This struct is used as argument to ath10k_core_register in order to
make it easier to add more bus parameters in the future.
Signed-off-by: Erik Stromdahl <erik.stromdahl@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Downloading firmware via BMI protocol takes too long time. For example,
a ~700K bytes firmware takes about 500ms to download via BMI protocol.
This is too long especially in suspend and resume scenario where firmware
is re-downloaded unless WoWLAN is enabled. Downloading firmware via diag CE
can reduce the time to ~40ms for a ~700K bytes firmware binary.
Ath10k driver parses the firmware to segments and downloads the segments
to the specified address directly. If the firmware is compressed or has
unsupported segments, ath10k driver will try BMI download again.
It's tested with QCA6174 hw3.2 and
firmware-6.bin_WLAN.RM.4.4.1-00111-QCARMSWP-1. QCA9377 is also affected.
Signed-off-by: Carl Huang <cjhuang@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
To download firmware via diag interface, driver needs to write the target
memory space below 1M. It means the bit20 should be zero for the converted
address if the target memory space is below 1M. Otherwise, bit20 is one if
the target address is larger or equal to 1M space.
As downloading firmware via diag interface is only required for qca6174
and qca9377, a new specific function is introduced to convert the target
address to ce address: ath10k_pci_qca6174_targ_cpu_to_ce_addri().
This function supports to convert any target address to ce address.
It's tested with QCA6174 hw3.2 and
firmware-6.bin_WLAN.RM.4.4.1-00111-QCARMSWP-1. QCA9377 is also affected.
Signed-off-by: Carl Huang <cjhuang@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Delay 1ms is too long for both diag read and write operations.
This is observed when writing a big memory buffer to target or
reading a big memory buffer from target. Take writing/reading
512k bytes as example, the delay itself is 256ms as the maximum
length of every write/read is 2k size.
Reduce the delay to 50us for read and write operations.
Take the ath10k_pci_targ_cpu_to_ce_addr() out of loop and put it
in the beginning of the loop for ath10k_pci_diag_read_mem().
The ath10k_pci_targ_cpu_to_ce_addr() is to convert the address
from target cpu's perspective to CE's perspective, so it makes
no sense to convert a CE's perspective address again in the loop.
It's a wrong implementation but happens to work.
If the target address is below 1M space, then the convert in the loop
from the second time becomes wrong because the previously converted address
is larger than 1M. The counterpart ath10k_pci_diag_write_mem() has the
correct implementation.
With this change, ath10k_pci_diage_read_mem() works correctly no matter
the target address is below 1M or above 1M.
It's tested with QCA6174 hw3.2 and
firmware-6.bin_WLAN.RM.4.4.1-00111-QCARMSWP-1. QCA9377 is also affected.
Signed-off-by: Carl Huang <cjhuang@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Average ack rssi value is weighted average of ack rssi for
no of msdu's has been sent.
This feature is enabled by the host driver if firmware is capable.
After receiving event from host, firmware allocates the necessary
memory to store the ack_rssi for data packets during the init time.
After each successful transmission, If tx completion status is OK
and 24th bit is set in HTT message header then host will fetch the
ack_rssi else host can ignore the ack_rssi field.
Signed-off-by: Balaji Pothunoori <bpothuno@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
array "ctl_power_table" access index "pream" is initialized with -1 and
is raised as a static analysis tool issue.
[drivers\net\wireless\ath\ath10k\wmi.c:4719] ->
[drivers\net\wireless\ath\ath10k\wmi.c:4730]: (error) Array index -1 is
out of bounds.
Since the "pream" index for accessing ctl_power_table array is initialized
with -1, there is a chance of memory access violation for the cases below.
1) wmi_pdev_tpc_final_table_event change frequency is between 2483 and 5180
2) pream_idx is out of the enumeration ranges of wmi_tpc_pream_2ghz,
wmi_tpc_pream_5ghz
Signed-off-by: K.T.VIJAYAKUMAAR <vijay.bvb@samsung.com>
[kvalo@codeaurora.org: clean up the warning message]
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
This patch adds 'tx_stats' in per station debugfs entry.
Use this command to dump tx_stats:
cat /sys/kernel/debug/ieee80211/phy0/netdev\:wlan0/
stations/<MACADDR>/tx_stats
Signed-off-by: Anilkumar Kolli <akolli@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
This patch adds per station tx statistics support.
Per station tx stats include
- pkts/bytes transmitted at all possible rates(mcs/nss/bw/gi).
