Hi,
Thomas mentioned:
"
That was the constraint of managed interrupts from the very beginning:
The driver/subsystem has to quiesce the interrupt line and the associated
queue _before_ it gets shutdown in CPU unplug and not fiddle with it
until it's restarted by the core when the CPU is plugged in again.
"
But no drivers or blk-mq do that before one hctx becomes inactive(all
CPUs for one hctx are offline), and even it is worse, blk-mq stills tries
to run hw queue after hctx is dead, see blk_mq_hctx_notify_dead().
This patchset tries to address the issue by the following approach:
- before the last cpu in hctx->cpumask is going to offline, mark this
hctx as inactive
- disable preempt during allocating tag for request, and after tag
is allocated, check if this hctx is inactive. If yes, give up this
allocation and try remote allocation from online CPUs
- before hctx becomes inactive, drain all allocated requests on this
hctx
Thanks John Garry for running lots of tests on arm64 with this previous
version patches and co-working on investigating all kinds of issues.
Thanks Christoph for review on old versions of this patches.
https://github.com/ming1/linux/commits/v5.7-rc-blk-mq-improve-cpu-hotplug
Ming Lei (6):
blk-mq: allocate request on cpu in hctx->cpumask for
blk_mq_alloc_request_hctx
blk-mq: set data->ctx and data->hctx explicitly in blk_mq_get_request
blk-mq: add blk_mq_all_tag_iter
blk-mq: prepare for draining IO when hctx's all CPUs are offline
blk-mq: disable preempt during allocating request tag
blk-mq: stop to allocate new requst and drain request before hctx
becomes inactive
block/blk-mq-debugfs.c | 2 +
block/blk-mq-tag.c | 47 +++++++++-
block/blk-mq-tag.h | 2 +
block/blk-mq.c | 184 ++++++++++++++++++++++++++++++++-----
block/blk-mq.h | 1 +
drivers/block/loop.c | 2 +-
drivers/md/dm-rq.c | 2 +-
include/linux/blk-mq.h | 10 ++
include/linux/cpuhotplug.h | 1 +
9 files changed, 221 insertions(+), 30 deletions(-)
Cc: John Garry <john.garry@xxxxxxxxxx>
Cc: Bart Van Assche <bvanassche@xxxxxxx>
Cc: Hannes Reinecke <hare@xxxxxxxx>
Cc: Christoph Hellwig <hch@xxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
--
2.25.2
