sdk/emulator/emulator-kernel.git
10 years agoKVM: s390: unintended fallthrough for external call
Christian Borntraeger [Wed, 3 Sep 2014 14:21:32 +0000 (16:21 +0200)]
KVM: s390: unintended fallthrough for external call

commit f346026e55f1efd3949a67ddd1dcea7c1b9a615e upstream.

We must not fallthrough if the conditions for external call are not met.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Reviewed-by: Thomas Huth <thuth@linux.vnet.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agokvm: fix potentially corrupt mmio cache
David Matlack [Mon, 18 Aug 2014 22:46:06 +0000 (15:46 -0700)]
kvm: fix potentially corrupt mmio cache

commit ee3d1570b58677885b4552bce8217fda7b226a68 upstream.

vcpu exits and memslot mutations can run concurrently as long as the
vcpu does not aquire the slots mutex. Thus it is theoretically possible
for memslots to change underneath a vcpu that is handling an exit.

If we increment the memslot generation number again after
synchronize_srcu_expedited(), vcpus can safely cache memslot generation
without maintaining a single rcu_dereference through an entire vm exit.
And much of the x86/kvm code does not maintain a single rcu_dereference
of the current memslots during each exit.

We can prevent the following case:

   vcpu (CPU 0)                             | thread (CPU 1)
--------------------------------------------+--------------------------
1  vm exit                                  |
2  srcu_read_unlock(&kvm->srcu)             |
3  decide to cache something based on       |
     old memslots                           |
4                                           | change memslots
                                            | (increments generation)
5                                           | synchronize_srcu(&kvm->srcu);
6  retrieve generation # from new memslots  |
7  tag cache with new memslot generation    |
8  srcu_read_unlock(&kvm->srcu)             |
...                                         |
   <action based on cache occurs even       |
    though the caching decision was based   |
    on the old memslots>                    |
...                                         |
   <action *continues* to occur until next  |
    memslot generation change, which may    |
    be never>                               |
                                            |

By incrementing the generation after synchronizing with kvm->srcu readers,
we ensure that the generation retrieved in (6) will become invalid soon
after (8).

Keeping the existing increment is not strictly necessary, but we
do keep it and just move it for consistency from update_memslots to
install_new_memslots.  It invalidates old cached MMIOs immediately,
instead of having to wait for the end of synchronize_srcu_expedited,
which makes the code more clearly correct in case CPU 1 is preempted
right after synchronize_srcu() returns.

To avoid halving the generation space in SPTEs, always presume that the
low bit of the generation is zero when reconstructing a generation number
out of an SPTE.  This effectively disables MMIO caching in SPTEs during
the call to synchronize_srcu_expedited.  Using the low bit this way is
somewhat like a seqcount---where the protected thing is a cache, and
instead of retrying we can simply punt if we observe the low bit to be 1.

Signed-off-by: David Matlack <dmatlack@google.com>
Reviewed-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Reviewed-by: David Matlack <dmatlack@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agokvm: x86: fix stale mmio cache bug
David Matlack [Mon, 18 Aug 2014 22:46:07 +0000 (15:46 -0700)]
kvm: x86: fix stale mmio cache bug

commit 56f17dd3fbc44adcdbc3340fe3988ddb833a47a7 upstream.

The following events can lead to an incorrect KVM_EXIT_MMIO bubbling
up to userspace:

(1) Guest accesses gpa X without a memory slot. The gfn is cached in
struct kvm_vcpu_arch (mmio_gfn). On Intel EPT-enabled hosts, KVM sets
the SPTE write-execute-noread so that future accesses cause
EPT_MISCONFIGs.

(2) Host userspace creates a memory slot via KVM_SET_USER_MEMORY_REGION
covering the page just accessed.

(3) Guest attempts to read or write to gpa X again. On Intel, this
generates an EPT_MISCONFIG. The memory slot generation number that
was incremented in (2) would normally take care of this but we fast
path mmio faults through quickly_check_mmio_pf(), which only checks
the per-vcpu mmio cache. Since we hit the cache, KVM passes a
KVM_EXIT_MMIO up to userspace.

This patch fixes the issue by using the memslot generation number
to validate the mmio cache.

Signed-off-by: David Matlack <dmatlack@google.com>
[xiaoguangrong: adjust the code to make it simpler for stable-tree fix.]
Signed-off-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Reviewed-by: David Matlack <dmatlack@google.com>
Reviewed-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Tested-by: David Matlack <dmatlack@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoBtrfs: send, fix data corruption due to incorrect hole detection
Filipe Manana [Sun, 30 Mar 2014 22:02:53 +0000 (23:02 +0100)]
Btrfs: send, fix data corruption due to incorrect hole detection

commit 766b5e5ae78dd04a93a275690a49e23d7dcb1f39 upstream.

During an incremental send, when we finish processing an inode (corresponding to
a regular file) we would assume the gap between the end of the last processed file
extent and the file's size corresponded to a file hole, and therefore incorrectly
send a bunch of zero bytes to overwrite that region in the file.

This affects only kernel 3.14.

Reproducer:

    mkfs.btrfs -f /dev/sdc
    mount /dev/sdc /mnt

    xfs_io -f -c "falloc -k 0 268435456" /mnt/foo

    btrfs subvolume snapshot -r /mnt /mnt/mysnap0

    xfs_io -c "pwrite -S 0x01 -b 9216 16190218 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x02 -b 1121 198720104 1121" /mnt/foo
    xfs_io -c "pwrite -S 0x05 -b 9216 107887439 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x06 -b 9216 225520207 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x07 -b 67584 102138300 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x08 -b 7000 94897484 7000" /mnt/foo
    xfs_io -c "pwrite -S 0x09 -b 113664 245083212 113664" /mnt/foo
    xfs_io -c "pwrite -S 0x10 -b 123 17937788 123" /mnt/foo
    xfs_io -c "pwrite -S 0x11 -b 39936 229573311 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x12 -b 67584 174792222 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x13 -b 9216 249253213 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x16 -b 67584 150046083 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x17 -b 39936 118246040 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x18 -b 67584 215965442 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x19 -b 33792 97096725 33792" /mnt/foo
    xfs_io -c "pwrite -S 0x20 -b 125952 166300596 125952" /mnt/foo
    xfs_io -c "pwrite -S 0x21 -b 123 1078957 123" /mnt/foo
    xfs_io -c "pwrite -S 0x25 -b 9216 212044492 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x26 -b 7000 265037146 7000" /mnt/foo
    xfs_io -c "pwrite -S 0x27 -b 42757 215922685 42757" /mnt/foo
    xfs_io -c "pwrite -S 0x28 -b 7000 69865411 7000" /mnt/foo
    xfs_io -c "pwrite -S 0x29 -b 67584 67948958 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x30 -b 39936 266967019 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x31 -b 1121 19582453 1121" /mnt/foo
    xfs_io -c "pwrite -S 0x32 -b 17408 257710255 17408" /mnt/foo
    xfs_io -c "pwrite -S 0x33 -b 39936 3895518 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x34 -b 125952 12045847 125952" /mnt/foo
    xfs_io -c "pwrite -S 0x35 -b 17408 19156379 17408" /mnt/foo
    xfs_io -c "pwrite -S 0x36 -b 39936 50160066 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x37 -b 113664 9549793 113664" /mnt/foo
    xfs_io -c "pwrite -S 0x38 -b 105472 94391506 105472" /mnt/foo
    xfs_io -c "pwrite -S 0x39 -b 23552 143632863 23552" /mnt/foo
    xfs_io -c "pwrite -S 0x40 -b 39936 241283845 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x41 -b 113664 199937606 113664" /mnt/foo
    xfs_io -c "pwrite -S 0x42 -b 67584 67380093 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x43 -b 67584 26793129 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x44 -b 39936 14421913 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x45 -b 123 253097405 123" /mnt/foo
    xfs_io -c "pwrite -S 0x46 -b 1121 128233424 1121" /mnt/foo
    xfs_io -c "pwrite -S 0x47 -b 105472 91577959 105472" /mnt/foo
    xfs_io -c "pwrite -S 0x48 -b 1121 7245381 1121" /mnt/foo
    xfs_io -c "pwrite -S 0x49 -b 113664 182414694 113664" /mnt/foo
    xfs_io -c "pwrite -S 0x50 -b 9216 32750608 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x51 -b 67584 266546049 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x52 -b 67584 87969398 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x53 -b 9216 260848797 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x54 -b 39936 119461243 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x55 -b 7000 200178693 7000" /mnt/foo
    xfs_io -c "pwrite -S 0x56 -b 9216 243316029 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x57 -b 7000 209658229 7000" /mnt/foo
    xfs_io -c "pwrite -S 0x58 -b 101376 179745192 101376" /mnt/foo
    xfs_io -c "pwrite -S 0x59 -b 9216 64012300 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x60 -b 125952 181705139 125952" /mnt/foo
    xfs_io -c "pwrite -S 0x61 -b 23552 235737348 23552" /mnt/foo
    xfs_io -c "pwrite -S 0x62 -b 113664 106021355 113664" /mnt/foo
    xfs_io -c "pwrite -S 0x63 -b 67584 135753552 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x64 -b 23552 95730888 23552" /mnt/foo
    xfs_io -c "pwrite -S 0x65 -b 11 17311415 11" /mnt/foo
    xfs_io -c "pwrite -S 0x66 -b 33792 120695553 33792" /mnt/foo
    xfs_io -c "pwrite -S 0x67 -b 9216 17164631 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x68 -b 9216 136065853 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x69 -b 67584 37752198 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x70 -b 101376 189717473 101376" /mnt/foo
    xfs_io -c "pwrite -S 0x71 -b 7000 227463698 7000" /mnt/foo
    xfs_io -c "pwrite -S 0x72 -b 9216 12655137 9216" /mnt/foo
    xfs_io -c "pwrite -S 0x73 -b 7000 7488866 7000" /mnt/foo
    xfs_io -c "pwrite -S 0x74 -b 113664 87813649 113664" /mnt/foo
    xfs_io -c "pwrite -S 0x75 -b 33792 25802183 33792" /mnt/foo
    xfs_io -c "pwrite -S 0x76 -b 39936 93524024 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x77 -b 33792 113336388 33792" /mnt/foo
    xfs_io -c "pwrite -S 0x78 -b 105472 184955320 105472" /mnt/foo
    xfs_io -c "pwrite -S 0x79 -b 101376 225691598 101376" /mnt/foo
    xfs_io -c "pwrite -S 0x80 -b 23552 77023155 23552" /mnt/foo
    xfs_io -c "pwrite -S 0x81 -b 11 201888192 11" /mnt/foo
    xfs_io -c "pwrite -S 0x82 -b 11 115332492 11" /mnt/foo
    xfs_io -c "pwrite -S 0x83 -b 67584 230278015 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x84 -b 11 120589073 11" /mnt/foo
    xfs_io -c "pwrite -S 0x85 -b 125952 202207819 125952" /mnt/foo
    xfs_io -c "pwrite -S 0x86 -b 113664 86672080 113664" /mnt/foo
    xfs_io -c "pwrite -S 0x87 -b 17408 208459603 17408" /mnt/foo
    xfs_io -c "pwrite -S 0x88 -b 7000 73372211 7000" /mnt/foo
    xfs_io -c "pwrite -S 0x89 -b 7000 42252122 7000" /mnt/foo
    xfs_io -c "pwrite -S 0x90 -b 23552 46784881 23552" /mnt/foo
    xfs_io -c "pwrite -S 0x91 -b 101376 63172351 101376" /mnt/foo
    xfs_io -c "pwrite -S 0x92 -b 23552 59341931 23552" /mnt/foo
    xfs_io -c "pwrite -S 0x93 -b 39936 239599283 39936" /mnt/foo
    xfs_io -c "pwrite -S 0x94 -b 67584 175643105 67584" /mnt/foo
    xfs_io -c "pwrite -S 0x97 -b 23552 105534880 23552" /mnt/foo
    xfs_io -c "pwrite -S 0x98 -b 113664 8236844 113664" /mnt/foo
    xfs_io -c "pwrite -S 0x99 -b 125952 144489686 125952" /mnt/foo
    xfs_io -c "pwrite -S 0xa0 -b 7000 73273112 7000" /mnt/foo
    xfs_io -c "pwrite -S 0xa1 -b 125952 194580243 125952" /mnt/foo
    xfs_io -c "pwrite -S 0xa2 -b 123 56296779 123" /mnt/foo
    xfs_io -c "pwrite -S 0xa3 -b 11 233066845 11" /mnt/foo
    xfs_io -c "pwrite -S 0xa4 -b 39936 197727090 39936" /mnt/foo
    xfs_io -c "pwrite -S 0xa5 -b 101376 53579812 101376" /mnt/foo
    xfs_io -c "pwrite -S 0xa6 -b 9216 85669738 9216" /mnt/foo
    xfs_io -c "pwrite -S 0xa7 -b 125952 21266322 125952" /mnt/foo
    xfs_io -c "pwrite -S 0xa8 -b 23552 125726568 23552" /mnt/foo
    xfs_io -c "pwrite -S 0xa9 -b 9216 18423680 9216" /mnt/foo
    xfs_io -c "pwrite -S 0xb0 -b 1121 165901483 1121" /mnt/foo

    btrfs subvolume snapshot -r /mnt /mnt/mysnap1

    xfs_io -c "pwrite -S 0xff -b 10 16190218 10" /mnt/foo

    btrfs subvolume snapshot -r /mnt /mnt/mysnap2

    md5sum /mnt/foo          # returns 79e53f1466bfc09fd82b450689e6119e
    md5sum /mnt/mysnap2/foo  # returns 79e53f1466bfc09fd82b450689e6119e too

    btrfs send /mnt/mysnap1 -f /tmp/1.snap
    btrfs send -p /mnt/mysnap1 /mnt/mysnap2 -f /tmp/2.snap

    mkfs.btrfs -f /dev/sdc
    mount /dev/sdc /mnt

    btrfs receive /mnt -f /tmp/1.snap
    btrfs receive /mnt -f /tmp/2.snap

    md5sum /mnt/mysnap2/foo  # returns 2bb414c5155767cedccd7063e51beabd !!

A testcase for xfstests follows soon too.

Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agopci_ids: Add support for Intel Quark ILB
Josef Ahmad [Tue, 2 Sep 2014 10:45:20 +0000 (13:45 +0300)]
pci_ids: Add support for Intel Quark ILB

commit bb048713bba3ead39f6112910906d9fe3f88ede7 upstream.

This patch adds the PCI id for Intel Quark ILB.
It will be used for GPIO and Multifunction device driver.

Signed-off-by: Josef Ahmad <josef.ahmad@intel.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Chang Rebecca Swee Fun <rebecca.swee.fun.chang@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agousb: pch_udc: usb gadget device support for Intel Quark X1000
Bryan O'Donoghue [Mon, 4 Aug 2014 17:22:54 +0000 (10:22 -0700)]
usb: pch_udc: usb gadget device support for Intel Quark X1000

commit a68df7066a6f974db6069e0b93c498775660a114 upstream.

This patch is to enable the USB gadget device for Intel Quark X1000

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@intel.com>
Signed-off-by: Bing Niu <bing.niu@intel.com>
Signed-off-by: Alvin (Weike) Chen <alvin.chen@intel.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Chang Rebecca Swee Fun <rebecca.swee.fun.chang@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agofs: Add a missing permission check to do_umount
Andy Lutomirski [Wed, 8 Oct 2014 19:32:47 +0000 (12:32 -0700)]
fs: Add a missing permission check to do_umount

commit a1480dcc3c706e309a88884723446f2e84fedd5b upstream.

