[lld] b5588ce - [LLD][MachO][NFC] Refactor LOH code (#141153)

[via llvm-commits]

Author: Ellis Hoag
Date: 2025-05-27T08:24:25-07:00
New Revision: b5588ce746755b4aaf071ab19a403099c8334f84

URL: https://github.com/llvm/llvm-project/commit/b5588ce746755b4aaf071ab19a403099c8334f84
DIFF: https://github.com/llvm/llvm-project/commit/b5588ce746755b4aaf071ab19a403099c8334f84.diff

LOG: [LLD][MachO][NFC] Refactor LOH code (#141153)

In `applyAdrpAddLdr()` we make a transformation that is identical to the
one in `applyAdrpAdd()`, so lets reuse that code. Also refactor
`forEachHint()` to use more `ArrayRef` and move around some lines for
consistancy.

Added: 
    

Modified: 
    lld/MachO/Arch/ARM64.cpp

Removed: 
    


################################################################################
diff  --git a/lld/MachO/Arch/ARM64.cpp b/lld/MachO/Arch/ARM64.cpp
index 9631269bf8f27..2fe96b26bfb55 100644
--- a/lld/MachO/Arch/ARM64.cpp
+++ b/lld/MachO/Arch/ARM64.cpp
@@ -392,25 +392,26 @@ static void writeImmediateLdr(void *loc, const Ldr &ldr) {
 // ->
 //   adr  xM, _foo
 //   nop
-static void applyAdrpAdd(uint8_t *buf, const ConcatInputSection *isec,
+static bool applyAdrpAdd(uint8_t *buf, const ConcatInputSection *isec,
                          uint64_t offset1, uint64_t offset2) {
   uint32_t ins1 = read32le(buf + offset1);
   uint32_t ins2 = read32le(buf + offset2);
   Adrp adrp;
   Add add;
   if (!parseAdrp(ins1, adrp) || !parseAdd(ins2, add))
-    return;
+    return false;
   if (adrp.destRegister != add.srcRegister)
-    return;
+    return false;
 
   uint64_t addr1 = isec->getVA() + offset1;
   uint64_t referent = pageBits(addr1) + adrp.addend + add.addend;
   int64_t delta = referent - addr1;
   if (!isValidAdrOffset(delta))
-    return;
+    return false;
 
   writeAdr(buf + offset1, add.destRegister, delta);
   writeNop(buf + offset2);
+  return true;
 }
 
 // Transforms two adrp instructions into a single adrp if their referent
@@ -495,16 +496,12 @@ static void applyAdrpAddLdr(uint8_t *buf, const ConcatInputSection *isec,
                             uint64_t offset1, uint64_t offset2,
                             uint64_t offset3) {
   uint32_t ins1 = read32le(buf + offset1);
-  Adrp adrp;
-  if (!parseAdrp(ins1, adrp))
-    return;
   uint32_t ins2 = read32le(buf + offset2);
-  Add add;
-  if (!parseAdd(ins2, add))
-    return;
   uint32_t ins3 = read32le(buf + offset3);
+  Adrp adrp;
+  Add add;
   Ldr ldr;
-  if (!parseLdr(ins3, ldr))
+  if (!parseAdrp(ins1, adrp) || !parseAdd(ins2, add) || !parseLdr(ins3, ldr))
     return;
   if (adrp.destRegister != add.srcRegister)
     return;
@@ -527,18 +524,8 @@ static void applyAdrpAddLdr(uint8_t *buf, const ConcatInputSection *isec,
     return;
   }
 
-  // Load the target address into a register and load from there indirectly.
-  //   adr x1, _foo
-  //   nop
-  //   ldr x2, [x1, #off]
-  int64_t adrOffset = referent - addr1;
-  if (isValidAdrOffset(adrOffset)) {
-    writeAdr(buf + offset1, ldr.baseRegister, adrOffset);
-    // Note: ld64 moves the offset into the adr instruction for AdrpAddLdr, but
-    // not for AdrpLdrGotLdr. Its effect is the same either way.
-    writeNop(buf + offset2);
+  if (applyAdrpAdd(buf, isec, offset1, offset2))
     return;
-  }
 
   // Move the target's page offset into the ldr's immediate offset.
   //   adrp x0, _foo at PAGE
@@ -572,67 +559,45 @@ static void applyAdrpLdrGotLdr(uint8_t *buf, const ConcatInputSection *isec,
     // adrp x1, _foo at GOTPAGE
     // ldr  x2, [x1, _foo at GOTPAGEOFF]
     // ldr  x3, [x2, #off]
-
-    uint32_t ins1 = read32le(buf + offset1);
-    Adrp adrp;
-    if (!parseAdrp(ins1, adrp))
-      return;
     uint32_t ins3 = read32le(buf + offset3);
     Ldr ldr3;
     if (!parseLdr(ins3, ldr3))
       return;
-
-    if (ldr2.baseRegister != adrp.destRegister)
-      return;
     if (ldr3.baseRegister != ldr2.destRegister)
       return;
     // Loads from the GOT must be pointer sized.
     if (ldr2.p2Size != 3 || ldr2.isFloat)
       return;
-
-    uint64_t addr1 = isec->getVA() + offset1;
-    uint64_t addr2 = isec->getVA() + offset2;
-    uint64_t referent = pageBits(addr1) + adrp.addend + ldr2.offset;
-    // Load the GOT entry's address directly.
-    //   nop
-    //   ldr x2, _foo at GOTPAGE + _foo at GOTPAGEOFF
-    //   ldr x3, [x2, #off]
-    Ldr literalLdr = ldr2;
-    literalLdr.offset = referent - addr2;
-    if (isLiteralLdrEligible(literalLdr)) {
-      writeNop(buf + offset1);
-      writeLiteralLdr(buf + offset2, literalLdr);
-    }
+    applyAdrpLdr(buf, isec, offset1, offset2);
   }
 }
 
-static uint64_t readValue(const uint8_t *&ptr, const uint8_t *end) {
-  unsigned int n = 0;
-  uint64_t value = decodeULEB128(ptr, &n, end);
-  ptr += n;
-  return value;
-}
-
 template <typename Callback>
 static void forEachHint(ArrayRef<uint8_t> data, Callback callback) {
   std::array<uint64_t, 3> args;
 
-  for (const uint8_t *p = data.begin(), *end = data.end(); p < end;) {
-    uint64_t type = readValue(p, end);
+  auto readNext = [&]() -> uint64_t {
+    unsigned int n = 0;
+    uint64_t value = decodeULEB128(data.data(), &n, data.end());
+    data = data.drop_front(n);
+    return value;
+  };
+
+  while (!data.empty()) {
+    uint64_t type = readNext();
     if (type == 0)
       break;
 
-    uint64_t argCount = readValue(p, end);
+    uint64_t argCount = readNext();
+    for (unsigned i = 0; i < argCount; ++i) {
+      uint64_t arg = readNext();
+      if (i < 3)
+        args[i] = arg;
+    }
     // All known LOH types as of 2022-09 have 3 or fewer arguments; skip others.
-    if (argCount > 3) {
-      for (unsigned i = 0; i < argCount; ++i)
-        readValue(p, end);
+    if (argCount > 3)
       continue;
-    }
-
-    for (unsigned i = 0; i < argCount; ++i)
-      args[i] = readValue(p, end);
-    callback(type, ArrayRef<uint64_t>(args.data(), argCount));
+    callback(type, ArrayRef(args.data(), argCount));
   }
 }