Takashi Iwai wrote: > At Wed, 16 Sep 2009 10:10:26 +0200, > Pavel Hofman wrote: >> Takashi Iwai wrote: >>> At Wed, 16 Sep 2009 09:15:56 +0200, >>> Pavel Hofman wrote: >>>> Takashi Iwai wrote: >>>>> At Wed, 16 Sep 2009 00:39:53 +0200, >>>>> pavel.hofman@xxxxxxxxxxx wrote: >>>>>> From: pavel <pavel@nahore.(none)> >>>>>> >>>>>> * get/set routines for GPIO mask and direction >>>>>> >>>>>> Signed-off-by: Pavel Hofman <pavel.hofman@xxxxxxxxxxx> >>>>> The get_mask callback is still undefined for ice1724... >>>> That is weird. I split the original commit using interactive git rebase >>>> + edit + add interactive into three new commits - get-mask, >>>> pro-rate-locked and all the external-clock. I commited all of them, >>>> rechecked git status before continuing with rebase - nothing uncommitted >>>> was left, no outstanding changes. But the get_mask commit did not show >>>> up in the subsequent interactive rebase list of commits. I was looking >>>> for it, I thought I found it in the main gpio-methods commit but >>>> overlooked that it is for ice1712.c only. >>>> >>>> The method and its callback has disappeared from my master after the >>>> rebase. Do you have any idea what could have gone wrong? >>> No idea... >>> But rebasing is a bit dangerous operation, and I often forget to finish >>> the rebase until the last commit, getting incomplete rebase in the >>> end. So, I usually copy the original branch first before rebase >>> explicitly, and check the difference from the original branch after >>> rebasing. >> Thanks for the info, I will do that next time. >> >>>> Yesterday I moved it from ice1712.h to quartet.h. The build succeeded >>>> but when loading the new snd-ice1724 module I got dmesg messages about >>>> unknown references to ak4113 methods used in the quartet code. That is >>>> why I left the include in ice1712.h which works fine. >>>> >>>> Unfortunately my knowledge of the build/load infrastructure is very >>>> limited, I can do just copy/paste in this area :) >> Please, how should I change the include to fix the module load? > > The inclusion of ak4113.h should be irrelevant with the build process > and the module dependency. > Is snd-ak4114.ko properly built and installed? > Complete gitcompile produced no errors when loading the module where ak4113.h was included directly in quartet.c. Thanks, Pavel. _______________________________________________ Alsa-devel mailing list Alsa-devel@xxxxxxxxxxxxxxxx http://mailman.alsa-project.org/mailman/listinfo/alsa-devel