Bug 15662 - Double brightness key event on Dell Vostro 1320
Summary: Double brightness key event on Dell Vostro 1320
Status: REJECTED WILL_NOT_FIX
Alias: None
Product: ACPI
Classification: Unclassified
Component: Power-Video (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: Zhang Rui
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-03-31 20:59 UTC by Felix Geyer
Modified: 2010-06-09 06:34 UTC (History)
2 users (show)

See Also:
Kernel Version: 2.6.32.10
Subsystem:
Regression: No
Bisected commit-id:


Attachments
Output of: grep . /sys/class/input/input*/* (7.40 KB, text/plain)
2010-03-31 21:04 UTC, Felix Geyer
Details
read-input-event output on /dev/input/event5 (215 bytes, text/plain)
2010-03-31 21:07 UTC, Felix Geyer
Details
read-input-event output on /dev/input/event6 (172 bytes, text/plain)
2010-03-31 21:08 UTC, Felix Geyer
Details
Output of acpidump (37.74 KB, application/gzipped-patch)
2010-04-01 16:43 UTC, Felix Geyer
Details

Description Felix Geyer 2010-03-31 20:59:53 UTC
The brightness keys don't work with KMS.
With the patch from https://bugzilla.kernel.org/show_bug.cgi?id=15054#c23 the keys work again but the brightness key events are emitted twice.
Comment 1 Felix Geyer 2010-03-31 21:04:09 UTC
Created attachment 25780 [details]
Output of: grep . /sys/class/input/input*/*
Comment 2 Felix Geyer 2010-03-31 21:07:11 UTC
Created attachment 25781 [details]
read-input-event output on /dev/input/event5

as requested in https://bugzilla.kernel.org/show_bug.cgi?id=15054#c32
Comment 3 Felix Geyer 2010-03-31 21:08:09 UTC
Created attachment 25782 [details]
read-input-event output on /dev/input/event6
Comment 4 Matthew Garrett 2010-04-01 14:12:18 UTC
Dell are sending the event via both the keyboard controller and ACPI. Can you attach the output of the acpidump command?
Comment 5 Felix Geyer 2010-04-01 16:43:49 UTC
Created attachment 25801 [details]
Output of acpidump
Comment 6 Felix Geyer 2010-04-19 09:24:12 UTC
Is there any more debug data you need?
Comment 7 Zhang Rui 2010-05-11 07:46:13 UTC
The hotkey works well in KMS mode, right?
                    Method (_Q11, 0, NotSerialized)
                    {
                        Store (0x87, P80H)
                        If (IGDS)
                        {
                            ^^^GFX0.BRTN (0x87)
                        }
                        Else
                        {
                            Notify (^^^P0P2.VGA.LCD, 0x87)
                        }
                    }

notification is sent to ACPI video device only if IGDS is not set by BIOS, which seems to be true on this laptop.
As IGDS stands for the IGD Opregion stuff which is supported by KMS only, I'm wondering why you want KMS disabled.
Comment 8 Felix Geyer 2010-05-11 08:56:28 UTC
No, it doesn't work well with KMS.

With kernel 2.6.34 rc7 it doesn't matter if KMS is enabled or not, the brightness key events are emitted twice.
Comment 9 Zhang Rui 2010-05-20 07:27:23 UTC
I see, BTRN also sends notifications to the ACPI video device.
So, IMO, the brightness hotkey control via ACPI works pretty well, except that there is an extra interrupt to the keyboard.

sounds like a firmware problem to me.
matthew, what do you think?
Comment 10 Zhang Rui 2010-06-09 06:34:06 UTC
bug closed as this is a firmware bug that we can not fix...
maybe you need some tricks to eliminate a hotkey event in user space.

Note You need to log in before you can comment on or make changes to this bug.