- ACK fails count
- ampdu bytes/pkts transmitted at all possible rates(mcs/nss/bw/gi).
- BA fails count
Tested on QCA9984/QCA4019/QCA988x
Firmware: 10.4-3.5.3-00057
10.2.4-1.0-00037
Signed-off-by: Anilkumar Kolli <akolli@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
This patch adds debugfs entry to enable/disable extended
tx statistics. Extended tx statistics are from peer stats
feature.
Signed-off-by: Anilkumar Kolli <akolli@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Get the legacy rate index to update the pkts/bytes counter
against each possible tx rate.
Signed-off-by: Anilkumar Kolli <akolli@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
The comparison of cid >= 0 is always true because cid is of type u8
(8 bits, unsigned).
Fix this by removing such comparison and updating the type of
variable cid to u8 in the caller function.
Addresses-Coverity-ID: 1473079 ("Unsigned compared against 0")
Fixes: b9010f105f ("wil6210: add FT roam support for AP and station")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
WCN3990 target uses separate htc service for pktlog.
Add pktlog service request and support for pktlog
rx path handling.
Testing:
Tested on WCN3990 and QCA6174 HW.
Tested FW: WLAN.HL.2.0-01192-QCAHLSWMTPLZ-1,
WLAN.RM.4.4.1-00109-QCARMSWPZ-1
Signed-off-by: Govind Singh <govinds@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Existing copy engine interrupt enable logic assumes that last
CE is using polling mode and due to this interrupt for last copy engine
are always disabled. WCN3990 uses last CE for pktlog and
interrupt remains disabled with existing logic.
To mitigate this issue, introduce CE_ATTR_POLL flag and control
the interrupt based on the flag which can be set in ce_attr.
Testing:
Tested on WCN3990 and QCA6174 HW.
Tested FW: WLAN.HL.2.0-01192-QCAHLSWMTPLZ-1,
WLAN.RM.4.4.1-00109-QCARMSWPZ-1
Signed-off-by: Govind Singh <govinds@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
To support dual-band variant of QCA9984, new extended board data (eBDF)
is introduced since existing board data ran out of space.
Below is the brief implementation & design detail,
----------------------------------------------------
1. New OTP changes to inform eBDF support in existing OTP download to
fetch board ID and chip ID. This is backward compatible and older
card sends 0 by default for eBDF support bit (bit 18 of OTP response) we
check in ath10k driver.
2. If eBDF is supported, then we need to fetch eBDF ID which is bundled
in downloaded board data. So again OTP is executed for knowing the eBDF ID.
This is done once we set 'board_data_initialized' bit. If eBDF ID
returned is zero, we continue booting with previous board data downloaded.
3. Based on the eBDF ID fetched, ath10k driver tries to download the
extended board data to a new offset ahead of already downloaded board
data address.
4. A new BD IE type, ATH10K_BD_IE_BOARD_EXT is added to differentiate in
bundling eBDF separately in board-2.bin and also to parse through
board bundle for eBDF download in ath10k boot.
5. If eBDF is not present in the board-2.bin bundle or when board ID is
zero, we do a fallback boot to "eboard.bin" in the same QCA9984/hw1.0 dir.
This is same as done to existing "board.bin" if board ID is not present
in board-2.bin bundle.
Current design is that eBDF size will be 2KB and eBDF ID will be
byte value.
Tested the above changes with dual-band variant of QCA9984 card. OTP
update needed for the test will be part of next FW release 10.4-3.6-xxxx.
Below are the logs with ath10k BOOT debugs enabled.
First OTP response :
---------------------
..
boot upload otp to 0x1234 len 9478 for board id
boot get otp board id result 0x00040400 board_id 1 chip_id 0 ext_bid_support 1
..
Second OTP response :
---------------------
..
boot upload otp to 0x1234 len 9478 for ext board id
boot get otp ext board id result 0x00000005 ext_board_id 5
boot using eboard name 'bus=pci,bmi-chip-id=0,bmi-eboard-id=5'
..
Extended board data download:
------------------------------
..
board name
00000000: 62 75 73 3d 70 63 69 2c 62 6d 69 2d 63 68 69 70 bus=pci,bmi-chip
00000010: 2d 69 64 3d 30 2c 62 6d 69 2d 65 62 6f 61 72 64 -id=0,bmi-eboard
00000020: 2d 69 64 3d 35 -id=5
boot found match for name 'bus=pci,bmi-chip-id=0,bmi-eboard-id=5'
boot found eboard data for 'bus=pci,bmi-chip-id=0,bmi-eboard-id=5'
using board api 2
boot writing ext board data to addr 0xc3000
..
