From: Halil Pasic <pasic@linux.ibm.com>
To: Tony Krowiak <akrowiak@linux.ibm.com>
Cc: Pierre Morel <pmorel@linux.ibm.com>,
borntraeger@de.ibm.com, alex.williamson@redhat.com,
cohuck@redhat.com, linux-kernel@vger.kernel.org,
linux-s390@vger.kernel.org, kvm@vger.kernel.org,
frankja@linux.ibm.com, david@redhat.com, schwidefsky@de.ibm.com,
heiko.carstens@de.ibm.com, freude@linux.ibm.com,
mimu@linux.ibm.com
Subject: Re: [PATCH v9 3/4] s390: ap: implement PAPQ AQIC interception in kernel
Date: Fri, 7 Jun 2019 16:29:03 +0200 [thread overview]
Message-ID: <20190607162903.22fd959f.pasic@linux.ibm.com> (raw)
In-Reply-To: <2ffee52b-5e7f-f52a-069f-0a43d6418341@linux.ibm.com>
On Tue, 4 Jun 2019 15:38:51 -0400
Tony Krowiak <akrowiak@linux.ibm.com> wrote:
> On 5/21/19 11:34 AM, Pierre Morel wrote:
> > We register a AP PQAP instruction hook during the open
> > of the mediated device. And unregister it on release.
[..]
> > +/**
> > + * vfio_ap_wait_for_irqclear
> > + * @apqn: The AP Queue number
> > + *
> > + * Checks the IRQ bit for the status of this APQN using ap_tapq.
> > + * Returns if the ap_tapq function succedded and the bit is clear.
>
> s/succedded/succeeded/
>
I'm gonna fix this up when picking.
> > + * Returns if ap_tapq function failed with invalid, deconfigured or
> > + * checkstopped AP.
> > + * Otherwise retries up to 5 times after waiting 20ms.
> > + *
> > + */
> > +static void vfio_ap_wait_for_irqclear(int apqn)
> > +{
> > + struct ap_queue_status status;
> > + int retry = 5;
> > +
> > + do {
> > + status = ap_tapq(apqn, NULL);
> > + switch (status.response_code) {
> > + case AP_RESPONSE_NORMAL:
> > + case AP_RESPONSE_RESET_IN_PROGRESS:
> > + if (!status.irq_enabled)
> > + return;
> > + /* Fall through */
> > + case AP_RESPONSE_BUSY:
> > + msleep(20);
> > + break;
> > + case AP_RESPONSE_Q_NOT_AVAIL:
> > + case AP_RESPONSE_DECONFIGURED:
> > + case AP_RESPONSE_CHECKSTOPPED:
> > + default:
> > + WARN_ONCE(1, "%s: tapq rc %02x: %04x\n", __func__,
> > + status.response_code, apqn);
> > + return;
>
> Why not just break out of the loop and just use the WARN_ONCE
> outside of the loop?
>
AFAIU the idea was to differentiate between got a strange response_code
and ran out of retires.
Actually I suspect that we are fine in case of AP_RESPONSE_Q_NOT_AVAIL,
AP_RESPONSE_DECONFIGURED and AP_RESPONSE_CHECKSTOPPED in a sense that
what should be the post-condition of this function is guaranteed to be
reached. What do you think?
While I think that we can do better here, I see this as something that
should be done on top.
> > + }
> > + } while (--retry);
> > +
> > + WARN_ONCE(1, "%s: tapq rc %02x: %04x could not clear IR bit\n",
> > + __func__, status.response_code, apqn);
> > +}
> > +
> > +/**
> > + * vfio_ap_free_aqic_resources
> > + * @q: The vfio_ap_queue
> > + *
> > + * Unregisters the ISC in the GIB when the saved ISC not invalid.
> > + * Unpin the guest's page holding the NIB when it exist.
> > + * Reset the saved_pfn and saved_isc to invalid values.
> > + * Clear the pointer to the matrix mediated device.
> > + *
> > + */
[..]
> > +struct ap_queue_status vfio_ap_irq_disable(struct vfio_ap_queue *q)
> > +{
> > + struct ap_qirq_ctrl aqic_gisa = {};
> > + struct ap_queue_status status;
> > + int retries = 5;
> > +
> > + do {
> > + status = ap_aqic(q->apqn, aqic_gisa, NULL);
> > + switch (status.response_code) {
> > + case AP_RESPONSE_OTHERWISE_CHANGED:
> > + case AP_RESPONSE_NORMAL:
> > + vfio_ap_wait_for_irqclear(q->apqn);
> > + goto end_free;
> > + case AP_RESPONSE_RESET_IN_PROGRESS:
> > + case AP_RESPONSE_BUSY:
> > + msleep(20);
> > + break;
> > + case AP_RESPONSE_Q_NOT_AVAIL:
> > + case AP_RESPONSE_DECONFIGURED:
> > + case AP_RESPONSE_CHECKSTOPPED:
> > + case AP_RESPONSE_INVALID_ADDRESS:
> > + default:
> > + /* All cases in default means AP not operational */
> > + WARN_ONCE(1, "%s: ap_aqic status %d\n", __func__,
> > + status.response_code);
> > + goto end_free;
>
> Why not just break out of the loop instead of repeating the WARN_ONCE
> message?
>
I suppose the reason is same as above. I'm not entirely happy with this
code myself. E.g. why do we do retries here -- shouldn't we just fail the
aqic by the guest?
[..]
> > +static int handle_pqap(struct kvm_vcpu *vcpu)
> > +{
> > + uint64_t status;
> > + uint16_t apqn;
> > + struct vfio_ap_queue *q;
> > + struct ap_queue_status qstatus = {
> > + .response_code = AP_RESPONSE_Q_NOT_AVAIL, };
> > + struct ap_matrix_mdev *matrix_mdev;
> > +
> > + /* If we do not use the AIV facility just go to userland */
> > + if (!(vcpu->arch.sie_block->eca & ECA_AIV))
> > + return -EOPNOTSUPP;
> > +
> > + apqn = vcpu->run->s.regs.gprs[0] & 0xffff;
> > + mutex_lock(&matrix_dev->lock);
> > +
> > + if (!vcpu->kvm->arch.crypto.pqap_hook)
>
> Wasn't this already checked in patch 2 prior to calling this
> function? In fact, doesn't the hook point to this function?
>
Let us benevolently call this defensive programming. We are actually
in that callback AFAICT, so it sure was set a moment ago, and I guess
the client code still holds the kvm.lock so it is guaranteed to stay
so unless somebody is playing foul.
We can address this with a patch on top.
Regards,
Halil
next prev parent reply other threads:[~2019-06-07 14:29 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-05-21 15:34 [PATCH v9 0/4] vfio: ap: AP Queue Interrupt Control Pierre Morel
2019-05-21 15:34 ` [PATCH v9 1/4] s390: ap: kvm: add PQAP interception for AQIC Pierre Morel
2019-06-12 13:55 ` Harald Freudenberger
2019-05-21 15:34 ` [PATCH v9 2/4] vfio: ap: register IOMMU VFIO notifier Pierre Morel
2019-06-12 13:56 ` Harald Freudenberger
2019-05-21 15:34 ` [PATCH v9 3/4] s390: ap: implement PAPQ AQIC interception in kernel Pierre Morel
2019-05-23 15:43 ` Tony Krowiak
2019-06-04 19:38 ` Tony Krowiak
2019-06-07 14:29 ` Halil Pasic [this message]
2019-06-11 14:37 ` Tony Krowiak
2019-06-11 15:00 ` Halil Pasic
2019-06-12 13:56 ` Harald Freudenberger
2019-05-21 15:34 ` [PATCH v9 4/4] s390: ap: kvm: Enable PQAP/AQIC facility for the guest Pierre Morel
2019-06-12 13:57 ` Harald Freudenberger
2019-06-25 20:13 ` Christian Borntraeger
2019-06-25 20:15 ` Christian Borntraeger
2019-06-26 21:12 ` Tony Krowiak
2019-06-27 6:54 ` Christian Borntraeger
2019-06-27 12:04 ` Halil Pasic
2019-05-23 15:36 ` [PATCH v9 0/4] vfio: ap: AP Queue Interrupt Control Tony Krowiak
2019-06-04 14:56 ` Halil Pasic
2019-07-01 12:00 ` Halil Pasic
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=20190607162903.22fd959f.pasic@linux.ibm.com \
--to=pasic@linux.ibm.com \
--cc=akrowiak@linux.ibm.com \
--cc=alex.williamson@redhat.com \
--cc=borntraeger@de.ibm.com \
--cc=cohuck@redhat.com \
--cc=david@redhat.com \
--cc=frankja@linux.ibm.com \
--cc=freude@linux.ibm.com \
--cc=heiko.carstens@de.ibm.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-s390@vger.kernel.org \
--cc=mimu@linux.ibm.com \
--cc=pmorel@linux.ibm.com \
--cc=schwidefsky@de.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;
as well as URLs for NNTP newsgroup(s).