Yixing Liu [Sat, 21 Aug 2021 09:53:26 +0000 (17:53 +0800)]
RDMA/hns: Enable stash feature of HIP09
The stash feature is enabled by default on HIP09.
Fixes: f93c39bc9547 ("RDMA/hns: Add support for QP stash") Fixes: bfefae9f108d ("RDMA/hns: Add support for CQ stash") Link: https://lore.kernel.org/r/1629539607-33217-3-git-send-email-liangwenpeng@huawei.com Signed-off-by: Yixing Liu <liuyixing1@huawei.com> Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The wrappers in include/linux/pci-dma-compat.h should go away.
The patch has been generated with the coccinelle script below.
It has been hand modified to use 'dma_set_mask_and_coherent()' instead of
'pci_set_dma_mask()/pci_set_consistent_dma_mask()' when applicable.
This is less verbose.
Gioh Kim [Fri, 6 Aug 2021 11:21:12 +0000 (13:21 +0200)]
RDMA/rtrs: Remove (void) casting for functions
Casting to (void) does nothing, remove them.
Link: https://lore.kernel.org/r/20210806112112.124313-7-haris.iqbal@ionos.com Suggested-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Gioh Kim <gi-oh.kim@ionos.com> Reviewed-by: Md Haris Iqbal <haris.iqbal@ionos.com> Signed-off-by: Jack Wang <jinpu.wang@ionos.com> Signed-off-by: Md Haris Iqbal <haris.iqbal@ionos.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Gioh Kim [Fri, 6 Aug 2021 11:21:11 +0000 (13:21 +0200)]
RDMA/rtrs-clt: Fix counting inflight IO
There are mis-match at counting inflight IO after changing the multipath
policy.
For example, we started fio test with round-robin policy and then we
changed the policy to min-inflight. IOs created under the RR policy is
finished under the min-inflight policy and inflight counter only
decreased. So the counter would be negative value. And also we started
fio test with min-inflight policy and changed the policy to the
round-robin. IOs created under the min-inflight policy increased the
inflight IO counter but the inflight IO counter was not decreased because
the policy was the round-robin when IO was finished.
So it should count IOs only if the IO is created under the min-inflight
policy. It should not care the policy when the IO is finished.
This patch adds a field mp_policy in struct rtrs_clt_io_req and stores the
multipath policy when an object of rtrs_clt_io_req is created. Then
rtrs-clt checks the mp_policy of only struct rtrs_clt_io_req instead of
the struct rtrs_clt.
Link: https://lore.kernel.org/r/20210806112112.124313-6-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>
Gioh Kim [Fri, 6 Aug 2021 11:21:10 +0000 (13:21 +0200)]
RDMA/rtrs: Remove all likely and unlikely
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>
Jack Wang [Fri, 6 Aug 2021 11:21:08 +0000 (13:21 +0200)]
RDMA/rtrs: Remove unused functions
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>
Jason Gunthorpe [Fri, 20 Aug 2021 14:26:43 +0000 (11:26 -0300)]
Merge branch 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux
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
Håkon Bugge [Wed, 11 Aug 2021 17:25:36 +0000 (19:25 +0200)]
RDMA/core/sa_query: Remove unused function
ib_sa_service_rec_query() was introduced in kernel v2.6.13 by
commit cbae32c56314 ("[PATCH] IB: Add Service Record support to SA client")
in 2005. It was not used then and have never been used since.
RDMA/qedr: Move variables reset to qedr_set_common_qp_params()
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 dc70f7c3ed34 ("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().
RDMA/hfi1: Stop using seq_get_buf in _driver_stats_seq_show
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>
Mark Bloch [Tue, 3 Aug 2021 23:19:59 +0000 (16:19 -0700)]
net/mlx5: Lag, Create shared FDB when in switchdev mode
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>
Mark Bloch [Tue, 3 Aug 2021 23:19:58 +0000 (16:19 -0700)]
net/mlx5: E-Switch, add logic to enable shared FDB
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>
Mark Bloch [Tue, 3 Aug 2021 23:19:57 +0000 (16:19 -0700)]
net/mlx5: Lag, move lag destruction to a workqueue
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>
Mark Bloch [Tue, 3 Aug 2021 23:19:56 +0000 (16:19 -0700)]
net/mlx5: Lag, properly lock eswitch if needed
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>
Mark Bloch [Tue, 3 Aug 2021 23:19:55 +0000 (16:19 -0700)]
net/mlx5: Add send to vport rules on paired device
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>
Mark Bloch [Tue, 3 Aug 2021 23:19:54 +0000 (16:19 -0700)]
net/mlx5: E-Switch, Add event callback for representors
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>
Roi Dayan [Tue, 3 Aug 2021 23:19:53 +0000 (16:19 -0700)]
net/mlx5e: Use shared mappings for restoring from metadata
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>
Ariel Levkovich [Tue, 3 Aug 2021 23:19:51 +0000 (16:19 -0700)]
net/mlx5: E-Switch, set flow source for send to uplink rule
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.
