public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Eric Farman <farman@linux.ibm.com>
To: Cornelia Huck <cohuck@redhat.com>
Cc: Jared Rossi <jrossi@linux.ibm.com>,
	Halil Pasic <pasic@linux.ibm.com>,
	linux-s390@vger.kernel.org, kvm@vger.kernel.org,
	Eric Farman <farman@linux.ibm.com>
Subject: [RFC PATCH v2 3/4] vfio-ccw: Expand SCSW usage to HALT and CLEAR
Date: Wed, 13 May 2020 16:29:33 +0200	[thread overview]
Message-ID: <20200513142934.28788-4-farman@linux.ibm.com> (raw)
In-Reply-To: <20200513142934.28788-1-farman@linux.ibm.com>

Expand the Activity Control flags to include HALT/CLEAR PENDING,
in the same way as is done for the START PENDING flag.

The POPS states that for HALT SUBCHANNEL:

> Condition code 2 is set, and no other action is
> taken, ... when a halt function or clear function is
> already in progress at the subchannel.

So take that into account in fsm_do_halt().

CLEAR SUBCHANNEL is the biggest hammer, and always gets to happen,
so no corresponding check is added to fsm_do_clear(). But it does
reset both START and HALT functions that may be active, which is
why it incorporates the resetting of the HALT bit on the interrupt
path.

FIXME: What happens if a guest hammers a bunch of CSCH in a row?
We clear this for the first interrupt; is that fine?

Signed-off-by: Eric Farman <farman@linux.ibm.com>
---
 drivers/s390/cio/vfio_ccw_drv.c |  6 ++++++
 drivers/s390/cio/vfio_ccw_fsm.c | 12 +++++++++++-
 2 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/drivers/s390/cio/vfio_ccw_drv.c b/drivers/s390/cio/vfio_ccw_drv.c
index ee153fa72a0f..55051972325f 100644
--- a/drivers/s390/cio/vfio_ccw_drv.c
+++ b/drivers/s390/cio/vfio_ccw_drv.c
@@ -101,6 +101,12 @@ static void vfio_ccw_sch_io_todo(struct work_struct *work)
 	if (private->mdev && scsw_is_solicited(&irb->scsw) && is_final) {
 		private->state = VFIO_CCW_STATE_IDLE;
 		private->scsw.cmd.actl &= ~SCSW_ACTL_START_PEND;
+		if (scsw_fctl(&irb->scsw) & SCSW_FCTL_HALT_FUNC)
+			private->scsw.cmd.actl &= ~SCSW_ACTL_HALT_PEND;
+		if (scsw_fctl(&irb->scsw) & SCSW_FCTL_CLEAR_FUNC) {
+			private->scsw.cmd.actl &= ~SCSW_ACTL_HALT_PEND;
+			private->scsw.cmd.actl &= ~SCSW_ACTL_CLEAR_PEND;
+		}
 	}
 
 	if (private->io_trigger)
diff --git a/drivers/s390/cio/vfio_ccw_fsm.c b/drivers/s390/cio/vfio_ccw_fsm.c
index 258ce32549f3..d8075a56eb9b 100644
--- a/drivers/s390/cio/vfio_ccw_fsm.c
+++ b/drivers/s390/cio/vfio_ccw_fsm.c
@@ -86,6 +86,14 @@ static int fsm_do_halt(struct vfio_ccw_private *private)
 
 	sch = private->sch;
 
+	if (scsw_actl(&private->scsw) & (SCSW_ACTL_HALT_PEND | SCSW_ACTL_CLEAR_PEND)) {
+		VFIO_CCW_MSG_EVENT(2,
+				   "%pUl: actl %x pending\n",
+				   mdev_uuid(private->mdev),
+				   scsw_actl(&private->scsw));
+		return -EBUSY;
+	}
+
 	spin_lock_irqsave(sch->lock, flags);
 
 	VFIO_CCW_TRACE_EVENT(2, "haltIO");
@@ -102,6 +110,7 @@ static int fsm_do_halt(struct vfio_ccw_private *private)
 		 * Initialize device status information
 		 */
 		sch->schib.scsw.cmd.actl |= SCSW_ACTL_HALT_PEND;
+		private->scsw.cmd.actl |= SCSW_ACTL_HALT_PEND;
 		ret = 0;
 		break;
 	case 1:		/* Status pending */
@@ -143,6 +152,7 @@ static int fsm_do_clear(struct vfio_ccw_private *private)
 		 * Initialize device status information
 		 */
 		sch->schib.scsw.cmd.actl = SCSW_ACTL_CLEAR_PEND;
+		private->scsw.cmd.actl |= SCSW_ACTL_CLEAR_PEND;
 		/* TODO: check what else we might need to clear */
 		ret = 0;
 		break;
@@ -246,7 +256,7 @@ static void fsm_io_request(struct vfio_ccw_private *private,
 	char *errstr = "request";
 	struct subchannel_id schid = get_schid(private);
 
-	if (scsw_actl(scsw) & SCSW_ACTL_START_PEND) {
+	if (scsw_actl(scsw) & (SCSW_ACTL_START_PEND | SCSW_ACTL_HALT_PEND | SCSW_ACTL_CLEAR_PEND)) {
 		io_region->ret_code = -EBUSY;
 		VFIO_CCW_MSG_EVENT(2,
 				   "%pUl (%x.%x.%04x): actl %x pending\n",
-- 
2.17.1


  parent reply	other threads:[~2020-05-13 14:29 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-13 14:29 [RFC PATCH v2 0/4] vfio-ccw: Fix interrupt handling for HALT/CLEAR Eric Farman
2020-05-13 14:29 ` [RFC PATCH v2 1/4] vfio-ccw: Do not reset FSM state for unsolicited interrupts Eric Farman
2020-05-13 14:29 ` [RFC PATCH v2 2/4] vfio-ccw: Utilize scsw actl to serialize start operations Eric Farman
2020-05-13 14:29 ` Eric Farman [this message]
2020-05-13 14:29 ` [RFC PATCH v2 4/4] vfio-ccw: Clean up how to react to a failed START Eric Farman
2020-05-14 13:46 ` [RFC PATCH v2 0/4] vfio-ccw: Fix interrupt handling for HALT/CLEAR Halil Pasic
2020-05-15 13:09   ` Eric Farman
2020-05-15 14:55     ` Halil Pasic
2020-05-15 15:58       ` Cornelia Huck
2020-05-15 17:41         ` Halil Pasic
2020-05-15 18:19           ` Eric Farman
2020-05-15 18:12       ` Eric Farman
2020-05-15 18:37         ` Halil Pasic
2020-05-18 22:01           ` Eric Farman
2020-05-15 19:35         ` Halil Pasic
2020-05-18 16:09 ` Cornelia Huck
2020-05-18 21:57   ` Eric Farman
2020-05-19 11:23     ` Cornelia Huck
2020-05-18 22:09   ` Halil Pasic
2020-05-19 11:36     ` Cornelia Huck
2020-05-19 12:10       ` Halil Pasic
2020-05-26  9:55 ` Cornelia Huck
2020-05-26 11:08   ` 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=20200513142934.28788-4-farman@linux.ibm.com \
    --to=farman@linux.ibm.com \
    --cc=cohuck@redhat.com \
    --cc=jrossi@linux.ibm.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