This patch increases checking CMQ status timeout value and
uses the same value with NIC driver to avoid deficiency of
time.
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>
This patch modified uar allocation algorithm in hns_roce_uar_alloc
function to avoid bitmap exhaust.
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>
mlx5 core dirver updates for both net-next and rdma-next branches.
From Christophe JAILLET, first three patche to use kvfree where needed.
From: Or Gerlitz <ogerlitz@mellanox.com>
Next six patches from Roi and Co adds support for merged
sriov e-switch which comes to serve cases where both PFs, VFs set
on them and both uplinks are to be used in single v-switch SW model.
When merged e-switch is supported, the per-port e-switch is logically
merged into one e-switch that spans both physical ports and all the VFs.
This model allows to offload TC eswitch rules between VFs belonging
to different PFs (and hence have different eswitch affinity), it also
sets the some of the foundations needed for uplink LAG support.
-----BEGIN PGP SIGNATURE-----
iQEcBAABAgAGBQJa/fLEAAoJEEg/ir3gV/o+7jUH/3n5/Uw1LLt3TfeKArx6i0F1
3G4U5B0ha03qiDqXprwhyQ3I6lgYmRBmjcxnqmvcqOAqO4/hSsjtTR+A/mgbEDhJ
YtdekFNEX+72h/N2GIpZwChIWSE3EcMPaLYnV8TwLUgh9YSust2sCLSBbJCjxOKc
j78M8ept/bXZwTm/iJhEjtmqw0xl91rl011chCAua0iEpH3wxteDARmKABFHMQxl
I3N/x/e/astgcSCNgpO4uDf9zEIRkNdzcHPzSMJ6C2Oo5W9XiZEekfw7WKj9nXfa
G+eGckkAyCOQ/r2lZ9nA0ZUvQ2X6JISvxgohuaCNwTgsz3acTxbLnQK4YWHzQCQ=
=iHi6
-----END PGP SIGNATURE-----
Merge tag 'mlx5-updates-2018-05-17' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux into for-next
mlx5-updates-2018-05-17
mlx5 core dirver updates for both net-next and rdma-next branches.
From Christophe JAILLET, first three patche to use kvfree where needed.
From: Or Gerlitz <ogerlitz@mellanox.com>
Next six patches from Roi and Co adds support for merged
sriov e-switch which comes to serve cases where both PFs, VFs set
on them and both uplinks are to be used in single v-switch SW model.
When merged e-switch is supported, the per-port e-switch is logically
merged into one e-switch that spans both physical ports and all the VFs.
This model allows to offload TC eswitch rules between VFs belonging
to different PFs (and hence have different eswitch affinity), it also
sets the some of the foundations needed for uplink LAG support.
* tag 'mlx5-updates-2018-05-17' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux:
net/mlx5e: Explicitly set source e-switch in offloaded TC rules
net/mlx5: Add source e-switch owner
net/mlx5e: Explicitly set destination e-switch in FDB rules
net/mlx5: Add destination e-switch owner
net/mlx5: Properly handle a vport destination when setting FTE
net/mlx5: Add merged e-switch cap
IB/mlx5: Use 'kvfree()' for memory allocated by 'kvzalloc()'
net/mlx5: Eswitch, Use 'kvfree()' for memory allocated by 'kvzalloc()'
net/mlx5: Vport, Use 'kvfree()' for memory allocated by 'kvzalloc()'
Instead of open coding memcmp() to check whether a given GID is zero or
not, use a helper function to do so, and replace instances of
memcpy(z,&zgid) with memset.
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
On fatal error the driver simulates CQE's for ULPs that rely on
completion of all their posted work-request.
For the GSI traffic, the mlx5 has its own mechanism that sends the
completions via software CQE's directly to the relevant CQ.
This should be kept in fatal error too, so the driver should simulate
such CQE's with the specified error state in order to complete GSI QP
work requests.
Without the fix the next deadlock might appears:
schedule_timeout+0x274/0x350
wait_for_common+0xec/0x240
mcast_remove_one+0xd0/0x120 [ib_core]
ib_unregister_device+0x12c/0x230 [ib_core]
mlx5_ib_remove+0xc4/0x270 [mlx5_ib]
mlx5_detach_device+0x184/0x1a0 [mlx5_core]
mlx5_unload_one+0x308/0x340 [mlx5_core]
mlx5_pci_err_detected+0x74/0xe0 [mlx5_core]
Cc: <stable@vger.kernel.org> # 4.7
Fixes: 89ea94a7b6 ("IB/mlx5: Reset flow support for IB kernel ULPs")
Signed-off-by: Erez Shitrit <erezsh@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Remove various prints of VMA pointers.
Reported-by: Michal Kalderon <Michal.Kalderon@cavium.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Reviewed-by: Michal Kalderon <michal.kalderon@cavium.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
The lower 15 bit of paramter of db structure means different
meanings when db type is sq, rq and srq.
Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Given we are dealing with nano-second level timers, when the timer
pops, ensure it happens on the CPU which caused the timer to be set
in the first place. This avoids excessive jitter from the desired
expiration time by avoiding the cost of switching our context to
another CPU that is cache cold for this given timer.
Reviewed-by: Kaike Wan <kaike.wan@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
For errorinfo MAD requests, the response has a 0 port number left over
from a memset. Instead we should always set the port number in the
response.
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
The knowledge of the internal workings of the expect receive
is too distributed.
Fix by:
- right size several rcd fields associated with
expect receive
- making an init entrance to init all the lists
- consolidate all the allocations into an array anchored
in the rcd
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Reviewed-by: Kaike Wan <kaike.wan@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Add trace support for 16B Management Packets.
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Don Hiatt <don.hiatt@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
16B Management Packets (L4=0x08) replace the BTH and DETH
of normal MAD packet packets with a header containing the
the source and destination queue pair numbers; fields that
were originally retrieved from the BTH/DETH are now populated
from this header as well as from the 16B LRH (e.g. pkey).
16B Management Packets are used as an optimized management
format on 16B fabrics.
These management packets have an opcode of IB_OPCODE_UD_SEND_ONLY,
a fixed 3Byte pad, and a header length of 24Bytes.
The decision as to when we send a management packet is based
upon either the source or destination queue pair number being
0 or 1.
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Don Hiatt <don.hiatt@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Add 16B Management Packet definition. This optimized packet
format replaces the ib_other_headers and BTH with a source
and destination QP number.
To support these packets we introduce struct opa_16b_mgmt
into the struct hfi1_16b_header.
This packet format is only used for MAD packets using the
IB_OPCODE_UD_SEND_ONLY opcode on QP0/1.
The original 16B implementation failed to use 16B management
packets so now we add their definition.
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Don Hiatt <don.hiatt@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Add a table of important fields from the fw_ri_tpte structure to the mr
resource tracking table. This is helpful in debugging.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Add a table of important fields from the c4iw_cq* structures to the cq
resource tracking table. This is helpful in debugging.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Add a table of important fields from the c4iw_ep* structures to the cm_id
resource tracking table. This is helpful in debugging.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
In active side connections, the provider_data field is not
getting set. This will be used in a subsequent patch to dump
state, so always set it.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
This patch adds the support of 64KB page size for hip08
in kernel.
Signed-off-by: Yixian Liu <liuyixian@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This patch reports the device's capbilities to offload
encapsulated MPLS tunnel protocols to user-space:
- Capability to offload MPLS over GRE.
- Capability to offload MPLS over UDP.
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Ariel Levkovich <lariel@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This patch introduces support for the MPLS flow spec and
allows the creation of rules that are matching on the
MPLS label.
Applying the rule matching depends on the flow specs order and
the location of the MPLS in the spec list as there are different
configurations to be made in the device in the cases of MPLSoGRE
and MPLSoUDP vs. non-encapsulated MPLS.
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Ariel Levkovich <lariel@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This patch introduces support for the GRE flow spec and
allowing the creation of rules based on the protocol and
key fields that are part of GRE protocol header.
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Ariel Levkovich <lariel@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
When 'kvzalloc()' is used to allocate memory, 'kvfree()' must be used to
free it.
Fixes: 1cbe6fc86c ("IB/mlx5: Add support for CQE compressing")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Acked-by: Jason Gunthorpe <jgg@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
If two listeners are created with different IP's but
same port, the second rdma_listen fails due to a
duplicate port entry being added from the CQP add
APBVT OP. commit f16dc0aa5e ("i40iw: Add support
for port reuse on active side connections") does not
account for listener side port reuse.
Check for duplicate port before invoking the CQP command
to add APBVT entry and delete the entry only if the port
is not in use. Additionally, consolidate all port-reuse
logic into i40iw_manage_apbvt.
Fixes: f16dc0aa5e ("i40iw: Add support for port reuse on active side connections")
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Remove sq/rq wr_id attributes because typically they are pointers and
we don't want to pass up kernel pointers.
