kernel-fxtec-pro1x/drivers/nvme/host
Anton Eidelman 6376736d01 nvme-multipath: fix possible io hang after ctrl reconnect
[ Upstream commit af8fd0424713a2adb812d10d55e86718152cf656 ]

The following scenario results in an IO hang:
1) ctrl completes a request with NVME_SC_ANA_TRANSITION.
   NVME_NS_ANA_PENDING bit in ns->flags is set and ana_work is triggered.
2) ana_work: nvme_read_ana_log() tries to get the ANA log page from the ctrl.
   This fails because ctrl disconnects.
   Therefore nvme_update_ns_ana_state() is not called
   and NVME_NS_ANA_PENDING bit in ns->flags is not cleared.
3) ctrl reconnects: nvme_mpath_init(ctrl,...) calls
   nvme_read_ana_log(ctrl, groups_only=true).
   However, nvme_update_ana_state() does not update namespaces
   because nr_nsids = 0 (due to groups_only mode).
4) scan_work calls nvme_validate_ns() finds the ns and re-validates OK.

Result:
The ctrl is now live but NVME_NS_ANA_PENDING bit in ns->flags is still set.
Consequently ctrl will never be considered a viable path by __nvme_find_path().
IO will hang if ctrl is the only or the last path to the namespace.

More generally, while ctrl is reconnecting, its ANA state may change.
And because nvme_mpath_init() requests ANA log in groups_only mode,
these changes are not propagated to the existing ctrl namespaces.
This may result in a mal-function or an IO hang.

Solution:
nvme_mpath_init() will nvme_read_ana_log() with groups_only set to false.
This will not harm the new ctrl case (no namespaces present),
and will make sure the ANA state of namespaces gets updated after reconnect.

Note: Another option would be for nvme_mpath_init() to invoke
nvme_parse_ana_log(..., nvme_set_ns_ana_state) for each existing namespace.

Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Anton Eidelman <anton@lightbitslabs.com>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
2019-11-12 19:21:11 +01:00
..
core.c nvme-pci: Fix a race in controller removal 2019-10-29 09:19:29 +01:00
fabrics.c nvme: call nvme_complete_rq when nvmf_check_ready fails for mpath I/O 2018-11-13 11:08:24 -08:00
fabrics.h nvme: if_ready checks to fail io to deleting controller 2018-07-24 13:44:40 +02:00
fault_inject.c nvme: Add fault injection feature 2018-03-26 08:53:43 -06:00
fc.c nvme-fc: use separate work queue to avoid warning 2019-09-16 08:21:44 +02:00
Kconfig IB: Revert "remove redundant INFINIBAND kconfig dependencies" 2018-05-28 10:40:16 -06:00
lightnvm.c lightnvm: remove minor version check for 2.0 2018-08-05 19:36:09 -06:00
Makefile nvme: Add fault injection feature 2018-03-26 08:53:43 -06:00
multipath.c nvme-multipath: fix possible io hang after ctrl reconnect 2019-11-12 19:21:11 +01:00
nvme.h nvme: fix a possible deadlock when passthru commands sent to a multipath device 2019-09-06 10:21:55 +02:00
pci.c nvme-pci: Fix async probe remove race 2019-09-06 10:21:55 +02:00
rdma.c nvme-rdma: fix a NULL deref when an admin connect times out 2019-05-31 06:46:15 -07:00
trace.c nvme: add disk name to trace events 2018-07-24 15:55:48 +02:00
trace.h nvme: add disk name to trace events 2018-07-24 15:55:48 +02:00