Mark Bloch [Tue, 3 Aug 2021 23:19:50 +0000 (16:19 -0700)]
RDMA/mlx5: Add shared FDB support
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>
Mark Bloch [Tue, 3 Aug 2021 23:19:49 +0000 (16:19 -0700)]
{net, RDMA}/mlx5: Extend send to vport rules
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>
Mark Bloch [Tue, 3 Aug 2021 23:19:48 +0000 (16:19 -0700)]
RDMA/mlx5: Fill port info based on the relevant eswitch
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>
Mark Bloch [Tue, 3 Aug 2021 23:19:47 +0000 (16:19 -0700)]
net/mlx5: Lag, add initial logic for shared FDB
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>
Leon Romanovsky [Tue, 3 Aug 2021 18:20:37 +0000 (21:20 +0300)]
RDMA/core: Properly increment and decrement QP usecnts
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.
Leon Romanovsky [Tue, 3 Aug 2021 18:20:36 +0000 (21:20 +0300)]
RDMA/core: Configure selinux QP during creation
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.
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.
Leon Romanovsky [Tue, 3 Aug 2021 18:20:34 +0000 (21:20 +0300)]
RDMA/core: Remove protection from wrong in-kernel API usage
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.
Leon Romanovsky [Tue, 3 Aug 2021 18:20:33 +0000 (21:20 +0300)]
RDMA/core: Delete duplicated and unreachable code
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.
Leon Romanovsky [Fri, 23 Jul 2021 11:39:51 +0000 (14:39 +0300)]
RDMA/mlx5: Drop in-driver verbs object creations
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.
Leon Romanovsky [Fri, 23 Jul 2021 11:39:50 +0000 (14:39 +0300)]
RDMA: Globally allocate and release QP memory
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>
Leon Romanovsky [Fri, 23 Jul 2021 11:39:49 +0000 (14:39 +0300)]
RDMA/rdmavt: Decouple QP and SGE lists allocations
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.
Leon Romanovsky [Fri, 23 Jul 2021 11:39:47 +0000 (14:39 +0300)]
RDMA/mlx5: Delete device resource mutex that didn't protect anything
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.
Leon Romanovsky [Fri, 23 Jul 2021 11:39:46 +0000 (14:39 +0300)]
RDMA/mlx5: Cancel pkey work before destroying device resources
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.
Leon Romanovsky [Fri, 23 Jul 2021 11:39:44 +0000 (14:39 +0300)]
RDMA/hns: Don't overwrite supplied QP attributes
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.
Leon Romanovsky [Wed, 28 Jul 2021 13:04:12 +0000 (16:04 +0300)]
docs: Fix infiniband uverbs minor number
Starting from the beginning of infiniband subsystem, the uverbs char
devices start from 192 as a minor number, see
commit bc38a6abdd5a ("[PATCH] IB uverbs: core implementation").
This patch updates the admin guide documentation to reflect it.
Leon Romanovsky [Fri, 23 Jul 2021 14:08:57 +0000 (17:08 +0300)]
RDMA/iwpm: Rely on the rdma_nl_[un]register() to ensure that requests are valid
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.
Leon Romanovsky [Fri, 23 Jul 2021 14:08:55 +0000 (17:08 +0300)]
RDMA/iwcm: Release resources if iw_cm module initialization fails
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.
