This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
Re: [RFA] deleting breakpoints inside of 'commands' [Repost]
Andrew Cagney wrote:
>
> >> Is it worth the effort? Is this duplication costly
> >> compared to everything else already being done by
> >> bpstat_do_actions? Or am I worrying over nothing?
>
> I think this is in the noise. GDB has performance problems with very
> large symbol files, it doesn't have problems with 3 line breakpoint scripts.
I know GDB has performance problems with symbols, but I do not
know that it doesn't have performance problems with executing
command lists. I know that when I used to work on the XRAY
debugger, macro performance was a really big issue, whereas
no one seems to have talked about it much in GDB...
> > I share your concerns. And I see no reason why this should be allowed
> > --
> > the script can always "disable" its own breakpoint with the same effect
> > for all practical purposes.
> >
> > A patch adding a "cannot delete self" error message would be nice.
>
> I would really rather not see GDB introduce, undocumented, edge
> conditions like this. I think the patch Don submitted had the very nice
> effect of eliminating the need for such a special case.
Obviously it would be bad for it to be undocumented.
But I do not agree that the restriction:
a breakpoint command set cannot delete itself
is particularly ugly. Obviously it would be (at least a little)
better to _not_ have such a restriction, but there is always
a cost/benefit analysis to these questions.