linux-kernel - Re: [RFC PATCH v2 1/9] vfio/pci: split vfio_pci

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: <20200207124831.391d5f70@w520.home>
Date:   Fri, 7 Feb 2020 12:48:31 -0700
From:   Alex Williamson <alex.williamson@...hat.com>
To:     Yan Zhao <yan.y.zhao@...el.com>
Cc:     kvm@...r.kernel.org, linux-kernel@...r.kernel.org,
        cohuck@...hat.com, zhenyuw@...ux.intel.com, zhi.a.wang@...el.com,
        kevin.tian@...el.com, shaopeng.he@...el.com, yi.l.liu@...el.com
Subject: Re: [RFC PATCH v2 1/9] vfio/pci: split vfio_pci_device into public
 and private parts

On Thu, 30 Jan 2020 21:09:56 -0500
Yan Zhao <yan.y.zhao@...el.com> wrote:

> split vfio_pci_device into two parts:
> (1) a public part,
>     including pdev, num_region, irq_type which are accessible from
>     outside of vfio.
> (2) a private part,
>     a pointer to vfio_pci_device_private, only accessible within vfio
> 
> Cc: Kevin Tian <kevin.tian@...el.com>
> Signed-off-by: Yan Zhao <yan.y.zhao@...el.com>
> ---
>  drivers/vfio/pci/vfio_pci.c         | 209 +++++++++++++++-------------
>  drivers/vfio/pci/vfio_pci_config.c  | 157 +++++++++++----------
>  drivers/vfio/pci/vfio_pci_igd.c     |  16 +--
>  drivers/vfio/pci/vfio_pci_intrs.c   | 171 ++++++++++++-----------
>  drivers/vfio/pci/vfio_pci_nvlink2.c |  16 +--
>  drivers/vfio/pci/vfio_pci_private.h |   5 +-
>  drivers/vfio/pci/vfio_pci_rdwr.c    |  36 ++---
>  include/linux/vfio.h                |   7 +
>  8 files changed, 321 insertions(+), 296 deletions(-)

I think the typical solution to something like this would be...

struct vfio_pci_device {
	...
};

struct vfio_pci_device_private {
	struct vfio_pci_device vdev;
	...
};

External code would be able to work with the vfio_pci_device and
internal code would do a container_of() to get access to the private
fields.  What's done here is pretty ugly and not very cache friendly.
Thanks,

Alex