[Bitcode] materialize Functions early when BlockAddress taken

IRLinker builds a work list of functions to materialize, then moves them
from a source module to a destination module one at a time.

This is a problem for blockaddress Constants, since they need not refer
to the function they are used in; IPSCCP is quite good at sinking these
constants deep into other functions when passed as arguments.

This would lead to curious errors during LTO:
  ld.lld: error: Never resolved function from blockaddress ...
based on the ordering of function definitions in IR.

The problem was that IRLinker would basically do:

  for function f in worklist:
    materialize f
    splice f from source module to destination module

in one pass, with Functions being lazily added to the running worklist.
This confuses BitcodeReader, which cannot disambiguate whether a
blockaddress is referring to a function which has not yet been parsed
("materialized") or is simply empty because its body was spliced out.
This causes BitcodeReader to insert Functions into its BasicBlockFwdRefs
list incorrectly, as it will never re-materialize an already
materialized (but spliced out) function.

Because of the possibility that blockaddress Constants may appear in
Functions other than the ones they reference, this patch adds a new
bitcode function code FUNC_CODE_BLOCKADDR_USERS that is a simple list of
Functions that contain BlockAddress Constants that refer back to this
Function, rather then the Function they are scoped in. We then
materialize those functions when materializing `f` from the example loop
above. This might over-materialize Functions should the user of
BitcodeReader ultimately decide not to link those Functions, but we can
at least now we can avoid this ordering related issue with blockaddresses.

Fixes: #52787
Fixes: ClangBuiltLinux/linux#1215

Reviewed By: dexonsmith

Differential Revision: https://reviews.llvm.org/D120781

Author: nickdesaulniers

llvm/include/llvm/Bitcode/LLVMBitCodes.h

@@ -585,14 +585,15 @@ enum FunctionCodes {
       52, // CATCHSWITCH: [num,args...] or [num,args...,bb]
   // 53 is unused.
   // 54 is unused.
-  FUNC_CODE_OPERAND_BUNDLE = 55, // OPERAND_BUNDLE: [tag#, value...]
-  FUNC_CODE_INST_UNOP = 56,      // UNOP:       [opcode, ty, opval]
-  FUNC_CODE_INST_CALLBR = 57,    // CALLBR:     [attr, cc, norm, transfs,
-                                 //              fnty, fnid, args...]
-  FUNC_CODE_INST_FREEZE = 58,    // FREEZE: [opty, opval]
-  FUNC_CODE_INST_ATOMICRMW = 59, // ATOMICRMW: [ptrty, ptr, valty, val,
-                                 //             operation, align, vol,
-                                 //             ordering, synchscope]
+  FUNC_CODE_OPERAND_BUNDLE = 55,  // OPERAND_BUNDLE: [tag#, value...]
+  FUNC_CODE_INST_UNOP = 56,       // UNOP:       [opcode, ty, opval]
+  FUNC_CODE_INST_CALLBR = 57,     // CALLBR:     [attr, cc, norm, transfs,
+                                  //              fnty, fnid, args...]
+  FUNC_CODE_INST_FREEZE = 58,     // FREEZE: [opty, opval]
+  FUNC_CODE_INST_ATOMICRMW = 59,  // ATOMICRMW: [ptrty, ptr, valty, val,
+                                  //             operation, align, vol,
+                                  //             ordering, synchscope]
+  FUNC_CODE_BLOCKADDR_USERS = 60, // BLOCKADDR_USERS: [value...]
 };
 
 enum UseListCodes {

llvm/lib/Bitcode/Reader/BitcodeAnalyzer.cpp

@@ -267,6 +267,7 @@ static Optional<const char *> GetCodeName(unsigned CodeID, unsigned BlockID,
       STRINGIFY_CODE(FUNC_CODE, INST_STOREATOMIC)
       STRINGIFY_CODE(FUNC_CODE, INST_CMPXCHG)
       STRINGIFY_CODE(FUNC_CODE, INST_CALLBR)
+      STRINGIFY_CODE(FUNC_CODE, BLOCKADDR_USERS)
     }
   case bitc::VALUE_SYMTAB_BLOCK_ID:
     switch (CodeID) {

llvm/lib/Bitcode/Reader/BitcodeReader.cpp

@@ -558,6 +558,13 @@ class BitcodeReader : public BitcodeReaderBase, public GVMaterializer {
   DenseMap<Function *, std::vector<BasicBlock *>> BasicBlockFwdRefs;
   std::deque<Function *> BasicBlockFwdRefQueue;
 
