[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1244796211.30512.32.camel@penberg-laptop>
Date: Fri, 12 Jun 2009 11:43:31 +0300
From: Pekka Enberg <penberg@...helsinki.fi>
To: Benjamin Herrenschmidt <benh@...nel.crashing.org>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>,
Linux Kernel list <linux-kernel@...r.kernel.org>,
linux-mm <linux-mm@...ck.org>, mingo@...e.hu,
cl@...ux-foundation.org, akpm@...ux-foundation.org, npiggin@...e.de
Subject: Re: slab: setup allocators earlier in the boot sequence
Hi Ben,
On Fri, 2009-06-12 at 18:40 +1000, Benjamin Herrenschmidt wrote:
> On Fri, 2009-06-12 at 10:45 +0300, Pekka Enberg wrote:
> > Hi Ben,
>
> > The call-sites I fixed up are all boot code AFAICT. And I like I said,
> > we can't really _miss_ any of those places, they must be checking for
> > slab_is_available() _anyway_; otherwise they have no business using
> > kmalloc(). And note: all call-sites that _unconditionally_ use
> > kmalloc(GFP_KERNEL) are safe because they worked before.
>
> No. The check for slab_is_available() can be levels higher, for example
> the vmalloc case. I'm sure I can find a whole bunch more :-) Besides
> I find the approach fragile, and it will suck for things that can be
> rightfully called also later on.
Yes, you're obviously right. I overlooked the fact that arch code have
their own special slab_is_available() heuristics (yikes!).
But are you happy with the two patches I posted so I can push them to
Linus?
Pekka
--
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