Fallback Extended board data download from "eboard.bin":
---------------------------------------------------------
..
board name
00000000: 62 75 73 3d 70 63 69 2c 62 6d 69 2d 63 68 69 70 bus=pci,bmi-chip
00000010: 2d 69 64 3d 30 2c 62 6d 69 2d 62 6f 61 72 64 2d -id=0,bmi-board-
00000020: 69 64 3d 31 30 id=10
failed to fetch board data for bus=pci,bmi-chip-id=0,bmi-eboard-id=5 from ath10k/QCA9984/hw1.0/board-2.bin
boot fw request 'ath10k/QCA9984/hw1.0/eboard.bin': 0
using board api 1
boot writing ext board data to addr 0xc3000
..
Signed-off-by: Sathishkumar Muruganandam <murugana@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Just moving functions down in the file, no functional changes.
Signed-off-by: Sathishkumar Muruganandam <murugana@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Add support for the Northstar Plus SerDes which is accessed through a
special page of the switch. Since this is something that most people
probably will not want to use, make it a configurable option with a
default on ARCH_BCM_NSP where it is the most useful currently.
The SerDes supports both SGMII and 1000baseX modes for both lanes, and
2500baseX for one of the lanes, and is internally looking like a
seemingly standard MII PHY, except for the few bits that got repurposed.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for PHYLINK, things are reasonably straight forward since we
do not yet support SerDes interfaces, that leaves us with just
MLO_AN_PHY and MLO_AN_FIXED to deal with.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extract the logic from b53_adjust_link() responsible for overriding a
given port's link, speed, duplex and pause settings and make two helper
functions to set the port's configuration and the port's link settings.
We will make use of both, as separate functions while adding PHYLINK
support next.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Update the SRAB driver to manage per-port interrupts. Since we cannot
sleep during b53_io_ops, schedule a workqueue whenever we get a port
specific interrupt. We will later make use of this to call back into
PHYLINK when there is e.g: a link state change.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some switches expose individual interrupt line(s) for port specific
event(s), allow configuring these interrupts at an appropriate time
during port_enable/disable callbacks where all port specific resources
are known to be set-up and ready for use.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds a new qed firmware with fixes and support for new features.
Fixes:
- Fix a rare case of device crash with iWARP, iSCSI or FCoE offload.
- Fix GRE tunneled traffic when iWARP offload is enabled.
- Fix RoCE failure in ib_send_bw when using inline data.
- Fix latency optimization flow for inline WQEs.
- BigBear 100G fix
RDMA:
- Reduce task context size.
- Application page sizes above 2GB support.
- Performance improvements.
ETH:
- Tenant DCB support.
- Replace RSS indirection table update interface.
Misc:
- Debug Tools changes.
Signed-off-by: Denis Bolotin <denis.bolotin@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adopt the SPDX license identifier headers to ease license compliance
management.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We need to bail out if lan78xx_get_endpoints() fails, otherwise the
result is overwritten.
Fixes: 55d7de9de6 ("Microchip's LAN7800 family USB 2/3 to 10/100/1000 Ethernet")
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Reviewed-by: Raghuram Chary Jallipalli <raghuramchary.jallipalli@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the accesses to rtsyms now all going via special helpers
we can easily make sure the driver is not reading past the
end of the symbol.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Francois H. Theron <francois.theron@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For ease of debug preface all error messages with the name
of the symbol which caused them. Use the same message format
for existing messages while at it.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Francois H. Theron <francois.theron@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/net/net_failover.c: In function 'net_failover_slave_unregister':
drivers/net/net_failover.c:598:35: warning:
variable 'primary_dev' set but not used [-Wunused-but-set-variable]
There should check the validity of 'slave_dev'.
Fixes: cfc80d9a11 ("net: Introduce net_failover driver")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
CHECKSUM_COMPLETE is not applicable to SCTP protocol.
Setting it for SCTP packets leads to CRC32c validation failure.
Fixes: bbceefce9a ("net/mlx5e: Support RX CHECKSUM_COMPLETE")
Signed-off-by: Alaa Hleihel <alaa@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
In multi-host (MH) NIC scheme, a single HW port serves multiple hosts
or sockets on the same host.
The HW uses a mechanism in the PCIe buffer which monitors
the amount of consumed PCIe buffers per host.
