This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
Re: nrun.c patch
- To: Stephane Carrez <Stephane dot Carrez at worldnet dot fr>
- Subject: Re: nrun.c patch
- From: Doug Evans <dje at transmeta dot com>
- Date: Sun, 10 Sep 2000 10:16:38 -0700 (PDT)
- Cc: cagney at redhat dot com, gdb-patches at sourceware dot cygnus dot com
- References: <200009100427.VAA27516@casey.transmeta.com><39BBADC1.94352650@worldnet.fr>
[sorry, resent 'cus I screwed up the cc list again]
Stephane Carrez writes:
> > Having to specify the -v with the -p is wrong.
>
> Sure!
>
> The sim_info() is used for several purposes. For several simulators, it is used
> to print various information (not only profiling as in your case). Having this
> information printed each time you run the simulator is not good too. In particular
> I don't want to have such information printed when I run the gcc testsuite.
>
> Now, I understand that -v is not the appropriate option. I guess a new flag
> is necessary in the SIM_DESC to control the final call to sim_info().
> Such flag should then be set when the profiling is activated, or when some
> -print-sim-info-at-the-end nice option is specified.
>
> I think this would solve your problem and mine.
The way things were intended to work, sim_info would always be run
and each target would have sim_info print the things the user
told it to print. If the user didn't tell the simulator to print
anything, then sim_info would print nothing. Up until your patch,
it has always worked this way. ie. your sim_info is supposed to
check various flags and print what the user told it to print.