[patch] Fix the 2012-01-26 regression by la_get_symbol_name_match_p [Re: Crash regression gdb.cp/no-dmgl-verbose.exp]

[Jan Kratochvil <jan dot kratochvil at redhat dot com>]

On Fri, 27 Jan 2012 06:15:09 +0100, Joel Brobecker wrote:
> What if you put strcmp_iw_ordered back in the following hunk?

No change.

It has started to expand all the symtabs now, it is nice it at least found the
bug
	PR symtab/13627 multiple .debug_types per objfile are not supported

but it should be fixed.  I will check in the patch below, it seems all the
functions return strcmp-like result.

No regressions on {x86_64,x86_64-m32,i686}-fedorarawhide-linux-gnu.

Your patch had:
## -328,9 +339,9 @@ static int
 iterate_name_matcher (const struct language_defn *language,
                      const char *name, void *d)
 {
-  const char **dname = d;
+  const struct symbol_matcher_data *data = d;

-  if (language->la_symbol_name_compare (name, *dname) == 0)
+  if (data->symbol_name_match_p (name, data->lookup_name))
     return 1;
   return 0;
 }


Thanks,
Jan


gdb/
2012-01-27  Jan Kratochvil  <jan.kratochvil@redhat.com>

	Fix the 2012-01-26 regression by la_get_symbol_name_match_p.
	* linespec.c (iterate_name_matcher): Negate the SYMBOL_NAME_MATCH_P
	result.

--- a/gdb/linespec.c
+++ b/gdb/linespec.c
@@ -340,7 +340,7 @@ iterate_name_matcher (const char *name, void *d)
 {
   const struct symbol_matcher_data *data = d;
 
-  if (data->symbol_name_match_p (name, data->lookup_name))
+  if (data->symbol_name_match_p (name, data->lookup_name) == 0)
     return 1;
   return 0;
 }