+  /// These are Functions that contain BlockAddresses which refer a different
+  /// Function. When parsing the different Function, queue Functions that refer
+  /// to the different Function. Those Functions must be materialized in order
+  /// to resolve their BlockAddress constants before the different Function
+  /// gets moved into another Module.
+  std::vector<Function *> BackwardRefFunctions;
+
   /// Indicates that we are using a new encoding for instruction operands where
   /// most operands in the current FUNCTION_BLOCK are encoded relative to the
   /// instruction number, for a more compact encoding.  Some instruction
@@ -881,6 +888,11 @@ Error BitcodeReader::materializeForwardReferencedFunctions() {
   }
   assert(BasicBlockFwdRefs.empty() && "Function missing from queue");
 
+  for (Function *F : BackwardRefFunctions)
+    if (Error Err = materialize(F))
+      return Err;
+  BackwardRefFunctions.clear();
+
   // Reset state.
   WillMaterializeAllForwardRefs = false;
   return Error::success();
@@ -4317,6 +4329,31 @@ Error BitcodeReader::parseFunctionBody(Function *F) {
       continue;
     }
 
+    case bitc::FUNC_CODE_BLOCKADDR_USERS: // BLOCKADDR_USERS: [vals...]
+      // The record should not be emitted if it's an empty list.
+      if (Record.empty())
+        return error("Invalid record");
+      // When we have the RARE case of a BlockAddress Constant that is not
+      // scoped to the Function it refers to, we need to conservatively
+      // materialize the referred to Function, regardless of whether or not
+      // that Function will ultimately be linked, otherwise users of
+      // BitcodeReader might start splicing out Function bodies such that we
+      // might no longer be able to materialize the BlockAddress since the
+      // BasicBlock (and entire body of the Function) the BlockAddress refers
+      // to may have been moved. In the case that the user of BitcodeReader
+      // decides ultimately not to link the Function body, materializing here
+      // could be considered wasteful, but it's better than a deserialization
+      // failure as described. This keeps BitcodeReader unaware of complex
+      // linkage policy decisions such as those use by LTO, leaving those
+      // decisions "one layer up."
+      for (uint64_t ValID : Record)
+        if (auto *F = dyn_cast<Function>(ValueList[ValID]))
+          BackwardRefFunctions.push_back(F);
+        else
+          return error("Invalid record");
+
+      continue;
+
     case bitc::FUNC_CODE_DEBUG_LOC_AGAIN:  // DEBUG_LOC_AGAIN
       // This record indicates that the last instruction is at the same
       // location as the previous instruction with a location.

llvm/lib/Bitcode/Writer/BitcodeWriter.cpp

@@ -19,6 +19,7 @@
 #include "llvm/ADT/None.h"
 #include "llvm/ADT/Optional.h"
 #include "llvm/ADT/STLExtras.h"
+#include "llvm/ADT/SmallPtrSet.h"
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringMap.h"
@@ -3359,8 +3360,10 @@ void ModuleBitcodeWriter::writeFunction(
   bool NeedsMetadataAttachment = F.hasMetadata();
 
   DILocation *LastDL = nullptr;
+  SmallPtrSet<Function *, 4> BlockAddressUsers;
+
   // Finally, emit all the instructions, in order.
-  for (const BasicBlock &BB : F)
+  for (const BasicBlock &BB : F) {
     for (const Instruction &I : BB) {
       writeInstruction(I, InstID, Vals);
 
@@ -3392,6 +3395,25 @@ void ModuleBitcodeWriter::writeFunction(
       LastDL = DL;
     }
 
+    if (BlockAddress *BA = BlockAddress::lookup(&BB)) {
+      for (User *U : BA->users()) {
+        if (auto *I = dyn_cast<Instruction>(U)) {
+          Function *P = I->getParent()->getParent();
+          if (P != &F)
+            BlockAddressUsers.insert(P);
+        }
+      }
+    }
+  }
+
+  if (!BlockAddressUsers.empty()) {
+    SmallVector<uint64_t, 4> Record;
+    Record.reserve(BlockAddressUsers.size());
+    for (Function *F : BlockAddressUsers)
+      Record.push_back(VE.getValueID(F));
+    Stream.EmitRecord(bitc::FUNC_CODE_BLOCKADDR_USERS, Record);
+  }
+
   // Emit names for all the instructions etc.
   if (auto *Symtab = F.getValueSymbolTable())
     writeFunctionLevelValueSymbolTable(*Symtab);

