Re: [9front] MB Pro early-2011 support status

[Nick Owens <mischief@offblast.org>]

On Sun, Apr 18, 2021 at 10:42 PM <unobe@cpan.org> wrote:
>
> I have written a description about what works and does not work on an
> early-2011 macbook pro.  The post is @
> https://blog.fallglow.com/2021/04-13/blog
>
> I have a few patches that have made it more reasonable to use 9front
> on a macbook from the get-go (most notably the new keymap and pc64mb
> conf file in the pc64 kernel directory) and have setup up a new branch
> to keep track of the changes.  It's not ready for merging to default
> but my hope is to get multi-touch working along with some of the other
> hardware (audio, camera, monitor blanking).  The branch is now hosted
> at https://fallglow.com/p9/mb : my plan is to track default and merge
> default into this branch every so often.
>
> One of my next steps that I'd like to do is to try to resolve is the
> lack of monitor blanking.  The current status of Sandy Bridge mentions
> it being unfinished (changeset 4713:3e7542933175 ).  There's code in
> the igfx for sandy bridge to blank, but it's not working for me when I
> set 'echo hwblank on >/dev/vgactl' and do 'echo blank >/dev/mousectl'.
> So I figure there's something that isn't implemented yet.  This is a
> new area of programming for me so I figure looking at the 9front code,
> the corresponding drivers for OpenBSD and Linux, and the programmer's
> manuals is what I should do.  Cinap--since you committed changeset
> 4713, do you recall anything I should be looking at for this?  (Or
> anyone else who is still reading this long, drawn-out message?) I am
> currently wading through the 2011 Intel Core Processor Family docs
> found at
> https://01.org/linuxgraphics/documentation/hardware-specification-prms
>

i added blanking for some cards in this commit:

changeset:   6686:2f61fba90cd1
user:        mischief <mischief@offblast.org>
date:        Tue Aug 21 11:31:45 2018 -0700
summary:     igfx: support blanking by turning off panel power and backlight

it should (in theory) work for your chip. you have the 0x0126 did
(from your sysinfo):

        case 0x0126:    /* SNB */
        case 0x0166:    /* IVB */
                off = 0xC7204;
                break;
        }

        /* toggle PP_CONTROL backlight & power state */
        if(blank)
                scr->mmio[off/4] &= ~0x5;
        else
                scr->mmio[off/4] |= 0x5;

you can also see the value of PP_CONTROL in the aux/vga -p output in
your sysinfo. maybe double check the programming manual for the
PP_CONTROL register location and value to see if those bits are
correct.

IIRC this code worked for me and others on thinkpads, but it's
possible that this register isn't hooked up in this hardware.