Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1643540

[PATCH 2/2] dax: Fix race between colliding PMD & PTE entries

From Ross Zwisler <ross.zwisler@linux.intel.com>
Newsgroups linux.kernel
Subject [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries
Date 2017-05-17 19:30 +0200
Message-ID <tIaVQ-5Z9-9@gated-at.bofh.it> (permalink)
References <tIaM9-5Vz-3@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw

We currently have two related PMD vs PTE races in the DAX code.  These can
both be easily triggered by having two threads reading and writing
simultaneously to the same private mapping, with the key being that private
mapping reads can be handled with PMDs but private mapping writes are
always handled with PTEs so that we can COW.

Here is the first race:

CPU 0					CPU 1

(private mapping write)
__handle_mm_fault()
  create_huge_pmd() - FALLBACK
  handle_pte_fault()
    passes check for pmd_devmap()

					(private mapping read)
					__handle_mm_fault()
					  create_huge_pmd()
					    dax_iomap_pmd_fault() inserts PMD

    dax_iomap_pte_fault() does a PTE fault, but we already have a DAX PMD
    			  installed in our page tables at this spot.

Here's the second race:

CPU 0					CPU 1

(private mapping write)
__handle_mm_fault()
  create_huge_pmd() - FALLBACK
					(private mapping read)
					__handle_mm_fault()
					  passes check for pmd_none()
					  create_huge_pmd()

  handle_pte_fault()
    dax_iomap_pte_fault() inserts PTE
					    dax_iomap_pmd_fault() inserts PMD,
					       but we already have a PTE at
					       this spot.

The core of the issue is that while there is isolation between faults to
the same range in the DAX fault handlers via our DAX entry locking, there
is no isolation between faults in the code in mm/memory.c.  This means for
instance that this code in __handle_mm_fault() can run:

	if (pmd_none(*vmf.pmd) && transparent_hugepage_enabled(vma)) {
		ret = create_huge_pmd(&vmf);

But by the time we actually get to run the fault handler called by
create_huge_pmd(), the PMD is no longer pmd_none() because a racing PTE
fault has installed a normal PMD here as a parent.  This is the cause of
the 2nd race.  The first race is similar - there is the following check in
handle_pte_fault():

	} else {
		/* See comment in pte_alloc_one_map() */
		if (pmd_devmap(*vmf->pmd) || pmd_trans_unstable(vmf->pmd))
			return 0;

So if a pmd_devmap() PMD (a DAX PMD) has been installed at vmf->pmd, we
will bail and retry the fault.  This is correct, but there is nothing
preventing the PMD from being installed after this check but before we
actually get to the DAX PTE fault handlers.

In my testing these races result in the following types of errors:

 BUG: Bad rss-counter state mm:ffff8800a817d280 idx:1 val:1
 BUG: non-zero nr_ptes on freeing mm: 15

Fix this issue by having the DAX fault handlers verify that it is safe to
continue their fault after they have taken an entry lock to block other
racing faults.

Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Reported-by: Pawel Lebioda <pawel.lebioda@intel.com>
Cc: stable@vger.kernel.org

---

I've written a new xfstest for this race, which I will send in response to
this patch series.  This series has also survived an xfstest run without
any new issues.

---
 fs/dax.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/fs/dax.c b/fs/dax.c
index c22eaf1..3cc02d1 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -1155,6 +1155,15 @@ static int dax_iomap_pte_fault(struct vm_fault *vmf,
 	}
 
 	/*
+	 * It is possible, particularly with mixed reads & writes to private
+	 * mappings, that we have raced with a PMD fault that overlaps with
+	 * the PTE we need to set up.  Now that we have a locked mapping entry
+	 * we can safely unmap the huge PMD so that we can install our PTE in
+	 * our page tables.
+	 */
+	split_huge_pmd(vmf->vma, vmf->pmd, vmf->address);
+
+	/*
 	 * Note that we don't bother to use iomap_apply here: DAX required
 	 * the file system block size to be equal the page size, which means
 	 * that we never have to deal with more than a single extent here.
@@ -1398,6 +1407,15 @@ static int dax_iomap_pmd_fault(struct vm_fault *vmf,
 		goto fallback;
 
 	/*
+	 * It is possible, particularly with mixed reads & writes to private
+	 * mappings, that we have raced with a PTE fault that overlaps with
+	 * the PMD we need to set up.  If so we just fall back to a PTE fault
+	 * ourselves.
+	 */
+	if (!pmd_none(*vmf->pmd))
+		goto unlock_entry;
+
+	/*
 	 * Note that we don't use iomap_apply here.  We aren't doing I/O, only
 	 * setting up a mapping, so really we're using iomap_begin() as a way
 	 * to look up our filesystem block.
-- 
2.9.4

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH 1/2] mm: avoid spurious 'bad pmd' warning messages Ross Zwisler <ross.zwisler@linux.intel.com> - 2017-05-17 19:20 +0200
  [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries Ross Zwisler <ross.zwisler@linux.intel.com> - 2017-05-17 19:30 +0200
    [PATCH] generic: add regression test for DAX PTE/PMD races Ross Zwisler <ross.zwisler@linux.intel.com> - 2017-05-17 19:30 +0200
    Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries Jan Kara <jack@suse.cz> - 2017-05-18 10:00 +0200
      Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries Ross Zwisler <ross.zwisler@linux.intel.com> - 2017-05-18 23:40 +0200
        Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries Jan Kara <jack@suse.cz> - 2017-05-22 16:40 +0200
          Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries Ross Zwisler <ross.zwisler@linux.intel.com> - 2017-05-22 21:50 +0200
    Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries Jan Kara <jack@suse.cz> - 2017-05-22 16:50 +0200
      Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries Ross Zwisler <ross.zwisler@linux.intel.com> - 2017-05-22 21:50 +0200
  Re: [PATCH 1/2] mm: avoid spurious 'bad pmd' warning messages Dave Hansen <dave.hansen@intel.com> - 2017-05-17 19:40 +0200
    Re: [PATCH 1/2] mm: avoid spurious 'bad pmd' warning messages Ross Zwisler <ross.zwisler@linux.intel.com> - 2017-05-17 20:30 +0200
  Re: [PATCH 1/2] mm: avoid spurious 'bad pmd' warning messages Jan Kara <jack@suse.cz> - 2017-05-22 16:50 +0200

csiph-web