[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20110715054512.GK3455@htj.dyndns.org>
Date: Fri, 15 Jul 2011 07:45:12 +0200
From: Tejun Heo <tj@...nel.org>
To: Yinghai Lu <yinghai@...nel.org>
Cc: mingo@...hat.com, hpa@...or.com, tglx@...utronix.de,
benh@...nel.crashing.org, davem@...emloft.net,
linux-kernel@...r.kernel.org, linux-arch@...r.kernel.org,
x86@...nel.org
Subject: Re: [PATCH 2.5/13] memblock: Use __meminit[data] instead of
__init[data]
Hello,
On Thu, Jul 14, 2011 at 02:00:51PM -0700, Yinghai Lu wrote:
> > #ifdef ARCH_DISCARD_MEMBLOCK
> > -#define __init_memblock __init
> > -#define __initdata_memblock __initdata
> > +#define __init_memblock __meminit
> > +#define __initdata_memblock __meminitdata
> > #else
> > #define __init_memblock
> > #define __initdata_memblock
>
> that is not good. It will keep memblock.reserved related data and
> functions after boot.
> and those data/func are not used after slab is ready.
Yeah, if MEMORY_HOTPLUG is enabled. It isn't too big area to begin
with and early_node_map[] and related code removal should offset it.
The problem is that refining different __init tags here is a bit
complex thanks to DISCARD_MEMBLOCK. We end up having to deal with the
combination of DISCARD_MEMBLOCK and MEMORY_HOTPLUG. I was hoping that
once we get memblock usage more consistent across different archs, we
could do away with DISCARD_MEMBLOCK.
Thank you.
--
tejun
--
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