[alsa-devel] intel-hda patch for Dell Latitude D531
Hi
I have attached three files to this mail:
1. patch_sigmatel.c.diff
I have a Latitude D531 which has a ATI SB600 with a STAC 9205. The chip is detected, but no config was read with the latest drivers. The problems are the dmic_nids arrays, which are ignored for auto config lookup. This calles is_in_nid_list, which expects a 0 terminated list, but these arrays are not.
Additionally I updated the dell_9205_m43_pin_configs with values from the latest win drivers.
2. create_dell_tables.pl
When looking into the ini files and comparing the nids and config values of the various tables, I got really confused, so I wrote a litte script to extract the values and build the table. This is the extended version, after I got my sound working. It extracts all pin infos from all ini files, generates the apropriate tables and prepends the Version.txt from the driver, resultiung in a kind of C header. Maybe it's usefull for you.
3. patch_sigmatel_dell.h
Output from running create_dell_tables.pl on the latest drivers for my D531 from Dell.
HTH
Jan-Marek Glogowski
---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.
At Thu, 11 Oct 2007 18:54:56 +0200, glogow@fbihome.de wrote:
Hi
I have attached three files to this mail:
- patch_sigmatel.c.diff
I have a Latitude D531 which has a ATI SB600 with a STAC 9205. The chip is detected, but no config was read with the latest drivers. The problems are the dmic_nids arrays, which are ignored for auto config lookup. This calles is_in_nid_list, which expects a 0 terminated list, but these arrays are not.
Oh thanks, I queued the fix on to my local tree. I'll apply it soon after releasing 1.0.15.
Additionally I updated the dell_9205_m43_pin_configs with values from the latest win drivers.
Hm, I'm not pretty about this. Does this change work for your machine well (I suppose you have a Dell Precision).
- create_dell_tables.pl
When looking into the ini files and comparing the nids and config values of the various tables, I got really confused, so I wrote a litte script to extract the values and build the table. This is the extended version, after I got my sound working. It extracts all pin infos from all ini files, generates the apropriate tables and prepends the Version.txt from the driver, resultiung in a kind of C header. Maybe it's usefull for you.
- patch_sigmatel_dell.h
Output from running create_dell_tables.pl on the latest drivers for my D531 from Dell.
That'll certainly help for debugging although we may need to tidy up a bit.
Thanks!
Takashi
- patch_sigmatel.c.diff
I have a Latitude D531 which has a ATI SB600 with a STAC 9205. The chip is detected, but no config was read with the latest drivers. The problems are the dmic_nids arrays, which are ignored for auto config lookup. This calles is_in_nid_list, which expects a 0 terminated list, but these arrays are not.
Oh thanks, I queued the fix on to my local tree. I'll apply it soon after releasing 1.0.15.
Any reason, why the array fix can't go into 1.0.15?
Additionally I updated the dell_9205_m43_pin_configs with values from the latest win drivers.
Hm, I'm not pretty about this. Does this change work for your machine well (I suppose you have a Dell Precision).
Should I try to get sound running with the old config? The real probem with the driver where the unterminated arrays, which I found after adapting the pin configs. I have no idea, if the driver can work with the old values. Is there real chance, even if the driver is working with my adapted values?
I have no knowledge about all the Dell product lines, but mine is a Dell Latitude D531 and is AMD + ATI based (it's a model from the german municipal administration page of Dell).
- patch_sigmatel_dell.h
Output from running create_dell_tables.pl on the latest drivers for my D531 from Dell.
That'll certainly help for debugging although we may need to tidy up a bit.
The perl script is GPL so just help yourself. The script tries to identify duplicateds and merge these into one value. I know there are also some GPIO information in the INI files of the driver.
BTW - the windows driver identifies the Dell notebook by the complete PCI ID of of the sound device, while the linux driver just uses the SUB IDs.
Jan-Marek
At Tue, 16 Oct 2007 09:38:03 +0200 (CEST), Jan-Marek Glogowski wrote:
- patch_sigmatel.c.diff
I have a Latitude D531 which has a ATI SB600 with a STAC 9205. The chip is detected, but no config was read with the latest drivers. The problems are the dmic_nids arrays, which are ignored for auto config lookup. This calles is_in_nid_list, which expects a 0 terminated list, but these arrays are not.
Oh thanks, I queued the fix on to my local tree. I'll apply it soon after releasing 1.0.15.
Any reason, why the array fix can't go into 1.0.15?
Because it's code-frozen right now.
Additionally I updated the dell_9205_m43_pin_configs with values from the latest win drivers.
Hm, I'm not pretty about this. Does this change work for your machine well (I suppose you have a Dell Precision).
Should I try to get sound running with the old config? The real probem with the driver where the unterminated arrays, which I found after adapting the pin configs. I have no idea, if the driver can work with the old values. Is there real chance, even if the driver is working with my adapted values?
Just try to remove the change for pin configs from your patch, and check whether it still works.
I have no knowledge about all the Dell product lines, but mine is a Dell Latitude D531 and is AMD + ATI based (it's a model from the german municipal administration page of Dell).
Interesting. In stac9205_cfg_tbl[], only Dell Precision is listed, so far.
- patch_sigmatel_dell.h
Output from running create_dell_tables.pl on the latest drivers for my D531 from Dell.
That'll certainly help for debugging although we may need to tidy up a bit.
The perl script is GPL so just help yourself. The script tries to identify duplicateds and merge these into one value. I know there are also some GPIO information in the INI files of the driver.
BTW - the windows driver identifies the Dell notebook by the complete PCI ID of of the sound device, while the linux driver just uses the SUB IDs.
PCI IDs are already used for the identifier of the driver itself. Only missing is the check of PCI revision ID.
Takashi
participants (3)
-
glogow＠fbihome.de -
Jan-Marek Glogowski
-
Takashi Iwai