Re: [PATCH RFC v2 02/24] scsi: allocate separate queue for reserved commands

[John Garry]

On 11/03/2020 06:58, Hannes Reinecke wrote:
> On 3/11/20 7:22 AM, Christoph Hellwig wrote:
> > On Tue, Mar 10, 2020 at 09:08:56PM +0000, John Garry wrote:
> > > On 10/03/2020 18:32, Christoph Hellwig wrote:
> > > > On Wed, Mar 11, 2020 at 12:25:28AM +0800, John Garry wrote:
> > > > > From: Hannes Reinecke <hare@xxxxxxxx>
> > > > >
> > > > > Allocate a separate 'reserved_cmd_q' for sending reserved commands.
> > > >
> > > > Why?  Reserved command specifically are not in any way tied to queues.
> > > > .
> > >
> > > So the v1 series used a combination of the sdev queue and the per-host
> > > reserved_cmd_q. Back then you questioned using the sdev queue for virtio
> > > scsi, and the unconfirmed conclusion was to use a common per-host q. This is
> > > the best link I can find now:
> > >
> > > https://www.mail-archive.com/linux-scsi@xxxxxxxxxxxxxxx/msg83177.html
> >
> > That was just a question on why virtio uses the per-device tags, which
> > didn't look like it made any sense.  What I'm worried about here is
> > mixing up the concept of reserved tags in the tagset, and queues to use
> > them.  Note that we already have the scsi_get_host_dev to allocate
> > a scsi_device and thus a request_queue for the host itself.  That seems
> > like the better interface to use a tag for a host wide command vs
> > introducing a parallel path.
> Ah. Right.
> Will be looking into that, and convert the patchset over to it.

Hannes,

I assume that you will take back over this series now. Let me know. The 
patches are here, if not in patchwork:
https://github.com/hisilicon/kernel-dev/tree/private-topic-sas-5.6-resv-commands-v2

> And the problem of the separate queue is the fact that I'll need a queue
> to reserve tags from; trying to allocate a tag directly from the bitmap
> turns out to be major surgery in the blocklayer with no immediate gain.
> And I can't use per-device queues as for some drivers the reserved
> commands are used to query the HBA itself to figure out how many devices
> are present.

Right, we still need to be able to allocate somewhere apart from sdev queue

>


Thanks