Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Keith Seitz <keiths@redhat.com>
To: Joel Brobecker <brobecker@adacore.com>
Cc: "gp >> \"gdb-patches@sourceware.org ml\"" <gdb-patches@sourceware.org>
Subject: Re: [RFA] Fix c++/16253 (tag/variable name collision)
Date: Tue, 25 Mar 2014 18:11:00 -0000	[thread overview]
Message-ID: <5331BB0D.4010606@redhat.com> (raw)
In-Reply-To: <20140324141527.GM4282@adacore.com>

[-- Attachment #1: Type: text/plain, Size: 765 bytes --]

On 03/24/2014 07:15 AM, Joel Brobecker wrote:
> This may not be directly related to your patch. I seem to have seen
> some unexplainable behavior in GDB occasionally in the recent past
> making me wonder whether there might be something fishy in the symbol
> lookup for Ada.

It is my patch that is causing the problems.

In this case, add_nonlocal_symbols is calling the qf 
map_matching_symbols method which is calling match_partial_symbol. Since 
this method now does strict matches against domain, an explicit check 
for STRUCT_DOMAIN matches must be added.

Give this amendment a shot and see how it goes. I suspect there are 
probably one or two more places where this occurs. [The easiest thing to 
do is audit any function which uses "VAR_DOMAIN".]

Keith


[-- Attachment #2: 16253-amended-1.patch --]
[-- Type: text/x-patch, Size: 1444 bytes --]

diff --git a/gdb/ada-lang.c b/gdb/ada-lang.c
index 00728bf..94c6cf8 100644
--- a/gdb/ada-lang.c
+++ b/gdb/ada-lang.c
@@ -5336,13 +5336,33 @@ add_nonlocal_symbols (struct obstack *obstackp, const char *name,
       data.objfile = objfile;
 
       if (is_wild_match)
-	objfile->sf->qf->map_matching_symbols (objfile, name, domain, global,
-					       aux_add_nonlocal_symbols, &data,
-					       wild_match, NULL);
+	{
+	  objfile->sf->qf->map_matching_symbols (objfile, name, domain, global,
+						 aux_add_nonlocal_symbols,
+						 &data, wild_match, NULL);
+	  if (domain == VAR_DOMAIN)
+	    {
+	      objfile->sf->qf->map_matching_symbols (objfile, name,
+						     STRUCT_DOMAIN, global,
+						     aux_add_nonlocal_symbols,
+						     &data, wild_match, NULL);
+	    }
+	}
       else
-	objfile->sf->qf->map_matching_symbols (objfile, name, domain, global,
-					       aux_add_nonlocal_symbols, &data,
-					       full_match, compare_names);
+	{
+	  objfile->sf->qf->map_matching_symbols (objfile, name, domain, global,
+						 aux_add_nonlocal_symbols,
+						 &data, full_match,
+						 compare_names);
+	  if (domain == VAR_DOMAIN)
+	    {
+	      objfile->sf->qf->map_matching_symbols (objfile, name,
+						     STRUCT_DOMAIN, global,
+						     aux_add_nonlocal_symbols,
+						     &data, full_match,
+						     compare_names);
+	    }
+	}
     }
 
   if (num_defns_collected (obstackp) == 0 && global && !is_wild_match)

  reply	other threads:[~2014-03-25 18:11 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-21 18:12 Keith Seitz
2014-03-24 14:15 ` Joel Brobecker
2014-03-25 18:11   ` Keith Seitz [this message]
2014-03-26 13:32     ` Joel Brobecker
2014-03-26 13:40 ` Joel Brobecker
2014-03-26 15:34   ` Keith Seitz
2014-04-10 18:54   ` Keith Seitz
2014-04-10 20:29     ` Tom Tromey
2014-05-28 17:55       ` Doug Evans
2014-04-17 19:03     ` Regression for gdb.cp/koenig.exp: p entry (c) [Re: [RFA] Fix c++/16253 (tag/variable name collision)] Jan Kratochvil

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5331BB0D.4010606@redhat.com \
    --to=keiths@redhat.com \
    --cc=brobecker@adacore.com \
    --cc=gdb-patches@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox