Port-arm archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Quartz64 firmware



On Mon, May 08, 2023 at 07:50:33PM +0100, Robert Swindells wrote:
> The times that it did work, at first NetBSD wouldn't detect the SSD, just
> ahcisata.
> I copied the wd config line from amd64 and added a "flags 0x0000" option,
> NetBSD
> then detected the drive and booted to multiuser.
> 
>   wd* at atabus? drive ? flags 0x0000
> 
> Maybe something similar to the "flags 0x0000" option needs to be added to
> the firmware to get it to use any device connected to SATA.

I don't understand. The config line in GENERIC64 is

	wd* at atabus? drive ?

and that translates to:

struct cfdata cfdata[] = {
    /* driver           attachment    unit state      loc   flags  pspec */
[..]
/*188: wd* at atabus? drive -1 */
    { "wd",             "wd",            0, STAR, loc+538,      0, &pspec33 },  

so (unsuprisingly) flags = 0 is the default. The wd(4) driver documents
the meaning of flags as individual disable bits, ending in:

	0x0000 means "use whatever the drive claims to support."

So theoretically both kernels should have behaved totally identical.

I'm really interested in understanding this, as my drive is not detected
as well (neither by the firmware nor by NetBSD). Only advise I got from
the pine chat was: use a SATA cable as short as possible, signals are not
really good with that hardware. So I replaced my standard cable with a
10 cm one (which did not fix the detection problem).

Martin


Home | Main Index | Thread Index | Old Index