[lkml]   [2011]   [Apr]   [1]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
Patch in this message
SubjectRe: [PATCH] thp+memcg-numa: fix BUG at include/linux/mm.h:370!
On Mon, Mar 14, 2011 at 08:58:23PM +0100, Johannes Weiner wrote:
> We don't care about the vma. It's all about assigning the physical
> page to the memcg that mm->owner belongs to.
> It would be the first callsite not holding the mmap_sem, but that is
> only because all existing sites are fault handlers that don't drop the
> lock for other reasons.

I was afraid it'd be the first callsite this is why I wasn't excited
to push it in 2.6.38, but Linus's right and we should micro-optimize
it for 2.6.39.

> I am not aware of anything that would rely on the lock in there, or
> would not deserve to break if it did.

Thanks for double checking.

What about this? (only problem is the thp-vmstat patch in -mm then
reject, maybe I should rediff it against -mm instead, as you wish)

Subject: thp: optimize memcg charge in khugepaged

From: Andrea Arcangeli <>

We don't need to hold the mmmap_sem through mem_cgroup_newpage_charge(), the
mmap_sem is only hold for keeping the vma stable and we don't need the vma
stable anymore after we return from alloc_hugepage_vma().

Signed-off-by: Andrea Arcangeli <>

diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 0a619e0..c61d9ad 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1762,12 +1762,9 @@ static void collapse_huge_page(struct mm_struct *mm,

+ up_read(&mm->mmap_sem);
new_page = *hpage;
- if (unlikely(mem_cgroup_newpage_charge(new_page, mm, GFP_KERNEL))) {
- up_read(&mm->mmap_sem);
- return;
- }
@@ -1782,20 +1779,20 @@ static void collapse_huge_page(struct mm_struct *mm,
new_page = alloc_hugepage_vma(khugepaged_defrag(), vma, address,
node, __GFP_OTHER_NODE);
+ /* after allocating the hugepage upgrade to mmap_sem write mode */
+ up_read(&mm->mmap_sem);
if (unlikely(!new_page)) {
- up_read(&mm->mmap_sem);
*hpage = ERR_PTR(-ENOMEM);
if (unlikely(mem_cgroup_newpage_charge(new_page, mm, GFP_KERNEL))) {
- up_read(&mm->mmap_sem);
- /* after allocating the hugepage upgrade to mmap_sem write mode */
- up_read(&mm->mmap_sem);

* Prevent all access to pagetables with the exception of

 \ /
  Last update: 2011-04-01 23:25    [W:0.050 / U:3.424 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site