All of
 help / color / mirror / Atom feed
From: Michal Hocko <>
Cc: Andrew Morton <>,
	Tetsuo Handa <>,
	Oleg Nesterov <>,
	David Rientjes <>,
	Vladimir Davydov <>,
	Michal Hocko <>
Subject: [PATCH v2 8/9] mm: make sure that kthreads will not refault oom reaped memory
Date: Thu, 25 Aug 2016 12:03:13 +0200	[thread overview]
Message-ID: <> (raw)
In-Reply-To: <>

From: Michal Hocko <>

There are only few use_mm() users in the kernel right now. Most
of them write to the target memory but vhost driver relies on
copy_from_user/get_user from a kernel thread context. This makes it
impossible to reap the memory of an oom victim which shares the mm with
the vhost kernel thread because it could see a zero page unexpectedly
and theoretically make an incorrect decision visible outside of the
killed task context.

To quote Michael S. Tsirkin:
: Getting an error from __get_user and friends is handled gracefully.
: Getting zero instead of a real value will cause userspace
: memory corruption.

The vhost kernel thread is bound to an open fd of the vhost device which
is not tight to the mm owner life cycle in general. The device fd can be
inherited or passed over to another process which means that we really
have to be careful about unexpected memory corruption because unlike for
normal oom victims the result will be visible outside of the oom victim

Make sure that no kthread context (users of use_mm) can ever see
corrupted data because of the oom reaper and hook into the page fault
path by checking MMF_UNSTABLE mm flag. __oom_reap_task_mm will set the
flag before it starts unmapping the address space while the flag is
checked after the page fault has been handled. If the flag is set
then SIGBUS is triggered so any g-u-p user will get a error code.

Regular tasks do not need this protection because all which share the mm
are killed when the mm is reaped and so the corruption will not outlive

This patch shouldn't have any visible effect at this moment because the
OOM killer doesn't invoke oom reaper for tasks with mm shared with
kthreads yet.

Acked-by: "Michael S. Tsirkin" <>
Signed-off-by: Michal Hocko <>
 include/linux/sched.h |  1 +
 mm/memory.c           | 13 +++++++++++++
 mm/oom_kill.c         |  8 ++++++++
 3 files changed, 22 insertions(+)

diff --git a/include/linux/sched.h b/include/linux/sched.h
index eda579f3283a..63acaf9cc51c 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -522,6 +522,7 @@ static inline int get_dumpable(struct mm_struct *mm)
 #define MMF_HAS_UPROBES		19	/* has uprobes */
 #define MMF_RECALC_UPROBES	20	/* MMF_HAS_UPROBES can be wrong */
 #define MMF_OOM_SKIP		21	/* mm is of no interest for the OOM killer */
+#define MMF_UNSTABLE		22	/* mm is unstable for copy_from_user */
diff --git a/mm/memory.c b/mm/memory.c
index 83be99d9d8a1..020226b4114b 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -3656,6 +3656,19 @@ int handle_mm_fault(struct vm_area_struct *vma, unsigned long address,
+	/*
+	 * This mm has been already reaped by the oom reaper and so the
+	 * refault cannot be trusted in general. Anonymous refaults would
+	 * lose data and give a zero page instead e.g. This is especially
+	 * problem for use_mm() because regular tasks will just die and
+	 * the corrupted data will not be visible anywhere while kthread
+	 * will outlive the oom victim and potentially propagate the date
+	 * further.
+	 */
+	if (unlikely((current->flags & PF_KTHREAD) && !(ret & VM_FAULT_ERROR)
+				&& test_bit(MMF_UNSTABLE, &vma->vm_mm->flags)))
 	return ret;
diff --git a/mm/oom_kill.c b/mm/oom_kill.c
index 3b990544db6d..5a3ba96c8338 100644
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -495,6 +495,14 @@ static bool __oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm)
 		goto unlock_oom;
+	/*
+	 * Tell all users of get_user/copy_from_user etc... that the content
+	 * is no longer stable. No barriers really needed because unmapping
+	 * should imply barriers already and the reader would hit a page fault
+	 * if it stumbled over a reaped memory.
+	 */
+	set_bit(MMF_UNSTABLE, &mm->flags);
 	tlb_gather_mmu(&tlb, mm, 0, -1);
 	for (vma = mm->mmap ; vma; vma = vma->vm_next) {
 		if (is_vm_hugetlb_page(vma))

To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to  For more info on Linux MM,
see: .
Don't email: <a href=mailto:""> </a>

  parent reply	other threads:[~2016-08-25 10:03 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-08-25 10:03 [PATCH v2 0/10] fortify oom killer even more Michal Hocko
2016-08-25 10:03 ` [PATCH v2 1/9] mm,oom_reaper: Reduce find_lock_task_mm() usage Michal Hocko
2016-08-25 10:03 ` [PATCH v2 2/9] mm,oom_reaper: Do not attempt to reap a task twice Michal Hocko
2016-08-25 10:03 ` [PATCH v2 3/9] oom: keep mm of the killed task available Michal Hocko
2016-08-25 10:03 ` [PATCH v2 4/9] kernel, oom: fix potential pgd_lock deadlock from __mmdrop Michal Hocko
2016-08-25 10:03 ` [PATCH v2 5/9] mm, oom: get rid of signal_struct::oom_victims Michal Hocko
2016-08-25 10:03 ` [PATCH v2 6/9] oom, suspend: fix oom_killer_disable vs. pm suspend properly Michal Hocko
2016-08-25 10:03 ` [PATCH v2 7/9] mm, oom: enforce exit_oom_victim on current task Michal Hocko
2016-08-25 10:03 ` Michal Hocko [this message]
2016-08-25 10:03 ` [PATCH v2 9/9] oom, oom_reaper: allow to reap mm shared by the kthreads Michal Hocko

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \ \ \ \ \ \ \ \ \ \
    --subject='Re: [PATCH v2 8/9] mm: make sure that kthreads will not refault oom reaped memory' \

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.