From: Karen Osmond <karen.osmond@gmail.com>
To: Tom Tromey <tromey@redhat.com>
Cc: gdb-patches@sourceware.org
Subject: Re: [patch] not able to set pending breakpoint at qualified C++ function names
Date: Wed, 03 Jun 2009 13:56:00 -0000 [thread overview]
Message-ID: <Pine.LNX.4.64.0906031434140.32368@localhost.localdomain> (raw)
In-Reply-To: <m3ws94fezf.fsf@fleche.redhat.com>
Hello!
Sorry about the excessive quoting below but both the original email and
its reply are rather old so thought full context might be helpful in this
case.
On Tue, 28 Apr 2009, Tom Tromey wrote:
> >>>>> "Karen" == Karen Osmond <karen.osmond@gmail.com> writes:
>
> Hi. I'm sorry nobody from GDB got back to you sooner.
>
> Karen> I noticed that it's not possible to set a pending breakpoint of
> Karen> the (C++) form "break X::Y::Z" (with breakpoint pending auto,
> Karen> gdb doesn't ask, and with it on, no breakpoint is added).
>
> Karen> Patch follows; I've only just met the gdb source so make of it
> Karen> what you will :)
>
> Karen> * Removed cplusplus_error helper function, as these were the only uses.
> Karen> (Though, it was adding an extra hint to the error, which perhaps should
> Karen> still happen. Erm, I must admit to not understanding the hint).
>
> I'd prefer not to remove this hint since it does help people, in my
> experience. What it means is that you can sometimes find valid
> completions by quoting the symbol.
>
> I think this could be done by just changing cplusplus_error to end
> with something like:
>
> message = ui_file_xstrdup (stream, &len);
> make_cleanup (xfree, message);
> throw_error (NOT_FOUND_ERROR, "%s", message);
Thanks for the explanation! I've changed this as you suggest.
> The rest of it seems ok to me.
>
> I assume you don't have a copyright assignment in place. I think you
> probably should; send me email off-list and I will get you started on
> that.
Emailed separately.
> A patch for gdb requires a ChangeLog entry. The GNU Coding Standards
> explain how to write one.
>
> Finally, I found a couple of formatting nits. GDB as a project is
> particular about this. The GNU standards also cover this pretty well.
>
> Karen> static struct symtabs_and_lines
> Karen> decode_compound (char **argptr, int funfirstline, char ***canonical,
> Karen> ! char *saved_arg, char *p, int* not_found_ptr)
>
> Use "int *", not "int* ".
>
> Karen> + if (not_found_ptr)
> Karen> + *not_found_ptr = 1;
>
> The indentation looks funny here; the second line should be 2 spaces
> in from the 'if'.
Sorry about this - ChangeLog added and nits addressed. Well, I hope the
indentation thing is addressed - I'm afraid this file contains a strange
mix of tabs and spaces, so I've generally copied the leading whitespace on
preceding lines in each bit.
New patch follows.
cheers,
-karen
2009-06-03 Karen Osmond <karen.osmond@gmail.com>
* linespec.c (find_method): Add new not_found_ptr parameter to
bring in line with symtab_from_filename, decode_variable.
(decode_compound): Likewise. Also propagate not_found_ptr to
find_method.
(decode_line_1): Propagage not_found_ptr to decode_compound.
(cplusplus_error): Now throws NOT_FOUND_ERROR.
Index: gdb/linespec.c
===================================================================
RCS file: /cvs/src/src/gdb/linespec.c,v
retrieving revision 1.85
diff -c -p -r1.85 linespec.c
*** gdb/linespec.c 23 May 2009 16:17:17 -0000 1.85
--- gdb/linespec.c 3 Jun 2009 12:56:34 -0000
*************** static struct symtabs_and_lines decode_c
*** 65,71 ****
int funfirstline,
char ***canonical,
char *saved_arg,
! char *p);
static struct symbol *lookup_prefix_sym (char **argptr, char *p);
--- 65,72 ----
int funfirstline,
char ***canonical,
char *saved_arg,
! char *p,
! int *not_found_ptr);
static struct symbol *lookup_prefix_sym (char **argptr, char *p);
*************** static struct symtabs_and_lines find_met
*** 74,80 ****
char *saved_arg,
char *copy,
struct type *t,
! struct symbol *sym_class);
static int collect_methods (char *copy, struct type *t,
struct symbol *sym_class,
--- 75,82 ----
char *saved_arg,
char *copy,
struct type *t,
! struct symbol *sym_class,
! int *not_found_ptr);
static int collect_methods (char *copy, struct type *t,
struct symbol *sym_class,
*************** static NORETURN void
*** 152,157 ****
--- 154,161 ----
cplusplus_error (const char *name, const char *fmt, ...)