Accessing do_remount_sb should require global CAP_SYS_ADMIN, but
only one of the two call sites was appropriately protected.

Fixes CVE-2014-7975.

Signed-off-by: Andy Lutomirski <luto@amacapital.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoBtrfs: fix race in WAIT_SYNC ioctl
Sage Weil [Fri, 26 Sep 2014 15:30:06 +0000 (08:30 -0700)]
Btrfs: fix race in WAIT_SYNC ioctl

commit 42383020beb1cfb05f5d330cc311931bc4917a97 upstream.

We check whether transid is already committed via last_trans_committed and
then search through trans_list for pending transactions.  If
last_trans_committed is updated by btrfs_commit_transaction after we check
it (there is no locking), we will fail to find the committed transaction
and return EINVAL to the caller.  This has been observed occasionally by
ceph-osd (which uses this ioctl heavily).

Fix by rechecking whether the provided transid <= last_trans_committed
after the search fails, and if so return 0.

Signed-off-by: Sage Weil <sage@redhat.com>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoBtrfs: fix build_backref_tree issue with multiple shared blocks
Josef Bacik [Fri, 19 Sep 2014 19:43:34 +0000 (15:43 -0400)]
Btrfs: fix build_backref_tree issue with multiple shared blocks

commit bbe9051441effce51c9a533d2c56440df64db2d7 upstream.

Marc Merlin sent me a broken fs image months ago where it would blow up in the
upper->checked BUG_ON() in build_backref_tree.  This is because we had a
scenario like this

block a -- level 4 (not shared)
   |
block b -- level 3 (reloc block, shared)
   |
block c -- level 2 (not shared)
   |
block d -- level 1 (shared)
   |
block e -- level 0 (shared)

We go to build a backref tree for block e, we notice block d is shared and add
it to the list of blocks to lookup it's backrefs for.  Now when we loop around
we will check edges for the block, so we will see we looked up block c last
time.  So we lookup block d and then see that the block that points to it is
block c and we can just skip that edge since we've already been up this path.
The problem is because we clear need_check when we see block d (as it is shared)
we never add block b as needing to be checked.  And because block c is in our
path already we bail out before we walk up to block b and add it to the backref
check list.

To fix this we need to reset need_check if we trip over a block that doesn't
need to be checked.  This will make sure that any subsequent blocks in the path
as we're walking up afterwards are added to the list to be processed.  With this
patch I can now mount Marc's fs image and it'll complete the balance without
panicing.  Thanks,

Reported-by: Marc MERLIN <marc@merlins.org>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoBtrfs: cleanup error handling in build_backref_tree
Josef Bacik [Fri, 19 Sep 2014 14:40:00 +0000 (10:40 -0400)]
Btrfs: cleanup error handling in build_backref_tree

commit 75bfb9aff45e44625260f52a5fd581b92ace3e62 upstream.

When balance panics it tends to panic in the

BUG_ON(!upper->checked);

test, because it means it couldn't build the backref tree properly.  This is
annoying to users and frankly a recoverable error, nothing in this function is
actually fatal since it is just an in-memory building of the backrefs for a
given bytenr.  So go through and change all the BUG_ON()'s to ASSERT()'s, and
fix the BUG_ON(!upper->checked) thing to just return an error.

This patch also fixes the error handling so it tears down the work we've done
properly.  This code was horribly broken since we always just panic'ed instead
of actually erroring out, so it needed to be completely re-worked.  With this
patch my broken image no longer panics when I mount it.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoBtrfs: try not to ENOSPC on log replay
Josef Bacik [Thu, 18 Sep 2014 15:30:44 +0000 (11:30 -0400)]
Btrfs: try not to ENOSPC on log replay

commit 1d52c78afbbf80b58299e076a159617d6b42fe3c upstream.

When doing log replay we may have to update inodes, which traditionally goes
through our delayed inode stuff.  This will try to move space over from the
trans handle, but we don't reserve space in our trans handle on replay since we
don't know how much we will need, so instead we try to flush.  But because we
have a trans handle open we won't flush anything, so if we are out of reserve
space we will simply return ENOSPC.  Since we know that if an operation made it
into the log then we definitely had space before the box bought the farm then we
don't need to worry about doing this space reservation.  Use the
fs_info->log_root_recovering flag to skip the delayed inode stuff and update the
item directly.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoBtrfs: fix up bounds checking in lseek
Liu Bo [Tue, 16 Sep 2014 09:49:30 +0000 (17:49 +0800)]
Btrfs: fix up bounds checking in lseek

commit 4d1a40c66bed0b3fa43b9da5fbd5cbe332e4eccf upstream.

An user reported this, it is because that lseek's SEEK_SET/SEEK_CUR/SEEK_END
allow a negative value for @offset, but btrfs's SEEK_DATA/SEEK_HOLE don't
prepare for that and convert the negative @offset into unsigned type,
so we get (end < start) warning.

[ 1269.835374] ------------[ cut here ]------------
[ 1269.836809] WARNING: CPU: 0 PID: 1241 at fs/btrfs/extent_io.c:430 insert_state+0x11d/0x140()
[ 1269.838816] BTRFS: end < start 4094 18446744073709551615
[ 1269.840334] CPU: 0 PID: 1241 Comm: a.out Tainted: G        W      3.16.0+ #306
[ 1269.858229] Call Trace:
[ 1269.858612]  [<ffffffff81801a69>] dump_stack+0x4e/0x68
[ 1269.858952]  [<ffffffff8107894c>] warn_slowpath_common+0x8c/0xc0
[ 1269.859416]  [<ffffffff81078a36>] warn_slowpath_fmt+0x46/0x50
[ 1269.859929]  [<ffffffff813b0fbd>] insert_state+0x11d/0x140
[ 1269.860409]  [<ffffffff813b1396>] __set_extent_bit+0x3b6/0x4e0
[ 1269.860805]  [<ffffffff813b21c7>] lock_extent_bits+0x87/0x200
[ 1269.861697]  [<ffffffff813a5b28>] btrfs_file_llseek+0x148/0x2a0
[ 1269.862168]  [<ffffffff811f201e>] SyS_lseek+0xae/0xc0
[ 1269.862620]  [<ffffffff8180b212>] system_call_fastpath+0x16/0x1b
[ 1269.862970] ---[ end trace 4d33ea885832054b ]---

This assumes that btrfs starts finding DATA/HOLE from the beginning of file
if the assigned @offset is negative.

Also we add alignment for lock_extent_bits 's range.

Reported-by: Toralf Förster <toralf.foerster@gmx.de>
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoBtrfs: add missing compression property remove in btrfs_ioctl_setflags
Filipe Manana [Thu, 11 Sep 2014 10:44:49 +0000 (11:44 +0100)]
Btrfs: add missing compression property remove in btrfs_ioctl_setflags

commit 78a017a2c92df9b571db0a55a016280f9019c65e upstream.

The behaviour of a 'chattr -c' consists of getting the current flags,
clearing the FS_COMPR_FL bit and then sending the result to the set
flags ioctl - this means the bit FS_NOCOMP_FL isn't set in the flags
passed to the ioctl. This results in the compression property not being
cleared from the inode - it was cleared only if the bit FS_NOCOMP_FL
was set in the received flags.

Reproducer:

    $ mkfs.btrfs -f /dev/sdd
    $ mount /dev/sdd /mnt && cd /mnt
    $ mkdir a
    $ chattr +c a
    $ touch a/file
    $ lsattr a/file
    --------c------- a/file
    $ chattr -c a
    $ touch a/file2
    $ lsattr a/file2
    --------c------- a/file2
    $ lsattr -d a
    ---------------- a

Reported-by: Andreas Schneider <asn@cryptomilk.org>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agobtrfs: wake up transaction thread from SYNC_FS ioctl
David Sterba [Wed, 23 Jul 2014 12:39:35 +0000 (14:39 +0200)]
btrfs: wake up transaction thread from SYNC_FS ioctl

commit 2fad4e83e12591eb3bd213875b9edc2d18e93383 upstream.

The transaction thread may want to do more work, namely it pokes the
cleaner ktread that will start processing uncleaned subvols.

This can be triggered by user via the 'btrfs fi sync' command, otherwise
there was a delay up to 30 seconds before the cleaner started to clean
old snapshots.

Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoLinux 3.14.22
Greg Kroah-Hartman [Wed, 15 Oct 2014 06:42:04 +0000 (08:42 +0200)]
Linux 3.14.22

10 years agoserial: 8250: Add Quark X1000 to 8250_pci.c
Bryan O'Donoghue [Tue, 23 Sep 2014 00:21:11 +0000 (01:21 +0100)]
serial: 8250: Add Quark X1000 to 8250_pci.c

commit 1ede7dcca3c4fa15a518ab0473126f9c3e621e4c upstream.

Quark X1000 contains two designware derived 8250 serial ports.
Each port has a unique PCI configuration space consisting of
BAR0:UART BAR1:DMA respectively.

Unlike the standard 8250 the register width is 32 bits for RHR,IER etc
The Quark UART has a fundamental clock @ 44.2368 MHz allowing for a
bitrate of up to about 2.76 megabits per second.

This patch enables standard 8250 mode

Signed-off-by: Bryan O'Donoghue <pure.logic@nexus-software.ie>
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agocrypto: caam - fix addressing of struct member
Cristian Stoica [Thu, 14 Aug 2014 10:51:57 +0000 (13:51 +0300)]
crypto: caam - fix addressing of struct member

commit 4451d494b1910bf7b7f8381a637d0fe6d2142467 upstream.

buf_0 and buf_1 in caam_hash_state are not next to each other.
Accessing buf_1 is incorrect from &buf_0 with an offset of only
size_of(buf_0). The same issue is also with buflen_0 and buflen_1

Signed-off-by: Cristian Stoica <cristian.stoica@freescale.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agousb: musb: dsps: kill OTG timer on suspend
Felipe Balbi [Mon, 15 Sep 2014 14:03:24 +0000 (09:03 -0500)]
usb: musb: dsps: kill OTG timer on suspend

commit 468bcc2a2ca071f652009d2d20d97f2437630cae upstream.

if we don't make sure to kill the timer, it could
expire after we have already gated our clocks.

That will trigger a Data Abort exception because
we would try to access register while clock is gated.

Fix that bug.

Fixes 869c597 (usb: musb: dsps: add support for suspend and resume)
Tested-by: Dave Gerlach <d-gerlach@ti.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoUSB: cp210x: add support for Seluxit USB dongle
Andreas Bomholtz [Mon, 22 Sep 2014 07:50:43 +0000 (09:50 +0200)]
USB: cp210x: add support for Seluxit USB dongle

commit dee80ad12d2b1b304286a707fde7ab05d1fc7bab upstream.

Added the Seluxit ApS USB Serial Dongle to cp210x driver.

Signed-off-by: Andreas Bomholtz <andreas@seluxit.com>
Signed-off-by: Johan Hovold <johan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoUSB: serial: cp210x: added Ketra N1 wireless interface support
Joe Savage [Sat, 20 Sep 2014 13:01:16 +0000 (08:01 -0500)]
USB: serial: cp210x: added Ketra N1 wireless interface support

commit bfc2d7dfdd761ae3beccdb26abebe03cef042f46 upstream.

Added support for Ketra N1 wireless interface, which uses the
Silicon Labs' CP2104 USB to UART bridge with customized PID 8946.

Signed-off-by: Joe Savage <joe.savage@goketra.com>
Signed-off-by: Johan Hovold <johan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoUSB: Add device quirk for ASUS T100 Base Station keyboard
Lu Baolu [Fri, 19 Sep 2014 02:13:50 +0000 (10:13 +0800)]
USB: Add device quirk for ASUS T100 Base Station keyboard

commit ddbe1fca0bcb87ca8c199ea873a456ca8a948567 upstream.

This full-speed USB device generates spurious remote wakeup event
as soon as USB_DEVICE_REMOTE_WAKEUP feature is set. As the result,
Linux can't enter system suspend and S0ix power saving modes once
this keyboard is used.

This patch tries to introduce USB_QUIRK_IGNORE_REMOTE_WAKEUP quirk.
With this quirk set, wakeup capability will be ignored during
device configure.

This patch could be back-ported to kernels as old as 2.6.39.

Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agotcp: fixing TLP's FIN recovery
Per Hurtig [Thu, 12 Jun 2014 15:08:32 +0000 (17:08 +0200)]
tcp: fixing TLP's FIN recovery

[ Upstream commit bef1909ee3ed1ca39231b260a8d3b4544ecd0c8f ]

Fix to a problem observed when losing a FIN segment that does not
contain data.  In such situations, TLP is unable to recover from
*any* tail loss and instead adds at least PTO ms to the
retransmission process, i.e., RTO = RTO + PTO.

Signed-off-by: Per Hurtig <per.hurtig@kau.se>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agosctp: handle association restarts when the socket is closed.
Vlad Yasevich [Fri, 3 Oct 2014 22:16:20 +0000 (18:16 -0400)]
sctp: handle association restarts when the socket is closed.

[ Upstream commit bdf6fa52f01b941d4a80372d56de465bdbbd1d23 ]

Currently association restarts do not take into consideration the
state of the socket.  When a restart happens, the current assocation
simply transitions into established state.  This creates a condition
where a remote system, through a the restart procedure, may create a
local association that is no way reachable by user.  The conditions
to trigger this are as follows:
  1) Remote does not acknoledge some data causing data to remain
     outstanding.
  2) Local application calls close() on the socket.  Since data
     is still outstanding, the association is placed in SHUTDOWN_PENDING
     state.  However, the socket is closed.
  3) The remote tries to create a new association, triggering a restart
     on the local system.  The association moves from SHUTDOWN_PENDING
     to ESTABLISHED.  At this point, it is no longer reachable by
     any socket on the local system.

This patch addresses the above situation by moving the newly ESTABLISHED
association into SHUTDOWN-SENT state and bundling a SHUTDOWN after
the COOKIE-ACK chunk.  This way, the restarted associate immidiately
enters the shutdown procedure and forces the termination of the
unreachable association.

Reported-by: David Laight <David.Laight@aculab.com>
Signed-off-by: Vlad Yasevich <vyasevich@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoteam: avoid race condition in scheduling delayed work
Joe Lawrence [Fri, 3 Oct 2014 13:58:34 +0000 (09:58 -0400)]
team: avoid race condition in scheduling delayed work

[ Upstream commit 47549650abd13d873fd2e5fc218db19e21031074 ]

When team_notify_peers and team_mcast_rejoin are called, they both reset
their respective .count_pending atomic variable. Then when the actual
worker function is executed, the variable is atomically decremented.
This pattern introduces a potential race condition where the
.count_pending rolls over and the worker function keeps rescheduling
until .count_pending decrements to zero again:

THREAD 1                           THREAD 2

========                           ========
team_notify_peers(teamX)
  atomic_set count_pending = 1
  schedule_delayed_work
                                   team_notify_peers(teamX)
                                   atomic_set count_pending = 1
team_notify_peers_work
  atomic_dec_and_test
    count_pending = 0
  (return)
                                   schedule_delayed_work
                                   team_notify_peers_work
                                   atomic_dec_and_test
                                     count_pending = -1
                                   schedule_delayed_work
                                   (repeat until count_pending = 0)

Instead of assigning a new value to .count_pending, use atomic_add to
tack-on the additional desired worker function invocations.