Mike Marciniszyn [Thu, 15 Jul 2021 16:04:45 +0000 (12:04 -0400)]
IB/hfi1: Adjust pkey entry in index 0
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: 7724105686e7 ("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>
Tal Gilboa [Sun, 18 Jul 2021 11:54:13 +0000 (14:54 +0300)]
IB/mlx5: Rename is_apu_thread_cq function to is_apu_cq
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>
Jason Gunthorpe [Tue, 20 Jul 2021 18:11:38 +0000 (15:11 -0300)]
Merge branch 'mlx5_dcs' into rdma.git for-next
Leon Romanovsky says:
====================
Add ConnectX DCS offload support
This patchset from Lior adds support of DCI stream channel (DCS) support.
DCS is an offload to SW load balancing of DC initiator work requests.
A single DC QP initiator (DCI) can be connected to only one target at the
time and can't start new connection until the previous work request is
completed.
This limitation causes to delays when the initiator process needs to
transfer data to multiple targets at the same time.
====================
* branch 'mlx5_dcs':
RDMA/mlx5: Add DCS offload support
RDMA/mlx5: Separate DCI QP creation logic
net/mlx5: Add DCS caps & fields support
Lior Nahmanson [Mon, 21 Jun 2021 07:06:16 +0000 (10:06 +0300)]
RDMA/mlx5: Add DCS offload support
DCS is an offload to SW load balancing of DC initiator work requests.
A single DCI can be connected to only one target at the time and can't
start new connection until the previous work request is completed. This
limitation will cause to delay when the initiator process needs to
transfer data to multiple targets at the same time. The SW solution is to
use a process that handling and spreading the work request on many DCIs
according to destinations.
This feature is an offload to this process and coming to reduce the load
from the CPU and improve the performance.
Lior Nahmanson [Mon, 21 Jun 2021 07:06:15 +0000 (10:06 +0300)]
RDMA/mlx5: Separate DCI QP creation logic
This patch isolates DCI QP creation logic to separate function, so this
change will reduce complexity when adding new features to DCI QP without
interfering with other QP types.
The code was copied from create_user_qp() while taking only DCI relevant bits.
Lior Nahmanson [Mon, 28 Dec 2020 08:38:12 +0000 (10:38 +0200)]
net/mlx5: Add DCS caps & fields support
This fields will be needed when adding a support for DCS offload
max_dci_stream_channels - maximum DCI stream channels supported per DCI.
max_dci_errored_streams - maximum DCI error stream channels
supported per DCI before a DCI move to error state.
Signed-off-by: Lior Nahmanson <liorna@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Bob Pearson [Wed, 7 Jul 2021 04:00:41 +0000 (23:00 -0500)]
RDMA/rxe: Fix types in rxe_icrc.c
Currently the ICRC is generated as a u32 type and then forced to a __be32
and stored into the ICRC field in the packet. The actual type of the ICRC
is __be32. This patch replaces u32 by __be32 and eliminates the casts.
The computation is exactly the same as the original but the types are more
consistent.
Bob Pearson [Wed, 7 Jul 2021 04:00:35 +0000 (23:00 -0500)]
RDMA/rxe: Fixup rxe_send and rxe_loopback
Fixup rxe_send() and rxe_loopback() in rxe_net.c to have the same calling
sequence. This patch makes them static and have the same parameter list
and return value.
IB/core: Read subnet_prefix in ib_query_port via cache.
ib_query_port() calls device->ops.query_port() to get the port
attributes. The method of querying is device driver specific. The same
function calls device->ops.query_gid() to get the GID and extract the
subnet_prefix (gid_prefix).
The GID and subnet_prefix are stored in a cache. But they do not get
read from the cache if the device is an Infiniband device. The
following change takes advantage of the cached subnet_prefix.
Testing with RDBMS has shown a significant improvement in performance
with this change.
IB/core: Shifting initialization of device->cache_lock
The lock cache_lock of struct ib_device is initialized in function
ib_cache_setup_one(). This is much later than the device initialization in
_ib_alloc_device().
This change shifts initialization of cache_lock in _ib_alloc_device().
IB/core: Updating cache for subnet_prefix in config_non_roce_gid_cache()
Currently, cache for subnet_prefix was getting updated by reading port
attributes via ib_query_port. ib_query_port() calls ops.query_gid() to get
subnet_prefix and returns it via port_attr.
In ib_cache_update(), config_non_roce_gid_cache() obtains GIDs by calling
ops.query_gid(). We utilize this to store subnet_prefix in cache.