{
struct ui_file *tmp_stream;
+ long len;
+ char *message;
tmp_stream = mem_fileopen ();
make_cleanup_ui_file_delete (tmp_stream);
*************** cplusplus_error (const char *name, const
*** 168,174 ****
("Hint: try '%s<TAB> or '%s<ESC-?>\n"
"(Note leading single quote.)"),
name, name);
! error_stream (tmp_stream);
}
/* Return the number of methods described for TYPE, including the
--- 172,181 ----
("Hint: try '%s<TAB> or '%s<ESC-?>\n"
"(Note leading single quote.)"),
name, name);
!
! message = ui_file_xstrdup (tmp_stream, &len);
! make_cleanup (xfree, message);
! throw_error (NOT_FOUND_ERROR, "%s", message);
}
/* Return the number of methods described for TYPE, including the
*************** decode_line_1 (char **argptr, int funfir
*** 765,771 ****
if (p[0] == '.' || p[1] == ':')
return decode_compound (argptr, funfirstline, canonical,
! saved_arg, p);
/* No, the first part is a filename; set file_symtab to be that file's
symtab. Also, move argptr past the filename. */
--- 772,778 ----
if (p[0] == '.' || p[1] == ':')
return decode_compound (argptr, funfirstline, canonical,
! saved_arg, p, not_found_ptr);
/* No, the first part is a filename; set file_symtab to be that file's
symtab. Also, move argptr past the filename. */
*************** decode_objc (char **argptr, int funfirst
*** 1195,1201 ****
static struct symtabs_and_lines
decode_compound (char **argptr, int funfirstline, char ***canonical,
! char *saved_arg, char *p)
{
struct symtabs_and_lines values;
char *p2;
--- 1202,1208 ----
static struct symtabs_and_lines
decode_compound (char **argptr, int funfirstline, char ***canonical,
! char *saved_arg, char *p, int *not_found_ptr)
{
struct symtabs_and_lines values;
char *p2;
*************** decode_compound (char **argptr, int funf
*** 1362,1368 ****
we'll lookup the whole string in the symbol tables. */
return find_method (funfirstline, canonical, saved_arg,
! copy, t, sym_class);
} /* End if symbol found */
--- 1369,1375 ----
we'll lookup the whole string in the symbol tables. */
return find_method (funfirstline, canonical, saved_arg,
! copy, t, sym_class, not_found_ptr);
} /* End if symbol found */
*************** decode_compound (char **argptr, int funf
*** 1385,1390 ****
--- 1392,1399 ----
/* Couldn't find any interpretation as classes/namespaces, so give
up. The quotes are important if copy is empty. */
+ if (not_found_ptr)
+ *not_found_ptr = 1;
cplusplus_error (saved_arg,
"Can't find member of namespace, class, struct, or union named \"%s\"\n",
copy);
*************** lookup_prefix_sym (char **argptr, char *
*** 1430,1436 ****
static struct symtabs_and_lines
find_method (int funfirstline, char ***canonical, char *saved_arg,
! char *copy, struct type *t, struct symbol *sym_class)
{
struct symtabs_and_lines values;
struct symbol *sym = NULL;
--- 1439,1445 ----
static struct symtabs_and_lines
find_method (int funfirstline, char ***canonical, char *saved_arg,
! char *copy, struct type *t, struct symbol *sym_class, int *not_found_ptr)
{
struct symtabs_and_lines values;
struct symbol *sym = NULL;
*************** find_method (int funfirstline, char ***c
*** 1481,1486 ****
--- 1490,1497 ----
}
else
tmp = copy;
+ if (not_found_ptr)
+ *not_found_ptr = 1;
if (tmp[0] == '~')
cplusplus_error (saved_arg,
"the class `%s' does not have destructor defined\n",
next prev parent reply other threads:[~2009-06-03 13:56 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-02-11 19:07 Karen Osmond
2009-04-29 1:03 ` Tom Tromey
2009-06-03 13:56 ` Karen Osmond [this message]
2009-06-04 21:50 ` Tom Tromey
2009-06-22 21:29 ` Karen Osmond
2009-06-25 18:09 ` Karen Osmond
2009-06-25 23:21 ` [patch] Test result regression for gdb.base/psymtab.exp [Re: [patch] not able to set pending breakpoint at qualified C++ function names] Jan Kratochvil
2009-06-26 14:50 ` Pedro Alves
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=Pine.LNX.4.64.0906031434140.32368@localhost.localdomain \
--to=karen.osmond@gmail.com \
--cc=gdb-patches@sourceware.org \
--cc=tromey@redhat.com \
/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