public inbox for linux-s390@vger.kernel.org
 help / color / mirror / Atom feed
From: Matthew Rosato <mjrosato@linux.ibm.com>
To: Eric Farman <farman@linux.ibm.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>,
	Alex Williamson <alex.williamson@redhat.com>,
	Cornelia Huck <cohuck@redhat.com>,
	Halil Pasic <pasic@linux.ibm.com>,
	kvm@vger.kernel.org, linux-s390@vger.kernel.org
Subject: Re: [PATCH v2 07/10] vfio/ccw: Create an OPEN FSM Event
Date: Thu, 16 Jun 2022 12:33:50 -0400	[thread overview]
Message-ID: <0816ab3a-8601-0462-6c2b-4ba7fa8a1e2b@linux.ibm.com> (raw)
In-Reply-To: <20220615203318.3830778-8-farman@linux.ibm.com>

On 6/15/22 4:33 PM, Eric Farman wrote:
> Move the process of enabling a subchannel for use by vfio-ccw
> into the FSM, such that it can manage the sequence of lifecycle
> events for the device.
> 
> That is, if the FSM state is NOT_OPER(erational), then do the work
> that would enable the subchannel and move the FSM to STANDBY state.
> An attempt to perform this event again from any of the other operating
> states (IDLE, CP_PROCESSING, CP_PENDING) will convert the device back
> to NOT_OPER so the configuration process can be started again.

Except STANDBY, which ignores the event via fsm_nop.  I wonder though, 
whether that's the right thing to do.  For each of the other states 
you're saying 'if it's already open, go back to NOT_OPER so we can start 
over' -- In this case a STANDBY->STANDBY is also a case of 'it's already 
open' so shouldn't we also go back to NOT_OPER so we can start over? 
Seems to me really we just don't expect to ever get an OPEN event unless 
we are in NOT_OPER.

If there's a reason to keep STANDBY->STANDBY as a nop, but we don't 
expect to see it and don't' want to WARN because of it, then maybe a log 
entry at least would make sense.

As for the IDLE/CP_PROCESSING/CP_PENDING cases, going fsm_notoper 
because this is unexpected probably makes sense, but the logging is 
going to be really confusing (before this change, you know that you 
called fsm_notoper because you got VFIO_CCW_EVENT_NOT_OPER -- now you'll 
see a log entry cut for NOT_OPER but won't be sure if it was for 
EVENT_NOT_OPER or EVENT_OPEN).  Maybe you can look at 'event' inside 
fsm_notoper and cut a slightly different trace entry when arriving here 
for EVENT_OPEN?

...

> +static void fsm_open(struct vfio_ccw_private *private,
> +		     enum vfio_ccw_event event)
> +{
> +	struct subchannel *sch = private->sch;
> +	int ret;
> +
> +	spin_lock_irq(sch->lock);
> +	sch->isc = VFIO_CCW_ISC;
> +	ret = cio_enable_subchannel(sch, (u32)(unsigned long)sch);
> +	if (!ret)
> +		private->state = VFIO_CCW_STATE_STANDBY;

nit: could get rid of 'ret' and just do

if (!cio_enable...)
      private->state = VFIO_CCW_STATE_STANDBY;

