linux-stable.git/drivers/infiniband, branch v5.2.3

IB/mlx5: Report correctly tag matching rendezvous capability

2019-07-26T07:11:05+00:00

commit 89705e92700170888236555fe91b45e4c1bb0985 upstream.

Userspace expects the IB_TM_CAP_RC bit to indicate that the device
supports RC transport tag matching with rendezvous offload. However the
firmware splits this into two capabilities for eager and rendezvous tag
matching.

Only if the FW supports both modes should userspace be told the tag
matching capability is available.

Cc:  # 4.13
Fixes: eb761894351d ("IB/mlx5: Fill XRQ capabilities")
Signed-off-by: Danit Goldberg 
Reviewed-by: Yishai Hadas 
Reviewed-by: Artemy Kovalyov 
Signed-off-by: Leon Romanovsky 
Signed-off-by: Jason Gunthorpe 
Signed-off-by: Greg Kroah-Hartman

RDMA/odp: Fix missed unlock in non-blocking invalidate_start

2019-07-26T07:11:04+00:00

commit 7608bf40cf2480057ec0da31456cc428791c32ef upstream.

If invalidate_start returns with EAGAIN then the umem_rwsem needs to be
unlocked as no invalidate_end will be called.

Cc: 
Fixes: ca748c39ea3f ("RDMA/umem: Get rid of per_mm->notifier_count")
Signed-off-by: Jason Gunthorpe 
Reviewed-by: Leon Romanovsky 
Signed-off-by: Doug Ledford 
Signed-off-by: Greg Kroah-Hartman

RDMA/srp: Accept again source addresses that do not have a port number

2019-07-26T07:11:04+00:00

commit bcef5b7215681250c4bf8961dfe15e9e4fef97d0 upstream.

The function srp_parse_in() is used both for parsing source address
specifications and for target address specifications. Target addresses
must have a port number. Having to specify a port number for source
addresses is inconvenient. Make sure that srp_parse_in() supports again
parsing addresses with no port number.

Cc: 
Fixes: c62adb7def71 ("IB/srp: Fix IPv6 address parsing")
Signed-off-by: Bart Van Assche 
Signed-off-by: Jason Gunthorpe 
Signed-off-by: Greg Kroah-Hartman

ipoib: correcly show a VF hardware address

2019-07-26T07:10:29+00:00

[ Upstream commit 64d701c608fea362881e823b666327f5d28d7ffd ]

in the case of IPoIB with SRIOV enabled hardware
ip link show command incorrecly prints
0 instead of a VF hardware address.

Before:
11: ib1:  mtu 2044 qdisc pfifo_fast
state UP mode DEFAULT group default qlen 256
    link/infiniband
80:00:00:66:fe:80:00:00:00:00:00:00:24:8a:07:03:00:a4:3e:7c brd
00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
    vf 0 MAC 00:00:00:00:00:00, spoof checking off, link-state disable,
trust off, query_rss off
...
After:
11: ib1:  mtu 2044 qdisc pfifo_fast
state UP mode DEFAULT group default qlen 256
    link/infiniband
80:00:00:66:fe:80:00:00:00:00:00:00:24:8a:07:03:00:a4:3e:7c brd
00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
    vf 0     link/infiniband
80:00:00:66:fe:80:00:00:00:00:00:00:24:8a:07:03:00:a4:3e:7c brd
00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff, spoof
checking off, link-state disable, trust off, query_rss off

v1->v2: just copy an address without modifing ifla_vf_mac
v2->v3: update the changelog

Signed-off-by: Denis Kirjanov 
Acked-by: Doug Ledford 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

RDMA/efa: Handle mmap insertions overflow

2019-06-18T20:27:24+00:00

When inserting a new mmap entry to the xarray we should check for
'mmap_page' overflow as it is limited to 32 bits.

Fixes: 40909f664d27 ("RDMA/efa: Add EFA verbs implementation")
Signed-off-by: Gal Pressman 
Signed-off-by: Doug Ledford

RDMA/efa: Fix success return value in case of error

2019-06-18T01:35:21+00:00

Existing code would mistakenly return success in case of error instead
of a proper return value.

Fixes: e9c6c5373088 ("RDMA/efa: Add common command handlers")
Reviewed-by: Firas JahJah 
Reviewed-by: Yossi Leybovich 
Signed-off-by: Gal Pressman 
Signed-off-by: Doug Ledford

IB/hfi1: Handle port down properly in pio

2019-06-18T01:15:40+00:00

The call to sc_buffer_alloc currently returns NULL (no buffer) or
a buffer descriptor.

There is a third case when the port is down.  Currently that
returns NULL and this prevents the caller from properly handling the
sc_buffer_alloc() failure.  A verbs code link test after the call is
racy so the indication needs to come from the state check inside the allocation
routine to be valid.

Fix by encoding the ECOMM failure like SDMA.   IS_ERR_OR_NULL() tests
are added at all call sites.  For verbs send, this needs to treat any
error by returning a completion without any MMIO copy.

Fixes: 7724105686e7 ("IB/hfi1: add driver files")
Reviewed-by: Dennis Dalessandro 
Signed-off-by: Mike Marciniszyn 
Signed-off-by: Dennis Dalessandro 
Signed-off-by: Doug Ledford

IB/hfi1: Handle wakeup of orphaned QPs for pio

2019-06-18T01:15:40+00:00

Once a send context is taken down due to a link failure, any QPs waiting
for pio credits will stay on the waitlist indefinitely.

Fix by wakeing up all QPs linked to piowait list.

Fixes: 7724105686e7 ("IB/hfi1: add driver files")
Reviewed-by: Dennis Dalessandro 
Signed-off-by: Mike Marciniszyn 
Signed-off-by: Dennis Dalessandro 
Signed-off-by: Doug Ledford

IB/hfi1: Wakeup QPs orphaned on wait list after flush

2019-06-18T01:15:40+00:00

Once an SDMA engine is taken down due to a link failure, any waiting QPs
that do not have outstanding descriptors in the ring will stay
on the dmawait list as long as the port is down.

Since there is no timer running, they will stay there for a long time.

The fix is to wake up all iowaits linked to dmawait. The send engine
will build and post packets that get flushed back.

Fixes: 7724105686e7 ("IB/hfi1: add driver files")
Reviewed-by: Kaike Wan 
Signed-off-by: Mike Marciniszyn 
Signed-off-by: Dennis Dalessandro 
Signed-off-by: Doug Ledford

IB/hfi1: Use aborts to trigger RC throttling

2019-06-18T01:15:40+00:00

SDMA and pio flushes will cause a lot of packets to be transmitted
after a link has gone down, using a lot of CPU to retransmit
packets.

Fix for RC QPs by recognizing the flush status and:
- Forcing a timer start
- Putting the QP into a "send one" mode

Fixes: 7724105686e7 ("IB/hfi1: add driver files")
Reviewed-by: Kaike Wan 
Signed-off-by: Mike Marciniszyn 
Signed-off-by: Dennis Dalessandro 
Signed-off-by: Doug Ledford