Many Thanks. ^^ -----Original Message----- From: Takashi Iwai <tiwai@xxxxxxx> Sent: Tuesday, January 29, 2019 11:37 PM To: Kailang <kailang@xxxxxxxxxxx> Cc: (alsa-devel@xxxxxxxxxxxxxxxx) <alsa-devel@xxxxxxxxxxxxxxxx> Subject: Re: hp_pin was NULL value On Tue, 29 Jan 2019 14:22:45 +0100, Takashi Iwai wrote: > > On Tue, 29 Jan 2019 09:39:56 +0100, > Kailang wrote: > > > > Hi Takashi, > > > > So I think put it under alc269 parser. Maybe it is the quickly method. > > > > err = alc269_parse_auto_config(codec); > > if (err < 0) > > goto error; > > + ..... > > + ..... > > Not really... The init sequence needs to be applied in two different > places: once in the init callback, and another in the resume callback > but only for the hibernation restore. > > The patches below are applied on top of yours, and this should make > things working. > > The first one lets the HD-audio core recording the currently processed > PM event, and the second one evaluates it and applies the missing init > sequence also for the hibernation resume. > > This isn't quite sexy, but it has the minimal change in the codec > driver side. If this requirement is more common, we can think of > splitting / reorganizing the codec callbacks to be more directly > called from the device pm ops. I did quick tests and the test result looks positive, at least, about the PM power_state change tracking. So I'm going to push your fix to for-linus for 5.0, while other two to for-next, for 5.1, as the S4 resume issue isn't any urgent bug and it's no regression, either. thanks, Takashi ------Please consider the environment before printing this e-mail. _______________________________________________ Alsa-devel mailing list Alsa-devel@xxxxxxxxxxxxxxxx http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
