PHYs using the vsc8574_probe fail to be initialized and their
config_init return -EIO leading to errors like:
"could not attach PHY: -5".
This is because when the conversion of the MSCC PHY driver to use the
shared PHY package helpers was done, the base address retrieval and the
base PHY read and write helpers in the driver were modified. In
particular, the base address retrieval logic was moved from the
config_init to the probe. But the vsc8574_probe was forgotten. This
patch fixes it.
Fixes: deb04e9c0f ("net: phy: mscc: use phy_package_shared")
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Reviewed-by: Michael Walle <michael@walle.cc>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The desc pointer is set but not used. Remove it.
Reported-by: kbuild test robot <lkp@intel.com>
Fixes: 8c7bd5a454 ("net: ethernet: mtk-star-emac: new driver")
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull drm fixes from Dave Airlie:
"A couple of amdgpu fixes and minor ingenic fixes:
amdgpu:
- display atomic test fix
- Fix soft hang in display vupdate code
ingenic:
- fix pointer cast
- fix crtc atomic check callback"
* tag 'drm-fixes-2020-05-29-1' of git://anongit.freedesktop.org/drm/drm:
drm/amd/display: Fix potential integer wraparound resulting in a hang
drm/amd/display: drop cursor position check in atomic test
gpu/drm: Ingenic: Fix opaque pointer casted to wrong type
gpu/drm: ingenic: Fix bogus crtc_atomic_check callback
Jeff Kirsher says:
====================
Intel Wired LAN Driver Updates 2020-05-28
This series contains updates to e1000, e1000e, igc, igb, ixgbe and i40e.
Takashi Iwai, from SUSE, replaces some uses of snprintf() with
scnprintf() since the succeeding calls may go beyond the given buffer
limit in i40e.
Jesper Dangaard Brouer fixes up code comments in i40e_xsk.c
Xie XiuQi, from Huawei, fixes a signed-integer-overflow warning ixgbe
driver.
Jason Yan, from Huawei, converts '==' expression to bool to resolve
warnings, also fixed a warning for assigning 0/1 to a bool variable in
the ixgbe driver. Converts functions that always return 0 to void in the
igb and i40e drivers.
YueHaibing, from Hauwei, cleans up dead code in ixgbe driver.
Sasha cleans up more dead code which is not used in the igc driver.
Added receive error counter to reflect the total number of non-filtered
packets received with errors. Fixed a register define name to properly
reflect the register address being used.
Andre updates the igc driver to reject NFC rules that have multiple
matches, which is not supported in i225 devices. Updates the total
number of NFC rules supported and added a code comment to explain what
is supported.
Punit Agrawal, from Toshiba, relaxes the condition to trigger a reset
for ME, which was leading to inconsistency between the state of the
hardware as expected by the driver in e1000e.
Hari, from the Linux community, cleaned up a code comment in the e1000
driver.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
'Commit dacce2be33 ("vmxnet3: add geneve and vxlan tunnel offload
support")' added support for encapsulation offload. However, while
preparing inner tso packet, it uses reference to outer ip headers.
This patch fixes this issue by using correct reference for inner
headers.
Fixes: dacce2be33 ("vmxnet3: add geneve and vxlan tunnel offload support")
Signed-off-by: Ronak Doshi <doshir@vmware.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Allow a ULP to ask the core to provide a completion queue based on a
least-used search on a per-device CQ pools. The device CQ pools grow in a
lazy fashion when more CQs are requested.
This feature reduces the amount of interrupts when using many QPs. Using
shared CQs allows for more effcient completion handling. It also reduces
the amount of overhead needed for CQ contexts.
Test setup:
Intel(R) Xeon(R) Platinum 8176M CPU @ 2.10GHz servers.
Running NVMeoF 4KB read IOs over ConnectX-5EX across Spectrum switch.
TX-depth = 32. The patch was applied in the nvme driver on both the target
and initiator. Four controllers are accessed from each core. In the
current test case we have exposed sixteen NVMe namespaces using four
different subsystems (four namespaces per subsystem) from one NVM port.
Each controller allocated X queues (RDMA QPs) and attached to Y CQs.
Before this series we had X == Y, i.e for four controllers we've created
total of 4X QPs and 4X CQs. In the shared case, we've created 4X QPs and
only X CQs which means that we have four controllers that share a
completion queue per core. Until fourteen cores there is no significant
change in performance and the number of interrupts per second is less than
a million in the current case.
