Add KASAN instrumentation of architecture-specific asm implementation
of bitops. It also covers s390 specific *_inv functions.
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Make s390/bitops test functions return bool values. That enforces return
value range to 0 and 1 and matches with asm-generic/bitops-instrumented.h
declarations as well as some other architectures implementations.
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Masahiro Yamada changed the zcrypt.h header file to use __u{16,32,64}
instead of uint{16,32,64}_t with ("s390: use __u{16,32,64} instead of
uint{16,32,64}_t in uapi header").
This makes all s390 header files pass - remove zcrypt.h from the blacklist.
Cc: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
When CONFIG_UAPI_HEADER_TEST=y, exported headers are compile-tested to
make sure they can be included from user-space.
Currently, zcrypt.h is excluded from the test coverage. To make it
join the compile-test, we need to fix the build errors attached below.
For a case like this, we decided to use __u{8,16,32,64} variable types
in this discussion:
https://lkml.org/lkml/2019/6/5/18
Build log:
CC usr/include/asm/zcrypt.h.s
In file included from <command-line>:32:0:
./usr/include/asm/zcrypt.h:163:2: error: unknown type name ‘uint16_t’
uint16_t cprb_len;
^~~~~~~~
./usr/include/asm/zcrypt.h:168:2: error: unknown type name ‘uint32_t’
uint32_t source_id;
^~~~~~~~
./usr/include/asm/zcrypt.h:169:2: error: unknown type name ‘uint32_t’
uint32_t target_id;
^~~~~~~~
./usr/include/asm/zcrypt.h:170:2: error: unknown type name ‘uint32_t’
uint32_t ret_code;
^~~~~~~~
./usr/include/asm/zcrypt.h:171:2: error: unknown type name ‘uint32_t’
uint32_t reserved1;
^~~~~~~~
./usr/include/asm/zcrypt.h:172:2: error: unknown type name ‘uint32_t’
uint32_t reserved2;
^~~~~~~~
./usr/include/asm/zcrypt.h:173:2: error: unknown type name ‘uint32_t’
uint32_t payload_len;
^~~~~~~~
./usr/include/asm/zcrypt.h:182:2: error: unknown type name ‘uint16_t’
uint16_t ap_id;
^~~~~~~~
./usr/include/asm/zcrypt.h:183:2: error: unknown type name ‘uint16_t’
uint16_t dom_id;
^~~~~~~~
./usr/include/asm/zcrypt.h:198:2: error: unknown type name ‘uint16_t’
uint16_t targets_num;
^~~~~~~~
./usr/include/asm/zcrypt.h:199:2: error: unknown type name ‘uint64_t’
uint64_t targets;
^~~~~~~~
./usr/include/asm/zcrypt.h:200:2: error: unknown type name ‘uint64_t’
uint64_t weight;
^~~~~~~~
./usr/include/asm/zcrypt.h:201:2: error: unknown type name ‘uint64_t’
uint64_t req_no;
^~~~~~~~
./usr/include/asm/zcrypt.h:202:2: error: unknown type name ‘uint64_t’
uint64_t req_len;
^~~~~~~~
./usr/include/asm/zcrypt.h:203:2: error: unknown type name ‘uint64_t’
uint64_t req;
^~~~~~~~
./usr/include/asm/zcrypt.h:204:2: error: unknown type name ‘uint64_t’
uint64_t resp_len;
^~~~~~~~
./usr/include/asm/zcrypt.h:205:2: error: unknown type name ‘uint64_t’
uint64_t resp;
^~~~~~~~
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
The IQD mcast queue doesn't support QAOB mode, so skip the
qdio_enable_async_operation() setup call for this queue. This avoids
the allocation of an unneeded QAOB pointer array, and sets up q->use_cq
properly so that drivers are prohibited from using QAOBs for mcast
traffic.
Take this opportunity to streamline the q->use_cq and aob != 0 checks.
The path to qdio_siga_output() is straight-forward, we don't need to
worry about being called with bad operands.
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
If the device driver were to send out a full queue's worth of SBALs,
current code would end up discovering the last of those SBALs as PRIMED
and erroneously skip the SIGA-w. This immediately stalls the queue.
