This is in symmetry to setup done in amdgpu_ring_init.
Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Otherwise we will look at the wrong place in the IB when GART
mappings are split into smaller updates.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Tom StDenis <Tom.StDenis@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If the vgic hasn't been created and initialized, we shouldn't attempt to
look at its data structures or flush/sync anything to the GIC hardware.
This fixes an issue reported by Alexander Graf when using a userspace
irqchip.
Fixes: 0919e84c0f ("KVM: arm/arm64: vgic-new: Add IRQ sync/flush framework")
Cc: stable@vger.kernel.org
Reported-by: Alexander Graf <agraf@suse.de>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
If userspace creates a PMU for the VCPU, but doesn't create an in-kernel
irqchip, then we end up in a nasty path where we try to take an
uninitialized spinlock, which can lead to all sorts of breakages.
Luckily, QEMU always creates the VGIC before the PMU, so we can
establish this as ABI and check for the VGIC in the PMU init stage.
This can be relaxed at a later time if we want to support PMU with a
userspace irqchip.
Cc: stable@vger.kernel.org
Cc: Shannon Zhao <shannon.zhao@linaro.org>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Sparse reports a below warning.
sound/soc/codecs/da7219.c:804:57: warning: dubious: x & !y
The line includes a condition statement; '(a < b) & !c'. Practically, the
evaluated value of this statement equals to the value of '(a < b) && !c'.
Although, it's not an usual way to use bitwise operations as logical
operations to several conditions.
This commit fixes the bug.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Acked-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
This hardware supports only 2-channel DAI, even mono ADC digital output
has two channels with the same data.
Having min_channels=1 results in broken playback of mono files in setups
where CPU DAI supports mono.
Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As long as reading datasheet of STAC9766/9767, this driver includes wrong
usage of DECLARE_TLV_DB_LINEAR().
In "8.1.2. Master Volume Registers", attenuation of lineout volumes is
represented in 5 bits by -1.5 dB/step from 0 to -46.5 dB. Thus,
'master_tlv' should be dB step representation.
In "8.1.14. Record Gain", gain of volumes is represented in 4 bits by
1.5 dB/step from 0 to 22.5 dB. Thus, 'record_tlv' should be dB step
representation.
In "8.1.5. PC BEEP Volume", attenuation of volume is represented in 4 bits
by -3 dB/step from 0 to 45 dB. Thus, 'beep_tlv' should be dB step
representation.
In "8.1.7. Stereo or Mic Volume" and so on, gain of volumes is represented
in 5 bits by -1.5 dB from 12 to -34.5 dB. Thus, 'mix_tlv' should be dB
step representation.
Totally, current implementation includes misuse of TLV-related macro.
This commit replaces usage of DECLARE_TLV_DB_LINEAR() with
SNDRV_CTL_TLVD_DECLARE_DB_SCALE(), to give proper information to
applications in user land.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Mark Brown <broonie@kernel.org>
Some code called by drm_crtc_force_disable_all() wants to wait for all
fences, so only do fence teardown after CRTCs are disabled.
Fixes: 84b89bdced ("drm/amdgpu: Turn off CRTCs on driver unload")
Cc: stable@vger.kernel.org # v4.8+
Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This driver has some unused variables. They should be removed.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Acked-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As long as reading datasheet of WM8991, this driver includes wrong usage
of DECLARE_TLV_DB_LINEAR().
In "Table 6 Input PGA Volume Range", volume is represented in 5 bits by
1.5 dB/step between -16.5/30.0 dB. Thus, 'in_pga_tlv' should be dB step
representation.
In "Table 34 LOMIX and ROMIX Volume Range", volume is represented in three
bits by -3 dB/step from 0 to -21 dB. Thus, 'out_mix_tlv' should be dB step
represenation.
In "Table 36 LOPGA, ROPGA, LOUT, ROUT and SPKVOL Volume Range", volume is
represented in 7 bits by 1 dB/step from -73 to 6 dB, including mute. Thus,
'out_pga_tlv' should be dB step representation.
In "Table 26 Digital Volume Range", volume is represented in 8 bits by
3/8 dB/step from -71.625 to 0 dB. Thus, 'out_dac_tlv' should be dB step
representation.
In "Table 16 ADC Digital Volume Range", volume is represented in 8 bits by
3/8 dB/step from -71.625 to 17.625 dB. Thus, 'in_adc_tlv' should be dB step
representation.
In "Table 23 Digital Sidetone Volume", volume is represented in 5 bits by
3 dB/step from -36 to 0 dB. Thus, 'out_sidetone_tlv' should be dB step
representation.