==================================================
|Cores|Current KIOPs |Shared KIOPs |improvement|
|-----|---------------|--------------|-----------|
|14 |2332 |2723 |16.7% |
|-----|---------------|--------------|-----------|
|20 |2086 |2712 |30% |
|-----|---------------|--------------|-----------|
|28 |1971 |2669 |35.4% |
|=================================================
|Cores|Current avg lat|Shared avg lat|improvement|
|-----|---------------|--------------|-----------|
|14 |767us |657us |14.3% |
|-----|---------------|--------------|-----------|
|20 |1225us |943us |23% |
|-----|---------------|--------------|-----------|
|28 |1816us |1341us |26.1% |
========================================================
|Cores|Current interrupts|Shared interrupts|improvement|
|-----|------------------|-----------------|-----------|
|14 |1.6M/sec |0.4M/sec |72% |
|-----|------------------|-----------------|-----------|
|20 |2.8M/sec |0.6M/sec |72.4% |
|-----|------------------|-----------------|-----------|
|28 |2.9M/sec |0.8M/sec |63.4% |
====================================================================
|Cores|Current 99.99th PCTL lat|Shared 99.99th PCTL lat|improvement|
|-----|------------------------|-----------------------|-----------|
|14 |67ms |6ms |90.9% |
|-----|------------------------|-----------------------|-----------|
|20 |5ms |6ms |-10% |
|-----|------------------------|-----------------------|-----------|
|28 |8.7ms |6ms |25.9% |
|===================================================================
Performance improvement with sixteen disks (sixteen CQs per core) is
comparable.
Link: https://lore.kernel.org/r/1590568495-101621-3-git-send-email-yaminf@mellanox.com
Signed-off-by: Yamin Friedman <yaminf@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Max Gurtovoy <maxg@mellanox.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
The 2afccbd283 ("spi: dw: Discard static DW DMA slave structures")
did a clean up of global variables, which is fine, but messed up with
the carefully provided information in the custom DMA slave structures.
There reader can find an assignment of the DMA request lines in use.
Partially revert the above mentioned commit to restore readability
and maintainability of the code.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20200529183150.44149-1-andriy.shevchenko@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
kobject_init_and_add() takes reference even when it fails. If this
function returns an error, kobject_put() must be called to properly clean
up the memory associated with the object. Previous
commit b8eb718348 ("net-sysfs: Fix reference count leak in
rx|netdev_queue_add_kobject") fixed a similar problem.
Link: https://lore.kernel.org/r/20200528030231.9082-1-wu000273@umn.edu
Signed-off-by: Qiushi Wu <wu000273@umn.edu>
Reviewed-by: Jason Gunthorpe <jgg@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
The platform vendor may expose an array of OEM-specific values to be used
in determining DPTF policy. These are obtained via the ODVP method, and
then simply exposed in sysfs. In addition, they are updated when a
notification is received or when the DPTF policy is changed by userland.
Conflicts:
drivers/thermal/intel/int340x_thermal/int3400_thermal.c
Signed-off-by: Matthew Garrett <mjg59@google.com>
Tested-by: Pandruvada, Srinivas <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Link: https://lore.kernel.org/r/20200414020953.255364-2-matthewgarrett@google.com
The mlx5 VF driver doesn't set QP tx port affinity because it doesn't know
if the lag is active or not, since the "lag_active" works only for PF
interfaces. In this case for VF interfaces only one lag is used which
brings performance issue.
Add a lag_tx_port_affinity CAP bit; When it is enabled and
"num_lag_ports > 1", then driver always set QP tx affinity, regardless
of lag state.
Link: https://lore.kernel.org/r/20200527055014.355093-1-leon@kernel.org
Signed-off-by: Mark Zhang <markz@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Return an error for sysfs and debugfs power interfaces during
gpu reset and suspend. Prevents access to the hw while it may
be in an unusable state.
v2: squash in fix to drop suspend check
Acked-by: Nirmoy Das <nirmoy.das@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Since the session name by itself is not sufficient to uniquely identify a
queue pair, include the queue pair number. Show the ASCII channel state
name instead of the numeric value. This change makes the ib_srpt debug
output more consistent.
Link: https://lore.kernel.org/r/20200525172212.14413-3-bvanassche@acm.org
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This reverts commit ecfbd3cf3b since Lukas Wunner noticed that we
start operating on the hardware before we check to see if this is a
spurious interrupt.
Reported-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
Third set of iwlwifi patches intended for v5.8
* Update range request API;
* Add ACPI DSM support;
* Support enabling 5.2GHz bands in Indonesia via ACPI;
* Bump FW API version to 56;
* TX queues refactoring started;
* Fix one memory leak;
* Some other small fixes and clean-ups;
# gpg: Signature made Fri 29 May 2020 10:38:28 AM EEST using RSA key ID 1A3CC5FA
# gpg: Good signature from "Luciano Roth Coelho (Luca) <luca@coelho.fi>"
# gpg: aka "Luciano Roth Coelho (Intel) <luciano.coelho@intel.com>"
mt76 patches for 5.8
* fixes for sparse warnings
* DBDC fixes
* mt7663 remain-on-channel support
* mt7915 spatial reuse support
* mt7915 radiotap fix
* station wcid allocation fix
* mt7663 powersave fix
* mt7663 scan fix
* mt7611n support
* cleanup
# gpg: Signature made Thu 28 May 2020 07:02:21 PM EEST using DSA key ID 02A76EF5
# gpg: Good signature from "Felix Fietkau <nbd@nbd.name>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg: There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 75D1 1A7D 91A7 710F 4900 42EF D77D 141D 02A7 6EF5
The commit 7b11eab041 ("blk-mq: blk-mq: provide forced completion
method") exports new API to force a request to complete without error
injection.
