* Expanding macros in breakpoint conditions
@ 2013-10-21 16:53 Eli Zaretskii
2013-10-21 17:09 ` Tom Tromey
0 siblings, 1 reply; 6+ messages in thread
From: Eli Zaretskii @ 2013-10-21 16:53 UTC (permalink / raw)
To: gdb
I was tripped by GDB today when I set a conditioned breakpoint, like
this:
(gdb) break foobar.c:1234 if FOO > 24.5
where FOO is a macro that expands to some_struct.foo. Then I ran the
program, and it didn't stop at the breakpoint, although the condition
surely held at some point, and that location is in a loop the program
goes through repeatedly, many times.
Then I did
(gdb) break foobar.c:1234 if some_struct.foo > 24.5
and re-ran the program. This time, it did stop. Moreover, GDB does
know about the macro, because right after it stop, I typed this:
(gdb) p FOO
$1 = 25.001
So my question is: is GDB supposed to expand macros in breakpoint
conditions?
This was with GDB 7.6.1 on Windows debugging a MinGW program compiled
with GCC 4.7.2.
TIA
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Expanding macros in breakpoint conditions
2013-10-21 16:53 Expanding macros in breakpoint conditions Eli Zaretskii
@ 2013-10-21 17:09 ` Tom Tromey
2013-10-23 17:04 ` Eli Zaretskii
0 siblings, 1 reply; 6+ messages in thread
From: Tom Tromey @ 2013-10-21 17:09 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: gdb
>>>>> "Eli" == Eli Zaretskii <eliz@gnu.org> writes:
Eli> So my question is: is GDB supposed to expand macros in breakpoint
Eli> conditions?
Yes, it should work under the conditions you describe -- namely when the
macro information is available.
I don't think there are any known macro expansion bugs here, either.
Tom
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Expanding macros in breakpoint conditions
2013-10-21 17:09 ` Tom Tromey
@ 2013-10-23 17:04 ` Eli Zaretskii
2013-10-23 17:26 ` Tom Tromey
2013-10-23 17:33 ` Keith Seitz
0 siblings, 2 replies; 6+ messages in thread
From: Eli Zaretskii @ 2013-10-23 17:04 UTC (permalink / raw)
To: Tom Tromey; +Cc: gdb
> From: Tom Tromey <tromey@redhat.com>
> Cc: gdb@sourceware.org
> Date: Mon, 21 Oct 2013 11:09:47 -0600
>
> >>>>> "Eli" == Eli Zaretskii <eliz@gnu.org> writes:
>
> Eli> So my question is: is GDB supposed to expand macros in breakpoint
> Eli> conditions?
>
> Yes, it should work under the conditions you describe -- namely when the
> macro information is available.
>
> I don't think there are any known macro expansion bugs here, either.
Thanks. I took a closer look, and discovered that the problem was
different. It started as a cockpit error: instead of typing
(gdb) break foo.c:707
I typed
(gdb) break foo:707
(i.e., I forgot the .c extension). File foo.c defines a function foo,
so what GDB did was put the breakpoint at entry to the function.
What I don't understand is why didn't GDB issue an error message in
this case. There's no label '707' in that function. Looks like the
linespec parser silently rejected the ":707" part, without telling
me. It would be nice if it didn't do this silently.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Expanding macros in breakpoint conditions
2013-10-23 17:04 ` Eli Zaretskii
@ 2013-10-23 17:26 ` Tom Tromey
2013-10-23 17:33 ` Keith Seitz
1 sibling, 0 replies; 6+ messages in thread
From: Tom Tromey @ 2013-10-23 17:26 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: gdb
Eli> What I don't understand is why didn't GDB issue an error message in
Eli> this case. There's no label '707' in that function. Looks like the
Eli> linespec parser silently rejected the ":707" part, without telling
Eli> me. It would be nice if it didn't do this silently.
It seems like a bug to me.
Tom
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Expanding macros in breakpoint conditions
2013-10-23 17:04 ` Eli Zaretskii
2013-10-23 17:26 ` Tom Tromey
@ 2013-10-23 17:33 ` Keith Seitz
2013-10-23 18:08 ` Eli Zaretskii
1 sibling, 1 reply; 6+ messages in thread
From: Keith Seitz @ 2013-10-23 17:33 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: gdb
On 10/23/2013 10:04 AM, Eli Zaretskii wrote:
> What I don't understand is why didn't GDB issue an error message in
> this case. There's no label '707' in that function. Looks like the
> linespec parser silently rejected the ":707" part, without telling
> me. It would be nice if it didn't do this silently.
My original submission for the linespec rewrite did actually do this,
but to maintain backward compatibility, I was asked to remove it.
Perhaps a warning or complaint might be a suitable compromise until this
is properly implemented?
Keith
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Expanding macros in breakpoint conditions
2013-10-23 17:33 ` Keith Seitz
@ 2013-10-23 18:08 ` Eli Zaretskii
0 siblings, 0 replies; 6+ messages in thread
From: Eli Zaretskii @ 2013-10-23 18:08 UTC (permalink / raw)
To: Keith Seitz; +Cc: gdb
> Date: Wed, 23 Oct 2013 10:33:18 -0700
> From: Keith Seitz <keiths@redhat.com>
> CC: gdb@sourceware.org
>
> My original submission for the linespec rewrite did actually do this,
> but to maintain backward compatibility, I was asked to remove it.
Makes sense.
> Perhaps a warning or complaint might be a suitable compromise until this
> is properly implemented?
Yes, a warning is certainly good enough. It would have caught my
attention right there and then.
Thanks.
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2013-10-23 18:08 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-10-21 16:53 Expanding macros in breakpoint conditions Eli Zaretskii
2013-10-21 17:09 ` Tom Tromey
2013-10-23 17:04 ` Eli Zaretskii
2013-10-23 17:26 ` Tom Tromey
2013-10-23 17:33 ` Keith Seitz
2013-10-23 18:08 ` Eli Zaretskii
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox