virtio-comment

Use device ID 41

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
content.tex | 2 ++
1 file changed, 2 insertions(+)

diff --git a/content.tex b/content.tex
index b9934b4fa05c..d9913d056317 100644
--- a/content.tex
+++ b/content.tex
@@ -2876,6 +2876,8 @@ \chapter{Device Types}\label{sec:Device Types}
\hline
40 & Bluetooth device \\
\hline
+41 & GPIO device \\
+\hline
\end{tabular}

Some of the devices above are unspecified by this document,
--
2.25.1

On 02-06-21, 12:12, Viresh Kumar wrote:
> Use device ID 41
>
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> ---
> content.tex | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/content.tex b/content.tex
> index b9934b4fa05c..d9913d056317 100644
> --- a/content.tex
> +++ b/content.tex
> @@ -2876,6 +2876,8 @@ \chapter{Device Types}\label{sec:Device Types}
> \hline
> 40 & Bluetooth device \\
> \hline
> +41 & GPIO device \\
> +\hline
> \end{tabular}
>
> Some of the devices above are unspecified by this document,

Opened issue for the same here:

https://github.com/oasis-tcs/virtio-spec/issues/109

--
viresh

On 02.06.21 08:42, Viresh Kumar wrote:
> Use device ID 41

Thanks for submitting that. I should have been my duty, but I've been
too busy w/ other things.

You've probably already seen my spec in the Linux virtio-gpio patch
queue. Feel free to texify it for submitting here. (*1)

Also please remind me to update the device id (need to fix kernel as
well as qemu side).


--mtx

*1)
https://github.com/metux/linux/commit/66c98eab3433a592b11d60d1a05c2609fee4811a
https://github.com/metux/linux/tree/submit/virtio-gpio-2020-12-07-v3

--
---
Hinweis: unverschlüsselte E-Mails können leicht abgehört und manipuliert
werden ! Für eine vertrauliche Kommunikation senden Sie bitte ihren
GPG/PGP-Schlüssel zu.
---
Enrico Weigelt, metux IT consult
Free software and Linux embedded engineering
info@metux.net -- +49-151-27565287

On 02-06-21, 13:57, Enrico Weigelt, metux IT consult wrote:
> On 02.06.21 08:42, Viresh Kumar wrote:
> > Use device ID 41
>
> Thanks for submitting that. I should have been my duty, but I've been
> too busy w/ other things.
>
> You've probably already seen my spec in the Linux virtio-gpio patch
> queue. Feel free to texify it for submitting here. (*1)

I am halfway through :)

> Also please remind me to update the device id (need to fix kernel as
> well as qemu side).

Sure, I would be fine to re-spin the kernel driver as well (we need to
really finish this stuff sooner) :)

--
viresh

On 2021/6/2 19:57, Enrico Weigelt, metux IT consult wrote:
> On 02.06.21 08:42, Viresh Kumar wrote:
>> Use device ID 41
>
> Thanks for submitting that. I should have been my duty, but I've been
> too busy w/ other things.
>
> You've probably already seen my spec in the Linux virtio-gpio patch
> queue. Feel free to texify it for submitting here. (*1)
>
> Also please remind me to update the device id (need to fix kernel as
> well as qemu side).
>
>
> --mtx
>
> *1)
> https://github.com/metux/linux/commit/66c98eab3433a592b11d60d1a05c2609fee4811a
> https://github.com/metux/linux/tree/submit/virtio-gpio-2020-12-07-v3
>

Has anyone ever tried to use GPIO in the guest to generate waveform ?

When the frequency reaches a certain level, the waveform will become
inaccurate
due to the delay and uncertainty of the communication time between
virtio frontend
and backend. I'm not sure if any good solution for this case.

Regards,
Jie

On 04-06-21, 10:28, Jie Deng wrote:
> Has anyone ever tried to use GPIO in the guest to generate waveform ?

I haven't tried the code yet, I just started where Enrico left
earlier. Done with Id allocation and specs (yes reviews pending), will
be looking into kernel now.

> When the frequency reaches a certain level, the waveform will become
> inaccurate
> due to the delay and uncertainty of the communication time between virtio
> frontend
> and backend. I'm not sure if any good solution for this case.

This is expected I would say. The interface is going to be slow and at
some frequency it will break the waveform.

--
viresh

On 04.06.21 04:28, Jie Deng wrote:

> Has anyone ever tried to use GPIO in the guest to generate waveform ?
>
> When the frequency reaches a certain level, the waveform will become
> inaccurate
> due to the delay and uncertainty of the communication time between
> virtio frontend
> and backend. I'm not sure if any good solution for this case.

Wooow, that's something I didn't ever dared considering. How well that
works heavily depends on actual hardware (even on bare metal) and lots
of other factors.

Even on a bare metal Linux system, I wouldn't want userland do such
things via raw gpio devices - instead have specific (RT capable) kernel
drivers that expose themselves via some fitting hilevel subsys (eg. iio
or pwm). For those use cases, we should have specific protocols.

If it helps, we can add certain metadata, eg. timing constraints to the
line descriptions.


--mtx

--
---
Hinweis: unverschlüsselte E-Mails können leicht abgehört und manipuliert
werden ! Für eine vertrauliche Kommunikation senden Sie bitte ihren
GPG/PGP-Schlüssel zu.
---
Enrico Weigelt, metux IT consult
Free software and Linux embedded engineering
info@metux.net -- +49-151-27565287