There should be no error injection when completing a request by timeout
handler.
Otherwise, the below would hang because timeout handler is failed.
echo 100 > /sys/kernel/debug/fail_io_timeout/probability
echo 1000 > /sys/kernel/debug/fail_io_timeout/times
echo 1 > /sys/block/nullb0/io-timeout-fail
dd if=/dev/zero of=/dev/nullb0 bs=512 count=1 oflag=direct
With this patch, the timeout handler is able to complete the IO.
Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
SD8997 firmware sends TLV_TYPE_MAX_CONN with struct hw_spec_max_conn to
inform kernel about maximum number of p2p connections and stations in AP
mode.
During initialization of SD8997 wifi chip kernel prints warning:
mwifiex_sdio mmc0:0001:1: Unknown GET_HW_SPEC TLV type: 0x217
This patch adds support for parsing TLV_TYPE_MAX_CONN (0x217) and sets
appropriate cfg80211 member 'max_ap_assoc_sta' from retrieved structure.
It allows userspace to retrieve NL80211_ATTR_MAX_AP_ASSOC_STA attribute.
Signed-off-by: Pali Rohár <pali@kernel.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20200521123559.29028-1-pali@kernel.org
The CCK TX setting when switch channel will fix the CCK to
path A only, so if the antenna is configured to path B
(e.g. iw phy set antenna 0x2 0x3 "TX B/RX AB"), then the CCK
packets can never be delivered to the air if only path B is
connected with an antenna (it can possibly be transmitted
through path A, but as path B is configured, the expected
behavior is incorrect).
This can also solve the racing issue of CCK TX setting between
driver and firmware. The CCK TX setting in driver should be
removed. Otherwise, the CCK TX setting would be wrong when the
racing occurs.
Fixes: 297bcf8222 ("rtw88: add support for set/get antennas")
Signed-off-by: Chien-Hsun Liao <ben.liao@realtek.com>
Signed-off-by: Yan-Hsuan Chuang <yhchuang@realtek.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20200522091234.24495-1-yhchuang@realtek.com
For older versions of gcc, the array = {0}; will cause warnings:
drivers/net/wireless/realtek/rtw88/rtw8822c.c: In function 'rtw8822c_power_trim':
>> drivers/net/wireless/realtek/rtw88/rtw8822c.c:1039:2: warning:
>> missing braces around initializer [-Wmissing-braces]
s8 bb_gain[2][8] = {0};
^
drivers/net/wireless/realtek/rtw88/rtw8822c.c:1039:2: warning: (near
initialization for 'bb_gain[0]') [-Wmissing-braces]
Fixes: 5ad4d8957b ("rtw88: set power trim according to efuse PG values")
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Yan-Hsuan Chuang <yhchuang@realtek.com>
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20200522035521.12295-1-yhchuang@realtek.com
Connecting to an AP with WPA2 security may fail. The IQK
and the EAPOL 4-way handshake may overlap because the
driver does IQK right after assoc success.
For 802.11n devices, the IQK is done in the driver and it
could require more than 100ms to complete. During IQK, any
TX/RX events are paused. So if the EAPOL 4-way handshake
started before IQK finished, then the 1/4 and 2/4 part of
the handshake could be dropped. The AP will then issue
deauth with reason IEEE8021X_FAILED (23).
To resolve this, move IQK routine into managed TX prepare
(ieee80211_ops::mgd_prepare_tx()). The callback is called
before the managed frames (auth/assoc) are sent. This will
make sure that the IQK is completed before the handshake
starts. But don't do IQK during scanning because doing it
on each channel will take too long.
For 802.11ac devices, the IQK is done in firmware and it
takes less time to complete. Therefore we don't see a
failure during the EAPOL 4-way handshake. But it is still
worth moving the IQK into ieee80211_ops::mgd_prepare_tx().
Fixes: f5df1a8b43 ("rtw88: 8723d: Add 8723DE to Kconfig and Makefile")
Tested-by: You-Sheng Yang <vicamo.yang@canonical.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Yan-Hsuan Chuang <yhchuang@realtek.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20200529025009.2468-4-yhchuang@realtek.com
Coex mechanism used to make BT have higher priority and more time to
transfer data when BT inquiry-page, which leads to poor WiFi performance.
Should take WiFi traffic into consideration. If the WiFi is having heavy
traffic, use another parameter to make sure WiFi has more chance to TX/RX,
while guarantee the priority of BT for inquiry. If the WiFi isn't busy
(connected or not), set proper parameter to fix originals.
Fixes: f5df1a8b43 ("rtw88: 8723d: Add 8723DE to Kconfig and Makefile")
Tested-by: You-Sheng Yang <vicamo.yang@canonical.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Yan-Hsuan Chuang <yhchuang@realtek.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20200529025009.2468-3-yhchuang@realtek.com
