Re: [PATCH V6 00/24] IOREQ feature (+ virtio-mmio) on Arm

[Julien Grall <julien@xxxxxxx>]

Hi Oleksandr,

On 29/01/2021 11:06, Oleksandr wrote:
> On 29.01.21 12:51, Oleksandr wrote:
> > On 29.01.21 10:49, Jan Beulich wrote:
> >
> > Hi Jan
> >
> > > On 29.01.2021 09:13, Wei Chen wrote:
> > > > I just tested the v6 and the latest backend service with the latest 
> > > > staging branch.
> > > > They work well.
> > > >
> > > > Tested-by: Wei Chen <Wei.Chen@xxxxxxx>
> > > An faod this was again Arm-only testing?
> > Yes, unfortunately I don't have a possibility to test on x86, if I had 
> > I would definitely perform testing.
> > I performed code analysis (the diff between V5 and V6 is small enough 
> > and these changes unlikely will add regression, so I was thinking that 
> > if basic x86 test passed on V5 (thanks to Julien) than it would likely 
> > pass on V6 as well),
> > but please don't get me wrong, I *completely* agree that even an 
> > obvious single patch must be tested.
> What worries me the most is:
>
>  >>> Unfortunately, I don't have a Windows VM in hand, so I can't 
> confirm if
>  >>> there is no regression there. Can anyone give a try?
>
> I am sorry, but would it be possible to make sure that current series 
> doesn't break that use-case? Or this could be tested (and fixed if there 
> is a need) after the feature freeze?
With Ian's help, I used Osstest to run a Windows job with your branch 
(see [1]).
The branch used is rebase version because there was a building issue 
with Mini-OS on your baseline (it is already fixed upstream).
The result are the same as the latest run with staging (see [2]). So 
there is no regression with your series applied!
I will go through the series one last time and then commit it.

Cheers,

[1] http://logs.test-lab.xenproject.org/osstest/logs/158785/
[2] 
http://logs.test-lab.xenproject.org/osstest/logs/158719/test-amd64-amd64-xl-qemuu-win7-amd64/info.html

>
--
Julien Grall