linux-kernel - Re: [PATCH] kernel:irq:manage: request threaded irq with a specified priority

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 for Android: free password hash cracker in your pocket

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

Message-ID: <875z0qzigk.ffs@nanos.tec.linutronix.de>
Date:   Tue, 13 Apr 2021 10:39:39 +0200
From:   Thomas Gleixner <tglx@...utronix.de>
To:     Song Chen <chensong_2000@....cn>
Cc:     linux-rt-users@...r.kernel.org, linux-kernel@...r.kernel.org,
        rostedt@...dmis.org, mingo@...hat.com, peterz@...radead.org,
        juri.lelli@...hat.com, vincent.guittot@...aro.org,
        dietmar.eggemann@....com, bsegall@...gle.com, mgorman@...e.de,
        bristot@...hat.com, keescook@...omium.org,
        gregkh@...uxfoundation.org, maz@...nel.org, joe@...ches.com,
        romain.perier@...il.com, john.garry@...wei.com
Subject: Re: [PATCH] kernel:irq:manage: request threaded irq with a specified priority

On Tue, Apr 13 2021 at 14:19, Song Chen wrote:
> In general, irq handler thread will be assigned a default priority which
> is MAX_RT_PRIO/2, as a result, no one can preempt others.
>
> Here is the case I found in a real project, an interrupt int_a is
> coming, wakes up its handler handler_a and handler_a wakes up a
> userspace RT process task_a.
>
> However, if another irq handler handler_b which has nothing to do
> with any RT tasks is running when int_a is coming, handler_a can't
> preempt handler_b, as a result, task_a can't be waken up immediately
> as expected until handler_b gives up cpu voluntarily. In this case,
> determinism breaks.

It breaks because the system designer failed to assign proper priorities
to the irq threads int_a, int_b and to the user space process task_a.

That's not solvable at the kernel level.

Thanks,

        tglx