On Wed, Aug 27, 2014 at 08:51:35AM +0100, Damien Lespiau wrote: > On Wed, Aug 27, 2014 at 08:47:54AM +0100, Damien Lespiau wrote: > > > An alternative would be to provide a second set of defines for eDP 1.4 > > > where the name implies the meaning and then use them as appropriate. > > > > We went through the idea as well and: > > > > I actually think the nominal voltage swing and pre-emph values are quite > > misleading. The hw is free to implement a wildly different set of voltage > > swing/pre-emph values. > > > > eDP 1.4 changes those nominal values as described in the cover letter, > > but there again, the actual hw implementation can choose fairly > > different values than the nominal ones. > > > > Also, the DP 1.2 spec documents this field as (see address 103h): > > > > TRAINING_LANE0_SET : Link Training Control_Lane0 > > Bits 1:0 = VOLTAGE SWING SET > > 00 –Voltage swing level 0 > > 01 –Voltage swing level 1 > > 10 –Voltage swing level 2 > > 11 –Voltage swing level 3 > > > > So, in that sense, we're closer to the latest spec with those LEVEL_X > > defines. > > I forgot to mention here that if we have separate defines for eDP 1.4, > then we lose the possibility to share training code with big DP and eDP > 1.3, not something desirable. Yeah, I'd like to see the training sequences extracted into common helpers. Thierry
Attachment:
pgpmrV0BO99Rn.pgp
Description: PGP signature
_______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/dri-devel
