linux-kernel - Re: [PATCH] irq: fix the interrupt trigger type override issue

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] [day] [month] [year] [list]

Message-ID: <87frqh82hr.ffs@tglx>
Date: Tue, 03 Sep 2024 12:18:08 +0200
From: Thomas Gleixner <tglx@...utronix.de>
To: Richard Clark <richard.xnu.clark@...il.com>
Cc: linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
 torvalds@...ux-foundation.org
Subject: Re: [PATCH] irq: fix the interrupt trigger type override issue

Richard!

On Tue, Sep 03 2024 at 15:55, Richard Clark wrote:
> On Mon, Sep 02, 2024 at 04:39:33PM +0200, Thomas Gleixner wrote:
>> So the mapping and the interrupt allocation persist even if nothing uses
>> them. What a waste.
>>
> I checked the code and found that it's not the kernel to create the mapping,
> it's by the driver calling platform_get_irq(...)/of_irq_get(...) to
> create.

:)
 
> Ah, the mapping is created from of_irq_get(...) by driver, the kernel also
> provides the mapping teardown interface - irq_dispose_mapping.
> IMO, the right way for the driver is:
> 	1) driver calls of_irq_get() to get the irq and create the mapping
> 	2) driver *should* call irq_dispose_mapping() as the teardown of step 1.
> 	3) free_irq is the teardown of the request_irq to free the irq and
> 	   its action.

Correct.

Thanks,

        tglx