On Thu, Mar 17, 2022 at 10:04:13PM +0200, Ville Syrjälä wrote: > On Thu, Mar 17, 2022 at 09:10:37PM +0200, Jani Nikula wrote: > > On Thu, 17 Mar 2022, Ville Syrjala <ville.syrjala@xxxxxxxxxxxxxxx> wrote: > > > From: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx> > > > > > > Currently get_lvds_fp_timing() still returns a pointer to the original > > > data block rather than our copy. Let's convert the data pointer offsets > > > to be relative to the data block rather than the whole BDB. With that > > > we can make get_lvds_fp_timing() return a pointer to the copy. > > > > Ugh, so just as I R-b'd the previous patch... I realize it's all broken > > without this, right? It does pointer arithmetics between bdb header and > > the allocated bdb for ptrs? > > > > Do we want a broken step? > > Probably not. Somehow I convinced myself that the artihmetic was being > done between the original block and bdb header. So now I need to figure > out how to get out of this mess, I guess ideally without having to just > squash the two patches together... I guess I could suck block_offset() from this patch into the previous patch and call it temporarily straight from get_lvds_fp_timing(). -- Ville Syrjälä Intel
