If there is a failure during probe of hfi1 before the sdma_map_lock is
initialized, the call to hfi1_free_devdata() will attempt to use a lock
that has not been initialized. If the locking correctness validator is on
then an INFO message and stack trace resembling the following may be seen:
INFO: trying to register non-static key.
The code is fine but needs lockdep annotation, or maybe
you didn't initialize this object before use?
turning off the locking correctness validator.
Call Trace:
register_lock_class+0x11b/0x880
__lock_acquire+0xf3/0x7930
lock_acquire+0xff/0x2d0
_raw_spin_lock_irq+0x46/0x60
sdma_clean+0x42a/0x660 [hfi1]
hfi1_free_devdata+0x3a7/0x420 [hfi1]
init_one+0x867/0x11a0 [hfi1]
pci_device_probe+0x40e/0x8d0
The use of sdma_map_lock in sdma_clean() is for freeing the sdma_map
memory, and sdma_map is not allocated/initialized until after
sdma_map_lock has been initialized. This code only needs to be run if
sdma_map is not NULL, and so checking for that condition will avoid trying
to use the lock before it is initialized.
Fixes: 473291b3ea ("IB/hfi1: Fix for early release of sdma context")
Fixes: 7724105686 ("IB/hfi1: add driver files")
Link: https://lore.kernel.org/r/20220520183701.48973.72434.stgit@awfm-01.cornelisnetworks.com
Reported-by: Zheyu Ma <zheyuma97@gmail.com>
Signed-off-by: Douglas Miller <doug.miller@cornelisnetworks.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
HW flushes after QP in error state is not reliable. This can lead to
application hang waiting on a completion for outstanding WRs. Implement a
SW mechanism to generate completions for any outstanding WR's after the QP
is modified to error.
This is accomplished by starting a delayed worker after the QP is modified
to error and the HW flush is performed. The worker will generate
completions that will be returned to the application when it polls the
CQ. This mechanism only applies to Kernel applications.
Link: https://lore.kernel.org/r/20220425181624.1617-1-shiraz.saleem@intel.com
Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Downstream patches will add support for hardware lag with
more than 2 ports. Add a way for users to query the number of lag ports.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Maor Gottlieb <maorg@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
In the tasklets (completer, responder, and requester) check the return
value from rxe_get() to detect failures to get a reference. This only
occurs if the qp has had its reference count drop to zero which indicates
that it no longer should be used.
The ref is never 0 today because the tasklets are flushed before the ref
is dropped. The next patch changes this so that the ref is dropped then
the tasklets are flushed.
Link: https://lore.kernel.org/r/20220421014042.26985-4-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Currently the #define IB_SRQ_INIT_MASK is used to distinguish the
rxe_create_srq verb from the rxe_modify_srq verb so that some code can be
shared between these two subroutines.
This commit splits rxe_srq_chk_attr into two subroutines: rxe_srq_chk_init
and rxe_srq_chk_attr which handle the create_srq and modify_srq verbs
separately.
Link: https://lore.kernel.org/r/20220421014042.26985-2-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The bt number of cqc_timer of HIP09 increases compared with that of HIP08.
Therefore, cqc_timer_bt_num and num_cqc_timer do not match. As a result,
the driver may fail to allocate cqc_timer. So the driver needs to uniquely
uses cqc_timer_bt_num to represent the bt number of cqc_timer.
Fixes: 0e40dc2f70 ("RDMA/hns: Add timer allocation support for hip08")
Link: https://lore.kernel.org/r/20220429093545.58070-1-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>
CMDQ may fail during HNS ROCEE initialization. The following is the log
when the execution fails:
hns3 0000:bd:00.2: In reset process RoCE client reinit.
hns3 0000:bd:00.2: CMDQ move tail from 840 to 839
hns3 0000:bd:00.2 hns_2: failed to set gid, ret = -11!
hns3 0000:bd:00.2: CMDQ move tail from 840 to 839
<...>
hns3 0000:bd:00.2: CMDQ move tail from 840 to 839
hns3 0000:bd:00.2: CMDQ move tail from 840 to 0
hns3 0000:bd:00.2: [cmd]token 14e mailbox 20 timeout.
hns3 0000:bd:00.2 hns_2: set HEM step 0 failed!
hns3 0000:bd:00.2 hns_2: set HEM address to HW failed!
hns3 0000:bd:00.2 hns_2: failed to alloc mtpt, ret = -16.
infiniband hns_2: Couldn't create ib_mad PD
infiniband hns_2: Couldn't open port 1
hns3 0000:bd:00.2: Reset done, RoCE client reinit finished.
However, even if ib_mad client registration failed, ib_register_device()
still returns success to the driver.
In the device initialization process, CMDQ execution fails because HW/FW
is abnormal. Therefore, if CMDQ fails, the initialization function should
set CMDQ to a fatal error state and return a failure to the caller.
Fixes: 9a4435375c ("IB/hns: Add driver files for hns RoCE driver")
Link: https://lore.kernel.org/r/20220429093104.26687-1-liangwenpeng@huawei.com
Signed-off-by: Yangyang Li <liyangyang20@huawei.com>
Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
rxe_mcast.c currently uses _irqsave spinlocks for rxe->mcg_lock while
rxe_recv.c uses _bh spinlocks for the same lock.
As there is no case where the mcg_lock can be taken from an IRQ, change
these all to bh locks so we don't have confusing mismatched lock types on
the same spinlock.