Fixes: 056f9c7f39 ("iw_cxgb4: dump detailed driver-specific QP information")
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
A recent patch set to rework the usage of debugfs and to add fault
injection capabilities via debugfs files to the hfi1 driver introduced a
build error that only shows up when debugfs is fully disabled. The
patchset mistakenly defines some empty stub functions in two different
headers when debugfs is disabled. Remove the set that shouldn't have
been there to resolve the issue.
Fixes: a74d5307ca ("IB/hfi1: Rework fault injection machinery")
Signed-off-by: Doug Ledford <dledford@redhat.com>
Moving receive-side WQE allocation logic into rdmavt will allow
further code reuse between qib and hfi1 drivers.
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Brian Welty <brian.welty@intel.com>
Signed-off-by: Harish Chegondi <harish.chegondi@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Currently the driver doesn't support completion vectors. These
are used to indicate which sets of CQs should be grouped together
into the same vector. A vector is a CQ processing thread that
runs on a specific CPU.
If an application has several CQs bound to different completion
vectors, and each completion vector runs on different CPUs, then
the completion queue workload is balanced. This helps scale as more
nodes are used.
Implement CQ completion vector support using a global workqueue
where a CQ entry is queued to the CPU corresponding to the CQ's
completion vector. Since the workqueue is global, it's guaranteed
to always be there when queueing CQ entries; Therefore, the RCU
locking for cq->rdi->worker in the hot path is superfluous.
Each completion vector is assigned to a different CPU. The number of
completion vectors available is computed by taking the number of
online, physical CPUs from the local NUMA node and subtracting the
CPUs used for kernel receive queues and the general interrupt.
Special use cases:
* If there are no CPUs left for completion vectors, the same CPU
for the general interrupt is used; Therefore, there would only
be one completion vector available.
* For multi-HFI systems, the number of completion vectors available
for each device is the total number of completion vectors in
the local NUMA node divided by the number of devices in the same
NUMA node. If there's a division remainder, the first device to
get initialized gets an extra completion vector.
Upon a CQ creation, an invalid completion vector could be specified.
Handle it as follows:
* If the completion vector is less than 0, set it to 0.
* Set the completion vector to the result of the passed completion
vector moded with the number of device completion vectors
available.
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
CPU masks are used to keep track of affinity assignments for IRQs
and processes. Operations performed on these affinity CPU masks are
duplicated throughout the code.
Create common functions for affinity CPU mask operations to remove
duplicate code.
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
When Hfi1 device is unresponsive, reading the RcvArrayCnt register
will return all 1's. This value is then used to remap chip's RcvArray.
The incorrect all ones value used in remapping RcvArray
will cause warn on as shown by trace below:
[<ffffffff81685eac>] dump_stack+0x19/0x1b
[<ffffffff81085820>] warn_slowpath_common+0x70/0xb0
[<ffffffff810858bc>] warn_slowpath_fmt+0x5c/0x80
[<ffffffff81065c29>] __ioremap_caller+0x279/0x320
[<ffffffff8142873c>] ? _dev_info+0x6c/0x90
[<ffffffffa021d155>] ? hfi1_pcie_ddinit+0x1d5/0x330 [hfi1]
[<ffffffff81065d62>] ioremap_wc+0x32/0x40
[<ffffffffa021d155>] hfi1_pcie_ddinit+0x1d5/0x330 [hfi1]
[<ffffffffa0204851>] hfi1_init_dd+0x1d1/0x2440 [hfi1]
[<ffffffff813503dc>] ? pci_write_config_word+0x1c/0x20
Read CCE revision register first to verify that WFR device is
responsive. If the read return "all ones", bail out from init
and fail the driver load.
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Kamenee Arumugam <kamenee.arumugam@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
The packet fault injection code present in the HFI1 driver had some
issues which not only fragment the code but also created user
confusion. Furthermore, it suffered from the following issues:
1. The fault_packet method only worked for received packets. This
meant that the only fault injection mode available for sent
packets is fault_opcode, which did not allow for random packet
drops on all egressing packets.
2. The mask available for the fault_opcode mode did not really work
due to the fact that the opcode values are not bits in a bitmask but
rather sequential integer values. Creating a opcode/mask pair that
would successfully capture a set of packets was nearly impossible.
3. The code was fragmented and used too many debugfs entries to
operate and control. This was confusing to users.
4. It did not allow filtering fault injection on a per direction basis -
egress vs. ingress.
