From a8abc77457cda320f7577c2da7d6d592aaac0ca6 Mon Sep 17 00:00:00 2001
From: Johannes Sixt <j6t@kdbg.org>
Date: Tue, 30 Nov 2010 22:47:40 +0100
Subject: Grok gdb 7's <incomplete sequence> marker in strings.

When a string ends in an incomplete multi-byte sequence, gdb adds an
<incomplete sequence...> fragment to the end of the string. Previously,
this resulted in a parse error and incomplete variable information.

Reported by Kevin Lemay.
(cherry picked from upstream commit 03da8a5ec97c8c7b125b2bd453d2f1c3a018d477)
---
 kdbg/gdbdriver.cpp          | 14 +++++++++++++-
 kdbg/testprogs/testfile.cpp | 19 +++++++++++++++++++
 2 files changed, 32 insertions(+), 1 deletion(-)

diff --git a/kdbg/gdbdriver.cpp b/kdbg/gdbdriver.cpp
index ac86016..16efe9a 100644
--- a/kdbg/gdbdriver.cpp
+++ b/kdbg/gdbdriver.cpp
@@ -1048,7 +1048,8 @@ moreStrings:
 	}
     }
     // is the string continued?
-    if (*p == ',') {
+    if (*p == ',')
+    {
 	// look ahead for another quote
 	const char* q = p+1;
 	while (isspace(*q))
@@ -1058,6 +1059,17 @@ moreStrings:
 	    p = q;
 	    goto moreStrings;
 	}
+
+	// some strings can end in <incomplete sequence ...>
+	if (strncmp(q, "<incomplete sequence", 20) == 0)
+	{
+	    p = q+20;
+	    while (*p != '\0' && *p != '>')
+		p++;
+	    if (*p != '\0') {
+		p++;		/* skip the '>' */
+	    }
+	}
     }
     /* very long strings are followed by `...' */
     if (*p == '.' && p[1] == '.' && p[2] == '.') {
diff --git a/kdbg/testprogs/testfile.cpp b/kdbg/testprogs/testfile.cpp
index 884ddad..242c504 100644
--- a/kdbg/testprogs/testfile.cpp
+++ b/kdbg/testprogs/testfile.cpp
@@ -101,6 +101,25 @@ void strtest(const char* t)
 template<typename F>
 void templated_strtest(F f, const char* t)
 {
+	// test <incomplete sequence> in various contexts
+	struct incomplete_seq_intern {
+		int val;
+		char is[4];
+		int val2;
+	};
+	struct incomplete_seq_end {
+		int val;
+		char is[4];
+	};
+	unsigned char a[4] = {',', 020, 021, 0325};
+	incomplete_seq_intern b = { 1, {',', 020, 021, 0325}, 2 };
+	incomplete_seq_end c = { 1, {',', 020, 021, 0325} };
+	unsigned char d[30][4] = { {',', 020, 021, 0325}, };
+	for (int i = 1; i < 30; i++)
+		memcpy(d[i], d[0], 4);
+	incomplete_seq_intern ba[30] = { { 1, {',', 020, 021, 0325}, 2 } };
+	incomplete_seq_end ca[30] = { { 1, {',', 020, 021, 0325} } };
+
 	f(t);
 }
 
-- 
cgit v1.2.3