[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <Zv2Dy7RST8Q-pzL5@gallifrey>
Date: Wed, 2 Oct 2024 17:32:59 +0000
From: "Dr. David Alan Gilbert" <linux@...blig.org>
To: Steven Rostedt <rostedt@...dmis.org>
Cc: Andy Shevchenko <andy@...nel.org>, kees@...nel.org,
akpm@...ux-foundation.org, pmladek@...e.com,
linux@...musvillemoes.dk, senozhatsky@...omium.org,
linux-hardening@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] printf: Remove unused 'bprintf'
* Steven Rostedt (rostedt@...dmis.org) wrote:
> On Wed, 2 Oct 2024 14:31:19 +0000
> "Dr. David Alan Gilbert" <linux@...blig.org> wrote:
>
> > > I am not familiar with tricks in BPF or ftrace code where this actually might
> > > be implicitly called via a macro, but brief grep gives nothing that might point
> > > to that.
> >
> > I've got an all-yes build (well, most after I took out broken stuff) booting
> > with it, and it has CONFIG_BINARY_PRINTF=y and CONFIG_FTRACE=y .
> >
> > trace_seq.c uses seq_buf_bprintf which uses bstr_printf rather than the plain
> > bprintf() that I've deleted.
> > Not sure where to dig in BPF, but I've had a fairly good grep around.
>
> I believe it is safe to delete. It looks like bprintf() was added for
> completeness, where as everything is just using the vbin_printf() directly.
> bprintf() is nothing more than a wrapper around it in case someone wanted
> to use binary prints directly. But I'm not sure there's a good use case for
> it, as all users would likely need to add some code around it for
> processing (like trace.c does).
>
> Send a v2 and I could take it for v6.13.
Thanks,
Just posted, message-id 20241002173147.210107-1-linux@...blig.org
Dave
> -- Steve
>
--
-----Open up your eyes, open up your mind, open up your code -------
/ Dr. David Alan Gilbert | Running GNU/Linux | Happy \
\ dave @ treblig.org | | In Hex /
\ _________________________|_____ http://www.treblig.org |_______/
Powered by blists - more mailing lists