lkml.org 
[lkml]   [2012]   [Jul]   [11]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: [PATCH] mm: hugetlb: flush dcache before returning zeroed huge page to userspace
On Tue, Jul 10, 2012 at 11:42:34AM +0100, Will Deacon wrote:
> On Tue, Jul 10, 2012 at 10:45:13AM +0100, Will Deacon wrote:
> > On Tue, Jul 10, 2012 at 12:57:14AM +0100, Hugh Dickins wrote:
> > > If I start to grep the architectures for non-empty flush_dcache_page(),
> > > I soon find things in arch/arm such as v4_mc_copy_user_highpage() doing
> > > if (!test_and_set_bit(PG_dcache_clean,)) __flush_dcache_page() - where
> > > the naming suggests that I'm right, it's the architecture's responsibility
> > > to arrange whatever flushing is needed in its copy and clear page functions.

[...]

> Ok, so this is exactly the problem. The hugetlb allocator uses its own
> pool of huge pages, so free_huge_page followed by a later alloc_huge_page
> will give you something where the page flags of the compound head do not
> guarantee that PG_arch_1 is clear.

Just to confirm, the following quick hack at least results in the correct
flushing for me (on ARM):


diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index e198831..7a7c9d3 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1141,6 +1141,7 @@ static struct page *alloc_huge_page(struct vm_area_struct *vma,
}

set_page_private(page, (unsigned long)spool);
+ clear_bit(PG_arch_1, &page->flags);

vma_commit_reservation(h, vma, addr);


The question is whether we should tidy that up for the core code or get
architectures to clear the bit in arch_make_huge_pte (which also seems to
work).

Will


\
 
 \ /
  Last update: 2012-07-11 20:41    [from the cache]
©2003-2011 Jasper Spaans