Signed-off-by: Joe Lawrence <joe.lawrence@stratus.com>
Acked-by: Jiri Pirko <jiri@resnulli.us>
Fixes: fc423ff00df3a19554414ee ("team: add peer notification")
Fixes: 492b200efdd20b8fcfdac87 ("team: add support for sending multicast rejoins")
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoip6_gre: fix flowi6_proto value in xmit path
Nicolas Dichtel [Thu, 2 Oct 2014 16:26:49 +0000 (18:26 +0200)]
ip6_gre: fix flowi6_proto value in xmit path

[ Upstream commit 3be07244b7337760a3269d56b2f4a63e72218648 ]

In xmit path, we build a flowi6 which will be used for the output route lookup.
We are sending a GRE packet, neither IPv4 nor IPv6 encapsulated packet, thus the
protocol should be IPPROTO_GRE.

Fixes: c12b395a4664 ("gre: Support GRE over IPv6")
Reported-by: Matthieu Ternisien d'Ouville <matthieu.tdo@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agohyperv: Fix a bug in netvsc_start_xmit()
KY Srinivasan [Mon, 29 Sep 2014 05:16:43 +0000 (22:16 -0700)]
hyperv: Fix a bug in netvsc_start_xmit()

[ Upstream commit dedb845ded56ded1c62f5398a94ffa8615d4592d ]

After the packet is successfully sent, we should not touch the skb
as it may have been freed. This patch is based on the work done by
Long Li <longli@microsoft.com>.

In this version of the patch I have fixed issues pointed out by David.
David, please queue this up for stable.

Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Tested-by: Long Li <longli@microsoft.com>
Tested-by: Sitsofe Wheeler <sitsofe@yahoo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agogro: fix aggregation for skb using frag_list
Eric Dumazet [Mon, 29 Sep 2014 17:34:29 +0000 (10:34 -0700)]
gro: fix aggregation for skb using frag_list

[ Upstream commit 73d3fe6d1c6d840763ceafa9afae0aaafa18c4b5 ]

In commit 8a29111c7ca6 ("net: gro: allow to build full sized skb")
I added a regression for linear skb that traditionally force GRO
to use the frag_list fallback.

Erez Shitrit found that at most two segments were aggregated and
the "if (skb_gro_len(p) != pinfo->gso_size)" test was failing.

This is because pinfo at this spot still points to the last skb in the
chain, instead of the first one, where we find the correct gso_size
information.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: 8a29111c7ca6 ("net: gro: allow to build full sized skb")
Reported-by: Erez Shitrit <erezsh@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoRevert "net/macb: add pinctrl consumer support"
Soren Brinkmann [Mon, 22 Sep 2014 23:49:08 +0000 (16:49 -0700)]
Revert "net/macb: add pinctrl consumer support"

[ Upstream commit 9026968abe7ad102f4ac5c6d96d733643f75399c ]

This reverts commit 8ef29f8aae524bd51298fb10ac6a5ce6c4c5a3d8.
The driver core already calls pinctrl_get() and claims the default
state. There is no need to replicate this in the driver.
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomacvtap: Fix race between device delete and open.
Vlad Yasevich [Mon, 22 Sep 2014 20:34:17 +0000 (16:34 -0400)]
macvtap: Fix race between device delete and open.

[ Upstream commit 40b8fe45d1f094e3babe7b2dc2b71557ab71401d ]

In macvtap device delete and open calls can race and
this causes a list curruption of the vlan queue_list.

The race intself is triggered by the idr accessors
that located the vlan device.  The device is stored
into and removed from the idr under both an rtnl and
a mutex.  However, when attempting to locate the device
in idr, only a mutex is taken.  As a result, once cpu
perfoming a delete may take an rtnl and wait for the mutex,
while another cput doing an open() will take the idr
mutex first to fetch the device pointer and later take
an rtnl to add a queue for the device which may have
just gotten deleted.

With this patch, we now hold the rtnl for the duration
of the macvtap_open() call thus making sure that
open will not race with delete.

CC: Michael S. Tsirkin <mst@redhat.com>
CC: Jason Wang <jasowang@redhat.com>
Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoxfrm: Generate queueing routes only from route lookup functions
Steffen Klassert [Tue, 16 Sep 2014 08:08:49 +0000 (10:08 +0200)]
xfrm: Generate queueing routes only from route lookup functions

[ Upstream commit b8c203b2d2fc961bafd53b41d5396bbcdec55998 ]

Currently we genarate a queueing route if we have matching policies
but can not resolve the states and the sysctl xfrm_larval_drop is
disabled. Here we assume that dst_output() is called to kill the
queued packets. Unfortunately this assumption is not true in all
cases, so it is possible that these packets leave the system unwanted.

We fix this by generating queueing routes only from the
route lookup functions, here we can guarantee a call to
dst_output() afterwards.

Fixes: a0073fe18e71 ("xfrm: Add a state resolution packet queue")
Reported-by: Konstantinos Kolelis <k.kolelis@sirrix.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoxfrm: Generate blackhole routes only from route lookup functions
Steffen Klassert [Tue, 16 Sep 2014 08:08:40 +0000 (10:08 +0200)]
xfrm: Generate blackhole routes only from route lookup functions

[ Upstream commit f92ee61982d6da15a9e49664ecd6405a15a2ee56 ]

Currently we genarate a blackhole route route whenever we have
matching policies but can not resolve the states. Here we assume
that dst_output() is called to kill the balckholed packets.
Unfortunately this assumption is not true in all cases, so
it is possible that these packets leave the system unwanted.

We fix this by generating blackhole routes only from the
route lookup functions, here we can guarantee a call to
dst_output() afterwards.

Fixes: 2774c131b1d ("xfrm: Handle blackhole route creation via afinfo.")
Reported-by: Konstantinos Kolelis <k.kolelis@sirrix.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agotg3: Allow for recieve of full-size 8021AD frames
Vlad Yasevich [Tue, 30 Sep 2014 23:39:36 +0000 (19:39 -0400)]
tg3: Allow for recieve of full-size 8021AD frames

[ Upstream commit 7d3083ee36b51e425b6abd76778a2046906b0fd3 ]

When receiving a vlan-tagged frame that still contains
a vlan header, the length of the packet will be greater
then MTU+ETH_HLEN since it will account of the extra
vlan header.  TG3 checks this for the case for 802.1Q,
but not for 802.1ad.  As a result, full sized 802.1ad
frames get dropped by the card.

Add a check for 802.1ad protocol when receving full
sized frames.

Suggested-by: Prashant Sreedharan <prashant@broadcom.com>
CC: Prashant Sreedharan <prashant@broadcom.com>
CC: Michael Chan <mchan@broadcom.com>
Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agotg3: Work around HW/FW limitations with vlan encapsulated frames
Vlad Yasevich [Thu, 18 Sep 2014 14:31:17 +0000 (10:31 -0400)]
tg3: Work around HW/FW limitations with vlan encapsulated frames

[ Upstream commit 476c18850c6cbaa3f2bb661ae9710645081563b9 ]

TG3 appears to have an issue performing TSO and checksum offloading
correclty when the frame has been vlan encapsulated (non-accelrated).
In these cases, tcp checksum is not correctly updated.

This patch attempts to work around this issue.  After the patch,
802.1ad vlans start working correctly over tg3 devices.

CC: Prashant Sreedharan <prashant@broadcom.com>
CC: Michael Chan <mchan@broadcom.com>
Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agonet: allow macvlans to move to net namespace
Francesco Ruggeri [Wed, 17 Sep 2014 17:40:44 +0000 (10:40 -0700)]
net: allow macvlans to move to net namespace

[ Upstream commit 0d0162e7a33d3710b9604e7c68c0f31f5c457428 ]

I cannot move a macvlan interface created on top of a bonding interface
to a different namespace:

% ip netns add dummy0
% ip link add link bond0 mac0 type macvlan
% ip link set mac0 netns dummy0
RTNETLINK answers: Invalid argument
%

The problem seems to be that commit f9399814927a ("bonding: Don't allow
bond devices to change network namespaces.") sets NETIF_F_NETNS_LOCAL
on bonding interfaces, and commit 797f87f83b60 ("macvlan: fix netdev
feature propagation from lower device") causes macvlan interfaces
to inherit its features from the lower device.

NETIF_F_NETNS_LOCAL should not be inherited from the lower device
by a macvlan.
Patch tested on 3.16.

Signed-off-by: Francesco Ruggeri <fruggeri@arista.com>
Acked-by: Cong Wang <cwang@twopensource.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agobridge: Fix br_should_learn to check vlan_enabled
Vlad Yasevich [Mon, 15 Sep 2014 19:24:26 +0000 (15:24 -0400)]
bridge: Fix br_should_learn to check vlan_enabled

[ Upstream commit c095f248e63ada504dd90c90baae673ae10ee3fe ]

As Toshiaki Makita pointed out, the BRIDGE_INPUT_SKB_CB will
not be initialized in br_should_learn() as that function
is called only from br_handle_local_finish().  That is
an input handler for link-local ethernet traffic so it perfectly
correct to check br->vlan_enabled here.

Reported-by: Toshiaki Makita<toshiaki.makita1@gmail.com>
Fixes: 20adfa1 bridge: Check if vlan filtering is enabled only once.
Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agobridge: Check if vlan filtering is enabled only once.
Vlad Yasevich [Fri, 12 Sep 2014 20:26:16 +0000 (16:26 -0400)]
bridge: Check if vlan filtering is enabled only once.

[ Upstream commit 20adfa1a81af00bf2027644507ad4fa9cd2849cf ]

The bridge code checks if vlan filtering is enabled on both
ingress and egress.   When the state flip happens, it
is possible for the bridge to currently be forwarding packets
and forwarding behavior becomes non-deterministic.  Bridge
may drop packets on some interfaces, but not others.

This patch solves this by caching the filtered state of the
packet into skb_cb on ingress.  The skb_cb is guaranteed to
not be over-written between the time packet entres bridge
forwarding path and the time it leaves it.  On egress, we
can then check the cached state to see if we need to
apply filtering information.

Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoipv6: restore the behavior of ipv6_sock_ac_drop()
WANG Cong [Fri, 5 Sep 2014 21:33:00 +0000 (14:33 -0700)]
ipv6: restore the behavior of ipv6_sock_ac_drop()

[ Upstream commit de185ab46cb02df9738b0d898b0c3a89181c5526 ]

It is possible that the interface is already gone after joining
the list of anycast on this interface as we don't hold a refcount
for the device, in this case we are safe to ignore the error.

What's more important, for API compatibility we should not
change this behavior for applications even if it were correct.

Fixes: commit a9ed4a2986e13011 ("ipv6: fix rtnl locking in setsockopt for anycast and multicast")
Cc: Sabrina Dubroca <sd@queasysnail.net>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agobonding: fix div by zero while enslaving and transmitting
Nikolay Aleksandrov [Fri, 12 Sep 2014 15:38:18 +0000 (17:38 +0200)]
bonding: fix div by zero while enslaving and transmitting

[ Upstream commit 9a72c2da690d78e93cff24b9f616412508678dd5 ]

The problem is that the slave is first linked and slave_cnt is
incremented afterwards leading to a div by zero in the modes that use it
as a modulus. What happens is that in bond_start_xmit()
bond_has_slaves() is used to evaluate further transmission and it becomes
true after the slave is linked in, but when slave_cnt is used in the xmit
path it is still 0, so fetch it once and transmit based on that. Since
it is used only in round-robin and XOR modes, the fix is only for them.
Thanks to Eric Dumazet for pointing out the fault in my first try to fix
this.

Call trace (took it out of net-next kernel, but it's the same with net):
[46934.330038] divide error: 0000 [#1] SMP
[46934.330041] Modules linked in: bonding(O) 9p fscache
snd_hda_codec_generic crct10dif_pclmul
[46934.330041] bond0: Enslaving eth1 as an active interface with an up
link
[46934.330051]  ppdev joydev crc32_pclmul crc32c_intel 9pnet_virtio
ghash_clmulni_intel snd_hda_intel 9pnet snd_hda_controller parport_pc
serio_raw pcspkr snd_hda_codec parport virtio_balloon virtio_console
snd_hwdep snd_pcm pvpanic i2c_piix4 snd_timer i2ccore snd soundcore
virtio_blk virtio_net virtio_pci virtio_ring virtio ata_generic
pata_acpi floppy [last unloaded: bonding]
[46934.330053] CPU: 1 PID: 3382 Comm: ping Tainted: G           O
3.17.0-rc4+ #27
[46934.330053] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
[46934.330054] task: ffff88005aebf2c0 ti: ffff88005b728000 task.ti:
ffff88005b728000
[46934.330059] RIP: 0010:[<ffffffffa0198c33>]  [<ffffffffa0198c33>]
bond_start_xmit+0x1c3/0x450 [bonding]
[46934.330060] RSP: 0018:ffff88005b72b7f8  EFLAGS: 00010246
[46934.330060] RAX: 0000000000000679 RBX: ffff88004b077000 RCX:
000000000000002a
[46934.330061] RDX: 0000000000000000 RSI: ffff88004b3f0500 RDI:
ffff88004b077940
[46934.330061] RBP: ffff88005b72b830 R08: 00000000000000c0 R09:
ffff88004a83e000
[46934.330062] R10: 000000000000ffff R11: ffff88004b1f12c0 R12:
ffff88004b3f0500
[46934.330062] R13: ffff88004b3f0500 R14: 000000000000002a R15:
ffff88004b077940
[46934.330063] FS:  00007fbd91a4c740(0000) GS:ffff88005f080000(0000)
knlGS:0000000000000000
[46934.330064] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[46934.330064] CR2: 00007f803a8bb000 CR3: 000000004b2c9000 CR4:
00000000000406e0
[46934.330069] Stack:
[46934.330071]  ffffffff811e6169 00000000e772fa05 ffff88004b077000
ffff88004b3f0500
[46934.330072]  ffffffff81d17d18 000000000000002a 0000000000000000
ffff88005b72b8a0
[46934.330073]  ffffffff81620108 ffffffff8161fe0e ffff88005b72b8c4
ffff88005b302000
[46934.330073] Call Trace:
[46934.330077]  [<ffffffff811e6169>] ?
__kmalloc_node_track_caller+0x119/0x300
[46934.330084]  [<ffffffff81620108>] dev_hard_start_xmit+0x188/0x410
[46934.330086]  [<ffffffff8161fe0e>] ? harmonize_features+0x2e/0x90
[46934.330088]  [<ffffffff81620b06>] __dev_queue_xmit+0x456/0x590
[46934.330089]  [<ffffffff81620c50>] dev_queue_xmit+0x10/0x20
[46934.330090]  [<ffffffff8168f022>] arp_xmit+0x22/0x60
[46934.330091]  [<ffffffff8168f090>] arp_send.part.16+0x30/0x40
[46934.330092]  [<ffffffff8168f1e5>] arp_solicit+0x115/0x2b0
[46934.330094]  [<ffffffff8160b5d7>] ? copy_skb_header+0x17/0xa0
[46934.330096]  [<ffffffff8162875a>] neigh_probe+0x4a/0x70
[46934.330097]  [<ffffffff8162979c>] __neigh_event_send+0xac/0x230
[46934.330098]  [<ffffffff8162a00b>] neigh_resolve_output+0x13b/0x220
[46934.330100]  [<ffffffff8165f120>] ? ip_forward_options+0x1c0/0x1c0
[46934.330101]  [<ffffffff81660478>] ip_finish_output+0x1f8/0x860
[46934.330102]  [<ffffffff81661f08>] ip_output+0x58/0x90
[46934.330103]  [<ffffffff81661602>] ? __ip_local_out+0xa2/0xb0
[46934.330104]  [<ffffffff81661640>] ip_local_out_sk+0x30/0x40
[46934.330105]  [<ffffffff81662a66>] ip_send_skb+0x16/0x50
[46934.330106]  [<ffffffff81662ad3>] ip_push_pending_frames+0x33/0x40
[46934.330107]  [<ffffffff8168854c>] raw_sendmsg+0x88c/0xa30
[46934.330110]  [<ffffffff81612b31>] ? skb_recv_datagram+0x41/0x60
[46934.330111]  [<ffffffff816875a9>] ? raw_recvmsg+0xa9/0x1f0
[46934.330113]  [<ffffffff816978d4>] inet_sendmsg+0x74/0xc0
[46934.330114]  [<ffffffff81697a9b>] ? inet_recvmsg+0x8b/0xb0
[46934.330115] bond0: Adding slave eth2
[46934.330116]  [<ffffffff8160357c>] sock_sendmsg+0x9c/0xe0
[46934.330118]  [<ffffffff81603248>] ?
move_addr_to_kernel.part.20+0x28/0x80
[46934.330121]  [<ffffffff811b4477>] ? might_fault+0x47/0x50
[46934.330122]  [<ffffffff816039b9>] ___sys_sendmsg+0x3a9/0x3c0
[46934.330125]  [<ffffffff8144a14a>] ? n_tty_write+0x3aa/0x530
[46934.330127]  [<ffffffff810d1ae4>] ? __wake_up+0x44/0x50
[46934.330129]  [<ffffffff81242b38>] ? fsnotify+0x238/0x310
[46934.330130]  [<ffffffff816048a1>] __sys_sendmsg+0x51/0x90
[46934.330131]  [<ffffffff816048f2>] SyS_sendmsg+0x12/0x20
[46934.330134]  [<ffffffff81738b29>] system_call_fastpath+0x16/0x1b
[46934.330144] Code: 48 8b 10 4c 89 ee 4c 89 ff e8 aa bc ff ff 31 c0 e9
1a ff ff ff 0f 1f 00 4c 89 ee 4c 89 ff e8 65 fb ff ff 31 d2 4c 89 ee 4c
89 ff <f7> b3 64 09 00 00 e8 02 bd ff ff 31 c0 e9 f2 fe ff ff 0f 1f 00
[46934.330146] RIP  [<ffffffffa0198c33>] bond_start_xmit+0x1c3/0x450
[bonding]
[46934.330146]  RSP <ffff88005b72b7f8>