In "Table 12 Left Input Mixer Volume Control", volume is represented in
3 bits by 3 dB/step from -12 to 6 dB
Totally, current implementation includes misuse of TLV-related macro.
This commit replaces usage of DECLARE_TLV_DB_LINEAR() with proper macros,
to give proper information to applications in user land.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Acked-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
When zeroing blocks for DAX allocations, we also have to unmap aliases
in the block device mappings. Otherwise writeback can overwrite zeros
with stale data from block device page cache.
Signed-off-by: Jan Kara <jack@suse.cz>
Sparse reports a below warning.
tpa6130a2.c:193:33: warning: symbol 'tpa6130a2_component_driver' was not declared. Should it be static?
The symbol is just used inner the file. Forthermore, it's constant. Thus,
it's better to add static and const qualifier.
This commit adds it.
Fixes: cb7e62256e (ASoC: tpa6130a2: Register component)
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Mark Brown <broonie@kernel.org>
Sparse reports below warnings.
rt5616.c:1270:24: warning: symbol 'rt5616_aif_dai_ops' was not declared. Should it be static?
rt5616.c:1277:27: warning: symbol 'rt5616_dai' was not declared. Should it be static?
These two symbols are just used inner the file, thus it's better to add
static qualifier.
This commit adds it.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Mark Brown <broonie@kernel.org>
The later Arizona parts do run write sequences to power up and down the
speaker path as such a delay needs to be inserted into the DAPM sequence
to allow this to run. This patch adds appropriate delays into the
existing coalesced delay scheme.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
When debugging it is useful to check the total power up/down delay that
is executed as part of the coalesced output delay. This patch adds some
debug prints for this.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Flowrings contain skbs waiting for transmission that were passed to us
by netif. It means we checked every one of them looking for 802.1x
Ethernet type. When deleting flowring we have to use freeing function
that will check for 802.1x type as well.
Freeing skbs without a proper check was leading to counter not being
properly decreased. This was triggering a WARNING every time
brcmf_netdev_wait_pend8021x was called.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Acked-by: Arend van Spriel <arend@broadcom.com>
Cc: stable@vger.kernel.org # 4.5+
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Even with timeout increased to 950 ms we get WARNINGs from time to time.
It mostly happens on A-MPDU stalls (e.g. when station goes out of
range). It may take up to 5-10 secods for the firmware to recover and
for that time it doesn't process packets.
It's still useful to have a message on time out as it may indicate some
firmware problem and incorrect key update. Raising a WARNING however
wasn't really that necessary, it doesn't point to any driver bug anymore
and backtrace wasn't much useful.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
The entries in this enum may be referenced in debug output. Adding explicit
values simplifies the lookup.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Only rtl8821ae implements WOWLAN; however, the other drivers may receive
a call requesting information about this mode. The other drivers need to
ignore the request rather than logging that the default branch of the
switch statement has been reached.
Reported by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Help along debugging by showing what switch/case variable is not
being processed in these messages.
Signed-off-by: Joe Perches <joe@perches.com>
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
* Small fix in CTDP error handling;
* Remove support for older firmwares;
* Clean-ups and a few simple fixes here and there.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=MdMZ
-----END PGP SIGNATURE-----
Merge tag 'iwlwifi-next-for-kalle-2015-09-26' of git://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/iwlwifi-next
* Some new HW IDs;
* Small fix in CTDP error handling;
* Remove support for older firmwares;
* Clean-ups and a few simple fixes here and there.
HP Spectre x360 with CX20724 codec has two speaker outputs while the
BIOS sets up only the bottom one (NID 0x17) and disables the top one
(NID 0x1d).
This patch adds a fixup simply defining the proper pincfg for NID 0x1d
so that the top speaker works as is.
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=169071
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The group list must be freed prior to freeing the command otherwise
we have a use-after-free.
Signed-off-by: Jason Baron <jbaron@akamai.com>
Cc: Yuval Mintz <Yuval.Mintz@qlogic.com>
Cc: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Yotam Gigi says:
====================
Fix tc-ife bugs
This patch-set contains two bugfixes in the tc-ife action, one fixing some
random behaviour in encode side, and one fixing the decode side packet
parsing logic.
v2->v3
- Fix the encode side instead of the decode side
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
On ife encode side, the action stores the different tlvs inside the ife
header, where each tlv length field should refer to the length of the
whole tlv (without additional padding) and not just the data length.
On ife decode side, the action iterates over the tlvs in the ife header
and parses them one by one, where in each iteration the current pointer is
advanced according to the tlv size.