Link: https://lore.kernel.org/r/20210712122625.1147-2-anand.a.khoje@oracle.com Suggested-by: Jason Gunthorpe <jgg@nvidia.com> Suggested-by: Aru Kolappan <aru.kolappan@oracle.com> Signed-off-by: Anand Khoje <anand.a.khoje@oracle.com> Signed-off-by: Haakon Bugge <haakon.bugge@oracle.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Ira Weiny [Thu, 24 Jun 2021 17:48:14 +0000 (10:48 -0700)]
RDMA/siw: Convert siw_tx_hdt() to kmap_local_page()
kmap() is being deprecated and will break uses of device dax after PKS
protection is introduced.[1]
The use of kmap() in siw_tx_hdt() is all thread local therefore
kmap_local_page() is a sufficient replacement and will work with pgmap
protected pages when those are implemented.
siw_tx_hdt() tracks pages used in a page_array. It uses that array to
unmap pages which were mapped on function exit. Not all entries in the
array are mapped and this is tracked in kmap_mask.
kunmap_local() takes a mapped address rather than a page. Alter
siw_unmap_pages() to take the iov array to reuse the iov_base address of
each mapping. Use PAGE_MASK to get the proper address for kunmap_local().
kmap_local_page() mappings are tracked in a stack and must be unmapped in
the opposite order they were mapped in. Because segments are mapped into
the page array in increasing index order, modify siw_unmap_pages() to
unmap pages in decreasing order.
Use kmap_local_page() instead of kmap() to map pages in the page_array.
Ira Weiny [Tue, 22 Jun 2021 06:14:21 +0000 (23:14 -0700)]
RDMA/siw: Remove kmap()
kmap() is being deprecated and will break uses of device dax after PKS
protection is introduced.[1]
These uses of kmap() in the SIW driver are thread local. Therefore
kmap_local_page() is sufficient to use and will work with pgmap protected
pages when those are implemnted.
There is one more use of kmap() in this driver which is split into its own
patch because kmap_local_page() has strict ordering rules and the use of
the kmap_mask over multiple segments must be handled carefully.
Therefore, that conversion is handled in a stand alone patch.
Use kmap_local_page() instead of kmap() in the 'easy' cases.
Jack Wang [Mon, 12 Jul 2021 06:07:50 +0000 (08:07 +0200)]
RDMA/rtrs: Move sq_wr_avail to rtrs_con
In order to account HB for sq_wr_avail properly, move sq_wr_avail from
rtrs_srv_con to rtrs_con.
Although rtrs-clt do not care sq_wr_avail, but still init it to
max_send_wr.
Fixes: b38041d50add ("RDMA/rtrs: Do not signal for heatbeat") Link: https://lore.kernel.org/r/20210712060750.16494-7-jinpu.wang@ionos.com Signed-off-by: Jack Wang <jinpu.wang@ionos.com> Reviewed-by: Aleksei Marov <aleksei.marov@ionos.com> Reviewed-by: Gioh Kim <gi-oh.kim@ionos.com> Reviewed-by: Md Haris Iqbal <haris.iqbal@ionos.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Jack Wang [Mon, 12 Jul 2021 06:07:49 +0000 (08:07 +0200)]
RDMA/rtrs: Remove unused flags parameter
flags is not used, so remove it from rtrs_post_rdma_write_imm_empty.
Link: https://lore.kernel.org/r/20210712060750.16494-6-jinpu.wang@ionos.com Signed-off-by: Jack Wang <jinpu.wang@ionos.com> Reviewed-by: Aleksei Marov <aleksei.marov@ionos.com> Reviewed-by: Gioh Kim <gi-oh.kim@ionos.com> Reviewed-by: Md Haris Iqbal <haris.iqbal@ionos.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Jack Wang [Mon, 12 Jul 2021 06:07:48 +0000 (08:07 +0200)]
RDMA/rtrs: Make rtrs_post_rdma_write_imm_empty static
It's only used in rtrs.c, so no need to export.
