Hi, On 05/14/2014 02:07 PM, Rafael J. Wysocki wrote: > On Wednesday, May 14, 2014 01:09:33 PM Hans de Goede wrote: >> Hi, >> >> On 05/12/2014 01:02 PM, Hans de Goede wrote: >>> Hi, >>> >>> On 05/12/2014 10:16 AM, Aaron Lu wrote: >>>> On 05/12/2014 03:57 PM, Hans de Goede wrote: >>>>> Like all of the other *30 ThinkPad models, the W530 has a broken acpi-video >>>>> backlight control. Note in order for this to actually fix things on the >>>>> ThinkPad W530 the commit titled: >>>>> "nouveau: Don't check acpi_video_backlight_support() before registering backlight" >>>>> is also needed. >>>> >>>> Note that the backlight_device_registered(raw) may return false as when >>>> acpi_video_verify_backlight_support is called, the nouveau driver may >>>> not run yet. >>>> >>>> Previously, we don't know anything about how laptops with nvidia graphics >>>> card alone control backlight in Win8, so the existing solution doesn't >>>> consider this case. If nvidia graphics system also should favour native >>>> backlight control interface in Win8, the current solution needs >>>> modifications. >>> >>> Hmm, how is this dealt with in the case of the intel gfx driver ? >> >> Ok, I've figured out now how this is dealt with in the case of the intel gfx >> drivers. That looks like something which will likely be hard to do for >> nouveau, since it relies on some intel gfx specific ACPI calls being there, >> which nouveau does not have. So this would require doing something like >> duplicating the nouveau pci-ids or some such, which would be far from ideal. >> >> Still this bug: https://bugzilla.redhat.com/show_bug.cgi?id=1093171 >> >> Shows that we need *working* video.use_native_brightness=1 support for >> non intel too. Which the current loading order issues caused by the >> backlight_device_registered(raw) breaks. >> >> So maybe we should simply drop the backlight_device_registered(raw) check? >> I don't think we really need it any win8 supporting laptop will have >> intel, nv or ati gfx, and all 3 of them have native backlight support, >> which if our experiences so far is any indication we will likely want >> to use instead of acpi_video. > > We can try that, but if it turns out to break things for someone, we'll need > a different approach. Ok, I'll write a patch for this tomorrow. As for trying a different approach, whatever the different approach may be, I strongly believe that we should get rid of the various module loading order dependencies we've now, not add more. I also have a plan now on how to cleanup the code for deciding which backlight control interface to use, and remove the various ordering issues we've there. But I've run out of steam for today, so I'll go and write that up and post it tomorrow. Regards, Hans -- To unsubscribe from this list: send the line "unsubscribe stable" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html