IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
The IO performance test with fio after swapping the likely and unlikely
macros in all if-statement shows no difference. They do not help for the
performance of rtrs.
Thanks to Haakon Bugge for the test scenario.
The fio test did random read on 32 rnbd devices and 64 processes.
Test environment:
- Intel(R) Xeon(R) Gold 6130 CPU @ 2.10GHz
- 376G memory
- kernel version: 5.4.86
- gcc version: gcc (Debian 8.3.0-6) 8.3.0
- Infiniband controller: Mellanox Technologies MT27800 Family [ConnectX-5]
Test result:
- before swapping: IOPS=829k, BW=3239MiB/s
- after swapping: IOPS=829k, BW=3238MiB/s
- remove all (un)likely: IOPS=829k, BW=3238MiB/s
Link: https://lore.kernel.org/r/20210806112112.124313-5-haris.iqbal@ionos.com
Signed-off-by: Gioh Kim <gi-oh.kim@ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@ionos.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Md Haris Iqbal <haris.iqbal@ionos.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The two functions are unused, so just remove them.
Link: https://lore.kernel.org/r/20210806112112.124313-3-haris.iqbal@ionos.com
Signed-off-by: Jack Wang <jinpu.wang@ionos.com>
Reviewed-by: Md Haris Iqbal <haris.iqbal@ionos.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Md Haris Iqbal <haris.iqbal@ionos.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
When all the paths are removed for a session, the addition of the first
path is like a new session for the storage server.
Hence, for_new_clt has to be set to 1.
Link: https://lore.kernel.org/r/20210806112112.124313-2-haris.iqbal@ionos.com
Signed-off-by: Md Haris Iqbal <haris.iqbal@ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@ionos.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Saeed Mahameed says:
====================
This pulls mlx5-next branch which includes patches already reviewed on
net-next and rdma mailing lists.
1) mlx5 single E-Switch FDB for lag
2) IB/mlx5: Rename is_apu_thread_cq function to is_apu_cq
3) Add DCS caps & fields support
We need this in net-next as multiple features are dependent on the
single FDB feature.
====================
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
* mellanox/mlx5-next:
net/mlx5: Lag, Create shared FDB when in switchdev mode
net/mlx5: E-Switch, add logic to enable shared FDB
net/mlx5: Lag, move lag destruction to a workqueue
net/mlx5: Lag, properly lock eswitch if needed
net/mlx5: Add send to vport rules on paired device
net/mlx5: E-Switch, Add event callback for representors
net/mlx5e: Use shared mappings for restoring from metadata
net/mlx5e: Add an option to create a shared mapping
net/mlx5: E-Switch, set flow source for send to uplink rule
RDMA/mlx5: Add shared FDB support
{net, RDMA}/mlx5: Extend send to vport rules
RDMA/mlx5: Fill port info based on the relevant eswitch
net/mlx5: Lag, add initial logic for shared FDB
net/mlx5: Return mdev from eswitch
IB/mlx5: Rename is_apu_thread_cq function to is_apu_cq
ib_sa_service_rec_query() was introduced in kernel v2.6.13 by
commit cbae32c563 ("[PATCH] IB: Add Service Record support to SA client")
in 2005. It was not used then and have never been used since.
Removing it and related functions/structs.
Link: https://lore.kernel.org/r/1628702736-12651-1-git-send-email-haakon.bugge@oracle.com
Signed-off-by: Håkon Bugge <haakon.bugge@oracle.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Qedr code is tightly coupled with existing both INIT transitions. Here,
during first INIT transition all variables are reset and the RESET state
is checked in post_recv() before any posting.
Commit dc70f7c3ed ("RDMA/cma: Remove unnecessary INIT->INIT transition")
exposed this bug.
So moving variables reset to qedr_set_common_qp_params() and also avoid
RESET state check for post_recv().
Link: https://lore.kernel.org/r/20210811051650.14914-1-pkushwaha@marvell.com
Signed-off-by: Michal Kalderon <mkalderon@marvell.com>
Signed-off-by: Ariel Elior <aelior@marvell.com>
Signed-off-by: Shai Malin <smalin@marvell.com>
Signed-off-by: Prabhakar Kushwaha <pkushwaha@marvell.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Just use seq_write to copy the stats into the seq_file buffer instead of
poking holes into the seq_file abstraction.
Link: https://lore.kernel.org/r/20210810151711.1795374-1-hch@lst.de
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>
Tested-by: Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
If re-registering an MR in hns_roce_rereg_user_mr(), we should return NULL
instead of passing 0 to ERR_PTR for clarity.
Fixes: 4e9fc1dae2 ("RDMA/hns: Optimize the MR registration process")
Link: https://lore.kernel.org/r/20210804125939.20516-1-yuehaibing@huawei.com
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
If both eswitches are in switchdev mode and the uplink representors
are enslaved to the same bond device create a shared FDB configuration.