In order to improve or fix the above issues, the following changes have
been made:
1. The fault injection methods have been combined into a single fault
injection facility. As such, the fault injection has been plugged
into both the send and receive code paths. Regardless of method used
the fault injection will operate on both egress and ingress packets.
2. The type of fault injection - by packet or by opcode - is now controlled
by changing the boolean value of the file "opcode_mode". When the value
is set to True, fault injection is done by opcode. Otherwise, by
packet.
2. The masking ability has been removed in favor of a bitmap that holds
opcodes of interest (one bit per opcode, a total of 256 bits). This
works in tandem with the "opcode_mode" value. When the value of
"opcode_mode" is False, this bitmap is ignored. When the value is
True, the bitmap lists all opcodes to be considered for fault injection.
By default, the bitmap is empty. When the user wants to filter by opcode,
the user sets the corresponding bit in the bitmap by echo'ing the bit
position into the 'opcodes' file. This gets around the issue that the set
of opcodes does not lend itself to effective masks and allow for extremely
fine-grained filtering by opcode.
4. fault_packet and fault_opcode methods have been combined. Hence, there
is only one debugfs directory controlling the entire operation of the
fault injection machinery. This reduces the number of debugfs entries
and provides a more unified user experience.
5. A new control files - "direction" - is provided to allow the user to
control the direction of packets, which are subject to fault injection.
6. A new control file - "skip_usec" - is added that would allow the user
to specify a "timeout" during which no fault injection will occur.
In addition, the following bug fixes have been applied:
1. The fault injection code has been split into its own header and source
files. This was done to better organize the code and support conditional
compilation without littering the code with #ifdef's.
2. The method by which the TX PIO packets were being marked for drop
conflicted with the way send contexts were being setup. As a result,
the send context was repeatedly being reset.
3. The fault injection only makes sense when the user can control it
through the debugfs entries. However, a kernel configuration can
enable fault injection but keep fault injection debugfs entries
disabled. Therefore, it makes sense that the HFI fault injection
code depends on both.
4. Error suppression did not take into account the method by which PIO
packets were being dropped. Therefore, even with error suppression
turned on, errors would still be displayed to the screen. A larger
enough packet drop percentage would case the kernel to crash because
the driver would be stuck printing errors.
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: Don Hiatt <don.hiatt@intel.com>
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Mitko Haralanov <mitko.haralanov@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
A warm restart will fail to unload the driver, leaving link state
potentially flapping up to the point the BIOS resets the adapter.
Correct the issue by hooking the shutdown pci method,
which will bring port down.
Cc: <stable@vger.kernel.org> # 4.9.x
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Alex Estrin <alex.estrin@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
User send context integrity bits are cleared before the context is
disabled. If the send context is still processing data, any packets
that need those integrity bits will cause an error and halt the send
context.
During the disable handling, the driver waits for the context to drain.
If the context is halted, the driver will eventually timeout because
the context won't drain and then incorrectly bounce the link.
Reorder the bit clearing and the context disable.
Examine the software state and send context status as well as the
egress status to determine if a send context is in the halted state.
Promote the check macros to static functions for consistency with the
new check and to follow kernel style.
Remove an unused define that refers to the egress timeout.
Cc: <stable@vger.kernel.org> # 4.9.x
Reviewed-by: Mitko Haralanov <mitko.haralanov@intel.com>
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
The driver_pstate() function is used to map internal driver state
information to externally defined states.
The VERIFY_CAP and GOING_UP states are config/training states, but
the mapping routing returns the POLLING value.
Update the return values for VERIFY_CAP and GOING_UP to return the
correct value: TRAINING.
Reviewed-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
For lid routed packets 'hop_cnt' is zero, therefore current
test is incomplete. Fix it by using local mad check for
both lid routed and direct routed MADs.
Reviewed-by: Mike Mariciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Alex Estrin <alex.estrin@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
A failure of program_rcvarray() is treated inconsistently by the
calling function. In one case the error is returned, in a second
case, the error is overwritten with EFAULT. In both cases the
code path is doing the same thing, allocating memory for groups,
so it should be consistent.
Make the error path consistent and return the error generated by
program_rcvarray().
Reviewed-by: Harish Chegondi <harish.chegondi@intel.com>
Fixes: 7e7a436ecb ("staging/hfi1: Add TID entry program function body")
Signed-off-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
When the LCB isn't able to get any lanes operational on the
first transition into mission mode, the link transfer active
never happens and the LNI stays in the polling state indefinitely.
