Hi Maxime,
On 28.04.2022 14:14, Maxime Ripard wrote:
> On Thu, Apr 28, 2022 at 03:18:08PM +0530, Jagan Teki wrote:
>> commit <711c7adc4687> ("drm: exynos: dsi: Use drm panel_bridge API")
>> added devm_drm_of_get_bridge for looking up if child node has panel
>> or bridge.
>>
>> However commit <b089c0a9b14c> ("Revert "drm: of: Lookup if child node
>> has panel or bridge") has reverted panel or bridge child node lookup
>> from devm_drm_of_get_bridge which eventually failed to find the DSI
>> devices in exynos drm dsi driver.
>>
>> So, use the conventional child panel bridge lookup helpers like it
>> does before.
>>
>> Signed-off-by: Jagan Teki <jagan@xxxxxxxxxxxxxxxxxxxx>
>> ---
>> drivers/gpu/drm/exynos/exynos_drm_dsi.c | 12 +++++++++++-
>> 1 file changed, 11 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/exynos/exynos_drm_dsi.c b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
>> index f067c86b0b12..ec673223d6b7 100644
>> --- a/drivers/gpu/drm/exynos/exynos_drm_dsi.c
>> +++ b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
>> @@ -25,6 +25,7 @@
>> #include <drm/drm_atomic_helper.h>
>> #include <drm/drm_bridge.h>
>> #include <drm/drm_mipi_dsi.h>
>> +#include <drm/drm_panel.h>
>> #include <drm/drm_print.h>
>> #include <drm/drm_probe_helper.h>
>> #include <drm/drm_simple_kms_helper.h>
>> @@ -1451,9 +1452,18 @@ static int exynos_dsi_host_attach(struct mipi_dsi_host *host,
>> struct device *dev = dsi->dev;
>> struct drm_encoder *encoder = &dsi->encoder;
>> struct drm_device *drm = encoder->dev;
>> + struct drm_panel *panel;
>> int ret;
>>
>> - dsi->out_bridge = devm_drm_of_get_bridge(dev, dev->of_node, 1, 0);
>> + panel = of_drm_find_panel(device->dev.of_node);
>> + if (!IS_ERR(panel)) {
>> + dsi->out_bridge = devm_drm_panel_bridge_add(dev, panel);
>> + } else {
>> + dsi->out_bridge = of_drm_find_bridge(device->dev.of_node);
>> + if (!dsi->out_bridge)
>> + dsi->out_bridge = ERR_PTR(-EINVAL);
>> + }
>> +
> We should just revert 711c7adc4687 entirely (and make it very explicit
> in the commit log). The out_bridge lifetime is completely screwed-up in
> that patch now. If there's a panel it will be managed, but it's not if
> there's a bridge.
Nope, this patch does exactly what is needed there.
If there is a panel, it creates a managed panel_bridge() proxy. If there
is a bridge, it simply gets a pointer to it. There is nothing to manage
if there is a bridge attached next to DSI. See devm_drm_of_get_bridge()
function, it does exactly the same things (managed panel_bridge or
simple pointer to the real next bridge). I agree that this is not very
obvious at the first sight, but that's how the code is now structured.
Please queue this to drm-misc-next, so the Exynos DSI will be again
operational after merging the drm-misc-fixes and drm-misc-next.
Best regards
--
Marek Szyprowski, PhD
Samsung R&D Institute Poland
