Re: [PATCH 2/5] blk-mq: rename hctx_lock & hctx_unlock

To: Ming Lei <ming.lei@xxxxxxxxxx>
Subject: Re: [PATCH 2/5] blk-mq: rename hctx_lock & hctx_unlock
From: Sagi Grimberg <sagi@xxxxxxxxxxx>
Date: Tue, 23 Nov 2021 10:54:30 +0200
Cc: Christoph Hellwig <hch@xxxxxx>, Jens Axboe <axboe@xxxxxxxxx>, "Martin K . Petersen" <martin.petersen@xxxxxxxxxx>, linux-block@xxxxxxxxxxxxxxx, linux-nvme@xxxxxxxxxxxxxxxxxxx, linux-scsi@xxxxxxxxxxxxxxx, Keith Busch <kbusch@xxxxxxxxxx>
In-reply-to: <YZww/1iBDbou1yQY@T590>
References: <20211119021849.2259254-1-ming.lei@redhat.com> <20211119021849.2259254-3-ming.lei@redhat.com> <ed13ee7f-a017-874a-cd28-e40b3aa6b4a7@grimberg.me> <YZuZCsCIyQrc+539@T590> <737e0543-9b7b-4872-082c-9ea51069d57f@grimberg.me> <YZww/1iBDbou1yQY@T590>
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.14.0

Then look at q->has_srcu that Bart suggested?


Bart just suggested to rename q->alloc_srcu as q->has_srcu.


Yea, is there a problem using that instead of having callers
pass a flag?

References:
- [PATCH 0/5] blk-mq: quiesce improvement
  - From: Ming Lei
- [PATCH 2/5] blk-mq: rename hctx_lock & hctx_unlock
  - From: Ming Lei
- Re: [PATCH 2/5] blk-mq: rename hctx_lock & hctx_unlock
  - From: Sagi Grimberg
- Re: [PATCH 2/5] blk-mq: rename hctx_lock & hctx_unlock
  - From: Ming Lei
- Re: [PATCH 2/5] blk-mq: rename hctx_lock & hctx_unlock
  - From: Sagi Grimberg
- Re: [PATCH 2/5] blk-mq: rename hctx_lock & hctx_unlock
  - From: Ming Lei