This is the mail archive of the
gdb-patches@sourceware.cygnus.com
mailing list for the GDB project.
Re: PATCH: restore value of remote_timeout
- To: jtc at redback dot com
- Subject: Re: PATCH: restore value of remote_timeout
- From: Andrew Cagney <ac131313 at cygnus dot com>
- Date: Wed, 09 Feb 2000 14:30:50 +1100
- CC: gdb-patches at sourceware dot cygnus dot com
- Organization: Cygnus Solutions
- References: <5m3dr7rwt5.fsf@jtc.redbacknetworks.com>
"J.T. Conklin" wrote:
>
> Back in August, I discovered that the default value of remote_timeout
> had been changed from 2 to 20 because GDB's Hitachi e-7000 ICE target
> used the remote debug protocol's timeout variable instead of creating
> a new variable specific for that target, and the e-7000 needed a long
> timeout.
>
> I complained that this change was particularly annoying to folks using
> the remote protocol. Stan replied:
>
> Stan> Since this is clearly a case of Cygnus making everbody's lives
> Stan> miserable for the sake of E7000-using customers :-), I'm
> Stan> amenable to bumping the default down to a more reasonable level.
> Stan> It would be even more clever to add the default timeout to the
> Stan> target vector, but I'll leave that as an exercise for the
> Stan> energetic.
>
> Not being energetic, I submitted the enclosed patch.
It's too hot to be energetic :-) I've applied the patch as is.
Down the track it would be good to see:
o target.[hc] to get a similar treatment
to gdbarch.[hc]. It would eliminate those
undebuggable macro's and automate the task
of adding new elements to the target vector.
o remote_timeout replaced with several
more precise protocol knobs that were
presumably implemented using a member
of the target vector.
thanks,
Andrew