Amrani, Ram [Wed, 24 Jan 2018 07:29:53 +0000 (09:29 +0200)]
MAINTAINERS: Remove Ram Amrani from Q-Logic RDMA driver
Remove myself from maintaining the qedr module as my period
of working with Cavium/Q-Logic has come to an end. I've had
a pleasure working with the community, cheers!
Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
The reason to it that init_rcu_head() is not exported and not supposed
to be used in modules. It is needed for dynamic initialization of
statically allocated rcu_head structures.
Fixes: 795bc112cd5a ("IB/srpt: Make it safe to use RCU for srpt_device.rch_list") Fixes: a11253142e6d ("IB/srpt: Rework multi-channel support") Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Mon, 22 Jan 2018 22:27:13 +0000 (14:27 -0800)]
IB/srp: Add target_can_queue login parameter
Although I'm not sure this parameter is useful for regular SRP users,
setting this parameter to 1 has shown to be invaluable for testing the
block layer core, SCSI core and device mapper queue running mechanisms.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Mon, 22 Jan 2018 22:27:12 +0000 (14:27 -0800)]
IB/srp: Add RDMA/CM support
Since the SRP_LOGIN_REQ defined in the SRP standard is larger than
what fits in the RDMA/CM login request private data, introduce a new
login request format for the RDMA/CM.
Note: since srp_daemon and ibsrpdm rely on the subnet manager and
since there is no equivalent of the IB subnet manager in non-IB
networks, login has to be performed manually for non-IB networks.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Mon, 22 Jan 2018 22:27:11 +0000 (14:27 -0800)]
kobject: Export kobj_ns_grab_current() and kobj_ns_drop()
Make it possible to call these two functions from a kernel module.
Note: despite their name, these two functions can be used meaningfully
independent of kobjects. A later patch will add calls to these
functions from the SRP driver because this patch series modifies the
SRP driver such that it can hold a reference to a namespace that can
last longer than the lifetime of the process through which the
namespace reference was obtained.
Parav Pandit [Tue, 9 Jan 2018 13:58:57 +0000 (15:58 +0200)]
RDMA/cma: Update RoCE multicast routines to use net namespace
rdma_dev_addr contains the net namespace pointer, while referring
bound_dev_if of the rdma_dev_addr, refer to the net namespace of
rdma_cm_id stored in rdma_dev_addr.
Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Tue, 9 Jan 2018 13:58:56 +0000 (15:58 +0200)]
RDMA/cma: Update cma_validate_port to honor net namespace
cma_validate_port uses rdma_dev_addr to validate the port of the cm_id.
It needs to honor the net namespace which is setup during cm_id creation
when finding netdevice.
Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Tue, 9 Jan 2018 13:58:54 +0000 (15:58 +0200)]
RDMA/cma: Check existence of netdevice during port validation
If valid netdevice is not found for RoCE, GID table should not be
searched with NULL netdevice.
Doing so causes the search routines to ignore the netdev argument and may
match the wrong GID table entry if the netdev is deleted.
Fixes: abae1b71dd37 ("IB/cma: cma_validate_port should verify the port and netdevice") Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Thu, 18 Jan 2018 08:11:19 +0000 (10:11 +0200)]
RDMA/core: Simplify rdma_addr_get_sgid() to not support RoCE
Now that all callers who care about RoCE addresses have been
converted to use rdma_read_gids() simplify rdma_addr_get_sgid()
to only support real GID addresses.
Callers should only use it for OPA and IB transports.
The now deleted implementation for RoCE has several bugs related to IPv6
support and incorrect/inconsistent 'GID' addresses compared to the CM
paths.
Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Thu, 18 Jan 2018 08:11:18 +0000 (10:11 +0200)]
net/rds: Use rdma_read_gids to read connection GIDs
Use the newly introduced rdma_read_gids() to read the SGID and DGID for
the connection which returns GID correctly for RoCE transport as well.
rdma_addr_get_dgid() for RoCE for client side connections returns MAC
address, instead of DGID.
rdma_addr_get_sgid() for RoCE doesn't return correct SGID for IPv6 and
when more than one IP address is assigned to the netdevice.
Therefore use transport agnostic rdma_read_gids() API provided by rdma_cm
module.
Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Thu, 18 Jan 2018 08:11:17 +0000 (10:11 +0200)]
RDMA/ucma: Use rdma cm API to query GID
Make use of rdma_read_gids() API to read SGID and DGID which returns
correct GIDs for RoCE and other transports.
rdma_addr_get_dgid() for RoCE for client side connections returns MAC
address, instead of DGID.
rdma_addr_get_sgid() for RoCE doesn't return correct SGID for IPv6 and
when more than one IP address is assigned to the netdevice.
Therefore use transport agnostic rdma_read_gids() API provided by rdma_cm
module.
Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Thu, 18 Jan 2018 08:11:16 +0000 (10:11 +0200)]
RDMA/cma: Introduce API to read GIDs for multiple transports
This patch introduces an API that allows legacy applications to query
GIDs for a rdma_cm_id which is used during connection establishment.
GIDs are stored and created differently for iWarp, IB and RoCE transports.
Therefore rdma_read_gids() returns GID for all the transports hiding
such internal details to caller.
It is usable for client side and server side connections.
In general continued use of GID based addressing outside of IB is
discouraged, so rdma_read_gids() should not be used by any new ULPs.
Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Bart Van Assche [Wed, 17 Jan 2018 00:14:17 +0000 (16:14 -0800)]
IB/srpt: Move the code for parsing struct ib_cm_req_event_param
This patch does not change any functionality but makes srpt_cm_req_recv()
independent of the IB/CM and hence simplifies the patch that introduces
RDMA/CM support.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Wed, 17 Jan 2018 00:14:16 +0000 (16:14 -0800)]
IB/srpt: Preparations for adding RDMA/CM support
Introduce a union in struct srpt_rdma_ch for member variables that
depend on the type of connection manager. Avoid that error messages
report the IB/CM ID.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Wed, 17 Jan 2018 00:14:15 +0000 (16:14 -0800)]
IB/srpt: Don't allow reordering of commands on wait list
If a receive I/O context is removed from the wait list and
srpt_handle_new_iu() fails to allocate a send I/O context then
re-adding the receive I/O context to the wait list can cause
reordering. Avoid this by only removing a receive I/O context
from the wait list after allocating a send I/O context succeeded.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Wed, 17 Jan 2018 00:14:14 +0000 (16:14 -0800)]
IB/srpt: Fix a race condition related to wait list processing
Wait list processing only occurs if the channel state >= CH_LIVE. Hence
set the channel state to CH_LIVE before triggering wait list processing
asynchronously.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Wed, 17 Jan 2018 00:14:13 +0000 (16:14 -0800)]
IB/srpt: Fix login-related race conditions
Make sure that sport->mutex is not released between the duplicate
channel check, adding a channel to the channel list and performing
the sport enabled check. Avoid that srpt_disconnect_ch() can be
invoked concurrently with the ib_send_cm_rep() call by
srpt_cm_req_recv().
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Wed, 17 Jan 2018 00:14:11 +0000 (16:14 -0800)]
IB/srpt: Simplify srpt_close_session()
Move a mutex lock and unlock statement from srpt_close_session()
into srpt_disconnect_ch_sync(). Since the previous patch removed
the last user of the return value of that function, change the
return value of srpt_disconnect_ch_sync() into void.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Wed, 17 Jan 2018 00:14:10 +0000 (16:14 -0800)]
IB/srpt: Rework multi-channel support
Store initiator and target port ID's once per nexus instead of in each
channel data structure. This change simplifies the duplicate connection
check in srpt_cm_req_recv().
