| Message ID | 20240816125408.310253-49-tzimmermann@suse.de (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | drm: Provide client setup helper and convert drivers | expand |
Hi Thomas, Thank you for the patch. On Fri, Aug 16, 2024 at 02:23:14PM +0200, Thomas Zimmermann wrote: > Call drm_client_setup() to run the kernel's default client setup > for DRM. Set fbdev_probe in struct drm_driver, so that the client > setup can start the common fbdev client. > > Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> > Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com> > Cc: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> > Cc: Michal Simek <michal.simek@amd.com> > --- > drivers/gpu/drm/xlnx/zynqmp_kms.c | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) > > diff --git a/drivers/gpu/drm/xlnx/zynqmp_kms.c b/drivers/gpu/drm/xlnx/zynqmp_kms.c > index bd1368df7870..f26b119322d5 100644 > --- a/drivers/gpu/drm/xlnx/zynqmp_kms.c > +++ b/drivers/gpu/drm/xlnx/zynqmp_kms.c > @@ -14,6 +14,7 @@ > #include <drm/drm_blend.h> > #include <drm/drm_bridge.h> > #include <drm/drm_bridge_connector.h> > +#include <drm/drm_client_setup.h> > #include <drm/drm_connector.h> > #include <drm/drm_crtc.h> > #include <drm/drm_device.h> > @@ -402,6 +403,7 @@ static const struct drm_driver zynqmp_dpsub_drm_driver = { > DRIVER_ATOMIC, > > DRM_GEM_DMA_DRIVER_OPS_WITH_DUMB_CREATE(zynqmp_dpsub_dumb_create), > + DRM_FBDEV_DMA_DRIVER_OPS, > > .fops = &zynqmp_dpsub_drm_fops, > > @@ -523,7 +525,7 @@ int zynqmp_dpsub_drm_init(struct zynqmp_dpsub *dpsub) > goto err_poll_fini; > > /* Initialize fbdev generic emulation. */ > - drm_fbdev_dma_setup(drm, 24); > + drm_client_setup(drm, drm_format_info(DRM_FORMAT_RGB888)); I know this would be a hassle to change, but do the majority of the callers of drm_client_setup() have a drm_format_info * already, or do they need to call drm_format_info() on a 4CC ? In the latter case, would it be better to pass the 4CC to drm_client_setup() ? As far as this patch goes, Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> > > return 0; >
Hi Am 18.08.24 um 22:19 schrieb Laurent Pinchart: > Hi Thomas, > > Thank you for the patch. > > On Fri, Aug 16, 2024 at 02:23:14PM +0200, Thomas Zimmermann wrote: >> Call drm_client_setup() to run the kernel's default client setup >> for DRM. Set fbdev_probe in struct drm_driver, so that the client >> setup can start the common fbdev client. >> >> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> >> Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com> >> Cc: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> >> Cc: Michal Simek <michal.simek@amd.com> >> --- >> drivers/gpu/drm/xlnx/zynqmp_kms.c | 4 +++- >> 1 file changed, 3 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/gpu/drm/xlnx/zynqmp_kms.c b/drivers/gpu/drm/xlnx/zynqmp_kms.c >> index bd1368df7870..f26b119322d5 100644 >> --- a/drivers/gpu/drm/xlnx/zynqmp_kms.c >> +++ b/drivers/gpu/drm/xlnx/zynqmp_kms.c >> @@ -14,6 +14,7 @@ >> #include <drm/drm_blend.h> >> #include <drm/drm_bridge.h> >> #include <drm/drm_bridge_connector.h> >> +#include <drm/drm_client_setup.h> >> #include <drm/drm_connector.h> >> #include <drm/drm_crtc.h> >> #include <drm/drm_device.h> >> @@ -402,6 +403,7 @@ static const struct drm_driver zynqmp_dpsub_drm_driver = { >> DRIVER_ATOMIC, >> >> DRM_GEM_DMA_DRIVER_OPS_WITH_DUMB_CREATE(zynqmp_dpsub_dumb_create), >> + DRM_FBDEV_DMA_DRIVER_OPS, >> >> .fops = &zynqmp_dpsub_drm_fops, >> >> @@ -523,7 +525,7 @@ int zynqmp_dpsub_drm_init(struct zynqmp_dpsub *dpsub) >> goto err_poll_fini; >> >> /* Initialize fbdev generic emulation. */ >> - drm_fbdev_dma_setup(drm, 24); >> + drm_client_setup(drm, drm_format_info(DRM_FORMAT_RGB888)); > I know this would be a hassle to change, but do the majority of the > callers of drm_client_setup() have a drm_format_info * already, or do > they need to call drm_format_info() on a 4CC ? In the latter case, would > it be better to pass the 4CC to drm_client_setup() ? Good idea. Most of the drivers have a 4cc code, but no format-info pointer. I think there are enough instances that it makes sense to add a little helper, drm_client_setup_with_fourcc(), for these drivers. There's also preferred_depth from drm_mode_config. It's already used in the client mode-setting code. That means that in some drivers, the format parameter to drm_client_setup() could likely be dropped. Best regards Thomas > > As far as this patch goes, > > Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> > >> >> return 0; >>
diff --git a/drivers/gpu/drm/xlnx/zynqmp_kms.c b/drivers/gpu/drm/xlnx/zynqmp_kms.c index bd1368df7870..f26b119322d5 100644 --- a/drivers/gpu/drm/xlnx/zynqmp_kms.c +++ b/drivers/gpu/drm/xlnx/zynqmp_kms.c @@ -14,6 +14,7 @@ #include <drm/drm_blend.h> #include <drm/drm_bridge.h> #include <drm/drm_bridge_connector.h> +#include <drm/drm_client_setup.h> #include <drm/drm_connector.h> #include <drm/drm_crtc.h> #include <drm/drm_device.h> @@ -402,6 +403,7 @@ static const struct drm_driver zynqmp_dpsub_drm_driver = { DRIVER_ATOMIC, DRM_GEM_DMA_DRIVER_OPS_WITH_DUMB_CREATE(zynqmp_dpsub_dumb_create), + DRM_FBDEV_DMA_DRIVER_OPS, .fops = &zynqmp_dpsub_drm_fops, @@ -523,7 +525,7 @@ int zynqmp_dpsub_drm_init(struct zynqmp_dpsub *dpsub) goto err_poll_fini; /* Initialize fbdev generic emulation. */ - drm_fbdev_dma_setup(drm, 24); + drm_client_setup(drm, drm_format_info(DRM_FORMAT_RGB888)); return 0;
Call drm_client_setup() to run the kernel's default client setup for DRM. Set fbdev_probe in struct drm_driver, so that the client setup can start the common fbdev client. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Cc: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Cc: Michal Simek <michal.simek@amd.com> --- drivers/gpu/drm/xlnx/zynqmp_kms.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)