Unset CONFIG_ADB_PMU_LED_DISK for ppc.
Why making this change? There is a LED driver that is using this config: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/macintosh/via-pmu-led.c It is exposed as /sys/class/leds/pmu-led:front. If CONFIG_ADB_PMU_LED_DISK is yes then the value of trigger will be "disk-activity" otherwise it will be "none" (the value can be overwritten by writing to /sys/class/leds/pmu-led:front/trigger after boot). I think "none" is a better default because the LED is not used as a disk activity indicator in Mac OS, rather it would pulse slowly when the notebook is in sleep mode. The LED is a bit too bright as a disk activity indicator, someone made a video showing this: https://youtu.be/cbmnuUc-G6c?t=55
For reference, Ubuntu, which used to ship a PowerPC version, is not enabling CONFIG_ADB_PMU_LED_DISK (formerly CONFIG_ADB_PMU_LED_IDE).
Should we change ppc64 as well? I don't know at this point. Note that in the LED driver code it appears that only PowerMac7,2 and PowerMac7,3 are allowed to drive the LED.
Merge request reports
Activity
@awilfox can you comment on please:
Should we change ppc64 as well? I don't know at this point. Note that in the LED driver code it appears that only PowerMac7,2 and PowerMac7,3 are allowed to drive the LED.
7,2 and 7,3 are G5 systems. I know that this LED configuration affects at least the iBook G3 and PowerBook G4, so it doesn't make sense that only the 7,2 and 7,3 can drive it, unless that was in reference to 64-bit systems.
The configuration should be consistent so I would tend to agree that if we change
ppc
we should changeppc64
as well. I don't have a hard preference on the configuration knob itself. We turned it on because it was the default in Gentoo and offered familiarity for people migrating off of Gentoo, but I don't think there are a great many people who are doing such things these days. Personally I find it vaguely annoying, since the LED on the PowerBook G4 is intensely bright, so I would tend to prefer=n
as well.mentioned in commit 1feea815
There is some LKML discussion on the matter here: https://lkml.org/lkml/2021/10/25/1997.