linux-kernel - Re: [PATCH] clocksource: timer-tegra186: Enable WDT at probe

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: <f39db414-1e51-45ea-bd5c-4c1d6fd9b87f@nvidia.com>
Date: Thu, 3 Jul 2025 15:42:57 +0100
From: Jon Hunter <jonathanh@...dia.com>
To: Thierry Reding <thierry.reding@...il.com>
Cc: Kartik Rajput <kkartik@...dia.com>, daniel.lezcano@...aro.org,
 tglx@...utronix.de, linux-kernel@...r.kernel.org, linux-tegra@...r.kernel.org
Subject: Re: [PATCH] clocksource: timer-tegra186: Enable WDT at probe


On 03/07/2025 15:19, Thierry Reding wrote:
> On Thu, Jul 03, 2025 at 03:11:57PM +0100, Jon Hunter wrote:
>>
>> On 03/07/2025 14:36, Thierry Reding wrote:
>>
>> ...
>>
>>> I clearly can't read code today. Seems generally fine, then, but I'm
>>> actually really enthused now about using a second watchdog for kernel
>>> petting. Since we don't use any of the other two watchdogs, is there
>>> any reason why we can't cleanly separate both use-cases? It would let
>>> us avoid some of these special cases that are not intuitive to
>>> understand.
>>
>> The only reason would be if for some reason the other are all allocated for
>> other uses outside of the kernel. We are currently only using the one for
>> the kernel so that it would mean updating all the device trees for all
>> platforms to support this too.
>>
>> I was also thinking about how do we identify/select if a watchdog is pet by
>> the kernel or userspace? I was thinking that the presence of the 'interrupt'
>> property in device-tree could be used; if present the kernel pets and if not
>> assume userspace pets. However, the 'interrupt' property is currently marked
>> as required and not optional.
> 
> The other two instances are part of the TKE block, too. It should be as
> simple as doing something like this:
> 
> 	tegra->wdt_kernel = tegra186_wdt_create(tegra, 1);
> 
> and using that instead of tegra->wdt.


OK, well that is much simpler than what I was thinking!

Jon

-- 
nvpublic