Fixes: 6090a0c4c7 ("RDMA/rxe: Cleanup rxe_mcast.c")
Link: https://lore.kernel.org/r/20220504202817.98247-1-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
These routines were not intended to be called under a spinlock and will
throw debugging warnings:
raw_local_irq_restore() called with IRQs enabled
WARNING: CPU: 13 PID: 3107 at kernel/locking/irqflag-debug.c:10 warn_bogus_irq_restore+0x2f/0x50
CPU: 13 PID: 3107 Comm: python3 Tainted: G E 5.18.0-rc1+ #7
Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006
RIP: 0010:warn_bogus_irq_restore+0x2f/0x50
Call Trace:
<TASK>
_raw_spin_unlock_irqrestore+0x75/0x80
rxe_attach_mcast+0x304/0x480 [rdma_rxe]
ib_attach_mcast+0x88/0xa0 [ib_core]
ib_uverbs_attach_mcast+0x186/0x1e0 [ib_uverbs]
ib_uverbs_handler_UVERBS_METHOD_INVOKE_WRITE+0xcd/0x140 [ib_uverbs]
ib_uverbs_cmd_verbs+0xdb0/0xea0 [ib_uverbs]
ib_uverbs_ioctl+0xd2/0x160 [ib_uverbs]
do_syscall_64+0x5c/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xae
Move them out of the spinlock, it is OK if there is some races setting up
the MC reception at the ethernet layer with rbtree lookups.
Fixes: 6090a0c4c7 ("RDMA/rxe: Cleanup rxe_mcast.c")
Link: https://lore.kernel.org/r/20220504202817.98247-1-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The calling of siw_cm_upcall and detaching new_cep with its listen_cep
should be atomistic semantics. Otherwise siw_reject may be called in a
temporary state, e,g, siw_cm_upcall is called but the new_cep->listen_cep
has not being cleared.
This fixes a WARN:
WARNING: CPU: 7 PID: 201 at drivers/infiniband/sw/siw/siw_cm.c:255 siw_cep_put+0x125/0x130 [siw]
CPU: 2 PID: 201 Comm: kworker/u16:22 Kdump: loaded Tainted: G E 5.17.0-rc7 #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.15.0-1 04/01/2014
Workqueue: iw_cm_wq cm_work_handler [iw_cm]
RIP: 0010:siw_cep_put+0x125/0x130 [siw]
Call Trace:
<TASK>
siw_reject+0xac/0x180 [siw]
iw_cm_reject+0x68/0xc0 [iw_cm]
cm_work_handler+0x59d/0xe20 [iw_cm]
process_one_work+0x1e2/0x3b0
worker_thread+0x50/0x3a0
? rescuer_thread+0x390/0x390
kthread+0xe5/0x110
? kthread_complete_and_exit+0x20/0x20
ret_from_fork+0x1f/0x30
</TASK>
Fixes: 6c52fdc244 ("rdma/siw: connection management")
Link: https://lore.kernel.org/r/d528d83466c44687f3872eadcb8c184528b2e2d4.1650526554.git.chengyou@linux.alibaba.com
Reported-by: Luis Chamberlain <mcgrof@kernel.org>
Reviewed-by: Bernard Metzler <bmt@zurich.ibm.com>
Signed-off-by: Cheng Xu <chengyou@linux.alibaba.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
In finish_packet() in rxe_req.c a variable was incorrectly called paylen
instead of payload. Elsewhere in the rxe source payload is always used for
the RoCE payload length and paylen is always used for the UDP payload
length. This will cause unnecessary confusion.
Replace paylen by payload in finish_packet().
Link: https://lore.kernel.org/r/20220420172316.5465-1-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
QP destroy is synchronous and waits for its refcnt to be decremented in
irdma_cm_node_free_cb (for iWARP) which fires after the RCU grace period
elapses.
Applications running a large number of connections are exposed to high
wait times on destroy QP for events like SIGABORT.
The long pole for this wait time is the firing of the call_rcu callback
during a CM node destroy which can be slow. It holds the QP reference
count and blocks the destroy QP from completing.
call_rcu only needs to make sure that list walkers have a reference to the
cm_node object before freeing it and thus need to wait for grace period
elapse. The rest of the connection teardown in irdma_cm_node_free_cb is
moved out of the grace period wait in irdma_destroy_connection. Also,
replace call_rcu with a simple kfree_rcu as it just needs to do a kfree on
the cm_node
Fixes: 146b9756f1 ("RDMA/irdma: Add connection manager")
Link: https://lore.kernel.org/r/20220425181703.1634-3-shiraz.saleem@intel.com
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
When connection establishment fails in iWARP mode, an app can drain the
QPs and hang because flush isn't issued when the QP is modified from RTR
state to error. Issue a flush in this case using function
irdma_cm_disconn().
Update irdma_cm_disconn() to do flush when cm_id is NULL, which is the
case when the QP is in RTR state and there is an error in the connection
establishment.
Fixes: b48c24c2d7 ("RDMA/irdma: Implement device supported verb APIs")
Link: https://lore.kernel.org/r/20220425181703.1634-2-shiraz.saleem@intel.com
Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Introduce mlx5_umr_post_send_wait() that uses a UMR adjusted flow for
posting WQEs. The next patches will gradually move UMR operations to use
this flow. Once done, will get rid of mlx5_ib_post_send_wait().
mlx5_umr_post_send_wait gets already written WQE segments and will only
memcpy it to the SQ. This way, we avoid packing all the data in a WR just
to unpack it into the WQE.
Link: https://lore.kernel.org/r/f027dd592fde62402b2d49efded8d1d22229d22b.1649747695.git.leonro@nvidia.com
Signed-off-by: Aharon Landau <aharonl@nvidia.com>
Reviewed-by: Michael Guralnik <michaelgur@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
