linux-kernel - Re: [PATCH] genirq: Deduplicate WARN_ON_ONCE() in generic_handle_domain

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <YoH29u6kNTXF/0Ew@linutronix.de>
Date:   Mon, 16 May 2022 09:02:14 +0200
From:   Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To:     Thomas Gleixner <tglx@...utronix.de>
Cc:     Lukas Wunner <lukas@...ner.de>, Marc Zyngier <maz@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        Jakub Kicinski <kuba@...nel.org>,
        Linus Walleij <linus.walleij@...aro.org>,
        Bartosz Golaszewski <brgl@...ev.pl>,
        linux-gpio@...r.kernel.org,
        Octavian Purdila <octavian.purdila@....com>,
        linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH] genirq: Deduplicate WARN_ON_ONCE() in
 generic_handle_domain_irq()

On 2022-05-16 08:53:29 [+0200], Thomas Gleixner wrote:
> > they tell you that the context is wrong.
> 
> Why? These handlers can be called from any context, really. Yes. They
> need to be called with interrupts disabled, but that's it. The warning
> is checking hard interrupt context unconditionally.

correct. If the context is wrong, the interrupts are usually not
disabled.

> > From looking at gpio-dln2 this is called from USB URB's callback which
> > is softirq. In the end dln2_gpio_event() is invoked while
> > dln2_dev::event_cb_lock is acquired.  That lock is acquired by
> > disabling interrupts which is what gets the locking right for
> > generic_handle_domain_irq(). If that lock lifted to spin_lock_bh()
> > (because it is always in urb's calback context and all HCDs complete
> > in one context unlike now) then this breaks.
> 
> Yes, but that's a different problem.
> 
> > And PREEMPT_RT is broken already. Therefore, last week, I've been
> > promoting generic_handle_domain_irq_safe()
> > https://lkml.kernel.org/r/YnkfWFzvusFFktSt@linutronix.de
> 
> Well, that's just a wrapper which adds the local_irq_save(), so it's not
> any different from having the local_irq_save() at the callsite, unless
> I'm missing something.

I haven't seen a local_irq_save() at the callsite. If it is, then it is
not any different.

> Thanks,
> 
>         tglx

Sebastian