* [PATCH 0/1] mm/hugetlb.c: teach follow_hugetlb_page() to handle FOLL_NOWAIT
@ 2019-01-09 2:02 Andrea Arcangeli
2019-01-09 2:02 ` [PATCH 1/1] " Andrea Arcangeli
0 siblings, 1 reply; 4+ messages in thread
From: Andrea Arcangeli @ 2019-01-09 2:02 UTC (permalink / raw)
To: Andrew Morton
Cc: linux-mm, linux-kernel, Mike Rapoport, Mike Kravetz, Peter Xu,
Dr. David Alan Gilbert
Hello,
this fixes a regression that resurfaced in hugetlbfs code after we
fixed it for the core VM a few months ago. This is only reproducible
doing postcopy live migration of KVM (the only user of FOLL_NOWAIT) if
backed by hugetlbfs memory. It's unrelated to userfaultfd, but
userfaultfd reproduces it easily because it's an heavy user of
VM_FAULT_RETRY retvals.
Thanks,
Andrea
Andrea Arcangeli (1):
mm/hugetlb.c: teach follow_hugetlb_page() to handle FOLL_NOWAIT
mm/hugetlb.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH 1/1] mm/hugetlb.c: teach follow_hugetlb_page() to handle FOLL_NOWAIT
2019-01-09 2:02 [PATCH 0/1] mm/hugetlb.c: teach follow_hugetlb_page() to handle FOLL_NOWAIT Andrea Arcangeli
@ 2019-01-09 2:02 ` Andrea Arcangeli
2019-01-09 2:53 ` Mike Kravetz
2019-01-09 5:05 ` Peter Xu
0 siblings, 2 replies; 4+ messages in thread
From: Andrea Arcangeli @ 2019-01-09 2:02 UTC (permalink / raw)
To: Andrew Morton
Cc: linux-mm, linux-kernel, Mike Rapoport, Mike Kravetz, Peter Xu,
Dr. David Alan Gilbert
hugetlb needs the same fix as faultin_nopage (which was applied in
96312e61282ae3f6537a562625706498cbc75594) or KVM hangs because it
thinks the mmap_sem was already released by hugetlb_fault() if it
returned VM_FAULT_RETRY, but it wasn't in the FOLL_NOWAIT case.
Fixes: ce53053ce378 ("kvm: switch get_user_page_nowait() to get_user_pages_unlocked()")
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Tested-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Reported-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
---
mm/hugetlb.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index e37efd5d8318..b3622d7888c8 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -4301,7 +4301,8 @@ long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
break;
}
if (ret & VM_FAULT_RETRY) {
- if (nonblocking)
+ if (nonblocking &&
+ !(fault_flags & FAULT_FLAG_RETRY_NOWAIT))
*nonblocking = 0;
*nr_pages = 0;
/*
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH 1/1] mm/hugetlb.c: teach follow_hugetlb_page() to handle FOLL_NOWAIT
2019-01-09 2:02 ` [PATCH 1/1] " Andrea Arcangeli
@ 2019-01-09 2:53 ` Mike Kravetz
2019-01-09 5:05 ` Peter Xu
1 sibling, 0 replies; 4+ messages in thread
From: Mike Kravetz @ 2019-01-09 2:53 UTC (permalink / raw)
To: Andrea Arcangeli, Andrew Morton
Cc: linux-mm, linux-kernel, Mike Rapoport, Peter Xu,
Dr. David Alan Gilbert
On 1/8/19 6:02 PM, Andrea Arcangeli wrote:
> hugetlb needs the same fix as faultin_nopage (which was applied in
> 96312e61282ae3f6537a562625706498cbc75594) or KVM hangs because it
> thinks the mmap_sem was already released by hugetlb_fault() if it
> returned VM_FAULT_RETRY, but it wasn't in the FOLL_NOWAIT case.
>
> Fixes: ce53053ce378 ("kvm: switch get_user_page_nowait() to get_user_pages_unlocked()")
> Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
> Tested-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
> Reported-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Thanks for fixing this.
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
--
Mike Kravetz
> ---
> mm/hugetlb.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index e37efd5d8318..b3622d7888c8 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -4301,7 +4301,8 @@ long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
> break;
> }
> if (ret & VM_FAULT_RETRY) {
> - if (nonblocking)
> + if (nonblocking &&
> + !(fault_flags & FAULT_FLAG_RETRY_NOWAIT))
> *nonblocking = 0;
> *nr_pages = 0;
> /*
>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH 1/1] mm/hugetlb.c: teach follow_hugetlb_page() to handle FOLL_NOWAIT
2019-01-09 2:02 ` [PATCH 1/1] " Andrea Arcangeli
2019-01-09 2:53 ` Mike Kravetz
@ 2019-01-09 5:05 ` Peter Xu
1 sibling, 0 replies; 4+ messages in thread
From: Peter Xu @ 2019-01-09 5:05 UTC (permalink / raw)
To: Andrea Arcangeli
Cc: Andrew Morton, linux-mm, linux-kernel, Mike Rapoport,
Mike Kravetz, Dr. David Alan Gilbert
On Tue, Jan 08, 2019 at 09:02:03PM -0500, Andrea Arcangeli wrote:
> hugetlb needs the same fix as faultin_nopage (which was applied in
> 96312e61282ae3f6537a562625706498cbc75594) or KVM hangs because it
> thinks the mmap_sem was already released by hugetlb_fault() if it
> returned VM_FAULT_RETRY, but it wasn't in the FOLL_NOWAIT case.
>
> Fixes: ce53053ce378 ("kvm: switch get_user_page_nowait() to get_user_pages_unlocked()")
> Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
> Tested-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
> Reported-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
FWIW:
Reviewed-by: Peter Xu <peterx@redhat.com>
> ---
> mm/hugetlb.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index e37efd5d8318..b3622d7888c8 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -4301,7 +4301,8 @@ long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
> break;
> }
> if (ret & VM_FAULT_RETRY) {
> - if (nonblocking)
> + if (nonblocking &&
> + !(fault_flags & FAULT_FLAG_RETRY_NOWAIT))
> *nonblocking = 0;
> *nr_pages = 0;
> /*
Regards,
--
Peter Xu
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2019-01-09 5:05 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-01-09 2:02 [PATCH 0/1] mm/hugetlb.c: teach follow_hugetlb_page() to handle FOLL_NOWAIT Andrea Arcangeli
2019-01-09 2:02 ` [PATCH 1/1] " Andrea Arcangeli
2019-01-09 2:53 ` Mike Kravetz
2019-01-09 5:05 ` Peter Xu
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).