Before, the encoding encoded only the data length inside the tlv, which led
to false parsing of ife the header. In addition, due to the fact that the
loop counter was unsigned, it could lead to infinite parsing loop.
This fix changes the loop counter to be signed and fixes the encoding to
take into account the tlv type and size.
Fixes: 28a10c426e ("net sched: fix encoding to use real length")
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
On ife encode side, external mac header is copied from the original packet
and may be overridden if the user requests. Before, the mac header copy
was done from memory region that might not be accessible anymore, as
skb_cow_head might free it and copy the packet. This led to random values
in the external mac header once the values were not set by user.
This fix takes the internal mac header from the packet, after the call to
skb_cow_head.
Fixes: ef6980b6be ("net sched: introduce IFE action")
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Nelson Chang says:
====================
net: ethernet: mediatek: some bug fixes for PDAM and HW LRO
1) Add to stop PDMA while stopping the frame engine
2) Modify the register settings for LRO relinquishments
3) Jump out from the waiting loop while LRO relinquishments are done
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
(1) Modify the register settings for LRO relinquishments
(2) Jump out from the waiting loop while LRO relinquishments are done
Signed-off-by: Nelson Chang <nelson.chang@mediatek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stop PDMA while the frame engine is going to stop.
Signed-off-by: Nelson Chang <nelson.chang@mediatek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The (1 << x) - 1 trick won't work when you
are trying to fill up all 64 bits, so add special
case for that.
Signed-off-by: Ben Greear <greearb@candelatech.com>
[kvalo@qca.qualcomm.com: remove the sentence about moving limits]
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
They are not necessarily named in an intuitive manner,
so at least add some comments to help the next person.
Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Chipset from QCA99X0 onwards (QCA99X0, QCA9984, QCA4019 & future)
rx_hdr_status is not padded to align in 4-byte boundary. Define a
new hw_params field to handle different alignment behaviour between
different hw. This patch fixes improper retrieval of rfc1042 header
with QCA4019. This patch along with "ath10k: Properly remove padding
from the start of rx payload" will fix traffic failure in ethernet
decap mode for QCA4019.
Signed-off-by: Vasanthakumar Thiagarajan <vthiagar@qti.qualcomm.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Since reset_control_get() never returns NULL, we can use IS_ERR()
instead of IS_ERR_OR_NULL(). The return statements can be simpler
as well.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Use the managed variant of reset_control_get() to simplify the
failure path and the .remove callback.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Use the managed variant of clk_get() to simplify the failure path
and the .remove callback.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Correct some trivial comment typos.
Remove unnecessary parentheses in a long line.
Signed-off-by: Joe Perches <joe@perches.com>
[kvalo@qca.qualcomm.com: drop the change for return]
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
If a pending socket is marked as rejected, we will decrease the
sk_ack_backlog twice. So don't decrement it for rejected sockets
in vsock_pending_work().
Testing of the rejected socket path was done through code
modifications.
Reported-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Jorgen Hansen <jhansen@vmware.com>
Reviewed-by: Adit Ranadive <aditr@vmware.com>
Reviewed-by: Aditya Sarwade <asarwade@vmware.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Issue reproduction procedure:
1. three or more fingers put on Touchpad.
2. release fingers from Touchpad.
3. move the cursor by one finger.
4. the cursor does not move.
Cause:
We do not notify multi fingers state correctly to input subsystem. For
example, when three fingers release from Touchpad, fingers state is 3 -> 0. It
needs to notify first, second and third finger's releasing state.
Fix this by not breaking out on z axis and move x,y,z input handling
code to the correct place so that it's in fact per-finger.
[jkosina@suse.cz: reword changelog]
Signed-off-by: Masaki Ota <masaki.ota@jp.alps.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Philippe Reynes says:
====================
net: bcmgenet: only use new api ethtool_{get|set}_link_ksettings
Some times ago, a serie of patches were committed :
- commit 62469c7600 ("net: ethernet: bcmgenet: use phydev from struct net_device")
- commit 6b352ebccb ("net: ethernet: broadcom: bcmgenet: use new api ethtool_{get|set}_link_ksettings")
The first patch add a regression on this driver, so it should be reverted.
As the second patch depend on the former, it should be reverted too.
The first patch is buggy because there is a "trick" in this driver.
The structure phydev is kept in the private data when the interface
go down, and used when the interface go up to enable the phy before
the function phy_connect is called.
I don't have this hardware, neither the datasheet. So I won't
update the driver to avoid this trick.
