lkml.org 
[lkml]   [2021]   [Aug]   [24]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [PATCH v5] virtio-blk: Add validation for block size in config space
From
Date

On 8/24/2021 5:47 AM, Jason Wang wrote:
> On Tue, Aug 24, 2021 at 6:31 AM Max Gurtovoy <mgurtovoy@nvidia.com> wrote:
>>
>> On 8/23/2021 3:13 PM, Michael S. Tsirkin wrote:
>>> On Mon, Aug 23, 2021 at 01:45:31PM +0300, Max Gurtovoy wrote:
>>>> It helpful if there is a justification for this.
>>>>
>>>> In this case, no such HW device exist and the only device that can cause
>>>> this trouble today is user space VDUSE device that must be validated by the
>>>> emulation VDUSE kernel driver.
>>>>
>>>> Otherwise, will can create 1000 commit like this in the virtio level (for
>>>> example for each feature for each virtio device).
>>> Yea, it's a lot of work but I don't think it's avoidable.
>>>
>>>>>>>>> And regardless of userspace device, we still need to fix it for other cases.
>>>>>>>> which cases ? Do you know that there is a buggy HW we need to workaround ?
>>>>>>>>
>>>>>>> No, there isn't now. But this could be a potential attack surface if
>>>>>>> the host doesn't trust the device.
>>>>>> If the host doesn't trust a device, why it continues using it ?
>>>>>>
>>>>> IIUC this is the case for the encrypted VMs.
>>>> what do you mean encrypted VM ?
>>>>
>>>> And how this small patch causes a VM to be 100% encryption supported ?
>>>>
>>>>>> Do you suggest we do these workarounds in all device drivers in the kernel ?
>>>>>>
>>>>> Isn't it the driver's job to validate some unreasonable configuration?
>>>> The check should be in different layer.
>>>>
>>>> Virtio blk driver should not cover on some strange VDUSE stuff.
>>> Yes I'm not convinced VDUSE is a valid use-case. I think that for
>>> security and robustness it should validate data it gets from userspace
>>> right there after reading it.
>>> But I think this is useful for the virtio hardening thing.
>>> https://lwn.net/Articles/865216/
>> I don't see how this change is assisting confidential computing.
>>
>> Confidential computingtalks about encrypting guest memory from the host,
>> and not adding some quirks to devices.
> In the case of confidential computing, the hypervisor and hard device
> is not in the trust zone. It means the guest doesn't trust the cloud
> vendor.

Confidential computing protects data during processing ("in-use" data).

Nothing to do with virtio feature negotiation.

>
> That's why we need to validate any input from them.
>
> Thanks
>
>>> Yongji - I think the commit log should be much more explicit that
>>> this is hardening. Otherwise people get confused and think this
>>> needs a CVE or a backport for security.
>>>

\
 
 \ /
  Last update: 2021-08-24 12:13    [W:0.089 / U:3.424 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site