11:58 <forkbomb> is there a reason exynos4412 has two nodes for the PMU? there's one under the fimc-is node, and the one inherited from exynos4.dtsi

12:04 <LiquidAcid> forkbomb, i don't see any fimc_is node in exynos4.dtsi

12:05 <forkbomb> yeah, the fimc-is node is in exynos4412.dtsi

12:05 <forkbomb> https://github.com/torvalds/linux/blob/master/arch/arm/boot/dts/exynos4412.dtsi#L314

12:06 <forkbomb> https://github.com/torvalds/linux/blob/master/arch/arm/boot/dts/exynos4.dtsi#L179

12:12 <LiquidAcid> forkbomb, i guess the driver writes to the pmu reg region directly

12:13 <LiquidAcid> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/media/platform/exynos4-is/fimc-is.c#n821

12:13 <LiquidAcid> afaik fimc clock and pmu setup is a bit complicated

12:13 <LiquidAcid> https://www.spinics.net/lists/linux-samsung-soc/msg60900.html

12:14 <LiquidAcid> you should still be able to hang the soc by printing the clock summary on mainline

12:17 <forkbomb> nice

12:18 <LiquidAcid> killswitch :D

12:20 <forkbomb> yep, that works :P

12:23 <LiquidAcid> good to know :D

12:29 <LiquidAcid> forkbomb, fimc_is_cpu_set_power() is using the reg directly

12:30 <LiquidAcid> not pretty, but i guess the fimc doesn't work without the a5

12:35 <forkbomb> yeah

12:37 <forkbomb> surely that could be refactored into a proper api though