Add a check to not attempt fast-requeue in this case. While at it also
make sure that the state of the previous SBAL was successfully extracted
before inspecting it.
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Reviewed-by: Jens Remus <jremus@linux.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Extend "parmarea" to include an offset of the version string, which is
stored as 8-byte big endian value.
To retrieve version string from bzImage reliably, one should check the
presence of "S390EP" ascii string at 0x10008 (available since v3.2),
then read the version string offset from 0x10428 (which has been 0
since v3.2 up to now). The string is null terminated.
Could be retrieved with the following "file" command magic (requires
file v5.34):
8 string \x02\x00\x00\x18\x60\x00\x00\x50\x02\x00\x00\x68\x60\x00\x00\x50\x40\x40\x40\x40\x40\x40\x40\x40 Linux S390
>0x10008 string S390EP
>>0x10428 bequad >0
>>>(0x10428.Q) string >\0 \b, version %s
Reported-by: Petr Tesarik <ptesarik@suse.com>
Suggested-by: Petr Tesarik <ptesarik@suse.com>
Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
We use "pmc->idx" and the "chained" bitmap to determine if the pmc is
chained, in kvm_pmu_pmc_is_chained(). But idx might be uninitialized
(and random) when we doing this decision, through a KVM_ARM_VCPU_INIT
ioctl -> kvm_pmu_vcpu_reset(). And the test_bit() against this random
idx will potentially hit a KASAN BUG [1].
In general, idx is the static property of a PMU counter that is not
expected to be modified across resets, as suggested by Julien. It
looks more reasonable if we can setup the PMU counter idx for a vcpu
in its creation time. Introduce a new function - kvm_pmu_vcpu_init()
for this basic setup. Oh, and the KASAN BUG will get fixed this way.
[1] https://www.spinics.net/lists/kvm-arm/msg36700.html
Fixes: 80f393a23b ("KVM: arm/arm64: Support chained PMU counters")
Suggested-by: Andrew Murray <andrew.murray@arm.com>
Suggested-by: Julien Thierry <julien.thierry@arm.com>
Acked-by: Julien Thierry <julien.thierry@arm.com>
Signed-off-by: Zenghui Yu <yuzenghui@huawei.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
I mistakenly dropped the inline while resolving the patch conflicts in
the previous fix patch. Without inline, we get compiler warnings wrt
unused functions.
Note that Mauro's original patch contained the correct changes; it's
all my fault to submit a patch before a morning coffee.
Fixes: c8917b8ff0 ("firmware: fix build errors in paged buffer handling code")
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20190723081159.22624-1-tiwai@suse.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The stub function for !CONFIG_IOMMU_IOVA needs to be
'static inline'.
Fixes: effa467870 ('iommu/vt-d: Don't queue_iova() if there is no flush queue')
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The cpu variable is still being used in the of_get_property() call
after the of_node_put() call, which may result in use-after-free.
Fixes: a9acc26b75 ("cpufreq/pasemi: fix possible object reference leak")
Signed-off-by: Wen Yang <wen.yang99@zte.com.cn>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The i.MX8M SAI block is not compatible with the i.MX6SX one, as the
register layout has changed due to two version registers being added
at the beginning of the address map. Remove the bogus compatible.
Fixes: 8c61538dc9 ("arm64: dts: imx8mq: Add SAI2 node")
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Passing 0 to cpuhp_remove_state() triggers the BUG_ON() in
__cpuhp_remove_state_cpuslocked() and the argument passed to
powercap_unregister_control_type() is expected to be a valid
pointer, so avoid calling these functions with incorrect
arguments from proc_thermal_rapl_remove().
Fixes: 555c45fe0d ("int340X/processor_thermal_device: add support for MMIO RAPL")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Zhang Rui <rui.zhang@intel.com>
According to i.MX8MM reference manual Rev.1, 03/2019:
SAI3_RXC pin's mux option #1 should be GPT1_CLK, NOT GPT1_CAPTURE2;
SAI3_TXFS pin's mux option #1 should be GPT1_CAPTURE2, NOT GPT1_CLK.
