linux-kernel - Re: [RFC v2 1/4] hisi-acc-vfio-pci: add new vfio

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]

Date:   Mon, 5 Jul 2021 15:32:47 -0300
From:   Jason Gunthorpe <jgg@...dia.com>
To:     Leon Romanovsky <leon@...nel.org>
Cc:     Shameerali Kolothum Thodi <shameerali.kolothum.thodi@...wei.com>,
        Max Gurtovoy <mgurtovoy@...dia.com>,
        "kvm@...r.kernel.org" <kvm@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "linux-crypto@...r.kernel.org" <linux-crypto@...r.kernel.org>,
        "alex.williamson@...hat.com" <alex.williamson@...hat.com>,
        Linuxarm <linuxarm@...wei.com>,
        liulongfang <liulongfang@...wei.com>,
        "Zengtao (B)" <prime.zeng@...ilicon.com>,
        yuzenghui <yuzenghui@...wei.com>,
        Jonathan Cameron <jonathan.cameron@...wei.com>,
        "Wangzhou (B)" <wangzhou1@...ilicon.com>
Subject: Re: [RFC v2 1/4] hisi-acc-vfio-pci: add new vfio_pci driver for
 HiSilicon ACC devices

On Mon, Jul 05, 2021 at 09:27:37PM +0300, Leon Romanovsky wrote:

> > I think, in any case, it would be good to update the Documentation based on
> > which way we end up doing this.
> 
> The request to update Documentation can be seen as an example of
> choosing not-good API decisions. Expectation to see all drivers to
> use same callbacks with same vfio-core function calls sounds strange
> to me.

It is not vfio-core, it is vfio-pci-core. It is similar to how some of
the fops stuff works, eg the generic_file whatever functions everyone
puts in.

It would be improved a bit by making the ops struct mutable and
populating it at runtime like we do in RDMA. Then the PCI ops and
driver ops could be merged together without the repetition.

Probably something that is more interesting if there are more drivers
as it is a fair amount of typing to make.

Jason