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Wed, 17 Jan 2018 00:14:09 +0000 (16:14 -0800)]
IB/srpt: Use the source GID as session name
Use the source GID as session name instead of the initiator port ID
from the SRP login request. The only functional change in this patch
is that it changes the session name shown in debug messages.
Note: the fifth argument that is passed to target_alloc_session() is
what the SCSI target core uses as key for lookups in the ACL (access
control list) information.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Wed, 17 Jan 2018 00:14:08 +0000 (16:14 -0800)]
IB/srpt: One target per port
In multipathing setups where a target system is equipped with
dual-port HCAs it is useful to have one connection per target port
instead of one connection per target HCA. Hence move the connection
list (rch_list) from struct srpt_device into struct srpt_port.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Feras Daoud [Tue, 16 Jan 2018 18:08:41 +0000 (20:08 +0200)]
IB/mlx5: Mmap the HCA's clock info to user-space
This patch maps the new page to user space applications to
allow converting a user space completion timestamp to system wall
time at the lowest possible latency cost.
By using a versioning scheme we allow compatibility between current
and future userspace libraries.
The change moves mlx5_ib_mmap_cmd enum from mlx5_ib.h to the
abi header file mlx5-abi.h.
Feras Daoud [Tue, 16 Jan 2018 18:08:40 +0000 (20:08 +0200)]
net/mlx5e: Add clock info page to mlx5 core devices
Adds a new page to mlx5 core containing clock info data that allows
user level applications to translate between cqe timestamp to
nanoseconds. The information stored into this page is represented
through mlx5_ib_clock_info.
In order to synchronize between kernel and user space a sequence
number is incremented at the beginning and end of each update.
An odd number means the data is being updated while an even means
the access was already done. To guarantee that the data structure
was accessed atomically user will:
repeat:
seq1 = <read sequence>
goto <repeate> while odd
<read data structure>
seq2 = <read sequence>
if seq1 != seq2 goto repeat
Reviewed-by: Yishai Hadas <yishaih@mellanox.com> Reviewed-by: Jason Gunthorpe <jgg@mellanox.com> Reviewed-by: Alex Vesker <valex@mellanox.com> Signed-off-by: Feras Daoud <ferasda@mellanox.com> Signed-off-by: Eitan Rabin <rabin@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
Sagi Grimberg [Sun, 14 Jan 2018 15:07:50 +0000 (17:07 +0200)]
IB/cq: Don't force IB_POLL_DIRECT poll context for ib_process_cq_direct
polling the completion queue directly does not interfere
with the existing polling logic, hence drop the requirement.
Be aware that running ib_process_cq_direct with non IB_POLL_DIRECT
CQ may trigger concurrent CQ processing.
This can be used for polling mode ULPs.
Cc: Bart Van Assche <bart.vanassche@wdc.com> Reported-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
[maxg: added wcs array argument to __ib_process_cq] Signed-off-by: Max Gurtovoy <maxg@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Fri, 12 Jan 2018 23:11:58 +0000 (15:11 -0800)]
RDMA/rxe: Fix a race condition in rxe_requester()
The rxe driver works as follows:
* The send queue, receive queue and completion queues are implemented as
circular buffers.
* ib_post_send() and ib_post_recv() calls are serialized through a spinlock.
* Removing elements from various queues happens from tasklet
context. Tasklets are guaranteed to run on at most one CPU. This serializes
access to these queues. See also rxe_completer(), rxe_requester() and
rxe_responder().
* rxe_completer() processes the skbs queued onto qp->resp_pkts.
* rxe_requester() handles the send queue (qp->sq.queue).
* rxe_responder() processes the skbs queued onto qp->req_pkts.
Since rxe_drain_req_pkts() processes qp->req_pkts, calling
rxe_drain_req_pkts() from rxe_requester() is racy. Hence this patch.
