[patch] unhide the flash chip on QDI KD7G

[Idwer Vollering]

Only works with "Flash Write Protect" set to "Disabled", in the BIOS menu.

With the attached patch write/erase work without the need of extra code.

Signed-off-by: Idwer Vollering

[Carl-Daniel Hailfinger]

On 14.05.2010 23:58, Idwer Vollering wrote:
> Only works with "Flash Write Protect" set to "Disabled", in the BIOS menu.
>
> With the attached patch write/erase work without the need of extra code.
>
> Signed-off-by: Idwer Vollering
> {0x1106, 0x3116, 0x1106, 0x3116,  0x1106, 0x3177, 0x1106, 0x3177, "^KM266-8235$", NULL,         NULL,          "QDI",         "KD7G",                  0,   OK, board_qdi_kd7g},

The big problem here is that DMI is ambiguous and PCI subsystem IDs are
ambiguous. Due to this, we can't merge the patch as is. Besides that, I
think you left out the w836xx_memw_enable change which was needed on
your machine according to our tests.

Michael knows the board enable table well, he probably has suggestions
on what to do.

The good news is that once we have generic superio probing, this board
enable can be killed completely.

Regards,
Carl-Daniel

[Michael Karcher]

Am Samstag, den 15.05.2010, 00:28 +0200 schrieb Carl-Daniel Hailfinger:
> On 14.05.2010 23:58, Idwer Vollering wrote:
> > {0x1106, 0x3116, 0x1106, 0x3116,  0x1106, 0x3177, 0x1106, 0x3177, "^KM266-8235$", NULL,         NULL,          "QDI",         "KD7G",                  0,   OK, board_qdi_kd7g},
>
> The big problem here is that DMI is ambiguous and PCI subsystem IDs are
> ambiguous. Due to this, we can't merge the patch as is.

Working hack: Write an enable function that checks for the presence of
the superio before poking around in 0x2e space.
> Michael knows the board enable table well, he probably has suggestions
> on what to do.

See suggestion above, but...
> The good news is that once we have generic superio probing, this board
> enable can be killed completely.

... can't we wait for this? I think a bit superio probing already is in.
Why not add the Winbond 836xx stuff now?

Finally: We don't need a special function board_qdi_kd7g, which is a
complete clone of w836xx_memw_enable_2e.

Regards,
  Michael Karcher