Fixes: c1c9d41319 ("dt-bindings: imx: Add pinctrl binding doc for imx8mm")
Signed-off-by: Anson Huang <Anson.Huang@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
We should only call dma_max_mapping_size for devices that have a DMA mask
set, otherwise we can run into a NULL pointer dereference that will crash
the system.
Also we need to do right shift to get the sectors from the size in bytes,
not a left shift.
Fixes: bdd17bdef7 ("scsi: core: take the DMA max mapping size into account")
Reported-by: Bart Van Assche <bvanassche@acm.org>
Reported-by: Ming Lei <tom.leiming@gmail.com>
Tested-by: Guilherme G. Piccoli <gpiccoli@canonical.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
We get a warning when CONFIG_SCSI_LOWLEVEL is disabled here:
WARNING: unmet direct dependencies detected for SCSI_FDOMAIN
Depends on [n]: SCSI_LOWLEVEL [=n] && SCSI [=y]
Selected by [m]:
- PCMCIA_FDOMAIN [=m] && SCSI_LOWLEVEL_PCMCIA [=y] && SCSI [=y] && PCMCIA [=y] && m && MODULES [=y]
Move all of SCSI_LOWLEVEL_PCMCIA inside of the existing SCSI_LOWLEVEL
section. Very few people use the PCMCIA support these days, and they likely
don't mind having to turn on SCSI_LOWLEVEL as well. This way we avoid the
link error and get a more sensible structure.
Fixes: 7d47fa065e62 ("scsi: fdomain: Add PCMCIA support")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Some glue logic drivers support 1G without having GMAC/GMAC4/XGMAC.
Let's allow this speed by default.
Reported-by: Ondrej Jirman <megi@xff.cz>
Tested-by: Ondrej Jirman <megi@xff.cz>
Fixes: 5b0d7d7da6 ("net: stmmac: Add the missing speeds that XGMAC supports")
Signed-off-by: Jose Abreu <joabreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jose Abreu says:
====================
net: stmmac: Two fixes
Two fixes targeting -net.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
We need the memory to be zeroed upon allocation so use kcalloc()
instead.
Signed-off-by: Jose Abreu <joabreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
RX Descriptors are being cleaned after setting the buffers which may
lead to buffer addresses being wiped out.
Fix this by clearing earlier the RX Descriptors.
Fixes: 2af6106ae9 ("net: stmmac: Introducing support for Page Pool")
Signed-off-by: Jose Abreu <joabreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit(net: phy: marvell: change default m88e1510 LED configuration),
the active LED of Hip07 devices is always off, because Hip07 just
use 2 LEDs.
This patch adds a phy_register_fixup_for_uid() for m88e1510 to
correct the LED configuration.
Fixes: 077772468e ("net: phy: marvell: change default m88e1510 LED configuration")
Signed-off-by: Yonglong Liu <liuyonglong@huawei.com>
Reviewed-by: linyunsheng <linyunsheng@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
PPv2's XLGMAC can wait for 3 idle frames before triggering a link up
event. This can cause the link to be stuck low when there's traffic on
the interface, so disable this feature.
Fixes: 4bb0432628 ("net: mvpp2: phylink support")
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The module reset code in the Renesas CPG/MSSR driver uses
read-modify-write (RMW) operations to write to a Software Reset Register
(SRCRn), and simple writes to write to a Software Reset Clearing
Register (SRSTCLRn), as was mandated by the R-Car Gen2 and Gen3 Hardware
User's Manuals.
However, this may cause a race condition when two devices are reset in
parallel: if the reset for device A completes in the middle of the RMW
operation for device B, device A may be reset again, causing subtle
failures (e.g. i2c timeouts):
thread A thread B
-------- --------
val = SRCRn
val |= bit A
SRCRn = val
delay
val = SRCRn (bit A is set)
SRSTCLRn = bit A
(bit A in SRCRn is cleared)
val |= bit B
SRCRn = val (bit A and B are set)
This can be reproduced on e.g. Salvator-XS using:
$ while true; do i2cdump -f -y 4 0x6A b > /dev/null; done &
$ while true; do i2cdump -f -y 2 0x10 b > /dev/null; done &
i2c-rcar e6510000.i2c: error -110 : 40000002
i2c-rcar e66d8000.i2c: error -110 : 40000002
According to the R-Car Gen3 Hardware Manual Errata for Rev.