Reported-by: Moni Shoua <monis@mellanox.com> Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Cc: stable@vger.kernel.org Signed-off-by: Doug Ledford <dledford@redhat.com>
Devesh Sharma [Thu, 11 Jan 2018 16:52:11 +0000 (11:52 -0500)]
RDMA/bnxt_re: Add SRQ support for Broadcom adapters
Shared receive queue (SRQ) is defined as a pool of
receive buffers shared among multiple QPs which belong
to same protection domain in a given process context.
Use of SRQ reduces the memory foot print of IB applications.
Broadcom adapters support SRQ, adding code-changes to enable
shared receive queue.
Selvin Xavier [Thu, 11 Jan 2018 16:52:10 +0000 (11:52 -0500)]
RDMA/bnxt_re: expose detailed stats retrieved from HW
Broadcom's adapter supports more granular statistics
to allow better understanding about the state of the
chip when data traffic is flowing.
Exposing the detailed stats to the consumer through
the standard hook available in the kverbs interface.
In order to retrieve all the information, driver
implements a firmware command.
Selvin Xavier [Thu, 11 Jan 2018 16:52:08 +0000 (11:52 -0500)]
RDMA/bnxt_re: Add support for query firmware version
The device now reports firmware version thus, removing
the hard coded values of the FW version string and
redundant fw_rev hook from sysfs. Adding code to query
firmware version from underlying device and report it
through the kernel verb to get firmware version string.
Selvin Xavier [Thu, 11 Jan 2018 16:52:07 +0000 (11:52 -0500)]
RDMA/bnxt_re: Enable RoCE on virtual functions
RoCE can be used by virtual functions (VFs) as well. Adding
code changes to allow resource reservation, initialization
and avail the resources to the RDMA applications running on
those VFs.
Currently, fifty percent of the total available resources
are reserved for PF and remaining are equally divided among
active VFs.
Sindhu Devale [Fri, 12 Jan 2018 00:10:52 +0000 (18:10 -0600)]
i40iw: Remove limit on re-posting AEQ entries to HW
Currently, if the number of processed Asynchronous Event Queue (AEQ)
entries exceeds 255, they are not returned to HW for re-use. During
scale-up, the unreturned AEQ entries can grow to the max AEQ size and
cause the HW to report an AEQ overflow.
Remove the check which limits the number of processed AEQ entries returned
to HW.
Shiraz Saleem [Fri, 12 Jan 2018 00:10:51 +0000 (18:10 -0600)]
i40iw: Zero-out consumer key on allocate stag for FMR
If the application invalidates the MR before the FMR WR, HW parses the
consumer key portion of the stag and returns an invalid stag key
Asynchronous Event (AE) that tears down the QP.
Fix this by zeroing-out the consumer key portion of the allocated stag
returned to application for FMR.
Fixes: ee855d3b93f3 ("RDMA/i40iw: Add base memory management extensions") Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
oulijun [Wed, 10 Jan 2018 06:39:53 +0000 (14:39 +0800)]
RDMA/hns: Set the guid for hip08 RoCE device
This patch assign a guid(Global Unique identifer) value to the hip08
device.
Signed-off-by: Lijun Ou <oulijun@huawei.com> Signed-off-by: Yixian Liu <liuyixian@huawei.com> Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
oulijun [Wed, 10 Jan 2018 06:39:52 +0000 (14:39 +0800)]
RDMA/hns: Update the verbs of polling for completion
If the port is a RoCEv2 port, the remote port address and QP information
which returned for UD will be modified.
Signed-off-by: Lijun Ou <oulijun@huawei.com> Signed-off-by: Yixian Liu <liuyixian@huawei.com> Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
oulijun [Wed, 10 Jan 2018 06:39:51 +0000 (14:39 +0800)]
RDMA/hns: Assign zero for pkey_index of wc in hip08
Because pkey is fixed for hip08 RoCE, it needs to assign zero for
pkey_index of wc. otherwise, it will happen an error when establishing
connection by communication management mechanism.
