARM_TLS_IE32:
   case R_ARM_TLS_LDM32:
   case R_ARM_TLS_LDO32:
   case R_ARM_TLS_LE32:
   case R_ARM_TLS_TPOFF32:

test/ELF/arm-sbrel32.s

+// RUN: llvm-mc -filetype=obj -triple=armv7a-none-linux-gnueabi %s -o %t
+// RUN: ld.lld %t -o %t2 2>&1
+// RUN: llvm-objdump -d -triple=armv7a-none-linux-gnueabi %t2 | FileCheck %s
+// REQUIRES: arm
+
+// Test the R_ARM_SBREL32 relocation which calculates the offset of the Symbol
+// from the static base. We define the static base to be the address of the
+// segment containing the symbol
+ .text
+ .syntax unified
+
+ .globl _start
+ .p2align       2
+ .type  _start,%function
+_start:
+        .fnstart
+        bx lr
+
+        .long   foo(sbrel)
+        .long   foo2(sbrel)
+        .long   foo3(sbrel)
+        .long   foo4(sbrel)
+// RW segment starts here
+        .data
+        .p2align 4
+foo:    .word 10
+foo2:   .word 20
+
+        .bss
+foo3:   .space 4
+foo4:   .space 4
+
+// CHECK: Disassembly of section .text:
+// CHECK-NEXT: _start:
+// CHECK-NEXT:    11000:        1e ff 2f e1     bx      lr
+// CHECK:         11004:        00 00 00 00     .word   0x00000000
+// CHECK-NEXT:    11008:        04 00 00 00     .word   0x00000004
+// CHECK-NEXT:    1100c:        08 00 00 00     .word   0x00000008
+// CHECK-NEXT:    11010:        0c 00 00 00     .word   0x0000000c