Re: [PATCH v2 2/2] HID: lenovo-legion: Add keyboard backlight driver for Legion Gen 10
From: Ciro Pedrini
Date: Wed Nov 11 2026 - 12:21:11 EST
Hi Rong,
Thanks for the review. I'll fix the rest in v3, but before I go on with
it I would like to hear from you on these.
On Wed, 2026-11-11 at 20:09 +0800, Rong Zhang wrote:
> On Tue, 2026-11-10 at 20:29 +0000, Ciro Pedrini via B4 Relay wrote:
> > Its hardware brightness, 0..9, is stored in the controller and applies to
> > every LED of the machine (keys, front strip, rear vents, lid logo).
>
> I don't know if it's appropriate to represent so many things with the
> name "legion::kbd_backlight".
I gave splitting it some thought, and looked (with the help of an LLM)
at how other drivers handle this. We could register one LED per zone,
but that seems to be usually rejected, because they would not be
independent: the controller has a single brightness for all zones, so
writing any of them would change all of them.
Per-zone control is only possible through the LampArray (c193). It has
319 lamps, each with its own colour, and they are grouped by zone (keys,
front strip, rear vents, lid logo), so a LampArray driver could light
or dim each zone on its own. That needs c197 to hand the LEDs over
first, as discussed in the other thread.
Also, going with kbd_backlight seems like a good idea so that UPower
and the desktops pick it up automatically.
Thoughts? I'd like a confirmation from someone who knows how this is
usually handled in Linux, as I don't have much experience with it.
> Just curious, can the software select which effect to use, or listen to
> an event when Fn+Space is pressed?
Yes. 0xCA reads the current effect and 0xC8 sets it. On Fn+Space the
controller sends report 5 as 05 01 P (P = the new effect), then
05 02 B (B = the brightness).
> I wonder, what if we set the report with a short report size, say,
> struct_size(msg, params, len)? Could the device handle it properly in
> this case? If so, there is a chance that we don't need to sleep for 100ms
> later (I guess the device is processing too much data, causing the
> massive delay).
You were right. I tried it on the laptop: it takes the short request,
and it only ever answers 5 bytes anyway, even when asked for 960. The
answer is there right away, I read it with delays from 100 ms down to 0
and it was always correct. So a brightness read goes from ~115 ms to
~2 ms.
Setting the brightness with a short request works too, and reading it
right after returns the new value.
In v3 I'll send and read short reports and drop the msleep().
> Ditto, what if we get the report with struct_size(msg, params, 1)?
Same, that works.
Thanks,
Ciro