4302208388
Not all interrupt sources that affect CIS bit were acknowledged. Specifically, bits in STATESTS (aka WAKESTS) were left set. The fix is to disable WAKEEN and clear STATESTS bits before the HDA interrupt is enabled. This way we should never get any STATESTS bits. I also added placeholders for all event bits that we currently do not enable, do not handle and do not clear. This might get useful when / if we enable any of them. Reported by: kib (Apollo Lake hardware) Tested by: kib (earlier, different change) MFC after: 2 weeks X-MFC with: r362294 |
||
|---|---|---|
| .. | ||
| hda | ||
| allegro_code.h | ||
| allegro_reg.h | ||
| als4000.c | ||
| als4000.h | ||
| atiixp.c | ||
| atiixp.h | ||
| aureal.c | ||
| aureal.h | ||
| cmi.c | ||
| cmireg.h | ||
| cs461x_dsp.h | ||
| cs4281.c | ||
| cs4281.h | ||
| csa.c | ||
| csamidi.c | ||
| csapcm.c | ||
| csareg.h | ||
| csavar.h | ||
| ds1-fw.h | ||
| ds1.c | ||
| ds1.h | ||
| emu10k1.c | ||
| emu10kx-midi.c | ||
| emu10kx-pcm.c | ||
| emu10kx.c | ||
| emu10kx.h | ||
| emuxkireg.h | ||
| envy24.c | ||
| envy24.h | ||
| envy24ht.c | ||
| envy24ht.h | ||
| es137x.c | ||
| es137x.h | ||
| fm801.c | ||
| hdspe-pcm.c | ||
| hdspe.c | ||
| hdspe.h | ||
| ich.c | ||
| ich.h | ||
| maestro3.c | ||
| maestro_reg.h | ||
| maestro.c | ||
| neomagic-coeff.h | ||
| neomagic.c | ||
| neomagic.h | ||
| solo.c | ||
| spicds.c | ||
| spicds.h | ||
| t4dwave.c | ||
| t4dwave.h | ||
| via82c686.c | ||
| via82c686.h | ||
| via8233.c | ||
| via8233.h | ||
| vibes.c | ||
| vibes.h | ||