CC: Eric Dumazet <eric.dumazet@gmail.com>
CC: Andy Gospodarek <andy@greyhouse.net>
CC: Jay Vosburgh <j.vosburgh@gmail.com>
CC: Veaceslav Falico <vfalico@gmail.com>
Fixes: 278b208375 ("bonding: initial RCU conversion")
Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoipv6: fix rtnl locking in setsockopt for anycast and multicast
Sabrina Dubroca [Tue, 2 Sep 2014 08:29:29 +0000 (10:29 +0200)]
ipv6: fix rtnl locking in setsockopt for anycast and multicast

[ Upstream commit a9ed4a2986e13011fcf4ed2d1a1647c53112f55b ]

Calling setsockopt with IPV6_JOIN_ANYCAST or IPV6_LEAVE_ANYCAST
triggers the assertion in addrconf_join_solict()/addrconf_leave_solict()

ipv6_sock_ac_join(), ipv6_sock_ac_drop(), ipv6_sock_ac_close() need to
take RTNL before calling ipv6_dev_ac_inc/dec. Same thing with
ipv6_sock_mc_join(), ipv6_sock_mc_drop(), ipv6_sock_mc_close() before
calling ipv6_dev_mc_inc/dec.

This patch moves ASSERT_RTNL() up a level in the call stack.

Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Reported-by: Tommi Rantala <tt.rantala@gmail.com>
Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agol2tp: fix race while getting PMTU on PPP pseudo-wire
Guillaume Nault [Wed, 3 Sep 2014 12:12:55 +0000 (14:12 +0200)]
l2tp: fix race while getting PMTU on PPP pseudo-wire

[ Upstream commit eed4d839b0cdf9d84b0a9bc63de90fd5e1e886fb ]

Use dst_entry held by sk_dst_get() to retrieve tunnel's PMTU.

The dst_mtu(__sk_dst_get(tunnel->sock)) call was racy. __sk_dst_get()
could return NULL if tunnel->sock->sk_dst_cache was reset just before the
call, thus making dst_mtu() dereference a NULL pointer:

[ 1937.661598] BUG: unable to handle kernel NULL pointer dereference at 0000000000000020
[ 1937.664005] IP: [<ffffffffa049db88>] pppol2tp_connect+0x33d/0x41e [l2tp_ppp]
[ 1937.664005] PGD daf0c067 PUD d9f93067 PMD 0
[ 1937.664005] Oops: 0000 [#1] SMP
[ 1937.664005] Modules linked in: l2tp_ppp l2tp_netlink l2tp_core ip6table_filter ip6_tables iptable_filter ip_tables ebtable_nat ebtables x_tables udp_tunnel pppoe pppox ppp_generic slhc deflate ctr twofish_generic twofish_x86_64_3way xts lrw gf128mul glue_helper twofish_x86_64 twofish_common blowfish_generic blowfish_x86_64 blowfish_common des_generic cbc xcbc rmd160 sha512_generic hmac crypto_null af_key xfrm_algo 8021q garp bridge stp llc tun atmtcp clip atm ext3 mbcache jbd iTCO_wdt coretemp kvm_intel iTCO_vendor_support kvm pcspkr evdev ehci_pci lpc_ich mfd_core i5400_edac edac_core i5k_amb shpchp button processor thermal_sys xfs crc32c_generic libcrc32c dm_mod usbhid sg hid sr_mod sd_mod cdrom crc_t10dif crct10dif_common ata_generic ahci ata_piix tg3 libahci libata uhci_hcd ptp ehci_hcd pps_core usbcore scsi_mod libphy usb_common [last unloaded: l2tp_core]
[ 1937.664005] CPU: 0 PID: 10022 Comm: l2tpstress Tainted: G           O   3.17.0-rc1 #1
[ 1937.664005] Hardware name: HP ProLiant DL160 G5, BIOS O12 08/22/2008
[ 1937.664005] task: ffff8800d8fda790 ti: ffff8800c43c4000 task.ti: ffff8800c43c4000
[ 1937.664005] RIP: 0010:[<ffffffffa049db88>]  [<ffffffffa049db88>] pppol2tp_connect+0x33d/0x41e [l2tp_ppp]
[ 1937.664005] RSP: 0018:ffff8800c43c7de8  EFLAGS: 00010282
[ 1937.664005] RAX: ffff8800da8a7240 RBX: ffff8800d8c64600 RCX: 000001c325a137b5
[ 1937.664005] RDX: 8c6318c6318c6320 RSI: 000000000000010c RDI: 0000000000000000
[ 1937.664005] RBP: ffff8800c43c7ea8 R08: 0000000000000000 R09: 0000000000000000
[ 1937.664005] R10: ffffffffa048e2c0 R11: ffff8800d8c64600 R12: ffff8800ca7a5000
[ 1937.664005] R13: ffff8800c439bf40 R14: 000000000000000c R15: 0000000000000009
[ 1937.664005] FS:  00007fd7f610f700(0000) GS:ffff88011a600000(0000) knlGS:0000000000000000
[ 1937.664005] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 1937.664005] CR2: 0000000000000020 CR3: 00000000d9d75000 CR4: 00000000000027e0
[ 1937.664005] Stack:
[ 1937.664005]  ffffffffa049da80 ffff8800d8fda790 000000000000005b ffff880000000009
[ 1937.664005]  ffff8800daf3f200 0000000000000003 ffff8800c43c7e48 ffffffff81109b57
[ 1937.664005]  ffffffff81109b0e ffffffff8114c566 0000000000000000 0000000000000000
[ 1937.664005] Call Trace:
[ 1937.664005]  [<ffffffffa049da80>] ? pppol2tp_connect+0x235/0x41e [l2tp_ppp]
[ 1937.664005]  [<ffffffff81109b57>] ? might_fault+0x9e/0xa5
[ 1937.664005]  [<ffffffff81109b0e>] ? might_fault+0x55/0xa5
[ 1937.664005]  [<ffffffff8114c566>] ? rcu_read_unlock+0x1c/0x26
[ 1937.664005]  [<ffffffff81309196>] SYSC_connect+0x87/0xb1
[ 1937.664005]  [<ffffffff813e56f7>] ? sysret_check+0x1b/0x56
[ 1937.664005]  [<ffffffff8107590d>] ? trace_hardirqs_on_caller+0x145/0x1a1
[ 1937.664005]  [<ffffffff81213dee>] ? trace_hardirqs_on_thunk+0x3a/0x3f
[ 1937.664005]  [<ffffffff8114c262>] ? spin_lock+0x9/0xb
[ 1937.664005]  [<ffffffff813092b4>] SyS_connect+0x9/0xb
[ 1937.664005]  [<ffffffff813e56d2>] system_call_fastpath+0x16/0x1b
[ 1937.664005] Code: 10 2a 84 81 e8 65 76 bd e0 65 ff 0c 25 10 bb 00 00 4d 85 ed 74 37 48 8b 85 60 ff ff ff 48 8b 80 88 01 00 00 48 8b b8 10 02 00 00 <48> 8b 47 20 ff 50 20 85 c0 74 0f 83 e8 28 89 83 10 01 00 00 89
[ 1937.664005] RIP  [<ffffffffa049db88>] pppol2tp_connect+0x33d/0x41e [l2tp_ppp]
[ 1937.664005]  RSP <ffff8800c43c7de8>
[ 1937.664005] CR2: 0000000000000020
[ 1939.559375] ---[ end trace 82d44500f28f8708 ]---

Fixes: f34c4a35d879 ("l2tp: take PMTU from tunnel UDP socket")
Signed-off-by: Guillaume Nault <g.nault@alphalink.fr>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agovxlan: fix incorrect initializer in union vxlan_addr
Gerhard Stenzel [Fri, 22 Aug 2014 19:34:16 +0000 (21:34 +0200)]
vxlan: fix incorrect initializer in union vxlan_addr

[ Upstream commit a45e92a599e77ee6a850eabdd0141633fde03915 ]

The first initializer in the following

        union vxlan_addr ipa = {
            .sin.sin_addr.s_addr = tip,
            .sa.sa_family = AF_INET,
        };

is optimised away by the compiler, due to the second initializer,
therefore initialising .sin.sin_addr.s_addr always to 0.
This results in netlink messages indicating a L3 miss never contain the
missed IP address. This was observed with GCC 4.8 and 4.9. I do not know about previous versions.
The problem affects user space programs relying on an IP address being
sent as part of a netlink message indicating a L3 miss.

Changing
            .sa.sa_family = AF_INET,
to
            .sin.sin_family = AF_INET,
fixes the problem.

Signed-off-by: Gerhard Stenzel <gerhard.stenzel@de.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoopenvswitch: fix panic with multiple vlan headers
Jiri Benc [Thu, 21 Aug 2014 19:33:44 +0000 (21:33 +0200)]
openvswitch: fix panic with multiple vlan headers

[ Upstream commit 2ba5af42a7b59ef01f9081234d8855140738defd ]

When there are multiple vlan headers present in a received frame, the first
one is put into vlan_tci and protocol is set to ETH_P_8021Q. Anything in the
skb beyond the VLAN TPID may be still non-linear, including the inner TCI
and ethertype. While ovs_flow_extract takes care of IP and IPv6 headers, it
does nothing with ETH_P_8021Q. Later, if OVS_ACTION_ATTR_POP_VLAN is
executed, __pop_vlan_tci pulls the next vlan header into vlan_tci.

This leads to two things:

1. Part of the resulting ethernet header is in the non-linear part of the
   skb. When eth_type_trans is called later as the result of
   OVS_ACTION_ATTR_OUTPUT, kernel BUGs in __skb_pull. Also, __pop_vlan_tci
   is in fact accessing random data when it reads past the TPID.

2. network_header points into the ethernet header instead of behind it.
   mac_len is set to a wrong value (10), too.

Reported-by: Yulong Pei <ypei@redhat.com>
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agopacket: handle too big packets for PACKET_V3
Eric Dumazet [Fri, 15 Aug 2014 16:16:04 +0000 (09:16 -0700)]
packet: handle too big packets for PACKET_V3

[ Upstream commit dc808110bb62b64a448696ecac3938902c92e1ab ]

af_packet can currently overwrite kernel memory by out of bound
accesses, because it assumed a [new] block can always hold one frame.

This is not generally the case, even if most existing tools do it right.

This patch clamps too long frames as API permits, and issue a one time
error on syslog.

[  394.357639] tpacket_rcv: packet too big, clamped from 5042 to 3966. macoff=82

In this example, packet header tp_snaplen was set to 3966,
and tp_len was set to 5042 (skb->len)

Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: f6fb8f100b80 ("af-packet: TPACKET_V3 flexible buffer implementation.")
Acked-by: Daniel Borkmann <dborkman@redhat.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agotcp: fix ssthresh and undo for consecutive short FRTO episodes
Neal Cardwell [Thu, 14 Aug 2014 20:13:07 +0000 (16:13 -0400)]
tcp: fix ssthresh and undo for consecutive short FRTO episodes

[ Upstream commit 0c9ab09223fe9922baeb22546c9a90d774a4bde6 ]

Fix TCP FRTO logic so that it always notices when snd_una advances,
indicating that any RTO after that point will be a new and distinct
loss episode.

Previously there was a very specific sequence that could cause FRTO to
fail to notice a new loss episode had started:

(1) RTO timer fires, enter FRTO and retransmit packet 1 in write queue
(2) receiver ACKs packet 1
(3) FRTO sends 2 more packets
(4) RTO timer fires again (should start a new loss episode)

The problem was in step (3) above, where tcp_process_loss() returned
early (in the spot marked "Step 2.b"), so that it never got to the
logic to clear icsk_retransmits. Thus icsk_retransmits stayed
non-zero. Thus in step (4) tcp_enter_loss() would see the non-zero
icsk_retransmits, decide that this RTO is not a new episode, and
decide not to cut ssthresh and remember the current cwnd and ssthresh
for undo.

There were two main consequences to the bug that we have
observed. First, ssthresh was not decreased in step (4). Second, when
there was a series of such FRTO (1-4) sequences that happened to be
followed by an FRTO undo, we would restore the cwnd and ssthresh from
before the entire series started (instead of the cwnd and ssthresh
from before the most recent RTO). This could result in cwnd and
ssthresh being restored to values much bigger than the proper values.

Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Fixes: e33099f96d99c ("tcp: implement RFC5682 F-RTO")
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agotcp: fix tcp_release_cb() to dispatch via address family for mtu_reduced()
Neal Cardwell [Thu, 14 Aug 2014 16:40:05 +0000 (12:40 -0400)]
tcp: fix tcp_release_cb() to dispatch via address family for mtu_reduced()

[ Upstream commit 4fab9071950c2021d846e18351e0f46a1cffd67b ]

Make sure we use the correct address-family-specific function for
handling MTU reductions from within tcp_release_cb().

Previously AF_INET6 sockets were incorrectly always using the IPv6
code path when sometimes they were handling IPv4 traffic and thus had
an IPv4 dst.

Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Diagnosed-by: Willem de Bruijn <willemb@google.com>
Fixes: 563d34d057862 ("tcp: dont drop MTU reduction indications")
Reviewed-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agosit: Fix ipip6_tunnel_lookup device matching criteria
Shmulik Ladkani [Thu, 14 Aug 2014 12:27:20 +0000 (15:27 +0300)]
sit: Fix ipip6_tunnel_lookup device matching criteria

[ Upstream commit bc8fc7b8f825ef17a0fb9e68c18ce94fa66ab337 ]

As of 4fddbf5d78 ("sit: strictly restrict incoming traffic to tunnel link device"),
when looking up a tunnel, tunnel's underlying interface (t->parms.link)
is verified to match incoming traffic's ingress device.

However the comparison was incorrectly based on skb->dev->iflink.

Instead, dev->ifindex should be used, which correctly represents the
interface from which the IP stack hands the ipip6 packets.

This allows setting up sit tunnels bound to vlan interfaces (otherwise
incoming ipip6 traffic on the vlan interface was dropped due to
ipip6_tunnel_lookup match failure).

Signed-off-by: Shmulik Ladkani <shmulik.ladkani@gmail.com>
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agotcp: don't use timestamp from repaired skb-s to calculate RTT (v2)
Andrey Vagin [Wed, 13 Aug 2014 12:03:10 +0000 (16:03 +0400)]
tcp: don't use timestamp from repaired skb-s to calculate RTT (v2)

[ Upstream commit 9d186cac7ffb1831e9f34cb4a3a8b22abb9dd9d4 ]

We don't know right timestamp for repaired skb-s. Wrong RTT estimations
isn't good, because some congestion modules heavily depends on it.

This patch adds the TCPCB_REPAIRED flag, which is included in
TCPCB_RETRANS.

Thanks to Eric for the advice how to fix this issue.

This patch fixes the warning:
[  879.562947] WARNING: CPU: 0 PID: 2825 at net/ipv4/tcp_input.c:3078 tcp_ack+0x11f5/0x1380()
[  879.567253] CPU: 0 PID: 2825 Comm: socket-tcpbuf-l Not tainted 3.16.0-next-20140811 #1
[  879.567829] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
[  879.568177]  0000000000000000 00000000c532680c ffff880039643d00 ffffffff817aa2d2
[  879.568776]  0000000000000000 ffff880039643d38 ffffffff8109afbd ffff880039d6ba80
[  879.569386]  ffff88003a449800 000000002983d6bd 0000000000000000 000000002983d6bc
[  879.569982] Call Trace:
[  879.570264]  [<ffffffff817aa2d2>] dump_stack+0x4d/0x66
[  879.570599]  [<ffffffff8109afbd>] warn_slowpath_common+0x7d/0xa0
[  879.570935]  [<ffffffff8109b0ea>] warn_slowpath_null+0x1a/0x20
[  879.571292]  [<ffffffff816d0a05>] tcp_ack+0x11f5/0x1380
[  879.571614]  [<ffffffff816d10bd>] tcp_rcv_established+0x1ed/0x710
[  879.571958]  [<ffffffff816dc9da>] tcp_v4_do_rcv+0x10a/0x370
[  879.572315]  [<ffffffff81657459>] release_sock+0x89/0x1d0
[  879.572642]  [<ffffffff816c81a0>] do_tcp_setsockopt.isra.36+0x120/0x860
[  879.573000]  [<ffffffff8110a52e>] ? rcu_read_lock_held+0x6e/0x80
[  879.573352]  [<ffffffff816c8912>] tcp_setsockopt+0x32/0x40
[  879.573678]  [<ffffffff81654ac4>] sock_common_setsockopt+0x14/0x20
[  879.574031]  [<ffffffff816537b0>] SyS_setsockopt+0x80/0xf0
[  879.574393]  [<ffffffff817b40a9>] system_call_fastpath+0x16/0x1b
[  879.574730] ---[ end trace a17cbc38eb8c5c00 ]---

v2: moving setting of skb->when for repaired skb-s in tcp_write_xmit,
    where it's set for other skb-s.

Fixes: 431a91242d8d ("tcp: timestamp SYN+DATA messages")
Fixes: 740b0f1841f6 ("tcp: switch rtt estimations to usec resolution")
Cc: Eric Dumazet <edumazet@google.com>
Cc: Pavel Emelyanov <xemul@parallels.com>
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoi40e: Don't stop driver probe when querying DCB config fails
Neerav Parikh [Wed, 13 Aug 2014 11:30:55 +0000 (04:30 -0700)]
i40e: Don't stop driver probe when querying DCB config fails

Commit id: 014269ff376f552363ecdab78d3d947fbe2237d9 in Linus's tree
should be queued up for stable 3.14 & 3.15 since the i40e driver will
not load when DCB is enabled, unless this patch is applied.

In case of any AQ command to query port's DCB configuration fails
during driver's probe time; the probe fails and returns an error.

This patch prevents this issue by continuing the driver probe even
when an error is returned.

Also, added an error message to dump the AQ error status to show what
error caused the failure to get the DCB configuration from firmware.

Change-ID: Ifd5663512588bca684069bb7d4fb586dd72221af
Signed-off-by: Neerav Parikh <neerav.parikh@intel.com>
Signed-off-by: Catherine Sullivan <catherine.sullivan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomyri10ge: check for DMA mapping errors
Stanislaw Gruszka [Tue, 12 Aug 2014 08:35:19 +0000 (10:35 +0200)]
myri10ge: check for DMA mapping errors

[ Upstream commit 10545937e866ccdbb7ab583031dbdcc6b14e4eb4 ]

On IOMMU systems DMA mapping can fail, we need to check for
that possibility.

Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agonet: Always untag vlan-tagged traffic on input.
Vlad Yasevich [Fri, 8 Aug 2014 18:42:13 +0000 (14:42 -0400)]
net: Always untag vlan-tagged traffic on input.

[ Upstream commit 0d5501c1c828fb97d02af50aa9d2b1a5498b94e4 ]

Currently the functionality to untag traffic on input resides
as part of the vlan module and is build only when VLAN support
is enabled in the kernel.  When VLAN is disabled, the function
vlan_untag() turns into a stub and doesn't really untag the
packets.  This seems to create an interesting interaction
between VMs supporting checksum offloading and some network drivers.

There are some drivers that do not allow the user to change
tx-vlan-offload feature of the driver.  These drivers also seem
to assume that any VLAN-tagged traffic they transmit will
have the vlan information in the vlan_tci and not in the vlan
header already in the skb.  When transmitting skbs that already
have tagged data with partial checksum set, the checksum doesn't
appear to be updated correctly by the card thus resulting in a
failure to establish TCP connections.

The following is a packet trace taken on the receiver where a
sender is a VM with a VLAN configued.  The host VM is running on
doest not have VLAN support and the outging interface on the
host is tg3:
10:12:43.503055 52:54:00:ae:42:3f > 28:d2:44:7d:c2:de, ethertype 802.1Q
(0x8100), length 78: vlan 100, p 0, ethertype IPv4, (tos 0x0, ttl 64, id 27243,
offset 0, flags [DF], proto TCP (6), length 60)
    10.0.100.1.58545 > 10.0.100.10.ircu-2: Flags [S], cksum 0xdc39 (incorrect
-> 0x48d9), seq 1069378582, win 29200, options [mss 1460,sackOK,TS val
4294837885 ecr 0,nop,wscale 7], length 0
10:12:44.505556 52:54:00:ae:42:3f > 28:d2:44:7d:c2:de, ethertype 802.1Q
(0x8100), length 78: vlan 100, p 0, ethertype IPv4, (tos 0x0, ttl 64, id 27244,
offset 0, flags [DF], proto TCP (6), length 60)
    10.0.100.1.58545 > 10.0.100.10.ircu-2: Flags [S], cksum 0xdc39 (incorrect
-> 0x44ee), seq 1069378582, win 29200, options [mss 1460,sackOK,TS val
4294838888 ecr 0,nop,wscale 7], length 0

This connection finally times out.

I've only access to the TG3 hardware in this configuration thus have
only tested this with TG3 driver.  There are a lot of other drivers
that do not permit user changes to vlan acceleration features, and
I don't know if they all suffere from a similar issue.

The patch attempt to fix this another way.  It moves the vlan header
stipping code out of the vlan module and always builds it into the
kernel network core.  This way, even if vlan is not supported on
a virtualizatoin host, the virtual machines running on top of such
host will still work with VLANs enabled.

CC: Patrick McHardy <kaber@trash.net>
CC: Nithin Nayak Sujir <nsujir@broadcom.com>
CC: Michael Chan <mchan@broadcom.com>
CC: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Acked-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agortnetlink: fix VF info size
Jiri Benc [Fri, 8 Aug 2014 14:44:32 +0000 (16:44 +0200)]
rtnetlink: fix VF info size

[ Upstream commit 945a36761fd7877660f630bbdeb4ff9ff80d1935 ]

Commit 1d8faf48c74b8 ("net/core: Add VF link state control") added new
attribute to IFLA_VF_INFO group in rtnl_fill_ifinfo but did not adjust size
of the allocated memory in if_nlmsg_size/rtnl_vfinfo_size. As the result, we
may trigger warnings in rtnl_getlink and similar functions when many VF
links are enabled, as the information does not fit into the allocated skb.

Fixes: 1d8faf48c74b8 ("net/core: Add VF link state control")
Reported-by: Yulong Pei <ypei@redhat.com>
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agonetlink: reset network header before passing to taps
Daniel Borkmann [Thu, 7 Aug 2014 20:22:47 +0000 (22:22 +0200)]
netlink: reset network header before passing to taps

[ Upstream commit 4e48ed883c72e78c5a910f8831ffe90c9b18f0ec ]

netlink doesn't set any network header offset thus when the skb is
being passed to tap devices via dev_queue_xmit_nit(), it emits klog
false positives due to it being unset like:

  ...
  [  124.990397] protocol 0000 is buggy, dev nlmon0
  [  124.990411] protocol 0000 is buggy, dev nlmon0
  ...

So just reset the network header before passing to the device; for
packet sockets that just means nothing will change - mac and net
offset hold the same value just as before.

Reported-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoLinux 3.14.21
Greg Kroah-Hartman [Thu, 9 Oct 2014 19:21:39 +0000 (12:21 -0700)]
Linux 3.14.21

10 years agomm: don't pointlessly use BUG_ON() for sanity check
Linus Torvalds [Mon, 28 Apr 2014 21:24:09 +0000 (14:24 -0700)]
mm: don't pointlessly use BUG_ON() for sanity check

commit 50f5aa8a9b248fa4262cf379863ec9a531b49737 upstream.

BUG_ON() is a big hammer, and should be used _only_ if there is some
major corruption that you cannot possibly recover from, making it
imperative that the current process (and possibly the whole machine) be
terminated with extreme prejudice.

The trivial sanity check in the vmacache code is *not* such a fatal
error.  Recovering from it is absolutely trivial, and using BUG_ON()
just makes it harder to debug for no actual advantage.

To make matters worse, the placement of the BUG_ON() (only if the range
check matched) actually makes it harder to hit the sanity check to begin
with, so _if_ there is a bug (and we just got a report from Srivatsa
Bhat that this can indeed trigger), it is harder to debug not just
because the machine is possibly dead, but because we don't have better
coverage.

BUG_ON() must *die*.  Maybe we should add a checkpatch warning for it,
because it is simply just about the worst thing you can ever do if you
hit some "this cannot happen" situation.

Reported-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Cc: Davidlohr Bueso <davidlohr@hp.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm: per-thread vma caching
Davidlohr Bueso [Mon, 7 Apr 2014 22:37:25 +0000 (15:37 -0700)]
mm: per-thread vma caching

commit 615d6e8756c87149f2d4c1b93d471bca002bd849 upstream.

This patch is a continuation of efforts trying to optimize find_vma(),
avoiding potentially expensive rbtree walks to locate a vma upon faults.
The original approach (https://lkml.org/lkml/2013/11/1/410), where the
largest vma was also cached, ended up being too specific and random,
thus further comparison with other approaches were needed.  There are
two things to consider when dealing with this, the cache hit rate and
the latency of find_vma().  Improving the hit-rate does not necessarily
translate in finding the vma any faster, as the overhead of any fancy
caching schemes can be too high to consider.

We currently cache the last used vma for the whole address space, which
provides a nice optimization, reducing the total cycles in find_vma() by
up to 250%, for workloads with good locality.  On the other hand, this
simple scheme is pretty much useless for workloads with poor locality.
Analyzing ebizzy runs shows that, no matter how many threads are
running, the mmap_cache hit rate is less than 2%, and in many situations
below 1%.

The proposed approach is to replace this scheme with a small per-thread
cache, maximizing hit rates at a very low maintenance cost.
Invalidations are performed by simply bumping up a 32-bit sequence
number.  The only expensive operation is in the rare case of a seq
number overflow, where all caches that share the same address space are
flushed.  Upon a miss, the proposed replacement policy is based on the
page number that contains the virtual address in question.  Concretely,
the following results are seen on an 80 core, 8 socket x86-64 box:

1) System bootup: Most programs are single threaded, so the per-thread
   scheme does improve ~50% hit rate by just adding a few more slots to
   the cache.

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 50.61%   | 19.90            |
| patched        | 73.45%   | 13.58            |
+----------------+----------+------------------+

2) Kernel build: This one is already pretty good with the current
   approach as we're dealing with good locality.

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 75.28%   | 11.03            |
| patched        | 88.09%   | 9.31             |
+----------------+----------+------------------+

3) Oracle 11g Data Mining (4k pages): Similar to the kernel build workload.

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 70.66%   | 17.14            |
| patched        | 91.15%   | 12.57            |
+----------------+----------+------------------+

4) Ebizzy: There's a fair amount of variation from run to run, but this
   approach always shows nearly perfect hit rates, while baseline is just
   about non-existent.  The amounts of cycles can fluctuate between
   anywhere from ~60 to ~116 for the baseline scheme, but this approach
   reduces it considerably.  For instance, with 80 threads:

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 1.06%    | 91.54            |
| patched        | 99.97%   | 14.18            |
+----------------+----------+------------------+

[akpm@linux-foundation.org: fix nommu build, per Davidlohr]
[akpm@linux-foundation.org: document vmacache_valid() logic]
[akpm@linux-foundation.org: attempt to untangle header files]
[akpm@linux-foundation.org: add vmacache_find() BUG_ON]
[hughd@google.com: add vmacache_valid_mm() (from Oleg)]
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: adjust and enhance comments]
Signed-off-by: Davidlohr Bueso <davidlohr@hp.com>
Reviewed-by: Rik van Riel <riel@redhat.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: Michel Lespinasse <walken@google.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Tested-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agovmscan: reclaim_clean_pages_from_list() must use mod_zone_page_state()
Christoph Lameter [Fri, 18 Apr 2014 22:07:10 +0000 (15:07 -0700)]
vmscan: reclaim_clean_pages_from_list() must use mod_zone_page_state()

commit 83da7510058736c09a14b9c17ec7d851940a4332 upstream.

