[llvm] Fix the dsymutil heuristic for excluding system interfaces. (PR #93745)

[Adrian Prantl via llvm-commits]

https://github.com/adrian-prantl updated https://github.com/llvm/llvm-project/pull/93745

>From ad3dce49a5880811686dd5118c9ac0fd41948172 Mon Sep 17 00:00:00 2001
From: Adrian Prantl <aprantl at apple.com>
Date: Wed, 29 May 2024 16:08:01 -0700
Subject: [PATCH] Fix the dsymutil heuristic for excluding system interfaces.

The function was meant to find the Developer/ dir, but it found a
Developer directory nested deep inside the top-level Developer dir.

The new implementation rejects everything in Xcode.app/Developer in
broad strokes.

rdar://128571037
---
 llvm/include/llvm/DWARFLinker/Utils.h         | 35 ++++++++++++++-----
 llvm/lib/DWARFLinker/Classic/DWARFLinker.cpp  |  4 +--
 .../Parallel/DWARFLinkerCompileUnit.cpp       |  4 +--
 .../DWARFLinkerParallel/DWARFLinkerTest.cpp   | 22 ++++++++++++
 4 files changed, 52 insertions(+), 13 deletions(-)

diff --git a/llvm/include/llvm/DWARFLinker/Utils.h b/llvm/include/llvm/DWARFLinker/Utils.h
index 23e59c967011a..693bb45abd840 100644
--- a/llvm/include/llvm/DWARFLinker/Utils.h
+++ b/llvm/include/llvm/DWARFLinker/Utils.h
@@ -37,17 +37,34 @@ inline Error finiteLoop(function_ref<Expected<bool>()> Iteration,
 }
 
 /// Make a best effort to guess the
-/// Xcode.app/Contents/Developer/Toolchains/ path from an SDK path.
-inline SmallString<128> guessToolchainBaseDir(StringRef SysRoot) {
+/// Xcode.app/Contents/Developer path from an SDK path.
+inline StringRef guessDeveloperDir(StringRef SysRoot) {
   SmallString<128> Result;
   // Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk
-  StringRef Base = sys::path::parent_path(SysRoot);
-  if (sys::path::filename(Base) != "SDKs")
-    return Result;
-  Base = sys::path::parent_path(Base);
-  Result = Base;
-  Result += "/Toolchains";
-  return Result;
+  auto it = sys::path::rbegin(SysRoot);
+  if (!it->ends_with(".sdk"))
+    return {};
+  ++it;
+  if (*it != "SDKs")
+    return {};
+  auto developerEnd = it;
+  ++it;
+  while (it != sys::path::rend(SysRoot)) {
+    if (*it != "Developer")
+      return {};
+    ++it;
+    if (*it == "Contents")
+      return StringRef(SysRoot.data(),
+                       developerEnd - sys::path::rend(SysRoot) - 1);
+    if (!it->ends_with(".platform"))
+      return {};
+    ++it;
+    if (*it != "Platforms")
+      return {};
+    developerEnd = it;
+    ++it;
+  }
+  return {};
 }
 
 inline bool isPathAbsoluteOnWindowsOrPosix(const Twine &Path) {
diff --git a/llvm/lib/DWARFLinker/Classic/DWARFLinker.cpp b/llvm/lib/DWARFLinker/Classic/DWARFLinker.cpp
index 3149d9b1d6624..84bbdfb5c3947 100644
--- a/llvm/lib/DWARFLinker/Classic/DWARFLinker.cpp
+++ b/llvm/lib/DWARFLinker/Classic/DWARFLinker.cpp
@@ -201,8 +201,8 @@ static void analyzeImportedModule(
     return;
   // Don't track interfaces that are part of the toolchain.
   // For example: Swift, _Concurrency, ...
-  SmallString<128> Toolchain = guessToolchainBaseDir(SysRoot);
-  if (!Toolchain.empty() && Path.starts_with(Toolchain))
+  StringRef DeveloperDir = guessDeveloperDir(SysRoot);
+  if (!DeveloperDir.empty() && Path.starts_with(DeveloperDir))
     return;
   std::optional<const char *> Name =
       dwarf::toString(DIE.find(dwarf::DW_AT_name));
diff --git a/llvm/lib/DWARFLinker/Parallel/DWARFLinkerCompileUnit.cpp b/llvm/lib/DWARFLinker/Parallel/DWARFLinkerCompileUnit.cpp
index 8c52040fdbc92..b50ea9ab49c1b 100644
--- a/llvm/lib/DWARFLinker/Parallel/DWARFLinkerCompileUnit.cpp
+++ b/llvm/lib/DWARFLinker/Parallel/DWARFLinkerCompileUnit.cpp
@@ -270,8 +270,8 @@ void CompileUnit::analyzeImportedModule(const DWARFDebugInfoEntry *DieEntry) {
     return;
   // Don't track interfaces that are part of the toolchain.
   // For example: Swift, _Concurrency, ...
-  SmallString<128> Toolchain = guessToolchainBaseDir(SysRoot);
-  if (!Toolchain.empty() && Path.starts_with(Toolchain))
+  StringRef DeveloperDir = guessDeveloperDir(SysRoot);
+  if (!DeveloperDir.empty() && Path.starts_with(DeveloperDir))
     return;
   if (std::optional<DWARFFormValue> Val = find(DieEntry, dwarf::DW_AT_name)) {
     Expected<const char *> Name = Val->getAsCString();
diff --git a/llvm/unittests/DWARFLinkerParallel/DWARFLinkerTest.cpp b/llvm/unittests/DWARFLinkerParallel/DWARFLinkerTest.cpp
index 0eb83d4b4fc9b..6439bf2cbc5f8 100644
--- a/llvm/unittests/DWARFLinkerParallel/DWARFLinkerTest.cpp
+++ b/llvm/unittests/DWARFLinkerParallel/DWARFLinkerTest.cpp
@@ -5,3 +5,25 @@
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
+
+#include "llvm/DWARFLinker/Utils.h"
+#include "gtest/gtest.h"
+
+using namespace llvm;
+using namespace dwarf_linker;
+
+#define DEVELOPER_DIR "/Applications/Xcode.app/Contents/Developer"
+
+namespace {
+
+TEST(DWARFLinker, PathTest) {
+  EXPECT_EQ(guessDeveloperDir("/Foo"), "");
+  EXPECT_EQ(guessDeveloperDir("Foo.sdk"), "");
+  EXPECT_EQ(guessDeveloperDir(
+                DEVELOPER_DIR
+                "/Platforms/MacOSX.platform/Developer/SDKs/MacOSX14.4.sdk"),
+            DEVELOPER_DIR);
+  EXPECT_EQ(guessDeveloperDir(DEVELOPER_DIR "/SDKs/MacOSX.sdk"), DEVELOPER_DIR);
+}
+
+} // anonymous namespace