On 9/9/25 19:44, Dmitry Baryshkov wrote:
On Tue, Sep 09, 2025 at 02:38:21PM +0530, Harikrishna Shenoy wrote:From: Jayesh Choudhary <j-choudh...@ti.com> In case if we get errors in cdns_mhdp_link_up() or cdns_mhdp_reg_read() in atomic_enable, we will go to cdns_mhdp_modeset_retry_fn() and will hit NULL pointer while trying to access the mutex. We need the connector to be set before that. Unlike in legacy !(DBANC) cases, we do not have connector initialised in bridge_attach(). So set the mhdp->connector in atomic_enable() earlier to avoid possible NULL pointer. Fixes: c932ced6b585 ("drm/tidss: Update encoder/bridge chain connect model") Signed-off-by: Jayesh Choudhary <j-choudh...@ti.com> --- .../drm/bridge/cadence/cdns-mhdp8546-core.c | 20 +++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-)It looks like you should reorder your commits: first apply the DBANC fixes, then drop support for !DBANC.
Before dropping !DBANC support, we can't change the connector to pointer cleanly
by cleanly I mean,the driver should be build correctly after applying each commit.
So, if the patches which fixes the bug of NULL pointer de reference due to DBNAC
are applied before dropping the code related to !DBANC code will result in build failure.
Hence the sequencing of commits.
