[RFC PATCH v2 1/3] drivers/accel: define kconfig and register a new major
Dave Airlie
airlied at gmail.com
Mon Nov 7 20:18:21 UTC 2022
On Mon, 7 Nov 2022 at 23:10, Jason Gunthorpe <jgg at nvidia.com> wrote:
>
> On Mon, Nov 07, 2022 at 03:01:08PM +0200, Oded Gabbay wrote:
> > I don't agree with your statement that it should be "a layer over top of DRM".
> > Anything on top of DRM is a device driver.
> > Accel is not a device driver, it is a new type of drm minor / drm driver.
>
> Yeah, I still think this is not the right way, you are getting almost
> nothing from DRM and making everything more complicated in the
> process.
You are looking at the small picture that is these patches, there are
just infrastructure to start the process of merging drivers and
reusing other parts of the drm code.
We aren't going to ever get anywhere if we start splitting code out of
drm just in case, we get this stuff rolling in the tree and if we have
a pressing need to refactor it out into separate libraries later then
we can address that from a more educated place, instead of just
throwing huge refactors around before we have any code to even use
them.
>
> IMHO this is much better, because accel has very little need of DRM to
> manage a struct device/cdev in the first place.
Right now it doesn't, but when drivers start leveraging the other code
it will reuse a lot more code.
I'm not going to spend too much time entertaining this, devm vs drmm
memory etc are real problems drm has already identified if not
completely solved.
Dave.
More information about the dri-devel
mailing list