llvm/test/Bitcode/blockaddress-users.ll

@@ -0,0 +1,38 @@
+; RUN: llvm-as %s -o %t.bc
+; RUN: llvm-bcanalyzer -dump %t.bc | FileCheck %s
+; RUN: llvm-dis %t.bc
+
+; There's a curious case where blockaddress constants may refer to functions
+; outside of the function they're used in. There's a special bitcode function
+; code, FUNC_CODE_BLOCKADDR_USERS, used to signify that this is the case.
+
+; The intent of this test is two-fold:
+; 1. Ensure we produce BLOCKADDR_USERS bitcode function code on the first fn,
+;    @repro, since @fun and @fun2 both refer to @repro via blockaddress
+;    constants.
+; 2. Ensure we can round-trip serializing+desearlizing such case.
+
+; CHECK: <FUNCTION_BLOCK
+; CHECK: <BLOCKADDR_USERS op0=2 op1=1
+; CHECK: <FUNCTION_BLOCK
+; CHECK: <FUNCTION_BLOCK
+
+define void @repro() {
+  br label %label
+
+label:
+  call void @fun()
+  ret void
+}
+
+define void @fun() noinline {
+  call void @f(i8* blockaddress(@repro, %label))
+  ret void
+}
+
+define void @fun2() noinline {
+  call void @f(i8* blockaddress(@repro, %label))
+  ret void
+}
+
+declare void @f(i8*)

llvm/test/Linker/blockaddress.ll

@@ -0,0 +1,125 @@
+; RUN: llvm-as %s -o %t.bc
+; RUN: llvm-link %t.bc -S | FileCheck %s
+
+declare void @f(i8*)
+
+; Test that a blockaddress in @y referring to %label in @x can be moved when @y
+; appears after @x.
+define void @x() {
+  br label %label
+
+label:
+  call void @y()
+  ret void
+}
+
+define void @y() {
+; CHECK: define void @y() {
+; CHECK-NEXT: call void @f(i8* blockaddress(@x, %label))
+  call void @f(i8* blockaddress(@x, %label))
+  ret void
+}
+
+; Test that a blockaddress in @a referring to %label in @b can be moved when @a
+; appears before @b.
+define void @a() {
+; CHECK: define void @a() {
+; CHECK-NEXT: call void @f(i8* blockaddress(@b, %label))
+  call void @f(i8* blockaddress(@b, %label))
+  ret void
+}
+
+define void @b() {
+  br label %label
+
+label:
+  call void @a()
+  ret void
+}
+
+; Test that @c and @d can both have blockaddress Constants that refer to one
+; another.
+
+define void @c() {
+; CHECK: define void @c() {
+; CHECK-NEXT:  br label %label
+; CHECK-EMPTY:
+; CHECK-NEXT: label:
+; CHECK-NEXT: call void @f(i8* blockaddress(@d, %label))
+  br label %label
+
+label:
+  call void @f(i8* blockaddress(@d, %label))
+  ret void
+}
+
+define void @d() {
+; CHECK: define void @d() {
+; CHECK-NEXT:  br label %label
+; CHECK-EMPTY:
+; CHECK-NEXT: label:
+; CHECK-NEXT: call void @f(i8* blockaddress(@c, %label))
+  br label %label
+
+label:
+  call void @f(i8* blockaddress(@c, %label))
+  ret void
+}
+
+; Test that Functions added to IRLinker's Worklist member lazily (linkonce_odr)
+; aren't susceptible to the the same issues as @x/@y above.
+define void @parsed() {
+  br label %label
+
+label:
+  ret void
+}
+
+define linkonce_odr void @lazy() {
+; CHECK: define linkonce_odr void @lazy() {
+; CHECK-NEXT: br label %label
+; CHECK-EMPTY:
+; CHECK-NEXT: label:
+; CHECK-NEXT: call void @f(i8* blockaddress(@parsed, %label))
+  br label %label
+
+label:
+  call void @f(i8* blockaddress(@parsed, %label))
+  ret void
+}
+
+define void @parsed2() {
+  call void @lazy()
+  ret void
+}
+
+; Same test as @lazy, just with one more level of lazy parsed functions.
+define void @parsed3() {
+  br label %label
+
+label:
+  ret void
+}
+
+define linkonce_odr void @lazy1() {
+; CHECK: define linkonce_odr void @lazy1() {
+; CHECK-NEXT: br label %label
+; CHECK-EMPTY:
+; CHECK-NEXT: label:
+; CHECK-NEXT: call void @f(i8* blockaddress(@parsed3, %label))
+  br label %label
+
+label:
+  call void @f(i8* blockaddress(@parsed3, %label))
+  ret void
+}
+
+define linkonce_odr void @lazy2() {
+  call void @lazy1()
+  ret void
+}
+
+define void @parsed4() {
+  call void @lazy2()
+  ret void
+}