Re: [PATCH v5] serial: 8250: fix use-after-free in IRQ chain handling

[Jing Wu]

From: Qiliang Yuan <realwujing@xxxxxxxxx>

On Thu, Jun 25, 2026 at 06:38:35AM +0200, Jiri Slaby wrote:
> I don't follow, the function now looks like this:
> ...
> What am I missing?

You're right, nothing is missing.  The function releases hash_mutex at
every return path, which is exactly what guard(mutex) provides.  And
i->lock is only held across the list manipulation, before request_irq(),
so scoped_guard(spinlock_irq, &i->lock) also fits.

I'll send a v6 keeping the guard style.

Thanks,
Qiliang