Signed-off-by: Lijun Ou <oulijun@huawei.com> Signed-off-by: Yixian Liu <liuyixian@huawei.com> Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
oulijun [Wed, 10 Jan 2018 06:39:50 +0000 (14:39 +0800)]
RDMA/hns: Fill sq wqe context of ud type in hip08
This patch mainly configure the fields of sq wqe of ud type when posting
wr of gsi qp type.
Signed-off-by: Lijun Ou <oulijun@huawei.com> Signed-off-by: Yixian Liu <liuyixian@huawei.com> Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
oulijun [Wed, 10 Jan 2018 06:39:49 +0000 (14:39 +0800)]
RDMA/hns: Add gsi qp support for modifying qp in hip08
It needs to Assign the values for some fields in qp context when qp type
is gsi qp type in hip08.
Signed-off-by: Lijun Ou <oulijun@huawei.com> Signed-off-by: Yixian Liu <liuyixian@huawei.com> Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
oulijun [Wed, 10 Jan 2018 06:39:48 +0000 (14:39 +0800)]
RDMA/hns: Create gsi qp in hip08
The gsi qp and rc qp use the same qp context structure and the created
flow, only differentiate them by qpn and qp type.
Signed-off-by: Lijun Ou <oulijun@huawei.com> Signed-off-by: Yixian Liu <liuyixian@huawei.com> Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
oulijun [Wed, 10 Jan 2018 06:39:47 +0000 (14:39 +0800)]
RDMA/hns: Assign the correct value for tx_cqn
When modifying qp from init to init, it need to assign the cqn of send cq
for tx cqn field of qp context. Otherwise, it will cause a mistake when
the send and recv cq sizes are different.
Signed-off-by: Lijun Ou <oulijun@huawei.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Xiongfeng Wang [Fri, 12 Jan 2018 07:56:05 +0000 (15:56 +0800)]
IB/cma: use strlcpy() instead of strncpy()
gcc-8 reports
drivers/infiniband/core/cma_configfs.c: In function 'make_cma_dev':
./include/linux/string.h:245:9: warning: '__builtin_strncpy' specified
bound 64 equals destination size [-Wstringop-truncation]
We need to use strlcpy() to make sure the string is nul-terminated.
Signed-off-by: Xiongfeng Wang <xiongfeng.wang@linaro.org> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Fri, 12 Jan 2018 05:58:42 +0000 (07:58 +0200)]
RDMA/core: Clarify rdma_ah_find_type
iWARP does not use rdma_ah_attr_type, and for this reason we do not have a
RDMA_AH_ATTR_TYPE_IWARP. rdma_ah_find_type should not even be called on iwarp
ports and for clarity it shouldn't have a special test for iWarp.
This changes the result from RDMA_AH_ATTR_TYPE_ROCE to RDMA_AH_ATTR_TYPE_IB
when wrongly called on an iWarp port.
Fixes: 44c58487d51a ("IB/core: Define 'ib' and 'roce' rdma_ah_attr types") Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Bodong Wang [Fri, 12 Jan 2018 05:58:41 +0000 (07:58 +0200)]
IB/core: Fix ib_wc structure size to remain in 64 bytes boundary
The change of slid from u16 to u32 results in sizeof(struct ib_wc)
cross 64B boundary, which causes more cache misses. This patch
rearranges the fields and remain the size to 64B.
Jack Morgenstein [Fri, 12 Jan 2018 05:58:40 +0000 (07:58 +0200)]
IB/mlx4: Fix incorrectly releasing steerable UD QPs when have only ETH ports
Allocating steerable UD QPs depends on having at least one IB port,
while releasing those QPs does not.
As a result, when there are only ETH ports, the IB (RoCE) driver
requests releasing a qp range whose base qp is zero, with
qp count zero.