When moving to shared FDB mode not only the hardware needs be configured
but the RDMA driver needs to reconfigure itself.
When such change is done, unload the RDMA devices, configure the hardware
and load the RDMA representors.
When destroying the lag (can happen if a PCI function is unbinded,
driver is unloaded or by just removing a netdev from the bond) make sure
to restore the system to the previous state only if possible.
For example, if a PCI function is unbinded there is no need to load the
representors as the device is going away.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Shared FDB allows to direct traffic from all the vports in the HCA to a
single eswitch. In order to do that three things are needed.
1) Point the ingress ACL of the slave uplink to that of the master.
With this, wire traffic from both uplinks will reach the same eswitch
with the same metadata where a single steering rule can catch traffic
from both ports.
2) Set the FDB root flow table of the slave's eswitch to that of the
master. As this flow table can change dynamically make sure to
sync it on any set root flow table FDB command.
This will make sure traffic from SFs, VFs, ECPFs and PFs reach the
master eswitch.
3) Split wire traffic at the eswitch manager egress ACL so that it's
directed to the native eswitch manager. We only treat wire traffic
from both ports the same at the eswitch level. If such traffic wasn't
handled in the eswitch it needs to reach the right representor to be
processed by software. For example LACP packets should *always*
reach the right uplink representor for correct operation.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
If a netdev is removed from the lag the lag should be destroyed.
With downstream patches this might trigger a reconfiguration of
representors on a different eswitch and such we don't have the proper
locking to so from this path. Move the destruction to be done by the
workqueue.
As the destruction won't affect the netdev side it okay to do so.
The RDMA side will be reconfigured and it already coded to handle such
reconfiguration.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Currently when doing hardware lag we check the eswitch mode
but as this isn't done under a lock the check isn't valid.
As the code needs to sync between two different devices an extra
care is needed.
- When going to change eswitch mode, if hardware lag is active destroy it.
- While changing eswitch modes block any hardware bond creation.
- Delay handling bonding events until there are no mode changes in
progress.
- When attaching a new mdev to lag, block until there is no mode change
in progress. In order for the mode change to finish the interface lock
will have to be taken. Release the lock and sleep for 100ms to
allow forward progress. As this is a very rare condition (can happen if
the user unbinds and binds a PCI function while also changing eswitch
mode of the other PCI function) it has no real world impact.
As taking multiple eswitch mode locks is now required lockdep will
complain about a possible deadlock. Register a key per eswitch to make
lockdep happy.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
When two mlx5 devices are paired in switchdev mode, always offload the
send-to-vport rule to the peer E-Switch. This allows to abstract
the logic when this is really necessary (single FDB) and combine
the logic of both cases into one.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
This callback will allow to notify representors about relevant events
when in OFFLOADS mode. In downstream patches, this will be used to notify
about PAIR/UNPAIR devcom events.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
FTEs are added with mapped metadata which is saved per eswitch.
When uplink reps are bonded and we are in a single FDB mode,
we could fail to find metadata which was stored on one eswitch mapping
but not the other or with a different id.
To resolve this issue use shared mapping between eswitch ports.
We do not have any conflict using a single mapping, for a type,
between the ports.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Set the flow source param to local vport for the uplink rep
send-to-vport rule.
This will comply with the recent changes in SW steering that
use the flow source as an indication for the rule type - rx or tx.
Since the uplink send-to-vport rule is forwarding traffic to the wire
it has to indicate that it is an sx rule and can't use the any port
value in the flow source.
Signed-off-by: Ariel Levkovich <lariel@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Shared FDB allows to create a single RDMA device that holds representors
from both eswitches. As shared FDB is only active when both uplink
representors are enslaved there is a single RDMA port that represents
both uplinks.
The number of ports is the number of vports on both eswitches minus one
as we only need 1 port for both uplinks.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
In shared FDB there is only one eswitch which is active and it receives
traffic from all representors and all vports in the HCA.
While the Ethernet representor will always reside on its native PF
the IB representor will not. Extend send to vport rule creation to
support such flows. Need to account for source vport that sends the
traffic (on which the representors resides) and the target eswitch
the traffic which reach.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
In shared FDB a single RDMA device can have representors that are
connected to two different eswitches. Use the right eswitch when
preparing the response to userspace.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
As shared FDB requires changes in two subsystems first expose the needed
core functions so the RDMA side can be changed.
mlx5_lag_is_master(): return true if a given mlx5 device is the lag master.
mlx5_lag_is_shared_fdb(): Returns true if the lag mode is shared FDB.
mlx5_lag_get_peer_mdev(): Return the peer mdev in lag.
