Le 27/08/24 - 16:39, Maxime Ripard a écrit :
> On Tue, Aug 27, 2024 at 03:24:10PM GMT, Louis Chauvet wrote:
> > Le 27/08/24 - 15:15, Maxime Ripard a écrit :
> > > Hi,
> > >
> > > On Tue, Aug 27, 2024 at 11:57:36AM GMT, Louis Chauvet wrote:
> > > > The current VKMS driver uses non-managed function to create connectors. It
> > > > is not an issue yet, but in order to support multiple devices easily,
> > > > convert this code to use drm and device managed helpers.
> > > >
> > > > Signed-off-by: Louis Chauvet <louis.chauvet@xxxxxxxxxxx>
> > > > ---
> > > > drivers/gpu/drm/vkms/vkms_drv.h | 1 -
> > > > drivers/gpu/drm/vkms/vkms_output.c | 22 ++++++++++++----------
> > > > 2 files changed, 12 insertions(+), 11 deletions(-)
> > > >
> > > > diff --git a/drivers/gpu/drm/vkms/vkms_drv.h b/drivers/gpu/drm/vkms/vkms_drv.h
> > > > index 5e46ea5b96dc..9a3c6c34d1f6 100644
> > > > --- a/drivers/gpu/drm/vkms/vkms_drv.h
> > > > +++ b/drivers/gpu/drm/vkms/vkms_drv.h
> > > > @@ -99,7 +99,6 @@ struct vkms_crtc_state {
> > > > struct vkms_output {
> > > > struct drm_crtc crtc;
> > > > struct drm_encoder encoder;
> > > > - struct drm_connector connector;
> > > > struct drm_writeback_connector wb_connector;
> > > > struct hrtimer vblank_hrtimer;
> > > > ktime_t period_ns;
> > > > diff --git a/drivers/gpu/drm/vkms/vkms_output.c b/drivers/gpu/drm/vkms/vkms_output.c
> > > > index 5ce70dd946aa..4fe6b88e8081 100644
> > > > --- a/drivers/gpu/drm/vkms/vkms_output.c
> > > > +++ b/drivers/gpu/drm/vkms/vkms_output.c
> > > > @@ -3,11 +3,11 @@
> > > > #include "vkms_drv.h"
> > > > #include <drm/drm_atomic_helper.h>
> > > > #include <drm/drm_edid.h>
> > > > +#include <drm/drm_managed.h>
> > > > #include <drm/drm_probe_helper.h>
> > > >
> > > > static const struct drm_connector_funcs vkms_connector_funcs = {
> > > > .fill_modes = drm_helper_probe_single_connector_modes,
> > > > - .destroy = drm_connector_cleanup,
> > > > .reset = drm_atomic_helper_connector_reset,
> > > > .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
> > > > .atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
> > > > @@ -50,7 +50,7 @@ int vkms_output_init(struct vkms_device *vkmsdev, int index)
> > > > {
> > > > struct vkms_output *output = &vkmsdev->output;
> > > > struct drm_device *dev = &vkmsdev->drm;
> > > > - struct drm_connector *connector = &output->connector;
> > > > + struct drm_connector *connector;
> > > > struct drm_encoder *encoder = &output->encoder;
> > > > struct drm_crtc *crtc = &output->crtc;
> > > > struct vkms_plane *primary, *cursor = NULL;
> > > > @@ -80,8 +80,15 @@ int vkms_output_init(struct vkms_device *vkmsdev, int index)
> > > > if (ret)
> > > > return ret;
> > > >
> > > > - ret = drm_connector_init(dev, connector, &vkms_connector_funcs,
> > > > - DRM_MODE_CONNECTOR_VIRTUAL);
> > > > + connector = drmm_kzalloc(dev, sizeof(*connector), GFP_KERNEL);
> > > > + if (!connector) {
> > > > + DRM_ERROR("Failed to allocate connector\n");
> > > > + ret = -ENOMEM;
> > > > + goto err_connector;
> > > > + }
> > > > +
> > >
> > > I think it would be worth explaining why you need to move to a separate
> > > allocation for the connector now.
> > >
> > > Maxime
> >
> > Hi,
> >
> > This is in preparation for ConfigFS implementation, as the number of
> > connector/encoders/crtc/planes... will be dynamic, we need to have
> > separate alloaction.
> >
> > If I add this paragraph in the commit message, is it sufficient?
> >
> > A specific allocation for the connector is not strictly necessary
> > at this point, but in order to implement dynamic configuration of
> > VKMS (configFS), it will be easier to have one allocation per
> > connector.
> >
> > (same for encoder & CRTC)
>
> Yeah, that's a good message, but it probably belongs in a separate patch
> then.
Can you explain what you mean by "in a separate patch"? I wanted to write
this paragraph in the commit log.
Louis Chauvet