When SR-IOV is enabled, and the VF driver is running on a VM over
a hypervisor which treats such qp release calls as errors
(rather than NOPs), we see lines in the VM message log like:
mlx4_core 0002:00:02.0: Failed to release qp range base:0 cnt:0
Fix this by adding a check for a zero count in mlx4_release_qp_range()
(which thus treats releasing 0 qps as a nop), and eliminating the
check for device managed flow steering when releasing steerable UD QPs.
(Freeing ib_uc_qpns_bitmap unconditionally is also OK, since it
remains NULL when steerable UD QPs are not allocated).
Cc: <stable@vger.kernel.org> Fixes: 4196670be786 ("IB/mlx4: Don't allocate range of steerable UD QPs for Ethernet-only device") Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Jason Gunthorpe [Thu, 11 Jan 2018 21:43:05 +0000 (14:43 -0700)]
RDMA: Mark imm_data as be32 in the verbs uapi header
This matches what the userspace copy of this header has been doing
for a while. imm_data is an opaque 4 byte array carried over the network,
and invalidate_rkey is in CPU byte order.
Parav Pandit [Tue, 9 Jan 2018 13:24:52 +0000 (15:24 +0200)]
IB/core: Attempt DMAC resolution for only RoCE
Instead of returning 0 (success) for RoCE scenarios where DMAC should
not be resolved, avoid such attempt and make code consistent with
ib_create_user_ah().
Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Tue, 9 Jan 2018 13:24:51 +0000 (15:24 +0200)]
IB/core: Limit DMAC resolution to userspace QPs
Currently ah_attr is initialized by the ib_cm layer for rdma_cm
based applications. For RoCE transport ah_attr.roce.dmac is already
initialized by ib_cm, rdma_cm either from wc, path record, route
resolve, explicit path record setting depending on active or passive
side QP. Therefore avoid resolving DMAC for QP of kernel consumers.
Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Tue, 9 Jan 2018 13:24:50 +0000 (15:24 +0200)]
IB/core: Perform modify QP on real one
Currently qp->port stores the port number whenever IB_QP_PORT
QP attribute mask is set (during QP state transition to INIT state).
This port number should be stored for the real QP when XRC target QP
is used.
Follow the ib_modify_qp() implementation and hide the access to ->real_qp.
Fixes: a512c2fbef9c ("IB/core: Introduce modify QP operation with udata") Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Randy Dunlap [Sat, 6 Jan 2018 00:22:32 +0000 (16:22 -0800)]
infiniband: fix sw/rdmavt/* kernel-doc notation
Use correct parameter names and formatting in function kernel-doc notation
to eliminate warnings from scripts/kernel-doc.
../drivers/infiniband/sw/rdmavt/mr.c:784: warning: Excess function parameter 'ibmfr' description in 'rvt_map_phys_fmr'
../drivers/infiniband/sw/rdmavt/vt.c:234: warning: Excess function parameter 'intex' description in 'rvt_query_pkey'
../drivers/infiniband/sw/rdmavt/vt.c:266: warning: Excess function parameter 'index' description in 'rvt_query_gid'
../drivers/infiniband/sw/rdmavt/vt.c:306: warning: Excess function parameter 'data' description in 'rvt_alloc_ucontext'
../drivers/infiniband/sw/rdmavt/cq.c:65: warning: Excess function parameter 'sig' description in 'rvt_cq_enter'
../drivers/infiniband/sw/rdmavt/qp.c:279: warning: Excess function parameter 'qpt' description in 'rvt_free_all_qps'
../drivers/infiniband/sw/rdmavt/mcast.c:282: warning: Excess function parameter 'igd' description in 'rvt_attach_mcast'
../drivers/infiniband/sw/rdmavt/mcast.c:345: warning: Excess function parameter 'igd' description in 'rvt_detach_mcast'
Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Cc: Doug Ledford <dledford@redhat.com> Cc: Jason Gunthorpe <jgg@mellanox.com> Cc: linux-doc@vger.kernel.org Acked-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Mon, 8 Jan 2018 15:04:48 +0000 (17:04 +0200)]
RDMA/cma: Fix rdma_cm path querying for RoCE
The 'if' logic in ucma_query_path was broken with OPA was introduced
and started to treat RoCE paths as as OPA paths. Invert the logic
of the 'if' so only OPA paths are treated as OPA paths.
