Re: [PATCH 2/2] Make a "gdb" dprintf that tells us load addresses.

[Glenn Washburn]

On Thu, 14 Oct 2021 13:09:21 -0400
Robbie Harwood <rharwood@redhat.com> wrote:

> Glenn Washburn <development@efficientek.com> writes:
> 
> > I don't agree with suppressing file:line for debug output. I have some
> > changes in the pipeline which will allow selectively disabling
> > conditions when also using "debug=all". Suppressing file:line makes it
> > harder to figure out where the debug output is coming from to find the
> > condition to disable it.
> 
> Grep for add-symbol-file or /usr/lib/debug etc. will get there.

I'm not following. Are you suggesting grepping for "add-symbol-file" in
the source? If so, I don't find that an acceptable answer to the
sysadmin trying to figure out hwo to disable that log message.

> > I think solution that we both might find agree able is to prepend "\n"
> > to your format string. That way "add-symbol-file" will still start on
> > a new-line, which seems to be what you're after.
> 
> During a boot with this enabled, you don't get just one line - you get
> one for each module (plus the initial one for kernel.exec).  So if
> nothing intersperses, you can do `set confirm off`, then copy/paste the
> block.

Yes I understood this is per module, and I'm also confused here as to
why this is your response because it doesn't address what its supposedly
responding to. You're just talking about how to use the functionality
that this patch provides. With the change that I suggest the actions in
your response would be the same.

> Be well,
> --Robbie

Glenn