[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1258118966.6167.21.camel@johannes.local>
Date: Fri, 13 Nov 2009 14:29:26 +0100
From: Johannes Berg <johannes@...solutions.net>
To: Andreas Schwab <schwab@...hat.com>
Cc: LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 2.6.32] ftrace: fix event format export
On Fri, 2009-11-13 at 14:22 +0100, Andreas Schwab wrote:
> Johannes Berg <johannes@...solutions.net> writes:
>
> > --- wireless-testing.orig/include/trace/ftrace.h 2009-11-13 13:15:21.000000000 +0100
> > +++ wireless-testing/include/trace/ftrace.h 2009-11-13 13:34:03.000000000 +0100
> > @@ -159,7 +159,7 @@
> > #undef __get_str
> >
> > #undef TP_printk
> > -#define TP_printk(fmt, args...) "%s, %s\n", #fmt, __stringify(args)
> > +#define TP_printk(fmt, args...) "\"%s\", %s\n", fmt, __stringify(args)
>
> Would using __stringify(fmt) work? If there are double quote characters
> in fmt your solution would produce output that is ambiguous.
No, for say
#define FOO_FMT "foo:%d"
and using
FOO_FMT ", %d"
that would give
""%s", %d"
instead of
"%s, %d"
However, isn't it already ambiguous that way? I fail to see why %s, #fmt
would preserve inside " properly.
johannes
Download attachment "signature.asc" of type "application/pgp-signature" (802 bytes)
Powered by blists - more mailing lists