On a certain configuration, under congestion,
the HW emulates a switch doing ECN marking on packets using ECN
indication on the completion descriptor (CQE).
The driver needs to set the ECN bits on the packet SKB,
such that the network stack can react on that, this commit does that.
Signed-off-by: Natali Shechtman <natali@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Changed "priv.clock.lock" lock from 'rw_lock' to 'seq_lock'
in order to improve packet rate performance.
Tested on Intel(R) Xeon(R) CPU E5-2660 v2 @ 2.20GHz.
Sent 64b packets between two peers connected by ConnectX-5,
and measured packet rate for the receiver in three modes:
no time-stamping (base rate)
time-stamping using rw_lock (old lock) for critical region
time-stamping using seq_lock (new lock) for critical region
Only the receiver time stamped its packets.
The measured packet rate improvements are:
Single flow (multiple TX rings to single RX ring):
without timestamping: 4.26 (M packets)/sec
with rw-lock (old lock): 4.1 (M packets)/sec
with seq-lock (new lock): 4.16 (M packets)/sec
1.46% improvement
Multiple flows (multiple TX rings to six RX rings):
without timestamping: 22 (M packets)/sec
with rw-lock (old lock): 11.7 (M packets)/sec
with seq-lock (new lock): 21.3 (M packets)/sec
82.05% improvement
The packet rate improvement is due to the lack of atomic operations
for the 'readers' by the seq-lock.
Since there are much more 'readers' than 'writers' contention
on this lock, almost all atomic operations are saved.
this results in a dramatic decrease in overall
cache misses.
Signed-off-by: Shay Agroskin <shayag@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Not all profiles query the HW Q counters in update_stats() callback.
HW Q couners are limited per device and in case of representors all
their Q counters are allocated on the parent PF device.
Avoid reundant allocation of HW Q counters by moving the allocation
to init_rx profile callback.
Signed-off-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Move the definition of mlx5e_priv_flags into en_ethtool.c because it's
only used there.
Fixes: 4e59e28881 ("net/mlx5e: Introduce net device priv flags infrastructure")
Signed-off-by: Kamal Heib <kamalheib1@gmail.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Previous patch in series changed flow counter storage structure from
rb_tree to linked list in order to improve flow counter traversal
performance. The drawback of such solution is that flow counter lookup by
id becomes linear in complexity.
Store pointers to flow counters in idr in order to improve lookup
performance to logarithmic again. Idr is non-intrusive data structure and
doesn't require extending flow counter struct with new elements. This means
that idr can be used for lookup, while linked list from previous patch is
used for traversal, and struct mlx5_fc size is <= 2 cache lines.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Amir Vadai <amir@vadai.me>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
In order to improve performance of flow counter stats query loop that
traverses all configured flow counters, replace rb_tree with double-linked
list. This change improves performance of traversing flow counters by
removing the tree traversal. (profiling data showed that call to rb_next
was most top CPU consumer)
However, lookup of flow flow counter in list becomes linear, instead of
logarithmic. This problem is fixed by next patch in series, which adds idr
for fast lookup. Idr is to be used because it is not an intrusive data
structure and doesn't require adding any new members to struct mlx5_fc,
which allows its control data part to stay <= 1 cache line in size.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Amir Vadai <amir@vadai.me>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
In order to prevent flow counters stats work function from traversing whole
flow counters tree while searching for deleted flow counters, new list to
store deleted flow counters is added to struct mlx5_fc_stats. Lockless
NULL-terminated single linked list data type is used due to following
reasons:
- This use case only needs to add single element to list and
remove/iterate whole list. Lockless list doesn't require any additional
synchronization for these operations.
- First cache line of flow counter data structure only has space to store
single additional pointer, which precludes usage of double linked list.
Remove flow counter 'deleted' flag that is no longer needed.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Amir Vadai <amir@vadai.me>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
In order to prevent flow counters stats work function from traversing whole
flow counters tree while searching for deleted flow counters, new list to
store deleted flow counters will be added to struct mlx5_fc_stats. However,
the flow counter structure itself has no space left to store any more data
in first cache line. To free space that is needed to store additional list
node, convert current addlist double linked list (two pointers per node) to
atomic single linked list (one pointer per node).
Lockless NULL-terminated single linked list data type doesn't require any
additional external synchronization for operations used by flow counters
module (add single new element, remove all elements from list and traverse
them). Remove addlist_lock that is no longer needed.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Amir Vadai <amir@vadai.me>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
This is a false positive report due to incorrect nested lock
annotations as we lock multiple fgs with the same subclass.