Seems to be called with preemption enabled.  Therefore it must use
mod_zone_page_state instead.

Signed-off-by: Christoph Lameter <cl@linux.com>
Reported-by: Grygorii Strashko <grygorii.strashko@ti.com>
Tested-by: Grygorii Strashko <grygorii.strashko@ti.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Santosh Shilimkar <santosh.shilimkar@ti.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm: vmscan: shrink_slab: rename max_pass -> freeable
Vladimir Davydov [Thu, 3 Apr 2014 21:47:32 +0000 (14:47 -0700)]
mm: vmscan: shrink_slab: rename max_pass -> freeable

commit d5bc5fd3fcb7b8dfb431694a8c8052466504c10c upstream.

The name `max_pass' is misleading, because this variable actually keeps
the estimate number of freeable objects, not the maximal number of
objects we can scan in this pass, which can be twice that.  Rename it to
reflect its actual meaning.

Signed-off-by: Vladimir Davydov <vdavydov@parallels.com>
Acked-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm: vmscan: respect NUMA policy mask when shrinking slab on direct reclaim
Vladimir Davydov [Thu, 3 Apr 2014 21:47:19 +0000 (14:47 -0700)]
mm: vmscan: respect NUMA policy mask when shrinking slab on direct reclaim

commit 99120b772b52853f9a2b829a21dd44d9b20558f1 upstream.

When direct reclaim is executed by a process bound to a set of NUMA
nodes, we should scan only those nodes when possible, but currently we
will scan kmem from all online nodes even if the kmem shrinker is NUMA
aware.  That said, binding a process to a particular NUMA node won't
prevent it from shrinking inode/dentry caches from other nodes, which is
not good.  Fix this.

Signed-off-by: Vladimir Davydov <vdavydov@parallels.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: Glauber Costa <glommer@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm/filemap.c: avoid always dirtying mapping->flags on O_DIRECT
Jens Axboe [Thu, 22 May 2014 18:54:16 +0000 (11:54 -0700)]
mm/filemap.c: avoid always dirtying mapping->flags on O_DIRECT

commit 7fcbbaf18392f0b17c95e2f033c8ccf87eecde1d upstream.

In some testing I ran today (some fio jobs that spread over two nodes),
we end up spending 40% of the time in filemap_check_errors().  That
smells fishy.  Looking further, this is basically what happens:

blkdev_aio_read()
    generic_file_aio_read()
        filemap_write_and_wait_range()
            if (!mapping->nr_pages)
                filemap_check_errors()

and filemap_check_errors() always attempts two test_and_clear_bit() on
the mapping flags, thus dirtying it for every single invocation.  The
patch below tests each of these bits before clearing them, avoiding this
issue.  In my test case (4-socket box), performance went from 1.7M IOPS
to 4.0M IOPS.

Signed-off-by: Jens Axboe <axboe@fb.com>
Acked-by: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm: optimize put_mems_allowed() usage
Mel Gorman [Thu, 3 Apr 2014 21:47:24 +0000 (14:47 -0700)]
mm: optimize put_mems_allowed() usage

commit d26914d11751b23ca2e8747725f2cae10c2f2c1b upstream.

Since put_mems_allowed() is strictly optional, its a seqcount retry, we
don't need to evaluate the function if the allocation was in fact
successful, saving a smp_rmb some loads and comparisons on some relative
fast-paths.

Since the naming, get/put_mems_allowed() does suggest a mandatory
pairing, rename the interface, as suggested by Mel, to resemble the
seqcount interface.

This gives us: read_mems_allowed_begin() and read_mems_allowed_retry(),
where it is important to note that the return value of the latter call
is inverted from its previous incarnation.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm/readahead.c: fix readahead failure for memoryless NUMA nodes and limit readahead...
Raghavendra K T [Thu, 3 Apr 2014 21:48:23 +0000 (14:48 -0700)]
mm/readahead.c: fix readahead failure for memoryless NUMA nodes and limit readahead pages

commit 6d2be915e589b58cb11418cbe1f22ff90732b6ac upstream.

Currently max_sane_readahead() returns zero on the cpu whose NUMA node
has no local memory which leads to readahead failure.  Fix this
readahead failure by returning minimum of (requested pages, 512).  Users
running applications on a memory-less cpu which needs readahead such as
streaming application see considerable boost in the performance.

Result:

fadvise experiment with FADV_WILLNEED on a PPC machine having memoryless
CPU with 1GB testfile (12 iterations) yielded around 46.66% improvement.

fadvise experiment with FADV_WILLNEED on a x240 machine with 1GB
testfile 32GB* 4G RAM numa machine (12 iterations) showed no impact on
the normal NUMA cases w/ patch.

  Kernel       Avg  Stddev
  base      7.4975   3.92%
  patched   7.4174   3.26%

[Andrew: making return value PAGE_SIZE independent]
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Raghavendra K T <raghavendra.kt@linux.vnet.ibm.com>
Acked-by: Jan Kara <jack@suse.cz>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm, compaction: ignore pageblock skip when manually invoking compaction
David Rientjes [Thu, 3 Apr 2014 21:47:23 +0000 (14:47 -0700)]
mm, compaction: ignore pageblock skip when manually invoking compaction

commit 91ca9186484809c57303b33778d841cc28f696ed upstream.

The cached pageblock hint should be ignored when triggering compaction
through /proc/sys/vm/compact_memory so all eligible memory is isolated.
Manually invoking compaction is known to be expensive, there's no need
to skip pageblocks based on heuristics (mainly for debugging).

Signed-off-by: David Rientjes <rientjes@google.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm, compaction: determine isolation mode only once
David Rientjes [Mon, 7 Apr 2014 22:37:34 +0000 (15:37 -0700)]
mm, compaction: determine isolation mode only once

commit da1c67a76f7cf2b3404823d24f9f10fa91aa5dc5 upstream.

The conditions that control the isolation mode in
isolate_migratepages_range() do not change during the iteration, so
extract them out and only define the value once.

This actually does have an effect, gcc doesn't optimize it itself because
of cc->sync.

Signed-off-by: David Rientjes <rientjes@google.com>
Cc: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm/compaction: clean-up code on success of ballon isolation
Joonsoo Kim [Mon, 7 Apr 2014 22:37:07 +0000 (15:37 -0700)]
mm/compaction: clean-up code on success of ballon isolation

commit b6c750163c0d138f5041d95fcdbd1094b6928057 upstream.

It is just for clean-up to reduce code size and improve readability.
There is no functional change.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm/compaction: check pageblock suitability once per pageblock
Joonsoo Kim [Mon, 7 Apr 2014 22:37:06 +0000 (15:37 -0700)]
mm/compaction: check pageblock suitability once per pageblock

commit c122b2087ab94192f2b937e47b563a9c4e688ece upstream.

isolation_suitable() and migrate_async_suitable() is used to be sure
that this pageblock range is fine to be migragted.  It isn't needed to
call it on every page.  Current code do well if not suitable, but, don't
do well when suitable.

1) It re-checks isolation_suitable() on each page of a pageblock that was
   already estabilished as suitable.
2) It re-checks migrate_async_suitable() on each page of a pageblock that
   was not entered through the next_pageblock: label, because
   last_pageblock_nr is not otherwise updated.

This patch fixes situation by 1) calling isolation_suitable() only once
per pageblock and 2) always updating last_pageblock_nr to the pageblock
that was just checked.

Additionally, move PageBuddy() check after pageblock unit check, since
pageblock check is the first thing we should do and makes things more
simple.

[vbabka@suse.cz: rephrase commit description]
Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm/compaction: change the timing to check to drop the spinlock
Joonsoo Kim [Mon, 7 Apr 2014 22:37:05 +0000 (15:37 -0700)]
mm/compaction: change the timing to check to drop the spinlock

commit be1aa03b973c7dcdc576f3503f7a60429825c35d upstream.

It is odd to drop the spinlock when we scan (SWAP_CLUSTER_MAX - 1) th
pfn page.  This may results in below situation while isolating
migratepage.

1. try isolate 0x0 ~ 0x200 pfn pages.
2. When low_pfn is 0x1ff, ((low_pfn+1) % SWAP_CLUSTER_MAX) == 0, so drop
   the spinlock.
3. Then, to complete isolating, retry to aquire the lock.

I think that it is better to use SWAP_CLUSTER_MAX th pfn for checking the
criteria about dropping the lock.  This has no harm 0x0 pfn, because, at
this time, locked variable would be false.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agodrbd: fix regression 'out of mem, failed to invoke fence-peer helper'
Lars Ellenberg [Wed, 9 Jul 2014 19:18:32 +0000 (21:18 +0200)]
drbd: fix regression 'out of mem, failed to invoke fence-peer helper'

commit bbc1c5e8ad6dfebf9d13b8a4ccdf66c92913eac9 upstream.

Since linux kernel 3.13, kthread_run() internally uses
wait_for_completion_killable().  We sometimes may use kthread_run()
while we still have a signal pending, which we used to kick our threads
out of potentially blocking network functions, causing kthread_run() to
mistake that as a new fatal signal and fail.

Fix: flush_signals() before kthread_run().

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm/compaction: do not call suitable_migration_target() on every page
Joonsoo Kim [Mon, 7 Apr 2014 22:37:04 +0000 (15:37 -0700)]
mm/compaction: do not call suitable_migration_target() on every page

commit 01ead5340bcf5f3a1cd2452c75516d0ef4d908d7 upstream.

suitable_migration_target() checks that pageblock is suitable for
migration target.  In isolate_freepages_block(), it is called on every
page and this is inefficient.  So make it called once per pageblock.

suitable_migration_target() also checks if page is highorder or not, but
it's criteria for highorder is pageblock order.  So calling it once
within pageblock range has no problem.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm/compaction: disallow high-order page for migration target
Joonsoo Kim [Mon, 7 Apr 2014 22:37:03 +0000 (15:37 -0700)]
mm/compaction: disallow high-order page for migration target

commit 7d348b9ea64db0a315d777ce7d4b06697f946503 upstream.

Purpose of compaction is to get a high order page.  Currently, if we
find high-order page while searching migration target page, we break it
to order-0 pages and use them as migration target.  It is contrary to
purpose of compaction, so disallow high-order page to be used for
migration target.

Additionally, clean-up logic in suitable_migration_target() to simplify
the code.  There is no functional changes from this clean-up.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm, compaction: avoid isolating pinned pages
David Rientjes [Thu, 3 Apr 2014 21:48:00 +0000 (14:48 -0700)]
mm, compaction: avoid isolating pinned pages

commit 119d6d59dcc0980dcd581fdadb6b2033b512a473 upstream.

Page migration will fail for memory that is pinned in memory with, for
example, get_user_pages().  In this case, it is unnecessary to take
zone->lru_lock or isolating the page and passing it to page migration
which will ultimately fail.

This is a racy check, the page can still change from under us, but in
that case we'll just fail later when attempting to move the page.

This avoids very expensive memory compaction when faulting transparent
hugepages after pinning a lot of memory with a Mellanox driver.

On a 128GB machine and pinning ~120GB of memory, before this patch we
see the enormous disparity in the number of page migration failures
because of the pinning (from /proc/vmstat):

compact_pages_moved 8450
compact_pagemigrate_failed 15614415

0.05% of pages isolated are successfully migrated and explicitly
triggering memory compaction takes 102 seconds.  After the patch:

compact_pages_moved 9197
compact_pagemigrate_failed 7

99.9% of pages isolated are now successfully migrated in this
configuration and memory compaction takes less than one second.

Signed-off-by: David Rientjes <rientjes@google.com>
Acked-by: Hugh Dickins <hughd@google.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoswap: change swap_list_head to plist, add swap_avail_head
Dan Streetman [Wed, 4 Jun 2014 23:09:59 +0000 (16:09 -0700)]
swap: change swap_list_head to plist, add swap_avail_head

commit 18ab4d4ced0817421e6db6940374cc39d28d65da upstream.

Originally get_swap_page() started iterating through the singly-linked
list of swap_info_structs using swap_list.next or highest_priority_index,
which both were intended to point to the highest priority active swap
target that was not full.  The first patch in this series changed the
singly-linked list to a doubly-linked list, and removed the logic to start
at the highest priority non-full entry; it starts scanning at the highest
priority entry each time, even if the entry is full.

Replace the manually ordered swap_list_head with a plist, swap_active_head.
Add a new plist, swap_avail_head.  The original swap_active_head plist
contains all active swap_info_structs, as before, while the new
swap_avail_head plist contains only swap_info_structs that are active and
available, i.e. not full.  Add a new spinlock, swap_avail_lock, to protect
the swap_avail_head list.

Mel Gorman suggested using plists since they internally handle ordering
the list entries based on priority, which is exactly what swap was doing
manually.  All the ordering code is now removed, and swap_info_struct
entries and simply added to their corresponding plist and automatically
ordered correctly.

Using a new plist for available swap_info_structs simplifies and
optimizes get_swap_page(), which no longer has to iterate over full
swap_info_structs.  Using a new spinlock for swap_avail_head plist
allows each swap_info_struct to add or remove themselves from the
plist when they become full or not-full; previously they could not
do so because the swap_info_struct->lock is held when they change
from full<->not-full, and the swap_lock protecting the main
swap_active_head must be ordered before any swap_info_struct->lock.

Signed-off-by: Dan Streetman <ddstreet@ieee.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: Shaohua Li <shli@fusionio.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Cc: Weijie Yang <weijieut@gmail.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Bob Liu <bob.liu@oracle.com>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agolib/plist: add plist_requeue
Dan Streetman [Wed, 4 Jun 2014 23:09:57 +0000 (16:09 -0700)]
lib/plist: add plist_requeue

commit a75f232ce0fe38bd01301899ecd97ffd0254316a upstream.

Add plist_requeue(), which moves the specified plist_node after all other
same-priority plist_nodes in the list.  This is essentially an optimized
plist_del() followed by plist_add().

This is needed by swap, which (with the next patch in this set) uses a
plist of available swap devices.  When a swap device (either a swap
partition or swap file) are added to the system with swapon(), the device
is added to a plist, ordered by the swap device's priority.  When swap
needs to allocate a page from one of the swap devices, it takes the page
from the first swap device on the plist, which is the highest priority
swap device.  The swap device is left in the plist until all its pages are
used, and then removed from the plist when it becomes full.

However, as described in man 2 swapon, swap must allocate pages from swap
devices with the same priority in round-robin order; to do this, on each
swap page allocation, swap uses a page from the first swap device in the
plist, and then calls plist_requeue() to move that swap device entry to
after any other same-priority swap devices.  The next swap page allocation
will again use a page from the first swap device in the plist and requeue
it, and so on, resulting in round-robin usage of equal-priority swap
devices.

Also add plist_test_requeue() test function, for use by plist_test() to
test plist_requeue() function.

Signed-off-by: Dan Streetman <ddstreet@ieee.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Shaohua Li <shli@fusionio.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Cc: Weijie Yang <weijieut@gmail.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Bob Liu <bob.liu@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agolib/plist: add helper functions
Dan Streetman [Wed, 4 Jun 2014 23:09:55 +0000 (16:09 -0700)]
lib/plist: add helper functions

commit fd16618e12a05df79a3439d72d5ffdac5d34f3da upstream.

Add PLIST_HEAD() to plist.h, equivalent to LIST_HEAD() from list.h, to
define and initialize a struct plist_head.

Add plist_for_each_continue() and plist_for_each_entry_continue(),
equivalent to list_for_each_continue() and list_for_each_entry_continue(),
to iterate over a plist continuing after the current position.

Add plist_prev() and plist_next(), equivalent to (struct list_head*)->prev
and ->next, implemented by list_prev_entry() and list_next_entry(), to
access the prev/next struct plist_node entry.  These are needed because
unlike struct list_head, direct access of the prev/next struct plist_node
isn't possible; the list must be navigated via the contained struct
list_head.  e.g.  instead of accessing the prev by list_prev_entry(node,
node_list) it can be accessed by plist_prev(node).

Signed-off-by: Dan Streetman <ddstreet@ieee.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Shaohua Li <shli@fusionio.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Cc: Weijie Yang <weijieut@gmail.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Bob Liu <bob.liu@oracle.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoswap: change swap_info singly-linked list to list_head
Dan Streetman [Wed, 4 Jun 2014 23:09:53 +0000 (16:09 -0700)]
swap: change swap_info singly-linked list to list_head

commit adfab836f4908deb049a5128082719e689eed964 upstream.

The logic controlling the singly-linked list of swap_info_struct entries
for all active, i.e.  swapon'ed, swap targets is rather complex, because:

 - it stores the entries in priority order
 - there is a pointer to the highest priority entry
 - there is a pointer to the highest priority not-full entry
 - there is a highest_priority_index variable set outside the swap_lock
 - swap entries of equal priority should be used equally

this complexity leads to bugs such as: https://lkml.org/lkml/2014/2/13/181
where different priority swap targets are incorrectly used equally.

That bug probably could be solved with the existing singly-linked lists,
but I think it would only add more complexity to the already difficult to
understand get_swap_page() swap_list iteration logic.

The first patch changes from a singly-linked list to a doubly-linked list
using list_heads; the highest_priority_index and related code are removed
and get_swap_page() starts each iteration at the highest priority
swap_info entry, even if it's full.  While this does introduce unnecessary
list iteration (i.e.  Schlemiel the painter's algorithm) in the case where
one or more of the highest priority entries are full, the iteration and
manipulation code is much simpler and behaves correctly re: the above bug;
and the fourth patch removes the unnecessary iteration.

The second patch adds some minor plist helper functions; nothing new
really, just functions to match existing regular list functions.  These
are used by the next two patches.

The third patch adds plist_requeue(), which is used by get_swap_page() in
the next patch - it performs the requeueing of same-priority entries
(which moves the entry to the end of its priority in the plist), so that
all equal-priority swap_info_structs get used equally.

The fourth patch converts the main list into a plist, and adds a new plist
that contains only swap_info entries that are both active and not full.
As Mel suggested using plists allows removing all the ordering code from
swap - plists handle ordering automatically.  The list naming is also
clarified now that there are two lists, with the original list changed
from swap_list_head to swap_active_head and the new list named
swap_avail_head.  A new spinlock is also added for the new list, so
swap_info entries can be added or removed from the new list immediately as
they become full or not full.

This patch (of 4):

Replace the singly-linked list tracking active, i.e.  swapon'ed,
swap_info_struct entries with a doubly-linked list using struct
list_heads.  Simplify the logic iterating and manipulating the list of
entries, especially get_swap_page(), by using standard list_head
functions, and removing the highest priority iteration logic.

The change fixes the bug:
https://lkml.org/lkml/2014/2/13/181
in which different priority swap entries after the highest priority entry
are incorrectly used equally in pairs.  The swap behavior is now as
advertised, i.e. different priority swap entries are used in order, and
equal priority swap targets are used concurrently.

Signed-off-by: Dan Streetman <ddstreet@ieee.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: Shaohua Li <shli@fusionio.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Cc: Weijie Yang <weijieut@gmail.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Bob Liu <bob.liu@oracle.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm: exclude memoryless nodes from zone_reclaim
Michal Hocko [Mon, 7 Apr 2014 22:37:01 +0000 (15:37 -0700)]
mm: exclude memoryless nodes from zone_reclaim

commit 70ef57e6c22c3323dce179b7d0d433c479266612 upstream.

We had a report about strange OOM killer strikes on a PPC machine
although there was a lot of swap free and a tons of anonymous memory
which could be swapped out.  In the end it turned out that the OOM was a
side effect of zone reclaim which wasn't unmapping and swapping out and
so the system was pushed to the OOM.  Although this sounds like a bug
somewhere in the kswapd vs.  zone reclaim vs.  direct reclaim
interaction numactl on the said hardware suggests that the zone reclaim
should not have been set in the first place:

  node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
  node 0 size: 0 MB
  node 0 free: 0 MB
  node 2 cpus:
  node 2 size: 7168 MB
  node 2 free: 6019 MB
  node distances:
  node   0   2
  0:  10  40
  2:  40  10

So all the CPUs are associated with Node0 which doesn't have any memory
while Node2 contains all the available memory.  Node distances cause an
automatic zone_reclaim_mode enabling.

Zone reclaim is intended to keep the allocations local but this doesn't
make any sense on the memoryless nodes.  So let's exclude such nodes for
init_zone_allows_reclaim which evaluates zone reclaim behavior and
suitable reclaim_nodes.

Signed-off-by: Michal Hocko <mhocko@suse.cz>
Acked-by: David Rientjes <rientjes@google.com>
Acked-by: Nishanth Aravamudan <nacc@linux.vnet.ibm.com>
Tested-by: Nishanth Aravamudan <nacc@linux.vnet.ibm.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agojiffies: Fix timeval conversion to jiffies
Andrew Hunter [Thu, 4 Sep 2014 21:17:16 +0000 (14:17 -0700)]
jiffies: Fix timeval conversion to jiffies

commit d78c9300c51d6ceed9f6d078d4e9366f259de28c upstream.

timeval_to_jiffies tried to round a timeval up to an integral number
of jiffies, but the logic for doing so was incorrect: intervals
corresponding to exactly N jiffies would become N+1. This manifested
itself particularly repeatedly stopping/starting an itimer:

setitimer(ITIMER_PROF, &val, NULL);
setitimer(ITIMER_PROF, NULL, &val);

would add a full tick to val, _even if it was exactly representable in
terms of jiffies_ (say, the result of a previous rounding.)  Doing
this repeatedly would cause unbounded growth in val.  So fix the math.

Here's what was wrong with the conversion: we essentially computed
(eliding seconds)

jiffies = usec  * (NSEC_PER_USEC/TICK_NSEC)

by using scaling arithmetic, which took the best approximation of
NSEC_PER_USEC/TICK_NSEC with denominator of 2^USEC_JIFFIE_SC =
x/(2^USEC_JIFFIE_SC), and computed:

jiffies = (usec * x) >> USEC_JIFFIE_SC

and rounded this calculation up in the intermediate form (since we
can't necessarily exactly represent TICK_NSEC in usec.) But the
scaling arithmetic is a (very slight) *over*approximation of the true
value; that is, instead of dividing by (1 usec/ 1 jiffie), we
effectively divided by (1 usec/1 jiffie)-epsilon (rounding
down). This would normally be fine, but we want to round timeouts up,
and we did so by adding 2^USEC_JIFFIE_SC - 1 before the shift; this
would be fine if our division was exact, but dividing this by the
slightly smaller factor was equivalent to adding just _over_ 1 to the
final result (instead of just _under_ 1, as desired.)

In particular, with HZ=1000, we consistently computed that 10000 usec
was 11 jiffies; the same was true for any exact multiple of
TICK_NSEC.

We could possibly still round in the intermediate form, adding
something less than 2^USEC_JIFFIE_SC - 1, but easier still is to
convert usec->nsec, round in nanoseconds, and then convert using
time*spec*_to_jiffies.  This adds one constant multiplication, and is
not observably slower in microbenchmarks on recent x86 hardware.

Tested: the following program:

int main() {
  struct itimerval zero = {{0, 0}, {0, 0}};
  /* Initially set to 10 ms. */
  struct itimerval initial = zero;
  initial.it_interval.tv_usec = 10000;
  setitimer(ITIMER_PROF, &initial, NULL);
  /* Save and restore several times. */
  for (size_t i = 0; i < 10; ++i) {
    struct itimerval prev;
    setitimer(ITIMER_PROF, &zero, &prev);
    /* on old kernels, this goes up by TICK_USEC every iteration */
    printf("previous value: %ld %ld %ld %ld\n",
           prev.it_interval.tv_sec, prev.it_interval.tv_usec,
           prev.it_value.tv_sec, prev.it_value.tv_usec);
    setitimer(ITIMER_PROF, &prev, NULL);
  }
    return 0;
}

Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Paul Turner <pjt@google.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Prarit Bhargava <prarit@redhat.com>
Reviewed-by: Paul Turner <pjt@google.com>
Reported-by: Aaron Jacobs <jacobsa@google.com>
Signed-off-by: Andrew Hunter <ahh@google.com>
[jstultz: Tweaked to apply to 3.17-rc]
Signed-off-by: John Stultz <john.stultz@linaro.org>
[bwh: Backported to 3.16: adjust filename]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomedia: vb2: fix VBI/poll regression
Hans Verkuil [Sat, 20 Sep 2014 19:16:35 +0000 (16:16 -0300)]
media: vb2: fix VBI/poll regression

commit 58d75f4b1ce26324b4d809b18f94819843a98731 upstream.

The recent conversion of saa7134 to vb2 unconvered a poll() bug that
broke the teletext applications alevt and mtt. These applications
expect that calling poll() without having called VIDIOC_STREAMON will
cause poll() to return POLLERR. That did not happen in vb2.

This patch fixes that behavior. It also fixes what should happen when
poll() is called when STREAMON is called but no buffers have been
queued. In that case poll() will also return POLLERR, but only for
capture queues since output queues will always return POLLOUT
anyway in that situation.

This brings the vb2 behavior in line with the old videobuf behavior.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm: numa: Do not mark PTEs pte_numa when splitting huge pages
Mel Gorman [Thu, 2 Oct 2014 18:47:42 +0000 (19:47 +0100)]
mm: numa: Do not mark PTEs pte_numa when splitting huge pages

commit abc40bd2eeb77eb7c2effcaf63154aad929a1d5f upstream.

This patch reverts 1ba6e0b50b ("mm: numa: split_huge_page: transfer the
NUMA type from the pmd to the pte"). If a huge page is being split due
a protection change and the tail will be in a PROT_NONE vma then NUMA
hinting PTEs are temporarily created in the protected VMA.

 VM_RW|VM_PROTNONE
|-----------------|
      ^
      split here

In the specific case above, it should get fixed up by change_pte_range()
but there is a window of opportunity for weirdness to happen. Similarly,
if a huge page is shrunk and split during a protection update but before
pmd_numa is cleared then a pte_numa can be left behind.

Instead of adding complexity trying to deal with the case, this patch
will not mark PTEs NUMA when splitting a huge page. NUMA hinting faults
will not be triggered which is marginal in comparison to the complexity
in dealing with the corner cases during THP split.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm, thp: move invariant bug check out of loop in __split_huge_page_map
Waiman Long [Wed, 6 Aug 2014 23:05:36 +0000 (16:05 -0700)]
mm, thp: move invariant bug check out of loop in __split_huge_page_map

commit f8303c2582b889351e261ff18c4d8eb197a77db2 upstream.

In __split_huge_page_map(), the check for page_mapcount(page) is
invariant within the for loop.  Because of the fact that the macro is
implemented using atomic_read(), the redundant check cannot be optimized
away by the compiler leading to unnecessary read to the page structure.

This patch moves the invariant bug check out of the loop so that it will
be done only once.  On a 3.16-rc1 based kernel, the execution time of a
microbenchmark that broke up 1000 transparent huge pages using munmap()
had an execution time of 38,245us and 38,548us with and without the
patch respectively.  The performance gain is about 1%.

Signed-off-by: Waiman Long <Waiman.Long@hp.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Cc: Scott J Norton <scott.norton@hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agohugetlb: ensure hugepage access is denied if hugepages are not supported
Nishanth Aravamudan [Tue, 6 May 2014 19:50:00 +0000 (12:50 -0700)]
hugetlb: ensure hugepage access is denied if hugepages are not supported

commit 457c1b27ed56ec472d202731b12417bff023594a upstream.

Currently, I am seeing the following when I `mount -t hugetlbfs /none
/dev/hugetlbfs`, and then simply do a `ls /dev/hugetlbfs`.  I think it's
related to the fact that hugetlbfs is properly not correctly setting
itself up in this state?:

  Unable to handle kernel paging request for data at address 0x00000031
  Faulting instruction address: 0xc000000000245710
  Oops: Kernel access of bad area, sig: 11 [#1]
  SMP NR_CPUS=2048 NUMA pSeries
  ....

In KVM guests on Power, in a guest not backed by hugepages, we see the
following:

  AnonHugePages:         0 kB
  HugePages_Total:       0
  HugePages_Free:        0
  HugePages_Rsvd:        0
  HugePages_Surp:        0
  Hugepagesize:         64 kB

HPAGE_SHIFT == 0 in this configuration, which indicates that hugepages
are not supported at boot-time, but this is only checked in
hugetlb_init().  Extract the check to a helper function, and use it in a
few relevant places.

This does make hugetlbfs not supported (not registered at all) in this
environment.  I believe this is fine, as there are no valid hugepages
and that won't change at runtime.

[akpm@linux-foundation.org: use pr_info(), per Mel]
[akpm@linux-foundation.org: fix build when HPAGE_SHIFT is undefined]
Signed-off-by: Nishanth Aravamudan <nacc@linux.vnet.ibm.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoCIFS: Fix SMB2 readdir error handling
Pavel Shilovsky [Mon, 18 Aug 2014 16:49:57 +0000 (20:49 +0400)]
CIFS: Fix SMB2 readdir error handling

commit 52755808d4525f4d5b86d112d36ffc7a46f3fb48 upstream.

SMB2 servers indicates the end of a directory search with
STATUS_NO_MORE_FILE error code that is not processed now.
This causes generic/257 xfstest to fail. Fix this by triggering
the end of search by this error code in SMB2_query_directory.

Also when negotiating CIFS protocol we tell the server to close
the search automatically at the end and there is no need to do
it itself. In the case of SMB2 protocol, we need to close it
explicitly - separate close directory checks for different
protocols.

Signed-off-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Steve French <smfrench@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoring-buffer: Fix infinite spin in reading buffer
Steven Rostedt (Red Hat) [Thu, 2 Oct 2014 20:51:18 +0000 (16:51 -0400)]
ring-buffer: Fix infinite spin in reading buffer

commit 24607f114fd14f2f37e3e0cb3d47bce96e81e848 upstream.

Commit 651e22f2701b "ring-buffer: Always reset iterator to reader page"
fixed one bug but in the process caused another one. The reset is to
update the header page, but that fix also changed the way the cached
reads were updated. The cache reads are used to test if an iterator
needs to be updated or not.

