Re: intel drmkms

[Robert Swindells <rjs%fdy2.co.uk@localhost>]

Taylor R Campbell wrote:
>   Date: Sat, 20 Sep 2014 16:20:25 +0100 (BST)
>   From: Robert Swindells <rjs%fdy2.co.uk@localhost>
>
>   The gmbus code does work when talking to the vga connector though.
>
>   I was wondering if it would work fully if the display probe was delayed
>   until interrupts are enabled.
>
>Unlikely -- when cold, the gmbus code uses polling rather than
>interrupts.

My thought was just that the timeout could be wrong when polling.

I don't know whether there really is anything for i2c to talk to.

>   It looks to me as if the driver needs to call the acpi _DDC method to
>   get the EDID if it can't map the ROM.
>
>Could be.  All of the ACPI code from upstream is currently detached.
>I will try to find a moment to take a look.  Can you file a PR and
>assign it to me?

There isn't anything in the upstream code to do this, you haven't left
out any features that would help.

The logic of intel_lvds_init() doesn't seem to me to match the ACPI
spec, the example of a laptop without a separate video BIOS ROM is
described as a valid configuration.

The strange bit is that it works with an older kernel, was there
another way of getting at the ROM contents without using the PCI
configuration ?