From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <jasowang@redhat.com>
MIME-Version: 1.0
References: <09b74816-8a1e-f993-640f-eb790a4a4698@redhat.com>
 <20210811100550.GC21582@quicinc.com> <2723b42a-f5d0-9c49-bf5c-302fbd4c947f@redhat.com>
 <20210816013138-mutt-send-email-mst@kernel.org> <20210816063550.GD5604@quicinc.com>
 <20210816074558-mutt-send-email-mst@kernel.org> <af73ed7b-b339-3ba2-87cb-176a9435fa2c@redhat.com>
 <20210817034312-mutt-send-email-mst@kernel.org> <20210817100333.GA9207@quicinc.com>
 <e07fd648-9044-d486-29b6-83ee64b6842a@redhat.com> <20210818051544.GC9207@quicinc.com>
In-Reply-To: <20210818051544.GC9207@quicinc.com>
From: Jason Wang <jasowang@redhat.com>
Date: Wed, 18 Aug 2021 13:40:45 +0800
Message-ID: <CACGkMEv11s=9z4gMKo_BcmVpoznEbaEOctDvp=NUK7UtrhgjjQ@mail.gmail.com>
Subject: Re: [virtio-dev] Re: [PATCH v1] virtio-mmio: Specify wait needed in
 driver during reset
Content-Type: text/plain; charset="UTF-8"
To: Srivatsa Vaddagiri <quic_svaddagi@quicinc.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>, Srivatsa Vaddagiri <svaddagi@qti.qualcomm.com>, Cornelia Huck <cohuck@redhat.com>, "virtio-dev@lists.oasis-open.org" <virtio-dev@lists.oasis-open.org>, Trilok Soni <quic_tsoni@quicinc.com>, Pratik Patel <quic_pratikp@quicinc.com>
List-ID: <virtio-dev.lists.oasis-open.org>

On Wed, Aug 18, 2021 at 1:15 PM Srivatsa Vaddagiri
<quic_svaddagi@quicinc.com> wrote:
>
> * Jason Wang <jasowang@redhat.com> [2021-08-18 10:54:37]:
>
> > >I think the option we discussed earlier of a new feature bit seems less
> > >intrusive than incrementing MMIO version?
> > >
> > >https://lists.oasis-open.org/archives/virtio-dev/202107/msg00168.html
> >
> >
> > Using features will result some interesting question:
> >
> > 1) the drivers usually reset the device before feature negotiation
> > 2) it means the driver must mandate this behavior even before feature
> > negotiation is done
> >
> > We don't have those issue if we increase the version (but it looks more
> > intrusive).
>
> Hmm ..are you suggesting that we read the Version register as part of reset code
> and do the wait part only for v3 device?

We only need to read the version once (see virtio_mmio_probe()). So it
could be something like

if (vm_dev->version == 0x3) {
    while (readl(vm_dev->base + VIRTIO_MMIO_STATUS))
        msleep(1);
}

> Otherwise we have the same issue as
> feature bit with version change also.
>
> Can't the driver do the same with feature register also? Do poll if
> WAIT_FOR_RESET_COMPLETION feature bit is set, which can happen before feature
> negotiation is complete.

The possible issue is that the feature negotiation is not done at that time.

Thanks

>
> - vatsa
>
> --
> Qualcomm Innovation Center, Inc. is submitting the attached "feedback" as a
> non-member to the virtio-dev mailing list for consideration and inclusion.
>