But the real goal of the first serie was to move to the new api
ethtool_{get|set}_link_ksettings. So I provide a new version of
the patch without the "cleaning" of driver to use the phydev
store in the net_device structure.
Changelog:
v3:
- use priv instead of dev (so all the code use the same phydev)
v2:
- use Florian Fainelli patches for the revert instead of Jaedon Shin
- simply use net: bcmgenet: for the prefix of the patch
====================
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ethtool api {get|set}_settings is deprecated.
We move this driver to new api {get|set}_link_ksettings.
Signed-off-by: Philippe Reynes <tremyfr@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit 62469c7600 ("net: ethernet: bcmgenet: use phydev
from struct net_device") because it causes GENETv1/2/3 adapters to
expose the following behavior after an ifconfig down/up sequence:
PING fainelli-linux (10.112.156.244): 56 data bytes
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.352 ms
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.472 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.496 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.517 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.536 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.557 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=752.448 ms (DUP!)
This was previously fixed by commit 5dbebbb44a ("net: bcmgenet:
Software reset EPHY after power on") but the commit we are reverting was
essentially making this previous commit void, here is why.
Without commit 62469c7600 we would have the following scenario after
an ifconfig down then up sequence:
- bcmgenet_open() calls bcmgenet_power_up() to make sure the PHY is
initialized *before* we get to initialize the UniMAC, this is
critical to ensure the PHY is in a correct state, priv->phydev is
valid, this code executes fine
- second time from bcmgenet_mii_probe(), through the normal
phy_init_hw() call (which arguably could be optimized out)
Everything is fine in that case. With commit 62469c7600, we would have
the following scenario to happen after an ifconfig down then up
sequence:
- bcmgenet_close() calls phy_disonnect() which makes dev->phydev become
NULL
- when bcmgenet_open() executes again and calls bcmgenet_mii_reset() from
bcmgenet_power_up() to initialize the internal PHY, the NULL check
becomes true, so we do not reset the PHY, yet we keep going on and
initialize the UniMAC, causing MAC activity to occur
- we call bcmgenet_mii_reset() from bcmgenet_mii_probe(), but this is
too late, the PHY is botched, and causes the above bogus pings/packets
transmission/reception to occur
Reported-by: Jaedon Shin <jaedon.shin@gmail.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Philippe Reynes <tremyfr@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit 6b352ebccb ("net: ethernet: broadcom: bcmgenet:
use new api ethtool_{get|set}_link_ksettings").
We needs to revert the commit 62469c7600 ("net: ethernet: bcmgenet:
use phydev from struct net_device"), because this commit add a
regression. As the commit 6b352ebccb ("net: ethernet: broadcom:
bcmgenet: use new api ethtool_{get|set}_link_ksettings") depend
on the first one, we also need to revert it first.
Signed-off-by: Philippe Reynes <tremyfr@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit 62469c7600 ("net: ethernet: bcmgenet: use phydev
from struct net_device") because it causes GENETv1/2/3 adapters to
expose the following behavior after an ifconfig down/up sequence:
PING fainelli-linux (10.112.156.244): 56 data bytes
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.352 ms
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.472 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.496 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.517 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.536 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=1.557 ms (DUP!)
64 bytes from 10.112.156.244: seq=1 ttl=61 time=752.448 ms (DUP!)
This was previously fixed by commit 5dbebbb44a ("net: bcmgenet:
Software reset EPHY after power on") but the commit we are reverting was
essentially making this previous commit void, here is why.
Without commit 62469c7600 we would have the following scenario after
an ifconfig down then up sequence:
- bcmgenet_open() calls bcmgenet_power_up() to make sure the PHY is
initialized *before* we get to initialize the UniMAC, this is
critical to ensure the PHY is in a correct state, priv->phydev is
valid, this code executes fine
- second time from bcmgenet_mii_probe(), through the normal
phy_init_hw() call (which arguably could be optimized out)
Everything is fine in that case. With commit 62469c7600, we would have
the following scenario to happen after an ifconfig down then up
sequence:
- bcmgenet_close() calls phy_disonnect() which makes dev->phydev become
NULL
- when bcmgenet_open() executes again and calls bcmgenet_mii_reset() from
bcmgenet_power_up() to initialize the internal PHY, the NULL check
becomes true, so we do not reset the PHY, yet we keep going on and
initialize the UniMAC, causing MAC activity to occur
- we call bcmgenet_mii_reset() from bcmgenet_mii_probe(), but this is
too late, the PHY is botched, and causes the above bogus pings/packets
transmission/reception to occur
Reported-by: Jaedon Shin <jaedon.shin@gmail.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