Otherwise the path records returned to rdma_cma users are mangled
when in RoCE mode.
Fixes: 57520751445b ("IB/SA: Add OPA path record type") Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Mon, 8 Jan 2018 15:04:47 +0000 (17:04 +0200)]
RDMA/cma: Fix rdma_cm raw IB path setting for RoCE
rdma_set_ib_path() missed setting path record fields for RoCE
transport when RoCE support was added.
This results in setting incorrect ndev, destination mac address,
incorrect GID type etc errors when user space attempts to set a raw
IB path using the roce IB path compatibility mapping from userspace.
Fixes: 3c86aa70bf67 ("RDMA/cm: Add RDMA CM support for IBoE devices") Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Huy Nguyen [Mon, 8 Jan 2018 10:15:38 +0000 (12:15 +0200)]
IB/core: Increase number of char device minors
There is a need to increase number of possible char devices to support
large number of SR-IOV instances. The current limit is in the range of
64-128 devices/ports. Increase it to support up to 1024.
The patch performs the following steps to refactor the code:
1. Removes the split bitmap for fixed and overflow dev numbers.
2. Pre-allocates the non-legacy major number range during driver
initialization, choosen for simplicity.
3. Add new define (RDMA_MAX_PORTS) that is shared between all drivers.
This is the maximum total number of ports on all struct ib_devices.
4. Set RDMA_MAX_PORTS to 1024.
Signed-off-by: Huy Nguyen <huyn@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Huy Nguyen [Mon, 8 Jan 2018 10:15:37 +0000 (12:15 +0200)]
IB/core: Remove the locking for character device bitmaps
Remove the locks that protect character device bitmaps of
uverbs, umad and issm.
The character device bitmaps are accessed in "client->add" and
"client->remove" calls from ib_register_device and ib_unregister_device
respectively. These calls are already protected by the "device_mutex"
mutex. Thus, the spinlocks are not needed.
Signed-off-by: Huy Nguyen <huyn@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Bart Van Assche [Tue, 9 Jan 2018 19:23:40 +0000 (11:23 -0800)]
RDMA/rxe: Fix a race condition related to the QP error state
The following sequence:
* Change queue pair state into IB_QPS_ERR.
* Post a work request on the queue pair.
Triggers the following race condition in the rdma_rxe driver:
* rxe_qp_error() triggers an asynchronous call of rxe_completer(), the function
that examines the QP send queue.
* rxe_post_send() posts a work request on the QP send queue.
If rxe_completer() runs prior to rxe_post_send(), it will drain the send
queue and the driver will assume no further action is necessary.
However, once we post the send to the send queue, because the queue is
in error, no send completion will ever happen and the send will get
stuck. In order to process the send, we need to make sure that
rxe_completer() gets run after a send is posted to a queue pair in an
error state. This patch ensures that happens.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Cc: Moni Shoua <monis@mellanox.com> Cc: <stable@vger.kernel.org> # v4.8 Signed-off-by: Doug Ledford <dledford@redhat.com>
Colin Ian King [Tue, 9 Jan 2018 15:55:43 +0000 (15:55 +0000)]
IB/mlx5: remove redundant assignment of mdev
The initial assignment to mdev is redundant as mdev is re-assigned
later and the first assigned value is never read. Remove this
redundant assignment.
Cleans up clang warning:
drivers/infiniband/hw/mlx5/main.c:359:24: warning: Value stored
to 'mdev' during its initialization is never read
Signed-off-by: Colin Ian King <colin.king@canonical.com> Acked-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Zhu Yanjun [Mon, 8 Jan 2018 05:14:25 +0000 (00:14 -0500)]
IB/rxe: remove unnecessary skb_clone in xmit
In xmit, there is a skb_clone. This function copies the struct sk_buff.