0.80 of Feb 28, 2018, reflected in Rev. 1.00 of the R-Car Gen3 Hardware
User's Manual, writes to SRCRn do not require read-modify-write cycles.
Note that the R-Car Gen2 Hardware User's Manual has not been updated
yet, and still says a read-modify-write sequence is required. According
to the hardware team, the reset hardware block is the same on both R-Car
Gen2 and Gen3, though.
Hence fix the issue by replacing the read-modify-write operations on
SRCRn by simple writes.
Reported-by: Yao Lihua <Lihua.Yao@desay-svautomotive.com>
Fixes: 6197aa65c4 ("clk: renesas: cpg-mssr: Add support for reset control")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Tested-by: Linh Phung <linh.phung.jy@renesas.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Make REGMAP_MMIO selected to avoid undefined reference to regmap symbols.
Fixes: d41f59fd92 ("clk: sprd: Add common infrastructure")
Signed-off-by: Chunyan Zhang <chunyan.zhang@unisoc.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
DT node for SiFive FU540-C000 GEMGXL Ethernet controller driver added
Signed-off-by: Yash Shah <yash.shah@sifive.com>
Reviewed-by: Sagar Kadam <sagar.kadam@sifive.com>
Cc: Andrew Lunn <andrew@lunn.ch>
[paul.walmsley@sifive.com: changed "phy1" to "phy0" at Andrew Lunn's
suggestion]
Signed-off-by: Paul Walmsley <paul.walmsley@sifive.com>
The 13MHz clock should be registered before clocksource driver is
initialized. Use CLK_OF_DECLARE_DRIVER() to guarantee.
Fixes: acddfc2c26 ("clk: mediatek: Add MT8183 clock support")
Cc: <stable@vger.kernel.org>
Signed-off-by: Weiyi Lu <weiyi.lu@mediatek.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Fix an incomplete devm_clk_bulk_get_optional() function documentation
by adding description of the num_clks argument as in other *clk_bulk*
functions.
Fixes: 9bd5ef0bd8 ("clk: Add devm_clk_bulk_get_optional() function")
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
In clk_generated_determine_rate(), if the divisor is greater than
GENERATED_MAX_DIV + 1, then the wrong best_rate will be returned.
If clk_generated_set_rate() will be called later with this wrong
rate, it will return -EINVAL, so the generated clock won't change
its value. Do no let the divisor be greater than GENERATED_MAX_DIV + 1.
Fixes: 8c7aa63289 ("clk: at91: clk-generated: remove useless divisor loop")
Signed-off-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Acked-by: Ludovic Desroches <ludovic.desroches@microchip.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
IEEE_8021QAZ_APP_SEL_STREAM is a valid selector for iSCSI connections, so
add code to use IEEE_8021QAZ_APP_SEL_STREAM selector to get priority mask.
Signed-off-by: Varun Prakash <varun@chelsio.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This patch add translations for:
- programming-languages
- kernel-docs (It is better to not translate this since English is
a requirement to get something useful out of it)
Signed-off-by: Federico Vaga <federico.vaga@vaga.pv.it>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Fix an off-by-one typo in the transparent huge pages admin
documentation.
Signed-off-by: Jeremy Cline <jcline@redhat.com>
Acked-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
On vega20, there is an SMU message to query it. On navi, it's fetched
from the metrics table.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v2:
add function smu_default_set_performance_level as default dpm level handler.
change function name smu_set_performance_level to smu_asic_set_performance_level
v1:
1.NAVI10_PEAK_SCLK_XTX 1830 Mhz
2.NAVI10_PEAK_SCLK_XT 1755 Mhz
3.NAVI10_PEAK_SCLK_XL 1625 Mhz
Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Jack Gui <Jack.Gui@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Some RISC-V systems include PCIe host controllers that support PCIe
message-signaled interrupts. For this to work on Linux, we need to
enable PCI_MSI_IRQ_DOMAIN and define struct msi_alloc_info. Support
for the latter is enabled by including the architecture-generic msi.h
include.
