[oi-dev] AMDgpu 1.1.2 and OpenChrome 0.5.159 drivers

Aurélien Larcher aurelien.larcher at gmail.com
Tue Sep 27 11:46:10 UTC 2016


On Tue, Sep 27, 2016 at 4:47 AM, Мартин Бохниг via oi-dev
<oi-dev at openindiana.org> wrote:
> Hi Ken,
>
> bills are _a_ damn urgent tyrannic core-problem, but still only a side-show.
> It is more like:
>
> Task: Build the 3 Pyramids of Gizeh in 1:1 size from original stones
> material as copy 1 mile south of the original Pyramids.
>
>
> Available funds: for food, staff and material: (0 (actually a large negative
> sum, but let's stick to calling it zero)
>
> Available workforce: 2 or 3 unpaid hungry slaves.
>
>
> Available time: 2 to 3 days
>
>
> Is that clearer now?
>
> As for the Via/OpenChrome (and other vendors') drivers: You need to check if
> they also require KMS now or not, if not then they may function on Solaris.
> But given that LinUX has DRM/KMS for almost everything such chances are
> getting slimmer by the year.
> Nevertheless - for Via/OpenChrome that may still be possible because it took
> very long until they first got KMS stable and into mainline:
>
> http://www.phoronix.com/scan.php?page=news_item&px=VIA-KMS-Revival
>
> But AMD/Ati KMS (or a more modern Intel "i915") : Our only chance in that
> direction would be, if Oracle itself actually gets this done *plus* decides
> to share it with us. I'm not convinced of that.
> Or if we make linuxkpi work.

The most active person regarding DRM/KMS drivers seems Francois Tigeot
from DragonflyBSD.

His presentation, that you and Randy Fishel pointed out, is pretty
clear on the way to follow.

https://www.x.org/wiki/Events/XDC2014/XDC2014TigeotDragonFlyBSD/XDC-2014_Porting_kms_drivers_to_DragonFly.pdf

I remember that Randy hinted that they would be moving away from the
"port to Solaris" approach of current DRM driver and follow the same
path as you recommended "port illumos to DRM drivers".

Additionally Jean-Sébastien Pédron would be the person to interact
with regarding radeon.

Regarding the Solaris port, it would be really nice to find a way to
actually interact in a mutually beneficial way with Randy.

The final point is that if we want more developments for OpenIndiana,
we need at least more people involved for packaging.
In my case, I certainly would like do more dev instead of packaging
stuff but it means that people should take over some pending tasks.
For instance, someone mentioned fonts recently: well they are not
packaged in oi-userland, so someone has to pick the ball.
Another thing is testing the pending X11 update branch and fixing a
few components with EGL.
I will not have the time to look at all of this on my own.

If that does not happen then I do not see the point in discussing anything.

Kind regards

Aurélien


>
>
> But a few too many "if"'s for such bold statements.
> Hence my shocked reaction.
>
>
>
> %martin
>
>
> Вторник, 27 сентября 2016, 0:08 UTC от ken mays <maybird1776 at yahoo.com>:
>
> Hi Martin,
>
> Right. The sole objective proved we can compile these 'current' 2D drivers
> on OI-Hipster today with the existing Solaris X11 infrastructure. Getting
> everything else working properly requires stupendous mind-boggling mental
> energy and funds. Everyone has bills to pay...
>
> ~ Ken
>
>
> _______________________________________________
> oi-dev mailing list
> oi-dev at openindiana.org
> https://openindiana.org/mailman/listinfo/oi-dev



-- 
---
Praise the Caffeine embeddings




More information about the oi-dev mailing list