The mentioned functions will be used by downstream patches in order
to add support for shared FDB for the RDMA side.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Export a function so users can retrieve the mellanox device that manages
the eswitch from the eswitch device.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
The QP usecnts were incremented through QP attributes structure while
decreased through QP itself. Rely on the ib_creat_qp_user() code that
initialized all QP parameters prior returning to the user and increment
exactly like destroy does.
Link: https://lore.kernel.org/r/25d256a3bb1fc480b77d7fe439817b993de48610.1628014762.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
All QP creation flows called ib_create_qp_security(), but differently.
This caused to the need to provide exclusion conditions for the XRC_TGT,
because such QP already had selinux configuration call.
In order to fix it, move ib_create_qp_security() to the general QP
creation routine.
Link: https://lore.kernel.org/r/4d7cd6f5828aca37fb62283e6b126b73ab86b18c.1628014762.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The low-level create QP function grew to be larger than any sensible
inline function should be. The inline attribute is not really needed for
that function and can be implemented as exported symbol.
Link: https://lore.kernel.org/r/2c08709d86f876c3dfb77684357b2a939e570ca4.1628014762.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The ib_create_named_qp() is kernel verb that is not used for user supplied
attributes. In such case, it is ULP responsibility to provide valid QP
attributes.
In-kernel API shouldn't check it, exactly like other functions that don't
check device capabilities.
Link: https://lore.kernel.org/r/b9b9e981d1af148b750750196e686199dbbf61f8.1628014762.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The ib_create_named_qp() is kernel verb and no kernel users exist that use
XRC_INI QP. Hence such QP path is not reachable. In addition, delete
duplicated assignments of QP attributes from the initialization structure.
Link: https://lore.kernel.org/r/1b4c0d1def5f8f6d26839e14d19da950cc4a0b05.1628014762.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
There is no real value in bypassing IB/core APIs for creating standard
objects with standard types. The open-coded variant didn't have any
restrack task management calls and caused to such objects to be not
present when running rdmatoool.
Link: https://lore.kernel.org/r/f745590e5fb7d56f90fdb25f64ee3983ba17e1e4.1627040189.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Convert QP object to follow IB/core general allocation scheme. That
change allows us to make sure that restrack properly kref the memory.
Link: https://lore.kernel.org/r/48e767124758aeecc433360ddd85eaa6325b34d9.1627040189.git.leonro@nvidia.com
Reviewed-by: Gal Pressman <galpress@amazon.com> #efa
Tested-by: Gal Pressman <galpress@amazon.com>
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com> #rdma and core
Tested-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Tested-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The rdmavt QP has fields that are both needed for the control and data
path. Such mixed declaration caused to the very specific allocation flow
with kzalloc_node and SGE list embedded into the struct rvt_qp.
This patch separates QP creation to two: regular memory allocation for the
control path and specific code for the SGE list, while the access to the
later is performed through derefenced pointer.
Such pointer and its context are expected to be in the cache, so
performance difference is expected to be negligible, if any exists.
Link: https://lore.kernel.org/r/f66c1e20ccefba0db3c69c58ca9c897f062b4d1c.1627040189.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Starting from commit 2b1f747071 ("RDMA/core: Allow drivers to disable
restrack DB") the restrack is able to handle non-standard QP types either.
That change allows us to rewrite custom QP calls to their IB/core
counterparts, so we will use general QP creation flow even for the driver
QP types.
Link: https://lore.kernel.org/r/51682ab82298748941f38bd23ee3bf77ef1cab7b.1627040189.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The dev->devr.mutex was intended to protect GSI QP pointer change in the
struct mlx5_ib_port_resources when it is accessed from the
pkey_change_work. However that pointer isn't changed during the runtime
and once IB/core adds MAD, it stays stable.
Link: https://lore.kernel.org/r/6e338c561033df20d92e1371fc6a7a0d93aad945.1627040189.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
In the driver release flow, we are ensuring that notifier is disabled and
no new works can be added to pkey_change_handler. It means that we can
cancel that handler before destroying resources to make sure that our
unwind routine is symmetrical to the allocation one.
Link: https://lore.kernel.org/r/f2b1ea1bad952e4e7a48a6f731de9e0344986b29.1627040189.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
QP attributes that were supplied by IB/core already have all parameters
set when they are passed to the driver. The drivers are not supposed to
change anything in struct ib_qp_init_attr.
Fixes: 66d86e529d ("RDMA/hns: Add UD support for HIP09")
Link: https://lore.kernel.org/r/5987138875e8ade9aa339d4db6e1bd9694ed4591.1627040189.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The call to internal QP creation function skips QP creation checks and
misses the addition of such device QPs to the restrack DB.
As a preparation to general allocation scheme, convert hns to use proper
API.