And some parameters are changed to the new skb. Then the new skb is sent
while the old skb is freed.
While the function skb_clone is removed, the parameter changes are made on
the old skb, then the old skb is sent. It can also work well.
The following tests are made.
server client
--------- ---------
|1.1.1.1|<----rxe-channel--->|1.1.1.2|
--------- ---------
On server: rping -s -a 1.1.1.1 -v -C 1000 -S 512
On client: rping -c -a 1.1.1.1 -v -C 1000 -S 512
The kernel config CONFIG_DEBUG_KMEMLEAK is enabled on both server
and client.
This test runs for several hours. There is no memory leak and the whole
system can work well.
As the above network, the following tests are made.
Bart Van Assche [Mon, 8 Jan 2018 19:00:51 +0000 (11:00 -0800)]
IB/srpt: Micro-optimize I/O context state manipulation
Since all I/O context state changes are already serialized, it is
not necessary to protect I/O context state changes with the I/O
context spinlock. Hence remove that spinlock.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Mon, 8 Jan 2018 19:00:50 +0000 (11:00 -0800)]
IB/srpt: Inline srpt_get_cmd_state()
It is not necessary to obtain ioctx->spinlock when reading the ioctx
state. Since after removal of this locking only a single line remains,
inline the srpt_get_cmd_state() function.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Mon, 8 Jan 2018 19:00:48 +0000 (11:00 -0800)]
IB/srpt: Reduce frequency of receive failure messages
Disabling an SRP target port causes the state of all QPs associated
with a port to be changed into IB_QPS_ERR. Avoid that this causes
one error message per I/O context to be reported.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Mon, 8 Jan 2018 19:00:47 +0000 (11:00 -0800)]
IB/srpt: Convert a warning into a debug message
At least when running the ib_srpt driver on top of the rdma_rxe
driver it is easy to trigger a zero-length write completion in
the CH_DISCONNECTED state. Hence make the message that reports
this less noisy.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Mon, 8 Jan 2018 19:00:43 +0000 (11:00 -0800)]
IB/srpt: Rename a local variable, a member variable and a constant
Rename rsp_size into max_rsp_size and SRPT_RQ_SIZE into MAX_SRPT_RQ_SIZE.
The new names better reflect the role of this member variable and constant.
Since the prefix "srp_" is superfluous in the context of the function
that creates an RDMA channel, rename srp_sq_size into sq_size.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Daniel Jurgens [Thu, 4 Jan 2018 15:25:43 +0000 (17:25 +0200)]
IB/mlx5: Don't advertise RAW QP support in dual port mode
When operating in dual port RoCE mode FW doesn't support steering for
raw QPs on the slave port. They still work on the master port, but
the user has no way of knowing which port is the master. The
capability is reported per device, not per port.
Signed-off-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Daniel Jurgens [Thu, 4 Jan 2018 15:25:41 +0000 (17:25 +0200)]
{net, IB}/mlx5: Change set_roce_gid to take a port number
When in dual port mode setting a RoCE GID for any port flows through the
master ports mlx5_core_dev. Provide an interface to set the port when
sending this command.
Signed-off-by: Daniel Jurgens <danielj@mellanox.com> Reviewed-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Daniel Jurgens [Thu, 4 Jan 2018 15:25:38 +0000 (17:25 +0200)]
IB/mlx5: Implement dual port functionality in query routines
Port operations must be routed to their native mlx5_core_dev. A
multiport RoCE device registers itself as having 2 ports even before a
2nd port is affiliated. If an unaffilated port is queried use capability
information from the master port, these values are the same.
Signed-off-by: Daniel Jurgens <danielj@mellanox.com> Reviewed-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>