* Stephen Boyd sboyd@kernel.org [221011 00:47]:
Quoting Mark Brown (2022-10-10 04:29:03)
The KernelCI bisection bot bisected a boot failure on the 5.10-rc stable tree on Panda to d86c6447ee250 ("clk: ti: Stop using legacy clkctrl names for omap4") in the v5.10 stable tree. There's a lot of clock related warnings/errors including:
<4>[ 0.000000] WARNING: CPU: 0 PID: 0 at drivers/clk/clk.c:3778 __clk_register+0x464/0x868 <4>[ 0.000000] ti_dt_clocks_register: failed to lookup clock node abe-clkctrl:0008:24, ret=-517 <4>[ 0.416076] omap_hwmod: debugss: cannot _init_clocks <4>[ 0.421447] ------------[ cut here ]------------ <4>[ 0.426513] WARNING: CPU: 0 PID: 1 at arch/arm/mach-omap2/omap_hwmod.c:2371 _init+0x428/0x488
(there's a *lot* of probe deferrals and hwmods that fail to init). The last output from the kernel is:
<3>[ 10.523590] twl6030_uv_to_vsel:OUT OF RANGE! non mapped vsel for 1410000 Vs max 1316660 <6>[ 10.531890] Power Management for TI OMAP4+ devices. <4>[ 10.537048] OMAP4 PM: u-boot >= v2012.07 is required for full PM support <5>[ 10.544555] Loading compiled-in X.509 certificates
I've left the full report from the bot with more information including full logs and a reported-by tag below:
I don't think we want that commit on stable. It depends on a DT change that may not be present. Tony?
Correct, that commit depends on devicetree data being available.
Regards,
Tony