Reset LCB upon receiving an 8051 interrupt for LCB to try to obtain
lanes with firmware version 1.25.0 or later. Also, update the LCB
reset value in other parts of the code with a macro defined to make
the code more maintainable and rename functions with the link_width
label to link_mode to reflect the fact that those functions set and
read link related data not just the link width.
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Several items of conflict have arisen between the RDMA stack's for-rc
branch and upcoming for-next work:
9fd4350ba8 ("IB/rxe: avoid double kfree_skb") directly conflicts with
2e47350789 ("IB/rxe: optimize the function duplicate_request")
Patches already submitted by Intel for the hfi1 driver will fail to
apply cleanly without this merge
Other people on the mailing list have notified that their upcoming
patches also fail to apply cleanly without this merge
Signed-off-by: Doug Ledford <dledford@redhat.com>
As most kernel RDMA ULPs, (e.g. NVMe over Fabrics in its default
"register_always=Y" mode) registers and invalidates user buffer
upon each IO.
Today the mlx5 driver is posting the registration work
request using scatter/gather entry for the MTT/KLM list.
The fetch of the MTT/KLM list becomes the bottleneck in
number of IO operation could be done by NVMe over Fabrics
host driver on a single adapter as shown below.
This patch is adding the support for inline registration
work request upon MTT/KLM list of size <=64B.
The result for NVMe over Fabrics is increase of > x3.5 for small
IOs as shown below, I expect other ULPs (e.g iSER, SRP, NFS over RDMA)
performance to be enhanced as well.
The following results were taken against a single NVMe-oF (RoCE link layer)
subsystem with a single namespace backed by null_blk using fio benchmark
(with rw=randread, numjobs=48, iodepth={16,64}, ioengine=libaio direct=1):
ConnectX-5 (pci Width x16)
---------------------------
Block Size s/g reg_wr inline reg_wr
++++++++++ +++++++++++++++ ++++++++++++++++
512B 1302.8K/34.82% 4951.9K/99.02%
1KB 1284.3K/33.86% 4232.7K/98.09%
2KB 1238.6K/34.1% 2797.5K/80.04%
4KB 1169.3K/32.46% 1941.3K/61.35%
8KB 1013.4K/30.08% 1236.6K/39.47%
16KB 695.7K/20.19% 696.9K/20.59%
32KB 350.3K/9.64% 350.6K/10.3%
64KB 175.86K/5.27% 175.9K/5.28%
ConnectX-4 (pci Width x8)
---------------------------
Block Size s/g reg_wr inline reg_wr
++++++++++ +++++++++++++++ ++++++++++++++++
512B 1285.8K/42.66% 4242.7K/98.18%
1KB 1254.1K/41.74% 3569.2K/96.00%
2KB 1185.9K/39.83% 2173.9K/75.58%
4KB 1069.4K/36.46% 1343.3K/47.47%
8KB 755.1K/27.77% 748.7K/29.14%
Tested-by: Nitzan Carmi <nitzanc@mellanox.com>
Signed-off-by: Idan Burstein <idanb@mellanox.com>
Signed-off-by: Max Gurtovoy <maxg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
The zgid is already provided by IB/core, so there is no need in locally
defined variable, let's drop it and reuse common one.
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Provide a cxgb4-specific function to fill in qp state details.
This allows dumping important c4iw_qp state useful for debugging.
Included in the dump are the t4_sq, t4_rq structs, plus a dump
of the t4_swsqe and t4swrqe descriptors for the first and last
pending entries.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
When the kernel was compiled using the UBSAN option,
we saw the following stack trace:
[ 1184.827917] UBSAN: Undefined behaviour in drivers/infiniband/hw/mlx4/mr.c:349:27
[ 1184.828114] signed integer overflow:
[ 1184.828247] -2147483648 - 1 cannot be represented in type 'int'
The problem was caused by calling round_up in procedure
mlx4_ib_umem_calc_optimal_mtt_size (on line 349, as noted in the stack
trace) with the second parameter (1 << block_shift) (which is an int).
The second parameter should have been (1ULL << block_shift) (which
is an unsigned long long).
(1 << block_shift) is treated by the compiler as an int (because 1 is
an integer).
Now, local variable block_shift is initialized to 31.
If block_shift is 31, 1 << block_shift is 1 << 31 = 0x80000000=-214748368.
This is the most negative int value.
Inside the round_up macro, there is a cast applied to ((1 << 31) - 1).
However, this cast is applied AFTER ((1 << 31) - 1) is calculated.
Since (1 << 31) is treated as an int, we get the negative overflow
identified by UBSAN in the process of calculating ((1 << 31) - 1).
The fix is to change (1 << block_shift) to (1ULL << block_shift) on
line 349.
Fixes: 9901abf583 ("IB/mlx4: Use optimal numbers of MTT entries")
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
When IRQ affinity is set and the interrupt type is unknown, a cpu
mask allocated within the function is never freed. Fix this memory
leak by allocating memory within the scope where it is used.
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
When allocating device data, if there's an allocation failure, the
already allocated memory won't be freed such as per-cpu counters.
Fix memory leaks in exception path by creating a common reentrant
clean up function hfi1_clean_devdata() to be used at driver unload
time and device data allocation failure.
To accomplish this, free_platform_config() and clean_up_i2c() are
changed to be reentrant to remove dependencies when they are called
in different order. This helps avoid NULL pointer dereferences
introduced by this patch if those two functions weren't reentrant.
In addition, set dd->int_counter, dd->rcv_limit,
dd->send_schedule and dd->tx_opstats to NULL after they're freed in
hfi1_clean_devdata(), so that hfi1_clean_devdata() is fully reentrant.
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
When an invalid num_vls is used as a module parameter, the code
execution follows an exception path where the macro dd_dev_err()
expects dd->pcidev->dev not to be NULL in hfi1_init_dd(). This
causes a NULL pointer dereference.
Fix hfi1_init_dd() by initializing dd->pcidev and dd->pcidev->dev
earlier in the code. If a dd exists, then dd->pcidev and
dd->pcidev->dev always exists.
BUG: unable to handle kernel NULL pointer dereference
at 00000000000000f0
IP: __dev_printk+0x15/0x90
Workqueue: events work_for_cpu_fn
RIP: 0010:__dev_printk+0x15/0x90
Call Trace:
dev_err+0x6c/0x90
? hfi1_init_pportdata+0x38d/0x3f0 [hfi1]
hfi1_init_dd+0xdd/0x2530 [hfi1]
? pci_conf1_read+0xb2/0xf0
? pci_read_config_word.part.9+0x64/0x80
? pci_conf1_write+0xb0/0xf0
? pcie_capability_clear_and_set_word+0x57/0x80
init_one+0x141/0x490 [hfi1]
local_pci_probe+0x3f/0xa0
work_for_cpu_fn+0x10/0x20
process_one_work+0x152/0x350
worker_thread+0x1cf/0x3e0
kthread+0xf5/0x130
? max_active_store+0x80/0x80
? kthread_bind+0x10/0x10
? do_syscall_64+0x6e/0x1a0
? SyS_exit_group+0x10/0x10
ret_from_fork+0x35/0x40
Cc: <stable@vger.kernel.org> # 4.9.x
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
AHG may be armed to use the stored header, which by design is limited
to edits in the PSN/A 32 bit word (bth2).
When the code is trying to send a BECN, the use of the stored header
will lose the BECN bit.
Fix by avoiding AHG when getting ready to send a BECN. This is
accomplished by always claiming the packet is not a middle packet which
is an AHG precursor. BECNs are not a normal case and this should not
hurt AHG optimizations.
Cc: <stable@vger.kernel.org> # 4.14.x
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
The module parameter num_user_context is defined as 'int' and
defaults to -1. The module_param_named() says that it is uint.
Correct module_param_named() type information and update the modinfo
text to reflect the default value.
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
The code for handling a marked UD packet unconditionally returns the
dlid in the header of the FECN marked packet. This is not correct
for multicast packets where the DLID is in the multicast range.
The subsequent attempt to send the CNP with the multicast lid will
cause the chip to halt the ack send context because the source
lid doesn't match the chip programming. The send context will
be halted and flush any other pending packets in the pio ring causing
the CNP to not be sent.
A part of investigating the fix, it was determined that the 16B work
broke the FECN routine badly with inconsistent use of 16 bit and 32 bits
types for lids and pkeys. Since the port's source lid was correctly 32
bits the type mixmatches need to be dealt with at the same time as
fixing the CNP header issue.
Fix these issues by:
- Using the ports lid for as the SLID for responding to FECN marked UD
packets
- Insure pkey is always 16 bit in this and subordinate routines
- Insure lids are 32 bits in this and subordinate routines
Cc: <stable@vger.kernel.org> # 4.14.x
Fixes: 88733e3b84 ("IB/hfi1: Add 16B UD support")
Reviewed-by: Don Hiatt <don.hiatt@intel.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
