[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [patch] remove hugetlb syscalls
- To: "'William Lee Irwin III'" <wli@holomorphy.com>,Rohit Seth <rseth@unix-os.sc.intel.com>
- Subject: RE: [patch] remove hugetlb syscalls
- From: "Seth, Rohit" <rohit.seth@intel.com>
- Date: Thu, 14 Nov 2002 14:12:06 -0800
- Cc: Benjamin LaHaise <bcrl@redhat.com>,"Seth, Rohit" <rohit.seth@intel.com>, dada1 <dada1@cosmosbay.com>,Christoph Hellwig <hch@infradead.org>,Rik van Riel <riel@conectiva.com.br>,Andrew Morton <akpm@digeo.com>, linux-mm@kvack.org,linux-kernel@vger.kernel.org
- Fake-Sender: owner-linux-mm@kvack.org
- Original-Recipient: rfc822;linux-mm-archive@humbolt.geo.uu.nl
- Sender: Rik van Riel <riel@nl.linux.org>
> -----Original Message-----
> From: William Lee Irwin III [mailto:wli@holomorphy.com]
> >
> On Thu, Nov 14, 2002 at 12:11:32PM -0800, Rohit Seth wrote:
> > This is not the problem with MAP_SHARED. It is the lack of (arch
> > specific) hugepage aligned function support in the kernel.
> You can use
> > the mmap on hugetlbfs using only MAP_FIXED with properly aligned
> > addresses (but then this also is only a hint to kernel).
> With addr ==
> > NULL in mmap, the function is bound to fail almost all the times.
>
> There's very little standing in the way of automatic
> placement. If in your opinion it should be implemented, I'll
> add that feature today.
>
mmap with addr==NULL is in my opinion a very useful thing.
> IIRC you mentioned you would like to export the arch-specific
> hugepage-aligned vma placement functions; once these are
> available, it should be trivial to reuse them.
>
>
I will export those functions from arch specific trees.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/