linux-kernel - Re: [PATCH 5/8] virt: vbox: Add support for the new VBG_IOCTL_ACQUIRE_GUEST

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] [day] [month] [year] [list]

Message-ID: <c81cd7ac-a698-8d9b-c5c1-6a4e0bc90966@redhat.com>
Date:   Tue, 23 Jun 2020 15:09:52 +0200
From:   Hans de Goede <hdegoede@...hat.com>
To:     Arnd Bergmann <arnd@...db.de>
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 5/8] virt: vbox: Add support for the new
 VBG_IOCTL_ACQUIRE_GUEST_CAPABILITIES ioctl

Hi,

Thank you for the review and sorry for being slow with
responding.

On 5/20/20 11:47 PM, Arnd Bergmann wrote:
> On Wed, May 20, 2020 at 9:55 PM Hans de Goede <hdegoede@...hat.com> wrote:
>>
>> Add support for the new VBG_IOCTL_ACQUIRE_GUEST_CAPABILITIES ioctl, this
>> is necessary for automatic resizing of the guest resolution to match the
>> VM-window size to work with the new VMSVGA virtual GPU which is now the
>> new default in VirtualBox.
>>
>> BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1789545
>> Signed-off-by: Hans de Goede <hdegoede@...hat.com>
> 
> This is not a great interface, but I see no bugs in the implementation
> or problems with portability.
> 
> If you want to improve it a little bit, note that spin_lock_irqsave() inside
> of mutex_lock() is a little bit silly since you know at this point that
> interrupts are enabled and you could use spin_lock_irq() instead.
> 
> I assume the driver does the same thing everywhere but I did not check.

Yes the driver does the same everywhere and TBH I would prefer to keep
it that way, this all is not that performance critical and in my experience
once a driver starts using the less save variants it is easy for some
mistakes to creep in.

As for the rate-limiting of the printk remark you made in reply to the
"virt: vbox: Log unknown ioctl requests as error" patch that is a valid
remark. I will prepare a v2 of this patch-set addressing that.

Regards,

Hans