Link: https://lore.kernel.org/r/20210712060750.16494-5-jinpu.wang@ionos.com Signed-off-by: Jack Wang <jinpu.wang@ionos.com> Reviewed-by: Aleksei Marov <aleksei.marov@ionos.com> Reviewed-by: Gioh Kim <gi-oh.kim@ionos.com> Reviewed-by: Md Haris Iqbal <haris.iqbal@ionos.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Jack Wang [Mon, 12 Jul 2021 06:07:47 +0000 (08:07 +0200)]
RDMA/rtrs: Enable the same selective signal for heartbeat and IO
On idle session, because we do not do signal for heartbeat, it will
overflow the send queue after sometime.
To avoid that, we need to enable the signal for heartbeat. To do that, add
a new member signal_interval in rtrs_path, which will set min of
queue_depth and SERVICE_CON_QUEUE_DEPTH, and track it for both heartbeat
and IO, so the sq queue full accounting is correct.
Fixes: b38041d50add ("RDMA/rtrs: Do not signal for heatbeat") Link: https://lore.kernel.org/r/20210712060750.16494-4-jinpu.wang@ionos.com Signed-off-by: Jack Wang <jinpu.wang@ionos.com> Reviewed-by: Aleksei Marov <aleksei.marov@ionos.com> Reviewed-by: Gioh Kim <gi-oh.kim@ionos.com> Reviewed-by: Md Haris Iqbal <haris.iqbal@ionos.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Jack Wang [Mon, 12 Jul 2021 06:07:46 +0000 (08:07 +0200)]
RDMA/rtrs: move wr_cnt from rtrs_srv_con to rtrs_con
We need to track also the wr used for heatbeat. This is a preparation for
that, will be used in later patch.
The io_cnt in rtrs_clt is removed, use wr_cnt instead.
Link: https://lore.kernel.org/r/20210712060750.16494-3-jinpu.wang@ionos.com Signed-off-by: Jack Wang <jinpu.wang@ionos.com> Reviewed-by: Aleksei Marov <aleksei.marov@ionos.com> Reviewed-by: Gioh Kim <gi-oh.kim@ionos.com> Reviewed-by: Md Haris Iqbal <haris.iqbal@ionos.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Jack Wang [Mon, 12 Jul 2021 06:07:45 +0000 (08:07 +0200)]
RDMA/rtrs: Add error messages for failed operations.
It could help debugging in case of error happens.
Link: https://lore.kernel.org/r/20210712060750.16494-2-jinpu.wang@ionos.com Signed-off-by: Jack Wang <jinpu.wang@ionos.com> Reviewed-by: Aleksei Marov <aleksei.marov@ionos.com> Reviewed-by: Gioh Kim <gi-oh.kim@ionos.com> Reviewed-by: Md Haris Iqbal <haris.iqbal@ionos.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
I know nothing about zone_device pages and !device_private pages; but if
try_to_migrate_one() will do nothing for them, then it's better that
try_to_migrate() filter them first, than trawl through all their vmas.
Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: Shakeel Butt <shakeelb@google.com> Reviewed-by: Alistair Popple <apopple@nvidia.com> Link: https://lore.kernel.org/lkml/1241d356-8ec9-f47b-a5ec-9b2bf66d242@google.com/ Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Jason Gunthorpe <jgg@nvidia.com> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Yang Shi <shy828301@gmail.com> Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/rmap: fix new bug: premature return from page_mlock_one()
In the unlikely race case that page_mlock_one() finds VM_LOCKED has been
cleared by the time it got page table lock, page_vma_mapped_walk_done()
must be called before returning, either explicitly, or by a final call
to page_vma_mapped_walk() - otherwise the page table remains locked.
Fixes: cd62734ca60d ("mm/rmap: split try_to_munlock from try_to_unmap") Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: Alistair Popple <apopple@nvidia.com> Reviewed-by: Shakeel Butt <shakeelb@google.com> Reported-by: kernel test robot <oliver.sang@intel.com> Link: https://lore.kernel.org/lkml/20210711151446.GB4070@xsang-OptiPlex-9020/ Link: https://lore.kernel.org/lkml/f71f8523-cba7-3342-40a7-114abc5d1f51@google.com/ Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Jason Gunthorpe <jgg@nvidia.com> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Yang Shi <shy828301@gmail.com> Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/rmap: fix old bug: munlocking THP missed other mlocks
The kernel recovers in due course from missing Mlocked pages: but there
was no point in calling page_mlock() (formerly known as
try_to_munlock()) on a THP, because nothing got done even when it was
found to be mapped in another VM_LOCKED vma.
