Re: [PATCH] timekeeping: Register default clocksource before taking tk_core.lock

[Mikhail Gavrilov]

On Tue, Jun 16, 2026 at 1:42 PM Breno Leitao <leitao@xxxxxxxxxx> wrote:
> I am seeing a similar issue, although from a different code path.
>
>          =============================
>          [ BUG: Invalid wait context ]
>          7.1.0-next-20260615upstream-14023-g8d6dbbbe3ba6 #32 Not tainted
>          -----------------------------
>          swapper/0/0 is trying to lock:
>          ffff800084e4bd80 (clocksource_mutex){....}-{4:4}, at: __clocksource_register_scale+0x5c4/0xa48
>          other info that might help us debug this:
>          context-{5:5}
>          1 lock held by swapper/0/0:
>           #0: ffff80008ca99930 (&tkd->lock){....}-{2:2}, at: timekeeping_init+0x158/0x290
>          stack backtrace:
>          CPU: 0 UID: 0 PID: 0 Comm: swapper/0 Not tainted 7.1.0-next-20260615upstream-14023-g8d6dbbbe3ba6 #32 PREEMPT(full)
>          Hardware name: linux,dummy-virt (DT)
>          Call trace:
>           show_stack+0x24/0x38 (C)
>           __dump_stack+0x28/0x38
>           dump_stack_lvl+0x94/0xe0
>           dump_stack+0x18/0x24
>           __lock_acquire+0xba4/0x3520
>           lock_acquire+0x150/0x390
>           __mutex_lock+0xe4/0x1240
>           mutex_lock_nested+0x30/0x48
>           __clocksource_register_scale+0x5c4/0xa48
>           clocksource_default_clock+0x40/0x68
>           timekeeping_init+0x164/0x290
>           start_kernel+0x2b0/0x4b8
>           __primary_switched+0x8c/0xa0
>
>
> Reviwed-by: Breno Leitao <leitao@xxxxxxxxxx>

Thanks for the Reviewed-by and the arm64 confirmation. Note that
f24df84cbe05, which this fixes, is itself Cc: stable, so this should
land in the same stable trees. I'm happy to resend with Cc:
stable@xxxxxxxxxxxxxxx added if you prefer; otherwise it can be picked
up via the Fixes: tag or added on apply.


-- 
Best Regards,
Mike Gavrilov.