Filter invalid encodings from Linux thread names

On Linux, a thread can only be 16 bytes (including the trailing \0).
A user sent in a test case where this causes a truncated UTF-8
sequence, causing gdbserver to create invalid XML.

I went back and forth about different ways to solve this, and in the
end decided to fix it in gdbserver, with the reason being that it
seems important to generate correct XML for the <thread> response.

I am not totally sure whether the call to setlocale could have
unplanned consequences.  This is needed, though, for nl_langinfo to
return the correct result.

Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=30618

2 files changed, 58 insertions, 2 deletions

diff --git a/gdbserver/linux-low.cc b/gdbserver/linux-low.cc
index f9001e2..eea2d8a 100644
--- a/gdbserver/linux-low.cc
+++ b/gdbserver/linux-low.cc
@@ -38,14 +38,16 @@
 #include <unistd.h>
 #include <sys/syscall.h>
 #include <sched.h>
-#include <ctype.h>
 #include <pwd.h>
 #include <sys/types.h>
 #include <dirent.h>
 #include <sys/stat.h>
 #include <sys/vfs.h>
 #include <sys/uio.h>
+#include <langinfo.h>
+#include <iconv.h>
 #include "gdbsupport/filestuff.h"
+#include "gdbsupport/gdb-safe-ctype.h"
 #include "tracepoint.h"
 #include <inttypes.h>
 #include "gdbsupport/common-inferior.h"
@@ -6999,10 +7001,63 @@ current_lwp_ptid (void)
   return ptid_of (current_thread);
 }
 
+/* A helper function that copies NAME to DEST, replacing non-printable
+   characters with '?'.  Returns DEST as a convenience.  */
+
+static const char *
+replace_non_ascii (char *dest, const char *name)
+{
+  while (*name != '\0')
+    {
+      if (!ISPRINT (*name))
+	*dest++ = '?';
+      else
+	*dest++ = *name;
+      ++name;
+    }
+  return dest;
+}
+
 const char *
 linux_process_target::thread_name (ptid_t thread)
 {
-  return linux_proc_tid_get_name (thread);
+  static char dest[100];
+
+  const char *name = linux_proc_tid_get_name (thread);
+  if (name == nullptr)
+    return nullptr;
+
+  /* Linux limits the comm file to 16 bytes (including the trailing
+     \0.  If the program or thread name is set when using a multi-byte
+     encoding, this might cause it to be truncated mid-character.  In
+     this situation, sending the truncated form in an XML <thread>
+     response will cause a parse error in gdb.  So, instead convert
+     from the locale's encoding (we can't be sure this is the correct
+     encoding, but it's as good a guess as we have) to UTF-8, but in a
+     way that ignores any encoding errors.  See PR remote/30618.  */
+  const char *cset = nl_langinfo (CODESET);
+  iconv_t handle = iconv_open ("UTF-8//IGNORE", cset);
+  if (handle == (iconv_t) -1)
+    return replace_non_ascii (dest, name);
+
+  size_t inbytes = strlen (name);
+  char *inbuf = const_cast<char *> (name);
+  size_t outbytes = sizeof (dest);
+  char *outbuf = dest;
+  size_t result = iconv (handle, &inbuf, &inbytes, &outbuf, &outbytes);
+
+  if (result == (size_t) -1)
+    {
+      if (errno == E2BIG)
+	outbuf = &dest[sizeof (dest) - 1];
+      else if ((errno == EILSEQ || errno == EINVAL)
+	       && outbuf < &dest[sizeof (dest) - 2])
+	*outbuf++ = '?';
+      *outbuf = '\0';
+    }
+
+  iconv_close (handle);
+  return *dest == '\0' ? nullptr : dest;
 }
 
 #if USE_THREAD_DB
diff --git a/gdbserver/server.cc b/gdbserver/server.cc
index a8e2356..c79b335 100644
--- a/gdbserver/server.cc
+++ b/gdbserver/server.cc
@@ -4197,6 +4197,7 @@ captured_main (int argc, char *argv[])
 int
 main (int argc, char *argv[])
 {
+  setlocale (LC_CTYPE, "");
 
   try
     {