This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] Implement way of checking if probe interface can evaluate arguments
- From: Tom Tromey <tromey at redhat dot com>
- To: Sergio Durigan Junior <sergiodj at redhat dot com>
- Cc: GDB Patches <gdb-patches at sourceware dot org>, Joel Brobecker <brobecker at adacore dot com>
- Date: Wed, 17 Jul 2013 11:05:01 -0600
- Subject: Re: [PATCH] Implement way of checking if probe interface can evaluate arguments
- References: <m3zjttesip dot fsf at redhat dot com> <20130711150047 dot GB11303 at blade dot nx> <m3k3kxc742 dot fsf at redhat dot com>
>>>>> "Sergio" == Sergio Durigan Junior <sergiodj@redhat.com> writes:
Sergio> Yes, I wanted to check only one probe.
I think the crux is this:
Sergio> + /* For SystemTap probes, we have to guarantee that the method
Sergio> + stap_is_single_operand is defined on gdbarch. If it is not, then it
Sergio> + means that argument evaluation is not implemented on this target. */
Sergio> + return gdbarch_stap_is_single_operand_p (gdbarch);
The question I have is whether this is sufficient.
I was picturing checking whether each particular probe can be parsed.
However, perhaps the above is enough -- you tell me.
Sergio> +"you will not be able to inspect probes's arguments.\n"
I think "the arguments of the probes" sounds less weird.
Sergio> +"Please, report a bug against GDB requesting for the implementation to be\n"
I would drop the comma after "Please" and write:
Please report a bug against GDB requesting a port to this target.
Also, does this warning show up if gdb tries to use the probes
automatically? I think for that case we will need a separate warning of
some kind, since the user probably doesn't even know probes are
involved.
Tom