Johannes Bauer schrieb: >> Any pointer could help me to add some support about it into >> scan-win-drivers.pl (I hope that I won't have to mimic the Windows PnP's >> job in perl ;-) > > OK, I'm fiddling with it right now to build a kind of lsbus - I've not > read anywhere such a program already exists.
So, I'm getting somewhere. libpci is pretty neat - however I think the problem is far too complex to solve it without kernel support: What we want is the HDA codec device/vendor ids (struct hda_codec*) Those are capsuled in HDA busses (struct hda_bus*) Those busses rely on a pci bus (struct pci_dev*) and provice HDA bus operations (struct hda_bus_ops*) Thsoe hda_bus_ops are the problem: They define the interface how to speak to a PCI device on the HDA level (send command, wait for response). The problem is: Access is not uniform. It varies for each and every device, sometimes even for subtypes. All those are specified in the patch_*.c files, which are 32 KLOC alone. What does this mean? We're screwed, pretty much. Unless the kernel provides some easy was of accessing the data. Which it does :-) I propose this: Import the HDA drivers into the initrd. Then do a scan of /proc/asound/card[0-9][0-9]? which lists the information we're looking for nicely: Codec: Realtek ALC888 Address: 0 Vendor Id: 0x10ec0888 Subsystem Id: 0x14627519 Revision Id: 0x100001 [...] (this is another PC, don't worry that the IDs are different than in my other posting). Kind regards, Johannes ------------------------------------------------------------------------------ This SF.net email is sponsored by: SourcForge Community SourceForge wants to tell your story. http://p.sf.net/sfu/sf-spreadtheword _______________________________________________ unattended-devel mailing list unattended-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/unattended-devel