It's true that we need to be careful: Mlocked accounting of pte-mapped
THPs is too difficult (so consistently avoided); but Mlocked accounting
of only-pmd-mapped THPs is supposed to work, even when multiple mappings
are mlocked and munlocked or munmapped. Refine the tests.
There is already a VM_BUG_ON_PAGE(PageDoubleMap) in page_mlock(), so
page_mlock_one() does not even have to worry about that complication.
(I said the kernel recovers: but would page reclaim be likely to split
THP before rediscovering that it's VM_LOCKED? I've not followed that up)
Fixes: 9a73f61bdb8a ("thp, mlock: do not mlock PTE-mapped file huge pages") Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: Shakeel Butt <shakeelb@google.com> Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Link: https://lore.kernel.org/lkml/cfa154c-d595-406-eb7d-eb9df730f944@google.com/ Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Alistair Popple <apopple@nvidia.com> Cc: Jason Gunthorpe <jgg@nvidia.com> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Yang Shi <shy828301@gmail.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/rmap: fix comments left over from recent changes
Parallel developments in mm/rmap.c have left behind some out-of-date
comments: try_to_migrate_one() also accepts TTU_SYNC (already commented
in try_to_migrate() itself), and try_to_migrate() returns nothing at
all.
TTU_SPLIT_FREEZE has just been deleted, so reword the comment about it
in mm/huge_memory.c; and TTU_IGNORE_ACCESS was removed in 5.11, so
delete the "recently referenced" comment from try_to_unmap_one() (once
upon a time the comment was near the removed codeblock, but they drifted
apart).
Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: Shakeel Butt <shakeelb@google.com> Reviewed-by: Alistair Popple <apopple@nvidia.com> Link: https://lore.kernel.org/lkml/563ce5b2-7a44-5b4d-1dfd-59a0e65932a9@google.com/ Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Jason Gunthorpe <jgg@nvidia.com> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Yang Shi <shy828301@gmail.com> Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge tag 'irq-urgent-2021-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull irq fixes from Ingo Molnar:
"Two fixes:
- Fix a MIPS IRQ handling RCU bug
- Remove a DocBook annotation for a parameter that doesn't exist
anymore"
* tag 'irq-urgent-2021-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irqchip/mips: Fix RCU violation when using irqdomain lookup on interrupt entry
genirq/irqdesc: Drop excess kernel-doc entry @lookup
Merge tag 'sched-urgent-2021-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler fixes from Ingo Molnar:
"Three fixes:
- Fix load tracking bug/inconsistency
- Fix a sporadic CFS bandwidth constraints enforcement bug
- Fix a uclamp utilization tracking bug for newly woken tasks"
* tag 'sched-urgent-2021-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/uclamp: Ignore max aggregation if rq is idle
sched/fair: Fix CFS bandwidth hrtimer expiry type
sched/fair: Sync load_sum with load_avg after dequeue
- Add cstate event support for Intel ICELAKE_X and ICELAKE_D"
* tag 'perf-urgent-2021-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/intel/uncore: Clean up error handling path of iio mapping
perf/x86/cstate: Add ICELAKE_X and ICELAKE_D support
Merge tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull more SCSI updates from James Bottomley:
"This is a set of minor fixes and clean ups in the core and various
drivers.
The only core change in behaviour is the I/O retry for spinup notify,
but that shouldn't impact anything other than the failing case"
* tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (23 commits)
scsi: virtio_scsi: Add validation for residual bytes from response
scsi: ipr: System crashes when seeing type 20 error
scsi: core: Retry I/O for Notify (Enable Spinup) Required error
scsi: mpi3mr: Fix warnings reported by smatch
scsi: qedf: Add check to synchronize abort and flush
scsi: MAINTAINERS: Add mpi3mr driver maintainers
scsi: libfc: Fix array index out of bound exception
scsi: mvsas: Use DEVICE_ATTR_RO()/RW() macro
scsi: megaraid_mbox: Use DEVICE_ATTR_ADMIN_RO() macro
scsi: qedf: Use DEVICE_ATTR_RO() macro
scsi: qedi: Use DEVICE_ATTR_RO() macro
scsi: message: mptfc: Switch from pci_ to dma_ API
scsi: be2iscsi: Fix some missing space in some messages
scsi: be2iscsi: Fix an error handling path in beiscsi_dev_probe()
scsi: ufs: Fix build warning without CONFIG_PM
scsi: bnx2fc: Remove meaningless bnx2fc_abts_cleanup() return value assignment
scsi: qla2xxx: Add heartbeat check
scsi: virtio_scsi: Do not overwrite SCSI status
scsi: libsas: Add LUN number check in .slave_alloc callback
scsi: core: Inline scsi_mq_alloc_queue()
...
