On 8/28/19 3:59 PM, Randy Dunlap wrote: > On 8/28/19 3:45 PM, Pierre-Louis Bossart wrote: >> >>>>> I just checked with Mark Brown's for-next tree 8aceffa09b4b9867153bfe0ff6f40517240cee12 >>>>> and things are fine in i386 mode, see below. >>>>> >>>>> next-20190828 also works fine for me in i386 mode. >>>>> >>>>> if you can point me to a tree and configuration that don't work I'll look into this, I'd need more info to progress. >>>> >>>> Please try the attached randconfig file. >>>> >>>> Thanks for looking. >>> >>> Ack, I see some errors as well with this config. Likely a missing dependency somewhere, working on this now. >> >> My bad, I added a fallback with static inline functions in the .h file when ACPI is not defined, but the .c file was still compiled. >> >> The diff below makes next-20190828 compile with Randy's config. >> >> It looks like the alsa-devel server is down btw? >> >> diff --git a/sound/hda/Makefile b/sound/hda/Makefile >> index 8560f6ef1b19..b3af071ce06b 100644 >> --- a/sound/hda/Makefile >> +++ b/sound/hda/Makefile >> @@ -14,5 +14,7 @@ obj-$(CONFIG_SND_HDA_CORE) += snd-hda-core.o >> #extended hda >> obj-$(CONFIG_SND_HDA_EXT_CORE) += ext/ >> >> +ifdef CONFIG_ACPI >> snd-intel-nhlt-objs := intel-nhlt.o >> obj-$(CONFIG_SND_INTEL_NHLT) += snd-intel-nhlt.o >> +endif >> > > works for me. Thanks. > Acked-by: Randy Dunlap <rdunlap@xxxxxxxxxxxxx> # build-tested > although this Makefile change should not be needed and the dependencies should be handled correctly in Kconfig files. -- ~Randy