A ring buffer iterator, when created, disables writes to the ring buffer
but does not stop other readers or consuming reads from happening.
Although all readers are synchronized via a lock, they are only
synchronized when in the ring buffer functions. Those functions may
be called by any number of readers. The iterator continues down when
its not interrupted by a consuming reader. If a consuming read
occurs, the iterator starts from the beginning of the buffer.

The way the iterator sees that a consuming read has happened since
its last read is by checking the reader "cache". The cache holds the
last counts of the read and the reader page itself.

Commit 651e22f2701b changed what was saved by the cache_read when
the rb_iter_reset() occurred, making the iterator never match the cache.
Then if the iterator calls rb_iter_reset(), it will go into an
infinite loop by checking if the cache doesn't match, doing the reset
and retrying, just to see that the cache still doesn't match! Which
should never happen as the reset is suppose to set the cache to the
current value and there's locks that keep a consuming reader from
having access to the data.

Fixes: 651e22f2701b "ring-buffer: Always reset iterator to reader page"
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoinit/Kconfig: Fix HAVE_FUTEX_CMPXCHG to not break up the EXPERT menu
Josh Triplett [Fri, 3 Oct 2014 23:19:24 +0000 (16:19 -0700)]
init/Kconfig: Fix HAVE_FUTEX_CMPXCHG to not break up the EXPERT menu

commit 62b4d2041117f35ab2409c9f5c4b8d3dc8e59d0f upstream.

commit 03b8c7b623c80af264c4c8d6111e5c6289933666 ("futex: Allow
architectures to skip futex_atomic_cmpxchg_inatomic() test") added the
HAVE_FUTEX_CMPXCHG symbol right below FUTEX.  This placed it right in
the middle of the options for the EXPERT menu.  However,
HAVE_FUTEX_CMPXCHG does not depend on EXPERT or FUTEX, so Kconfig stops
placing items in the EXPERT menu, and displays the remaining several
EXPERT items (starting with EPOLL) directly in the General Setup menu.

Since both users of HAVE_FUTEX_CMPXCHG only select it "if FUTEX", make
HAVE_FUTEX_CMPXCHG itself depend on FUTEX.  With this change, the
subsequent items display as part of the EXPERT menu again; the EMBEDDED
menu now appears as the next top-level item in the General Setup menu,
which makes General Setup much shorter and more usable.

Signed-off-by: Josh Triplett <josh@joshtriplett.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoFix problem recognizing symlinks
Steve French [Thu, 25 Sep 2014 06:26:55 +0000 (01:26 -0500)]
Fix problem recognizing symlinks

commit 19e81573fca7b87ced7701e01ba164b968d929bd upstream.

Changeset eb85d94bd introduced a problem where if a cifs open
fails during query info of a file we
will still try to close the file (happens with certain types
of reparse points) even though the file handle is not valid.

In addition for SMB2/SMB3 we were not mapping the return code returned
by Windows when trying to open a file (like a Windows NFS symlink)
which is a reparse point.

Signed-off-by: Steve French <smfrench@gmail.com>
Reviewed-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agodrm/i915: Flush the PTEs after updating them before suspend
Chris Wilson [Thu, 25 Sep 2014 09:13:12 +0000 (10:13 +0100)]
drm/i915: Flush the PTEs after updating them before suspend

commit 91e56499304f3d612053a9cf17f350868182c7d8 upstream.

As we use WC updates of the PTE, we are responsible for notifying the
hardware when to flush its TLBs. Do so after we zap all the PTEs before
suspend (and the BIOS tries to read our GTT).

Fixes a regression from

commit 828c79087cec61eaf4c76bb32c222fbe35ac3930
Author: Ben Widawsky <benjamin.widawsky@intel.com>
Date:   Wed Oct 16 09:21:30 2013 -0700

    drm/i915: Disable GGTT PTEs on GEN6+ suspend

that survived and continue to cause harm even after

commit e568af1c626031925465a5caaab7cca1303d55c7
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Wed Mar 26 20:08:20 2014 +0100

    drm/i915: Undo gtt scratch pte unmapping again

v2: Trivial rebase.
v3: Fixes requires pointer dances.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=82340
Tested-by: ming.yao@intel.com
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Takashi Iwai <tiwai@suse.de>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Todd Previte <tprevite@gmail.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomd/raid5: disable 'DISCARD' by default due to safety concerns.
NeilBrown [Thu, 2 Oct 2014 03:45:00 +0000 (13:45 +1000)]
md/raid5: disable 'DISCARD' by default due to safety concerns.

commit 8e0e99ba64c7ba46133a7c8a3e3f7de01f23bd93 upstream.

It has come to my attention (thanks Martin) that 'discard_zeroes_data'
is only a hint.  Some devices in some cases don't do what it
says on the label.

The use of DISCARD in RAID5 depends on reads from discarded regions
being predictably zero.  If a write to a previously discarded region
performs a read-modify-write cycle it assumes that the parity block
was consistent with the data blocks.  If all were zero, this would
be the case.  If some are and some aren't this would not be the case.
This could lead to data corruption after a device failure when
data needs to be reconstructed from the parity.

As we cannot trust 'discard_zeroes_data', ignore it by default
and so disallow DISCARD on all raid4/5/6 arrays.

As many devices are trustworthy, and as there are benefits to using
DISCARD, add a module parameter to over-ride this caution and cause
DISCARD to work if discard_zeroes_data is set.

If a site want to enable DISCARD on some arrays but not on others they
should select DISCARD support at the filesystem level, and set the
raid456 module parameter.
    raid456.devices_handle_discard_safely=Y

As this is a data-safety issue, I believe this patch is suitable for
-stable.
DISCARD support for RAID456 was added in 3.7

Cc: Shaohua Li <shli@kernel.org>
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Heinz Mauelshagen <heinzm@redhat.com>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Acked-by: Mike Snitzer <snitzer@redhat.com>
Fixes: 620125f2bf8ff0c4969b79653b54d7bcc9d40637
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agocpufreq: integrator: fix integrator_cpufreq_remove return type
Arnd Bergmann [Fri, 26 Sep 2014 20:19:12 +0000 (22:19 +0200)]
cpufreq: integrator: fix integrator_cpufreq_remove return type

commit d62dbf77f7dfaa6fb455b4b9828069a11965929c upstream.

When building this driver as a module, we get a helpful warning
about the return type:

drivers/cpufreq/integrator-cpufreq.c:232:2: warning: initialization from incompatible pointer type
  .remove = __exit_p(integrator_cpufreq_remove),

If the remove callback returns void, the caller gets an undefined
value as it expects an integer to be returned. This fixes the
problem by passing down the value from cpufreq_unregister_driver.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agomm: migrate: Close race between migration completion and mprotect
Mel Gorman [Thu, 2 Oct 2014 18:47:41 +0000 (19:47 +0100)]
mm: migrate: Close race between migration completion and mprotect

commit d3cb8bf6081b8b7a2dabb1264fe968fd870fa595 upstream.

A migration entry is marked as write if pte_write was true at the time the
entry was created. The VMA protections are not double checked when migration
entries are being removed as mprotect marks write-migration-entries as
read. It means that potentially we take a spurious fault to mark PTEs write
again but it's straight-forward. However, there is a race between write
migrations being marked read and migrations finishing. This potentially
allows a PTE to be write that should have been read. Close this race by
double checking the VMA permissions using maybe_mkwrite when migration
completes.

[torvalds@linux-foundation.org: use maybe_mkwrite]
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoperf: fix perf bug in fork()
Peter Zijlstra [Thu, 2 Oct 2014 23:17:02 +0000 (16:17 -0700)]
perf: fix perf bug in fork()

commit 6c72e3501d0d62fc064d3680e5234f3463ec5a86 upstream.

Oleg noticed that a cleanup by Sylvain actually uncovered a bug; by
calling perf_event_free_task() when failing sched_fork() we will not yet
have done the memset() on ->perf_event_ctxp[] and will therefore try and
'free' the inherited contexts, which are still in use by the parent
process.  This is bad..

Suggested-by: Oleg Nesterov <oleg@redhat.com>
Reported-by: Oleg Nesterov <oleg@redhat.com>
Reported-by: Sylvain 'ythier' Hitier <sylvain.hitier@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoudf: Avoid infinite loop when processing indirect ICBs
Jan Kara [Thu, 4 Sep 2014 12:06:55 +0000 (14:06 +0200)]
udf: Avoid infinite loop when processing indirect ICBs

commit c03aa9f6e1f938618e6db2e23afef0574efeeb65 upstream.

We did not implement any bound on number of indirect ICBs we follow when
loading inode. Thus corrupted medium could cause kernel to go into an
infinite loop, possibly causing a stack overflow.

Fix the possible stack overflow by removing recursion from
__udf_read_inode() and limit number of indirect ICBs we follow to avoid
infinite loops.

Signed-off-by: Jan Kara <jack@suse.cz>
Cc: Chuck Ebbert <cebbert.lkml@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoLinux 3.14.20
Greg Kroah-Hartman [Sun, 5 Oct 2014 21:52:37 +0000 (14:52 -0700)]
Linux 3.14.20

10 years agoARM: DRA7: Add support for soc_is_dra74x() and soc_is_dra72x() variants
Rajendra Nayak [Thu, 28 Aug 2014 01:38:22 +0000 (19:38 -0600)]
ARM: DRA7: Add support for soc_is_dra74x() and soc_is_dra72x() variants

commit af438fec6cb99fc2e2faf8b16b865af26ce722e6 upstream.

Use the corresponding compatibles to identify the devices.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Acked-by: Nishanth Menon <nm@ti.com>
Tested-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoclk: qcom: mdp_lut_clk is a child of mdp_src
Stephen Boyd [Wed, 9 Jul 2014 01:36:06 +0000 (18:36 -0700)]
clk: qcom: mdp_lut_clk is a child of mdp_src

commit f87dfcabc6f173cc811d185d33327f50a8c88399 upstream.

The mdp_lut_clk isn't a child of the mdp_clk. Instead it's the
child of the mdp_src clock. Fix it.

Fixes: 6d00b56fe "clk: qcom: Add support for MSM8960's multimedia clock controller (MMCC)"
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoclk: qcom: Fix MN frequency tables, parent map, and jpegd
Stephen Boyd [Wed, 9 Jul 2014 01:36:06 +0000 (18:36 -0700)]
clk: qcom: Fix MN frequency tables, parent map, and jpegd

commit ff20783f7b9f35b29e768d8ecc7076c1ca1a60ca upstream.

Clocks that don't have a pre-divider don't list any pre-divider
in their frequency tables, but their tables are initialized using
aggregate initializers. Use tagged initializers so we properly
assign the m and n values for each frequency. Furthermore, the
mmcc_pxo_pll8_pll2_pll3 array improperly mapped the second
element to pll2 instead of pll8, causing the clock driver to
recalculate the wrong rate for any clocks using this array along
with a rate that uses pll2. Plus the .num_parents field is 3
instead of 4 so you can't even switch the parent to pll3. Finally
I noticed that the jpegd clock improperly indicates that the
pre-divider width is only 2, when it's actually 4 bits wide.

Fixes: 6d00b56fe "clk: qcom: Add support for MSM8960's multimedia clock controller (MMCC)"
Tested-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agostaging/lustre: disable virtual block device for 64K pages
Arnd Bergmann [Fri, 20 Jun 2014 12:23:28 +0000 (14:23 +0200)]
staging/lustre: disable virtual block device for 64K pages

commit 0bf22be0da8ea74bc7ccc5b07d7855830be16eca upstream.

The lustre virtual block device cannot handle 64K pages and fails at compile
time. To avoid running into this error, let's disable the Kconfig option
for this driver in cases it doesn't support.

Reported-by: Dann Frazier <dann.frazier@canonical.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoaio: block exit_aio() until all context requests are completed
Gu Zheng [Wed, 3 Sep 2014 09:45:44 +0000 (17:45 +0800)]
aio: block exit_aio() until all context requests are completed

commit 6098b45b32e6baeacc04790773ced9340601d511 upstream.

It seems that exit_aio() also needs to wait for all iocbs to complete (like
io_destroy), but we missed the wait step in current implemention, so fix
it in the same way as we did in io_destroy.

Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
10 years agoclk: prevent erronous parsing of children during rate change
Tero Kristo [Thu, 21 Aug 2014 13:47:45 +0000 (16:47 +0300)]
clk: prevent erronous parsing of children during rate change

commit 067bb1741c27c8d3b74ac98c0b8fc12b31e67005 upstream.

In some cases, clocks can switch their parent with clk_set_rate, for
example clk_mux can do this in some cases. Current implementation of
clk_change_rate uses un-safe list iteration on the clock children, which
will cause wrong clocks to be parsed in case any of the clock children
change their parents during the change rate operation. Fixed by using
the safe list iterator instead.

The problem was detected due to some divide by zero errors generated
by clock init on dra7-evm board, see discussion under
http://article.gmane.org/gmane.linux.ports.arm.kernel/349180 for details.

Fixes: 71472c0c06cf ("clk: add support for clock reparent on set_rate")
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Reported-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agoperf/x86/intel: Use rdmsrl_safe() when initializing RAPL PMU
Venkatesh Srinivas [Thu, 13 Mar 2014 19:36:26 +0000 (12:36 -0700)]
perf/x86/intel: Use rdmsrl_safe() when initializing RAPL PMU

commit 24223657806a0ebd0ae5c9caaf7b021091889cf2 upstream.

CPUs which should support the RAPL counters according to
Family/Model/Stepping may still issue #GP when attempting to access
the RAPL MSRs. This may happen when Linux is running under KVM and
we are passing-through host F/M/S data, for example. Use rdmsrl_safe
to first access the RAPL_POWER_UNIT MSR; if this fails, do not
attempt to use this PMU.

Signed-off-by: Venkatesh Srinivas <venkateshs@google.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1394739386-22260-1-git-send-email-venkateshs@google.com
Cc: zheng.z.yan@intel.com
Cc: eranian@google.com
Cc: ak@linux.intel.com
Cc: linux-kernel@vger.kernel.org
[ The patch also silently fixes another bug: rapl_pmu_init() didn't handle the memory alloc failure case previously. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
[backport by whissi]
Cc: Thomas D <whissi@whissi.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agopartitions: aix.c: off by one bug
Dan Carpenter [Tue, 5 Aug 2014 08:09:59 +0000 (11:09 +0300)]
partitions: aix.c: off by one bug

commit d97a86c170b4e432f76db072a827fe30b4d6f659 upstream.

The lvip[] array has "state->limit" elements so the condition here
should be >= instead of >.

Fixes: 6ceea22bbbc8 ('partitions: add aix lvm partition support files')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Philippe De Muyter <phdm@macqel.be>
Signed-off-by: Jens Axboe <axboe@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
10 years agodmaengine: dw: don't perform DMA when dmaengine_submit is called
Andy Shevchenko [Wed, 18 Jun 2014 09:15:38 +0000 (12:15 +0300)]
dmaengine: dw: don't perform DMA when dmaengine_submit is called

commit dd8ecfcac66b4485416b2d1df0ec4798b198d7d6 upstream.

Accordingly to discussion [1] and followed up documentation the DMA controller
driver shouldn't start any DMA operations when dmaengine_submit() is called.

This patch fixes the workflow in dw_dmac driver to follow the documentation.

[1] http://www.spinics.net/lists/arm-kernel/msg125987.html

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Cc: "Petallo, MauriceX R" <mauricex.r.petallo@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>