At Tue, 10 Feb 2009 15:31:22 -0200, Herton Ronaldo Krzesinski wrote:
Em Segunda-feira 09 Fevereiro 2009, às 17:27:55, Takashi Iwai escreveu:
At Mon, 9 Feb 2009 17:04:44 -0200, Herton Ronaldo Krzesinski wrote:
Em Segunda-feira 09 Fevereiro 2009, às 14:42:55, Takashi Iwai escreveu:
At Mon, 09 Feb 2009 08:32:58 +0100,
I wrote:
At Sun, 8 Feb 2009 19:50:34 -0200,
Herton Ronaldo Krzesinski wrote:
This adds a new sound quirk entry (model=ecs202) for an ecs
motherboard
with IDT STAC9221 codec (1019:2950).
Signed-off-by: Herton Ronaldo Krzesinski herton@mandriva.com.br
Applied. Thanks.
BTW, I changed the quirk listing and introduced new macros SND_PCI_QUIRK_MASK() and SND_PCI_QUIRK_VENDOR(). The latter is what it was with subdevice = 0, and the former is an entry with bitmask to match. This should simplify the list in future.
The topic/hda branch already includes this change: git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6.git topic/hda
This is nice. I saw now another cleanup that can be done, what about this? (default to STAC_HP_DV5 looks a good thing to be done):
Yes, I originally thought of these HP series. But, 103c:361a is HP mini 1000, and this seems to be working better with hp-m4 model. So, we should be a bit careful about this change.
The alsa-info of that device is found in https://bugzilla.novell.com/show_bug.cgi?id=467381
From my understanding the STAC_HP_DV5 only exists to enable gpio0 by default, while STAC_HP_M4 will besides enabling it (passing the "if (spec->board_config
STAC_92HD71BXX_REF)" check), will also do extra setup.
Right. In addition, hp-m4 sets the pin 0x0e statically as the internal mic. This seems mismatching with some HP dv5, as far as I know.
I don't know if enabling gpio0 with dir=1 by default will introduce some bug with some machine, but the idea is that all HP laptops with these codecs looks to need gpio0 enabled; even if some of them needs hp-m4 because of extra setup needed (besides gpio initialization), hp-dv5 would be safe for them as both enable gpio the same way. But I agree we should take care in case initializing gpio0 by default can introduce bugs with another unlisted hp laptops.
I guess the GPIO0 is OK for all HP laptops. But, the fixed 0x0e pin isn't.
Takashi
thanks,
Takashi
[PATCH] ALSA: hda - Cleanup quirk list for HP dv laptops with
stac92hd71bxx
Cleanup with SND_PCI_QUIRK_MASK entries for model=hp-dv5.
Signed-off-by: Herton Ronaldo Krzesinski herton@mandriva.com.br
sound/pci/hda/patch_sigmatel.c | 12 +++++------- 1 files changed, 5 insertions(+), 7 deletions(-)
diff --git a/sound/pci/hda/patch_sigmatel.c
b/sound/pci/hda/patch_sigmatel.c
index ddb10bb..0f8a27f 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -1812,16 +1812,14 @@ static struct snd_pci_quirk
stac92hd71bxx_cfg_tbl[] = {
"HP dv5", STAC_HP_M4),SND_PCI_QUIRK(PCI_VENDOR_ID_HP, 0x30f4, "HP dv7", STAC_HP_M4),
- SND_PCI_QUIRK(PCI_VENDOR_ID_HP, 0x30f7,
"HP dv4", STAC_HP_DV5),- SND_PCI_QUIRK(PCI_VENDOR_ID_HP, 0x3600,
"HP dv5", STAC_HP_DV5),- SND_PCI_QUIRK(PCI_VENDOR_ID_HP, 0x3603,
"HP dv5", STAC_HP_DV5),"unknown HP", STAC_HP_M4),
"unknown HP", STAC_HP_M4),- SND_PCI_QUIRK_MASK(PCI_VENDOR_ID_HP, 0xff00, 0x3000,
"HP dv laptop series", STAC_HP_DV5),- SND_PCI_QUIRK_MASK(PCI_VENDOR_ID_HP, 0xff00, 0x3600,
"HP dv laptop series", STAC_HP_DV5),-- 1.6.1.3
Takashi
-- []'s Herton