> +	spin_unlock_irq(sch->lock);
> +}
> +
>   /*
>    * Device statemachine
>    */
> @@ -373,29 +389,34 @@ fsm_func_t *vfio_ccw_jumptable[NR_VFIO_CCW_STATES][NR_VFIO_CCW_EVENTS] = {
>   		[VFIO_CCW_EVENT_IO_REQ]		= fsm_io_error,
>   		[VFIO_CCW_EVENT_ASYNC_REQ]	= fsm_async_error,
>   		[VFIO_CCW_EVENT_INTERRUPT]	= fsm_disabled_irq,
> +		[VFIO_CCW_EVENT_OPEN]		= fsm_open,
>   	},
>   	[VFIO_CCW_STATE_STANDBY] = {
>   		[VFIO_CCW_EVENT_NOT_OPER]	= fsm_notoper,
>   		[VFIO_CCW_EVENT_IO_REQ]		= fsm_io_error,
>   		[VFIO_CCW_EVENT_ASYNC_REQ]	= fsm_async_error,
>   		[VFIO_CCW_EVENT_INTERRUPT]	= fsm_irq,
> +		[VFIO_CCW_EVENT_OPEN]		= fsm_nop,
>   	},
>   	[VFIO_CCW_STATE_IDLE] = {
>   		[VFIO_CCW_EVENT_NOT_OPER]	= fsm_notoper,
>   		[VFIO_CCW_EVENT_IO_REQ]		= fsm_io_request,
>   		[VFIO_CCW_EVENT_ASYNC_REQ]	= fsm_async_request,
>   		[VFIO_CCW_EVENT_INTERRUPT]	= fsm_irq,
> +		[VFIO_CCW_EVENT_OPEN]		= fsm_notoper,
>   	},
>   	[VFIO_CCW_STATE_CP_PROCESSING] = {
>   		[VFIO_CCW_EVENT_NOT_OPER]	= fsm_notoper,
>   		[VFIO_CCW_EVENT_IO_REQ]		= fsm_io_retry,
>   		[VFIO_CCW_EVENT_ASYNC_REQ]	= fsm_async_retry,
>   		[VFIO_CCW_EVENT_INTERRUPT]	= fsm_irq,
> +		[VFIO_CCW_EVENT_OPEN]		= fsm_notoper,
>   	},
>   	[VFIO_CCW_STATE_CP_PENDING] = {
>   		[VFIO_CCW_EVENT_NOT_OPER]	= fsm_notoper,
>   		[VFIO_CCW_EVENT_IO_REQ]		= fsm_io_busy,
>   		[VFIO_CCW_EVENT_ASYNC_REQ]	= fsm_async_request,
>   		[VFIO_CCW_EVENT_INTERRUPT]	= fsm_irq,
> +		[VFIO_CCW_EVENT_OPEN]		= fsm_notoper,
>   	},
>   };
> diff --git a/drivers/s390/cio/vfio_ccw_private.h b/drivers/s390/cio/vfio_ccw_private.h
> index 4cfdd5fc0961..8dff1699a7d9 100644
> --- a/drivers/s390/cio/vfio_ccw_private.h
> +++ b/drivers/s390/cio/vfio_ccw_private.h
> @@ -142,6 +142,7 @@ enum vfio_ccw_event {
>   	VFIO_CCW_EVENT_IO_REQ,
>   	VFIO_CCW_EVENT_INTERRUPT,
>   	VFIO_CCW_EVENT_ASYNC_REQ,
> +	VFIO_CCW_EVENT_OPEN,
>   	/* last element! */
>   	NR_VFIO_CCW_EVENTS
>   };


  reply	other threads:[~2022-06-16 16:34 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-15 20:33 [PATCH v2 00/10] s390/vfio-ccw rework Eric Farman
2022-06-15 20:33 ` [PATCH v2 01/10] vfio/ccw: Remove UUID from s390 debug log Eric Farman
2022-06-15 20:57   ` Kirti Wankhede
2022-06-15 20:33 ` [PATCH v2 02/10] vfio/ccw: Fix FSM state if mdev probe fails Eric Farman
2022-06-15 20:33 ` [PATCH v2 03/10] vfio/ccw: Do not change FSM state in subchannel event Eric Farman
2022-06-16 15:35   ` Matthew Rosato
2022-06-16 17:13     ` Eric Farman
2022-06-15 20:33 ` [PATCH v2 04/10] vfio/ccw: Remove private->mdev Eric Farman
2022-06-16 15:36   ` Matthew Rosato
2022-06-23 14:59   ` Jason Gunthorpe
2022-06-15 20:33 ` [PATCH v2 05/10] vfio/ccw: Pass enum to FSM event jumptable Eric Farman
2022-06-15 20:33 ` [PATCH v2 06/10] vfio/ccw: Flatten MDEV device (un)register Eric Farman
2022-06-15 20:33 ` [PATCH v2 07/10] vfio/ccw: Create an OPEN FSM Event Eric Farman
2022-06-16 16:33   ` Matthew Rosato [this message]
2022-06-16 17:14     ` Eric Farman
2022-06-16 17:18       ` Matthew Rosato
2022-06-15 20:33 ` [PATCH v2 08/10] vfio/ccw: Create a CLOSE FSM event Eric Farman
2022-06-16 16:48   ` Matthew Rosato
2022-06-15 20:33 ` [PATCH v2 09/10] vfio/ccw: Refactor vfio_ccw_mdev_reset Eric Farman
2022-06-15 20:33 ` [PATCH v2 10/10] vfio/ccw: Move FSM open/close to MDEV open/close Eric Farman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=0816ab3a-8601-0462-6c2b-4ba7fa8a1e2b@linux.ibm.com \
    --to=mjrosato@linux.ibm.com \
    --cc=alex.williamson@redhat.com \
    --cc=cohuck@redhat.com \
    --cc=farman@linux.ibm.com \
    --cc=jgg@nvidia.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=pasic@linux.ibm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox