[RemoveDIs] Add flag to preserve the debug info format of input IR (#87379)

This patch adds a new flag: `--preserve-input-debuginfo-format`

This flag instructs the tool to not convert the debug info format
(intrinsics/records) of input IR, but to instead determine the format of
the input IR and overwrite the other format-determining flags so that we
process and output the file in the same format that we received it in.
This flag is turned off by llvm-link, llvm-lto, and llvm-lto2, and
should be turned off by any other tool that expects to parse multiple IR
modules and have their debug info formats match.

The motivation for this flag is to allow tools to not convert the debug
info format - verify-uselistorder and llvm-reduce, and any downstream
tools that seek to test or mutate IR as-is, without applying extraneous
modifications to the input. This is a necessary step to using debug
records by default in all (other) LLVM tools.

3 files changed, 24 insertions, 6 deletions

diff --git a/llvm/lib/IR/BasicBlock.cpp b/llvm/lib/IR/BasicBlock.cpp
index ae99267..6e62767 100644
--- a/llvm/lib/IR/BasicBlock.cpp
+++ b/llvm/lib/IR/BasicBlock.cpp
@@ -30,11 +30,19 @@ using namespace llvm;
 #define DEBUG_TYPE "ir"
 STATISTIC(NumInstrRenumberings, "Number of renumberings across all blocks");
 
-cl::opt<bool>
-    UseNewDbgInfoFormat("experimental-debuginfo-iterators",
-                        cl::desc("Enable communicating debuginfo positions "
-                                 "through iterators, eliminating intrinsics"),
-                        cl::init(true));
+cl::opt<bool> UseNewDbgInfoFormat(
+    "experimental-debuginfo-iterators",
+    cl::desc("Enable communicating debuginfo positions through iterators, "
+             "eliminating intrinsics. Has no effect if "
+             "--preserve-input-debuginfo-format=true."),
+    cl::init(true));
+cl::opt<cl::boolOrDefault> PreserveInputDbgFormat(
+    "preserve-input-debuginfo-format", cl::Hidden,
+    cl::desc("When set to true, IR files will be processed and printed in "
+             "their current debug info format, regardless of default behaviour "
+             "or other flags passed. Has no effect if input IR does not "
+             "contain debug records or intrinsics. Ignored in llvm-link, "
+             "llvm-lto, and llvm-lto2."));
 
 bool WriteNewDbgInfoFormatToBitcode /*set default value in cl::init() below*/;
 cl::opt<bool, true> WriteNewDbgInfoFormatToBitcode2(
@@ -147,6 +155,9 @@ void BasicBlock::setIsNewDbgInfoFormat(bool NewFlag) {
   else if (!NewFlag && IsNewDbgInfoFormat)
     convertFromNewDbgValues();
 }
+void BasicBlock::setNewDbgInfoFormatFlag(bool NewFlag) {
+  IsNewDbgInfoFormat = NewFlag;
+}
 
 ValueSymbolTable *BasicBlock::getValueSymbolTable() {
   if (Function *F = getParent())
diff --git a/llvm/lib/IR/Function.cpp b/llvm/lib/IR/Function.cpp
index eb126f1..b5fda9b 100644
--- a/llvm/lib/IR/Function.cpp
+++ b/llvm/lib/IR/Function.cpp
@@ -103,6 +103,12 @@ void Function::setIsNewDbgInfoFormat(bool NewFlag) {
   else if (!NewFlag && IsNewDbgInfoFormat)
     convertFromNewDbgValues();
 }
+void Function::setNewDbgInfoFormatFlag(bool NewFlag) {
+  for (auto &BB : *this) {
+    BB.setNewDbgInfoFormatFlag(NewFlag);
+  }
+  IsNewDbgInfoFormat = NewFlag;
+}
 
 //===----------------------------------------------------------------------===//
 // Argument Implementation
diff --git a/llvm/lib/IR/IRPrintingPasses.cpp b/llvm/lib/IR/IRPrintingPasses.cpp
index 84fb8e6..43252c5 100644
--- a/llvm/lib/IR/IRPrintingPasses.cpp
+++ b/llvm/lib/IR/IRPrintingPasses.cpp
@@ -25,7 +25,8 @@ using namespace llvm;
 
 cl::opt<bool> WriteNewDbgInfoFormat(
     "write-experimental-debuginfo",
-    cl::desc("Write debug info in the new non-intrinsic format"),
+    cl::desc("Write debug info in the new non-intrinsic format. Has no effect "
+             "if --preserve-input-debuginfo-format=true."),
     cl::init(false));
 
 namespace {