Instead of locking all fgs only lock the one being used as was
done before.
Fixes: bd71b08ec2 ("net/mlx5: Support multiple updates of steering rules in parallel")
Signed-off-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Copy and paste bug was introduced in the offending patch.
We need to write udp source port value into the headers value and not
headers criteria "mask".
Fixes: 142644f8a1 ("net/mlx5e: Ethtool steering flow parsing refactoring")
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Currently, mlx5_attach_interface does not check for error
after calling intf->attach or intf->add. When these two calls
fails, the client is not initialized and will cause issues such as
kernel panic on invalid address in the teardown path (mlx5_detach_interface)
Fixes: 737a234bb6 ("net/mlx5: Introduce attach/detach to interface API")
Signed-off-by: Huy Nguyen <huyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
The PCI BDF is not unique. PCI domain must also be considered when
searching for the next physical device during lag setup. Example below:
mlx5_core 0000:01:00.0: MLX5E: StrdRq(1) RqSz(8) StrdSz(128) RxCqeCmprss(0)
mlx5_core 0000:01:00.1: MLX5E: StrdRq(1) RqSz(8) StrdSz(128) RxCqeCmprss(0)
mlx5_core 0001:01:00.0: MLX5E: StrdRq(1) RqSz(8) StrdSz(128) RxCqeCmprss(0)
mlx5_core 0001:01:00.1: MLX5E: StrdRq(1) RqSz(8) StrdSz(128) RxCqeCmprss(0)
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Aviv Heller <avivh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
If building match list fg fails and we never jumped to
search_again_locked label then the function returned without
unlocking the read lock.
Fixes: bd71b08ec2 ("net/mlx5: Support multiple updates of steering rules in parallel")
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
The memory allocated for the slow path table flow group input structure
was not freed upon successful return, fix that.
Fixes: 1967ce6ea5 ("net/mlx5: E-Switch, Refactor fast path FDB table creation in switchdev mode")
Signed-off-by: Raed Salem <raeds@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Minimal stride size is 16.
Hence, the number of strides in a fragment (of PAGE_SIZE)
is <= PAGE_SIZE / 16 <= 4K.
u16 is sufficient to represent this.
Fixes: d7037ad73d ("net/mlx5: Fix QP fragmented buffer allocation")
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Minimal stride size is 16.
Hence, the number of strides in a fragment (of PAGE_SIZE)
is <= PAGE_SIZE / 16 <= 4K.
u16 is sufficient to represent this.
Fixes: 388ca8be00 ("IB/mlx5: Implement fragmented completion queue (CQ)")
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
When initializing the device (procedure init_one), the driver
calls mlx5_pci_init to perform pci initialization. As part of this
initialization, mlx5_pci_init creates a debugfs directory.
If this creation fails, init_one aborts, returning failure to
the caller (which is the probe method caller).
The main reason for such a failure to occur is if the debugfs
directory already exists. This can happen if the last time
mlx5_pci_close was called, debugfs_remove (silently) failed due
to the debugfs directory not being empty.
Guarantee that such a debugfs_remove failure will not occur by
instead calling debugfs_remove_recursive in procedure mlx5_pci_close.
Fixes: 59211bd3b6 ("net/mlx5: Split the load/unload flow into hardware and software flows")
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Reviewed-by: Daniel Jurgens <danielj@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
When the mlx5 health mechanism detects a problem while the driver
is in the middle of init_one or remove_one, the driver needs to prevent
the health mechanism from scheduling future work; if future work
is scheduled, there is a problem with use-after-free: the system WQ
tries to run the work item (which has been freed) at the scheduled
future time.
Prevent this by disabling work item scheduling in the health mechanism
when the driver is in the middle of init_one() or remove_one().
Fixes: e126ba97db ("mlx5: Add driver for Mellanox Connect-IB adapters")
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Reviewed-by: Feras Daoud <ferasda@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
With performance optimization the spi transfer and messages of basic
register operations like qcaspi_read_register moved into the private
driver structure. But they weren't protected against mutual access
(e.g. between driver kthread and ethtool). So dumping the QCA7000
registers via ethtool during network traffic could make spi_sync
hang forever, because the completion in spi_message is overwritten.
So revert the optimization completely.
Fixes: 291ab06ecf ("net: qualcomm: new Ethernet over SPI driver for QCA700")
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
