linux-kernel - Re: [PATCH] virtio_pci: Wait for legacy device to be reset

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20230411071111-mutt-send-email-mst@kernel.org>
Date:   Tue, 11 Apr 2023 07:13:37 -0400
From:   "Michael S. Tsirkin" <mst@...hat.com>
To:     Angus Chen <angus.chen@...uarmicro.com>
Cc:     Jason Wang <jasowang@...hat.com>,
        "virtualization@...ts.linux-foundation.org" 
        <virtualization@...ts.linux-foundation.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] virtio_pci: Wait for legacy device to be reset

On Tue, Apr 11, 2023 at 10:57:51AM +0000, Angus Chen wrote:
> > > > Yes, not very clear still. What are the device and vendor ID of the
> > > > card? Does it have the virtio capabilities?
> > > > Thanks
> > > It's have the virtio capabilities.
> > > When it is used in bare metal ，
> > >    it just use PCI_VENDOR_ID_REDHAT_QUMRANET.
> > > When it use in vm,
> > >  It is just like :
> > > { PCI_DEVICE_SUB(PCI_VENDOR_ID_REDHAT_QUMRANET,
> > >              VIRTIO_TRANS_ID_NET,
> > >              PCI_VENDOR_ID_JAGUARMICRO,
> > >              VIRTIO_ID_NET) },
> > >     { PCI_DEVICE_SUB(PCI_VENDOR_ID_REDHAT_QUMRANET,
> > >              VIRTIO_TRANS_ID_BLOCK,
> > >              PCI_VENDOR_ID_JAGUARMICRO,
> > >              VIRTIO_ID_BLOCK) },
> > >
> > 
> > Okay. So it's a transitional device. The question, then,
> > is how come you end up with drivers/virtio/virtio_pci_legacy.c
> > with a modern linux at all?
> > Did you by chance set force_legacy = true ?
> Yes,I can provide more information about it.
> We can test it use force_legacy ,or we use a server with multi os but it use only one card.
> Thanks.

Okay, so I don't yet see lots of value of merging this patch upstream.
The problematic code does not run by default as the modern driver is
used, and applying the patch upstream does nothing to fix the multi-os
case.

What did I miss?


-- 
MST