Re: [PATCH 3/6] INIT: Limit the number of per cpu calibration bootupmessages

[Mike Travis]

David Miller wrote:
> From: Cyrill Gorcunov <gorcunov@xxxxxxxxx>
> Date: Tue, 17 Nov 2009 00:50:52 +0300
>
> > On Mon, Nov 16, 2009 at 01:46:07PM -0800, H. Peter Anvin wrote:
> > > On 11/16/2009 01:43 PM, Cyrill Gorcunov wrote:
> > > > It seems we have one
> > > >
> > > > 	arch/x86/kernel/setup.c:125:unsigned int boot_cpu_id __read_mostly;
> > > >
> > > > 	-- Cyrill
> > > We probably should make it an inline function so that if other arches
> > > want to define it to be a constant or some other kind of special thing
> > > they can.
> > IA-64 and SPARC already has this variable. But boot_cpu_id() as an
> > inline function seem to be more natural/portable ineed.
>
> Only 32-bit SPARC actually has it.  On sparc64 we have no reason to
> remember which processor was the boot cpu, and remembering it merely
> for the sake of only printing out the bogomips message once seems a
> bit excessive?
>
> How about:
>
> 	static bool printed;
>
> 	if (!printed) {
> 		printk(...);
> 		printed = true;
> 	}
>
> Or, alternatively, use an atomic_t instead of a bool if you think
> races matter this early in the boot process.

Yeah, I was thinking along these same lines.  Thanks for the feedback!

Mike
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/