Merge tag 'perf-tools-for-v5.14-2021-07-10' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux
Pull more perf tool updates from Arnaldo Carvalho de Melo:
"New features:
- Enable use of BPF counters with 'perf stat --for-each-cgroup',
using per-CPU 'cgroup-switch' events with an attached BPF program
that does aggregation per-cgroup in the kernel instead of using
per-cgroup perf events.
- Add Topdown metrics L2 events as default events in 'perf stat' for
systems having those events.
Hardware tracing:
- Add a config for max loops without consuming a packet in the Intel
PT packet decoder, set via 'perf config intel-pt.max-loops=N'
Hardware enablement:
- Disable misleading NMI watchdog message in 'perf stat' on hybrid
systems such as Intel Alder Lake.
- Add a dummy event on hybrid systems to collect metadata records.
- Add 24x7 nest metric events for the Power10 platform.
Fixes:
- Fix event parsing for PMUs starting with the same prefix.
- Fix the 'perf trace' 'trace' alias installation dir.
- Fix buffer size to report iregs in perf script python scripts,
supporting the extended registers in PowerPC.
- Fix overflow in elf_sec__is_text().
- Fix 's' on source line when disasm is empty in the annotation TUI,
accessible via 'perf annotate', 'perf report' and 'perf top'.
- Plug leaks in scandir() returned dirent entries in 'perf test' when
sorting the shell tests.
- Fix --task and --stat with pipe input in 'perf report'.
- Fix 'perf probe' use of debuginfo files by build id.
- If a DSO has both dynsym and symtab ELF sections, read from both
when loading the symbol table, fixing a problem processing Fedora
32 glibc DSOs.
Libraries:
- Add grouping of events to libperf, from code in tools/perf,
allowing libperf users to use that mode.
Misc:
- Filter plt stubs from the 'perf probe --functions' output.
- Update UAPI header copies for asound, DRM, mman-common.h and the
ones affected by the quotactl_fd syscall"
* tag 'perf-tools-for-v5.14-2021-07-10' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux: (29 commits)
perf test: Add free() calls for scandir() returned dirent entries
libperf: Add tests for perf_evlist__set_leader()
libperf: Remove BUG_ON() from library code in get_group_fd()
libperf: Add group support to perf_evsel__open()
perf tools: Fix pattern matching for same substring in different PMU type
perf record: Add a dummy event on hybrid systems to collect metadata records
perf stat: Add Topdown metrics L2 events as default events
libperf: Adopt evlist__set_leader() from tools/perf as perf_evlist__set_leader()
libperf: Move 'nr_groups' from tools/perf to evlist::nr_groups
libperf: Move 'leader' from tools/perf to perf_evsel::leader
libperf: Move 'idx' from tools/perf to perf_evsel::idx
libperf: Change tests to single static and shared binaries
perf intel-pt: Add a config for max loops without consuming a packet
perf stat: Disable the NMI watchdog message on hybrid
perf vendor events power10: Adds 24x7 nest metric events for power10 platform
perf script python: Fix buffer size to report iregs in perf script
perf trace: Fix the perf trace link location
perf top: Fix overflow in elf_sec__is_text()
perf annotate: Fix 's' on source line when disasm is empty
perf probe: Do not show @plt function by default
...
Commit dbbee9d5cd83 ("mm/page_alloc: convert per-cpu list protection to
local_lock") folded in a workaround patch for pahole that was unable to
deal with zero-sized percpu structures.
A superior workaround is achieved with commit a0b8200d06ad ("kbuild:
skip per-CPU BTF generation for pahole v1.18-v1.21").
This patch reverts the dummy field and the pahole version check.
Fixes: dbbee9d5cd83 ("mm/page_alloc: convert per-cpu list protection to local_lock") Signed-off-by: Mel Gorman <mgorman@techsingularity.net> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>