[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Wed, 02 Feb 2011 09:28:13 -0500
From: Steven Rostedt <rostedt@...dmis.org>
To: David Laight <David.Laight@...LAB.COM>
Cc: Ian Munsie <imunsie@....ibm.com>,
Andreas Dilger <adilger@...ger.ca>,
Andreas Schwab <schwab@...ux-m68k.org>,
Dave Kleikamp <shaggy@...ux.vnet.ibm.com>,
"open list:DOCUMENTATION" <linux-doc@...r.kernel.org>,
Namhyung Kim <namhyung@...il.com>,
Jiri Kosina <jkosina@...e.cz>, Jason Baron <jbaron@...hat.com>,
linux-kernel@...r.kernel.org, Alexander Graf <agraf@...e.de>,
Ingo Molnar <mingo@...hat.com>,
Paul Mackerras <paulus@...ba.org>,
KOSAKI Motohiro <kosaki.motohiro@...fujitsu.com>,
Frederic Weisbecker <fweisbec@...il.com>,
Scott Wood <scottwood@...escale.com>,
Nathan Lynch <ntl@...ox.com>,
Andrew Morton <akpm@...ux-foundation.org>,
David Gibson <dwg@....ibm.com>,
linuxppc-dev <linuxppc-dev@...ts.ozlabs.org>,
Avantika Mathur <mathur@...ux.vnet.ibm.com>
Subject: RE: [PATCH 4/6] ftrace syscalls: Allow arch specific syscallsymbol
matching
On Wed, 2011-02-02 at 14:15 +0000, David Laight wrote:
> > +#define arch_syscall_match_sym_name(sym, name) !strcmp(sym + 3, name
> + 3)
>
> Whenever you use a #define macro arg, you should enclose it in ().
> About the only time you don't need to is when it is being
> passed as an argument to another function
> (ie when it's use is also ',' separated).
>
> So the above ought to be:
> #define arch_syscall_match_sym_name(sym, name) (!strcmp((sym) + 3,
> (name) + 3))
I would have mentioned this if I wanted it to stay a macro ;)
>
> Whether an inline function is better or worse is much more subtle!
> For instance I've used:
> asm volatile ( "# line " STR(__LINE__) :: )
> to stop gcc merging the tails of conditionals.
> Useful when the conditional is at the end of a loop (etc),
> it might increase code size slightly, but removes a branch.
>
> If I put one of those in an 'inline' function separate copies
> of the function end up sharing code.
> With a #define __LINE__ differs so they don't.
>
> (I had some code to get below 190 clocks, these changes
> were significant!)
For what you were doing, this may have helped. But the code in question
is the "default" version of the function. I much more prefer it to be a
static inline. The issues you experience could change from gcc to gcc.
But static inlined functions are much cleaner and easier to read than
macros.
Using a macro for this purpose is just too messy.
Again, look at include/trace/ftrace.h. If I'm saying using a macro is
ugly, then don't use it! Listen to me, because I'm Mr. Ugly Macro Man.
-- Steve
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists