From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Reply-To: pmorel@linux.ibm.com Subject: Re: [PATCH v2 5/8] s390: vfio-ap: wait for queue empty on queue reset References: <1555796980-27920-1-git-send-email-akrowiak@linux.ibm.com> <1555796980-27920-6-git-send-email-akrowiak@linux.ibm.com> From: Pierre Morel Date: Tue, 23 Apr 2019 14:50:26 +0200 MIME-Version: 1.0 In-Reply-To: <1555796980-27920-6-git-send-email-akrowiak@linux.ibm.com> Content-Type: text/plain; charset="utf-8"; format="flowed" Content-Language: en-US Content-Transfer-Encoding: 8bit Message-Id: <9a23da6d-89a8-227d-b756-6e002b0199e3@linux.ibm.com> Sender: kvm-owner@vger.kernel.org List-Archive: List-Post: To: Tony Krowiak , linux-s390@vger.kernel.org, linux-kernel@vger.kernel.org, kvm@vger.kernel.org Cc: freude@linux.ibm.com, borntraeger@de.ibm.com, cohuck@redhat.com, frankja@linux.ibm.com, david@redhat.com, schwidefsky@de.ibm.com, heiko.carstens@de.ibm.com, pasic@linux.ibm.com, alex.williamson@redhat.com, kwankhede@nvidia.com List-ID: On 20/04/2019 23:49, Tony Krowiak wrote: > Refactors the AP queue reset function to wait until the queue is empty > after the PQAP(ZAPQ) instruction is executed to zero out the queue as > required by the AP architecture. > > Signed-off-by: Tony Krowiak > --- > drivers/s390/crypto/vfio_ap_ops.c | 35 ++++++++++++++++++++++++++++++++--- > 1 file changed, 32 insertions(+), 3 deletions(-) > > diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c > index 091804317c5e..31ce30ee784d 100644 > --- a/drivers/s390/crypto/vfio_ap_ops.c > +++ b/drivers/s390/crypto/vfio_ap_ops.c > @@ -744,15 +744,44 @@ static int vfio_ap_mdev_group_notifier(struct notifier_block *nb, > return NOTIFY_OK; > } > > -static int vfio_ap_mdev_reset_queue(unsigned int apid, unsigned int apqi, > - unsigned int retry) > +static void vfio_ap_mdev_wait_for_qempty(unsigned long apid, unsigned long apqi) > { > struct ap_queue_status status; > + ap_qid_t qid = AP_MKQID(apid, apqi); > + int retry = 5; > + > + do { > + status = ap_tapq(qid, NULL); > + switch (status.response_code) { > + case AP_RESPONSE_NORMAL: > + if (status.queue_empty) > + return; > + msleep(20); > + break; > + case AP_RESPONSE_RESET_IN_PROGRESS: > + case AP_RESPONSE_BUSY: > + msleep(20); > + break; > + default: > + pr_warn("%s: tapq err %02x: 0x%04x may not be empty\n", > + __func__, status.response_code, q->apqn); q is not defined. use qid ? > + return; > + } > + } while (--retry); > +} > + > +int vfio_ap_mdev_reset_queue(unsigned int apid, unsigned int apqi) > +{ > + struct ap_queue_status status; > + int retry = 5; > > do { > status = ap_zapq(AP_MKQID(apid, apqi)); > switch (status.response_code) { > case AP_RESPONSE_NORMAL: > + vfio_ap_mdev_wait_for_qempty(apid, apqi); > + return 0; > + case AP_RESPONSE_DECONFIGURED: > return 0; > case AP_RESPONSE_RESET_IN_PROGRESS: > case AP_RESPONSE_BUSY: > @@ -778,7 +807,7 @@ static int vfio_ap_mdev_reset_queues(struct mdev_device *mdev) > matrix_mdev->matrix.apm_max + 1) { > for_each_set_bit_inv(apqi, matrix_mdev->matrix.aqm, > matrix_mdev->matrix.aqm_max + 1) { > - ret = vfio_ap_mdev_reset_queue(apid, apqi, 1); > + ret = vfio_ap_mdev_reset_queue(apid, apqi); > /* > * Regardless whether a queue turns out to be busy, or > * is not operational, we need to continue resetting > -- Pierre Morel Linux/KVM/QEMU in Böblingen - Germany