2007-08-14 19:41:51

by Adam Litke

Subject: [PATCH] Fix VM_FAULT flags conversion for hugetlb

It seems a simple mistake was made when converting follow_hugetlb_page()
over to the VM_FAULT flags bitmask stuff:
(commit 83c54070ee1a2d05c89793884bea1a03f2851ed4).

By using the wrong bitmask, hugetlb_fault() failures are not being
recognized. This results in an infinite loop whenever
follow_hugetlb_page is involved in a failed fault.

Signed-off-by: Adam Litke <[email protected]>

diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index d7ca59d..de4cf45 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -643,7 +643,7 @@ int follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
spin_unlock(&mm->page_table_lock);
ret = hugetlb_fault(mm, vma, vaddr, 0);
spin_lock(&mm->page_table_lock);
- if (!(ret & VM_FAULT_MAJOR))
+ if (!(ret & VM_FAULT_ERROR))
continue;

remainder = 0;

--
Adam Litke - (agl at us.ibm.com)
IBM Linux Technology Center

2007-08-16 02:51:22

by Nick Piggin

Subject: Re: [PATCH] Fix VM_FAULT flags conversion for hugetlb

On Tue, Aug 14, 2007 at 02:41:21PM -0500, Adam Litke wrote:
> It seems a simple mistake was made when converting follow_hugetlb_page()
> over to the VM_FAULT flags bitmask stuff:
> (commit 83c54070ee1a2d05c89793884bea1a03f2851ed4).
>
> By using the wrong bitmask, hugetlb_fault() failures are not being
> recognized. This results in an infinite loop whenever
> follow_hugetlb_page is involved in a failed fault.
>
> Signed-off-by: Adam Litke <[email protected]>

Thanks Adam. Looks good.

>
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index d7ca59d..de4cf45 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -643,7 +643,7 @@ int follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
> spin_unlock(&mm->page_table_lock);
> ret = hugetlb_fault(mm, vma, vaddr, 0);
> spin_lock(&mm->page_table_lock);
> - if (!(ret & VM_FAULT_MAJOR))
> + if (!(ret & VM_FAULT_ERROR))
> continue;
>
> remainder = 0;
>
> --
> Adam Litke - (agl at us.ibm.com)
> IBM Linux Technology Center