commit 4f56cc431b5b458ab118b553f2937e0b40dbf448
author Manuel Klimek <klimek@google.com> Wed Dec 01 11:56:51 2021 +0100
committer Copybara-Service <copybara-worker@google.com> Wed Dec 01 03:03:17 2021 -0800
tree 2491de633d826a267e04d11a0147b6b510c7315a
parent c2cf39cba94340da1e0da69dd32f725aebe601cb

Fix segfault in clang-format.

Fix bug where we'd read past the end of the tokens after merging _T
macro strings.

GitOrigin-RevId: d688b31628004ed9e109e63c2b43701f97252ff3

lib/Format/FormatTokenLexer.cpp

diff --git a/lib/Format/FormatTokenLexer.cpp b/lib/Format/FormatTokenLexer.cpp
index 972f59e..64fbd2d 100644
--- a/lib/Format/FormatTokenLexer.cpp
+++ b/lib/Format/FormatTokenLexer.cpp
@@ -752,6 +752,8 @@
   Tokens.pop_back();
   Tokens.pop_back();
   Tokens.back() = String;
+  if (FirstInLineIndex >= Tokens.size())
+    FirstInLineIndex = Tokens.size() - 1;
   return true;
 }
 

unittests/Format/FormatTest.cpp

diff --git a/unittests/Format/FormatTest.cpp b/unittests/Format/FormatTest.cpp
index 65809ab..9b6607b 100644
--- a/unittests/Format/FormatTest.cpp
+++ b/unittests/Format/FormatTest.cpp
@@ -12879,6 +12879,12 @@
             format("f(\n"
                    "\n"
                    "_T(\"XXXXXXXXXXXXXXXXXXXXXXXXXXXXXn\"));"));
+  // Regression test for accessing tokens past the end of a vector in the
+  // TokenLexer.
+  verifyNoCrash(R"(_T(
+"
+)
+)");
 }
 
 TEST_F(FormatTest, BreaksStringLiteralOperands) {