per-tagset SRCU struct and quiesce v2

[Christoph Hellwig]

Hi all,

this series moves the SRCU struct used for quiescing to the tag_set
as the SRCU critical sections for dispatch take about the same time
on all queues anyway, and then adopts the series from Chao that provides
tagset-wide quiesce to use that infrastructure.

To do so it does a fair amount of cleanups in how nvme quiesces queues.
Even with that nvme is still a bit of a mess in how it quiesces queues,
so further work is probably needed later.

Note: testing is a bit limited as I'm travelling.  Before this proceeds
it should get surprise removal testing for nvme-pci.

Changes since v1:
 - a whole bunch of new nvme cleanups
 - use the notify version of set_capacity in blk_mark_disk_dead
 - improve the documentation for blk_mq_wait_quiesce_done
 - dynamically allocate the srcu_struct in struct blk_mq_tag_set
 - minor spelling fixes