Update names field in source maps (#8068)

This scans the code after optimizations and removes unused function
names from 'names' field in the source map, reducing its size.
Emscripten has not been generating 'names' field so far, but after
emscripten-core/emscripten#25870, it will
generate the field in case `llvm-dwarfdump` supports a new option
`--filter-child-tag`.

Author: aheejin

src/wasm/wasm-binary.cpp

@@ -16,16 +16,13 @@
 
 #include <algorithm>
 #include <fstream>
-#include <iomanip>
 
-#include "ir/eh-utils.h"
 #include "ir/module-utils.h"
 #include "ir/names.h"
 #include "ir/table-utils.h"
 #include "ir/type-updating.h"
 #include "pass.h"
 #include "support/bits.h"
-#include "support/debug.h"
 #include "support/stdckdint.h"
 #include "support/string.h"
 #include "wasm-annotations.h"
@@ -1242,6 +1239,43 @@ void WasmBinaryWriter::writeSourceMapProlog() {
     }
   }
 
+  // Remove unused function names from 'names' field.
+  if (!wasm->debugInfoSymbolNames.empty()) {
+    std::vector<std::string> newSymbolNames;
+    std::map<Index, Index> oldToNewIndex;
+
+    // Collect all used symbol name indexes.
+    for (auto& func : wasm->functions) {
+      for (auto& [_, location] : func->debugLocations) {
+        if (location && location->symbolNameIndex) {
+          uint32_t oldIndex = *location->symbolNameIndex;
+          assert(oldIndex < wasm->debugInfoSymbolNames.size());
+          oldToNewIndex[oldIndex] = 0; // placeholder
+        }
+      }
+    }
+
+    // Create the new list of names and the mapping from old to new indices.
+    uint32_t index = 0;
+    for (auto& [oldIndex, newIndex] : oldToNewIndex) {
+      newSymbolNames.push_back(wasm->debugInfoSymbolNames[oldIndex]);
+      newIndex = index++;
+    }
+
+    // Update all debug locations to point to the new indices.
+    for (auto& func : wasm->functions) {
+      for (auto& [_, location] : func->debugLocations) {
+        if (location && location->symbolNameIndex) {
+          uint32_t oldIndex = *location->symbolNameIndex;
+          location->symbolNameIndex = oldToNewIndex[oldIndex];
+        }
+      }
+    }
+
+    // Replace the old symbol names with the new, pruned list.
+    wasm->debugInfoSymbolNames = std::move(newSymbolNames);
+  }
+
   auto writeOptionalString = [&](const char* name, const std::string& str) {
     if (!str.empty()) {
       *sourceMap << "\"" << name << "\":\"" << str << "\",";
@@ -1269,10 +1303,6 @@ void WasmBinaryWriter::writeSourceMapProlog() {
     writeStringVector("sourcesContent", wasm->debugInfoSourcesContent);
   }
 
-  // TODO: This field is optional; maybe we should omit if it's empty.
-  // TODO: Binaryen actually does not correctly preserve symbol names when it
-  // rewrites the mappings. We should maybe just drop them, or else handle
-  // them correctly.
   writeStringVector("names", wasm->debugInfoSymbolNames);
 
   *sourceMap << "\"mappings\":\"";

test/lit/source-map-names.wast

@@ -0,0 +1,32 @@
+;; RUN: wasm-opt %s --remove-unused-module-elements -o %t.wasm -osm %t.map
+;; RUN: wasm-dis %t.wasm --source-map %t.map | filecheck %s --check-prefix OUT-WAST
+;; RUN: cat %t.map | filecheck %s --check-prefix OUT-MAP
+
+;; After --remove-unused-module-elements, the output source map's 'names' field
+;; should NOT contain 'unused'
+
+;; OUT-MAP: "names":["used","used2"]
+
+(module
+  (export "used" (func $used))
+  (export "used2" (func $used2))
+
+  (func $unused
+    ;;@ src.cpp:1:1:unused
+    (nop)
+  )
+
+  (func $used
+    ;; OUT-WAST:      ;;@ src.cpp:2:1:used
+    ;; OUT-WAST-NEXT: (nop)
+    ;;@ src.cpp:2:1:used
+    (nop)
+  )
+
+  (func $used2
+    ;; OUT-WAST:      ;;@ src.cpp:3:1:used
+    ;; OUT-WAST-NEXT: (nop)
+    ;;@ src.cpp:3:1:used2
+    (nop)
+  )
+)