Displaying 3 results from an estimated 3 matches for "argubaly".
Did you mean:
arguably
2024 Jul 28
1
[PATCH v2 02/37] drm/nouveau: handle pci/tegra drm_dev_{alloc, register} from common code
...share what other places are doing this stuff.
To recap, when converting a legacy driver into an aux split we've
found in several places that there is existing userspace that has
hardwired certain sysfs paths. ie an assumption that an infiniband
device appears under the sys/../pci/ directory.
Argubaly this userspace is not in good shape, but we have to preserve
it.
So the approach is to make the sysfs visible elements tied to the
original sysfs location (ie the pci device) and continue to use aux
otherwise for discovery, probing and tying subsystems together.
Obviously you have to be careful a...
2024 Jul 28
1
[PATCH v2 02/37] drm/nouveau: handle pci/tegra drm_dev_{alloc, register} from common code
...this stuff.
>
> To recap, when converting a legacy driver into an aux split we've
> found in several places that there is existing userspace that has
> hardwired certain sysfs paths. ie an assumption that an infiniband
> device appears under the sys/../pci/ directory.
>
> Argubaly this userspace is not in good shape, but we have to preserve
> it.
>
> So the approach is to make the sysfs visible elements tied to the
> original sysfs location (ie the pci device) and continue to use aux
> otherwise for discovery, probing and tying subsystems together.
>
>...
2024 Jul 26
2
[PATCH v2 02/37] drm/nouveau: handle pci/tegra drm_dev_{alloc, register} from common code
On 27/7/24 01:41, Danilo Krummrich wrote:
> On Fri, Jul 26, 2024 at 02:27:53PM +1000, Ben Skeggs wrote:
>>>>>> +
>>>>>> +static struct nouveau_drm *
>>>>>> +nouveau_drm_device_new(const struct drm_driver *drm_driver, struct device *parent,
>>>>>> + struct nvkm_device *device)
>>>>>> +{