Link: https://lore.kernel.org/r/7b236c15f7d5abb368958297ac6962d8459cb824.1627040189.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Use -EINVAL return type to identify whether error is returned because of
"Out of MR resources" or any other error types.
Link: https://lore.kernel.org/r/20210729151732.30995-2-pkushwaha@marvell.com
Signed-off-by: Shai Malin <smalin@marvell.com>
Signed-off-by: Ariel Elior <aelior@marvell.com>
Signed-off-by: Prabhakar Kushwaha <pkushwaha@marvell.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
To have more accurate error return type use -EOPNOTSUPP instead of
-EINVAL.
Link: https://lore.kernel.org/r/20210729151732.30995-1-pkushwaha@marvell.com
Signed-off-by: Shai Malin <smalin@marvell.com>
Signed-off-by: Ariel Elior <aelior@marvell.com>
Signed-off-by: Prabhakar Kushwaha <pkushwaha@marvell.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Starting from the beginning of infiniband subsystem, the uverbs char
devices start from 192 as a minor number, see
commit bc38a6abdd ("[PATCH] IB uverbs: core implementation").
This patch updates the admin guide documentation to reflect it.
Fixes: 9d85025b04 ("docs-rst: create an user's manual book")
Link: https://lore.kernel.org/r/bad03e6bcde45550c01e12908a6fe7dfa4770703.1627477347.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The core netlink code alread guarentees that no netlink callback can be
running outside the rdma_nl_register/unregister() region and this
registration happens during module init/exit. Thus it is already prevented
that iwpm_valid_client() can ever fail. Remove it.
Link: https://lore.kernel.org/r/a9f05a78f9996bf6ea47099b5e02671bf742f5ab.1627048781.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The failure during iw_cm module initialization partially left the system
with unreleased memory and other resources. Rewrite the module init/exit
routines in such way that netlink commands will be opened only after
successful initialization.
Fixes: b493d91d33 ("iwcm: common code for port mapper")
Link: https://lore.kernel.org/r/b01239f99cb1a3e6d2b0694c242d89e6410bcd93.1627048781.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
refcount_t type and corresponding API can protect refcounters from
accidental underflow and overflow and further use-after-free situations.
Link: https://lore.kernel.org/r/1626674454-56075-1-git-send-email-xiyuyang19@fudan.edu.cn
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn>
Signed-off-by: Xin Tan <tanxin.ctf@gmail.com>
Tested-by: Josh Fisher <josh.fisher@cornelisnetworks.com>
Acked-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
It is possible for the primary IPoIB network device associated with any
RDMA device to fail to join certain multicast groups preventing IPv6
neighbor discovery and possibly other network ULPs from working
correctly. The IPv4 broadcast group is not affected as the IPoIB network
device handles joining that multicast group directly.
This is because the primary IPoIB network device uses the pkey at ndex 0
in the associated RDMA device's pkey table. Anytime the pkey value of
index 0 changes, the primary IPoIB network device automatically modifies
it's broadcast address (i.e. /sys/class/net/[ib0]/broadcast), since the
broadcast address includes the pkey value, and then bounces carrier. This
includes initial pkey assignment, such as when the pkey at index 0
transitions from the opa default of invalid (0x0000) to some value such as
the OPA default pkey for Virtual Fabric 0: 0x8001 or when the fabric
manager is restarted with a configuration change causing the pkey at index
0 to change. Many network ULPs are not sensitive to the carrier bounce and
are not expecting the broadcast address to change including the linux IPv6
stack. This problem does not affect IPoIB child network devices as their
pkey value is constant for all time.
To mitigate this issue, change the default pkey in at index 0 to 0x8001 to
cover the predominant case and avoid issues as ipoib comes up and the FM
sweeps.
At some point, ipoib multicast support should automatically fix
non-broadcast addresses as it does with the primary broadcast address.
Fixes: 7724105686 ("IB/hfi1: add driver files")
Link: https://lore.kernel.org/r/20210715160445.142451.47651.stgit@awfm-01.cornelisnetworks.com
Suggested-by: Josh Collier <josh.d.collier@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
There is no counter for dmawait in AIP, which hampers debugging
performance issues.
Add the counter increment when the txq is queued.
Fixes: d99dc602e2 ("IB/hfi1: Add functions to transmit datagram ipoib packets")
Link: https://lore.kernel.org/r/20210715160440.142451.8278.stgit@awfm-01.cornelisnetworks.com
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
is_apu_thread_cq() used to detect CQs which are attached to APU
threads. This was extended to support other elements as well,
so the function was renamed to is_apu_cq().
c_eqn_or_apu_element was extended from 8 bits to 32 bits, which wan't
reflected when the APU support was first introduced.
Acked-by: Michael S. Tsirkin <mst@redhat.com> # vdpa
Signed-off-by: Tal Gilboa <talgi@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