Signed-off-by: Wesley Terpstra <wesley@sifive.com>
[paul.walmsley@sifive.com: split initial patch into one arch/riscv
patch and one drivers/pci patch]
Signed-off-by: Paul Walmsley <paul.walmsley@sifive.com>
The "struct drm_connector" iteration cursor from
"for_each_new_connector_in_state" is never used in atomic_remove_fb()
which generates a compilation warning,
drivers/gpu/drm/drm_framebuffer.c: In function 'atomic_remove_fb':
drivers/gpu/drm/drm_framebuffer.c:838:24: warning: variable 'conn' set
but not used [-Wunused-but-set-variable]
Silence it by marking "conn" __maybe_unused.
Signed-off-by: Qian Cai <cai@lca.pw>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/1563822886-13570-1-git-send-email-cai@lca.pw
The RISC-V port has grown significantly over the past year. Paul's been
helping out for a while ago. We agreed in person that he'd take over
collecting the patches and submitting the PRs, but it looks like I
forgot to make it official.
Signed-off-by: Palmer Dabbelt <palmer@sifive.com>
Signed-off-by: Paul Walmsley <paul.walmsley@sifive.com>
The statement sounds more like a literal translation
Signed-off-by: Federico Vaga <federico.vaga@vaga.pv.it>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
The patch translates the following patches in Italian:
d9d7c0c497 docs: Note that :c:func: should no longer be used
83e8b971f8 sphinx.rst: Add note about code snippets embedded in the text
cca5e0b8a4 Documentation: PGP: update for newer HW devices
Signed-off-by: Federico Vaga <federico.vaga@vaga.pv.it>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
The following scenario was encountered during testing of logical
partition mobility on pseries partitions with bonded ibmvnic
adapters in LACP mode.
1. Driver receives a signal that the device has been
swapped, and it needs to reset to initialize the new
device.
2. Driver reports loss of carrier and begins initialization.
3. Bonding driver receives NETDEV_CHANGE notifier and checks
the slave's current speed and duplex settings. Because these
are unknown at the time, the bond sets its link state to
BOND_LINK_FAIL and handles the speed update, clearing
AD_PORT_LACP_ENABLE.
4. Driver finishes recovery and reports that the carrier is on.
5. Bond receives a new notification and checks the speed again.
The speeds are valid but miimon has not altered the link
state yet. AD_PORT_LACP_ENABLE remains off.
Because the slave's link state is still BOND_LINK_FAIL,
no further port checks are made when it recovers. Though
the slave devices are operational and have valid speed
and duplex settings, the bond will not send LACPDU's. The
simplest fix I can see is to force another speed check
in bond_miimon_commit. This way the bond will update
AD_PORT_LACP_ENABLE if needed when transitioning from
BOND_LINK_FAIL to BOND_LINK_UP.
CC: Jarod Wilson <jarod@redhat.com>
CC: Jay Vosburgh <j.vosburgh@gmail.com>
CC: Veaceslav Falico <vfalico@gmail.com>
CC: Andy Gospodarek <andy@greyhouse.net>
Signed-off-by: Thomas Falcon <tlfalcon@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a TID sequence error occurs while receiving TID RDMA READ RESP
packets, all packets after flow->flow_state.r_next_psn should be dropped,
including those response packets for subsequent segments.
The current implementation will drop the subsequent response packets for
the segment to complete next, but may accept packets for subsequent
segments and therefore mistakenly advance the r_next_psn fields for the
corresponding software flows. This may result in failures to complete
subsequent segments after the current segment is completed.
The fix is to only use the flow pointed by req->clear_tail for checking
KDETH PSN instead of finding a flow from the request's flow array.
Fixes: b885d5be9c ("IB/hfi1: Unify the software PSN check for TID RDMA READ/WRITE")
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20190715164540.74174.54702.stgit@awfm-01.aw.intel.com
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Kaike Wan <kaike.wan@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
