linux-kernel - Re: [PATCH v2] thermal: tegra-bpmp: Handle offline zones

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]

Date:   Thu, 30 Mar 2023 16:11:50 +0300
From:   Mikko Perttunen <cyndis@...si.fi>
To:     Daniel Lezcano <daniel.lezcano@...aro.org>,
        "Rafael J. Wysocki" <rafael@...nel.org>,
        Amit Kucheria <amitk@...nel.org>,
        Zhang Rui <rui.zhang@...el.com>,
        Thierry Reding <thierry.reding@...il.com>,
        Jonathan Hunter <jonathanh@...dia.com>
Cc:     Mikko Perttunen <mperttunen@...dia.com>, linux-pm@...r.kernel.org,
        linux-tegra@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] thermal: tegra-bpmp: Handle offline zones

On 3/30/23 15:36, Daniel Lezcano wrote:
> On 30/03/2023 12:06, Mikko Perttunen wrote:
>> On 3/30/23 13:03, Daniel Lezcano wrote:
>>> On 30/03/2023 11:49, Mikko Perttunen wrote:
>>>> From: Mikko Perttunen <mperttunen@...dia.com>
>>>>
>>>> Thermal zones located in power domains may not be accessible when
>>>> the domain is powergated. In this situation, reading the temperature
>>>> will return -BPMP_EFAULT. When evaluating trips, BPMP will internally
>>>> use -256C as the temperature for offline zones.
>>>
>>>> For smooth operation, for offline zones, return -EAGAIN when reading
>>>> the temperature and allow registration of zones even if they are
>>>> offline during probe.
>>>
>>> I think it makes more sense to check if the power domain associated 
>>> with the device is powered up and if not return -EPROBE_DEFER.
>>
>> The power domains in question are related to computer vision engines 
>> that only get powered on when in use, possibly never if the user 
>> doesn't run a computer vision workload on the system. We still want 
>> other thermal zones to be available.
> 
> Ok, I see the point.
> 
> I'm worried about the semantic of the errors returned, the translation 
> from BPMP_EFAULT to EAGAIN and the assumption it is a disabled (may be 
> forever) thermal zone.
> 
> What does the documentation say for the error msg.rx.ret == -BPMP_EFAULT?
> 

The documentation says

Value          | Description
-------------- | -----------------------------------------
0              | Temperature query succeeded.
-#BPMP_EINVAL  | Invalid request parameters.
-#BPMP_ENOENT  | No driver registered for thermal zone.
-#BPMP_EFAULT  | Problem reading temperature measurement.

In practice, what BPMP_EFAULT means here is that the hardware has no 
indicated temperature for the zone, which really only happens if the 
power domain is powered off.

Mikko

> 
>>>> Signed-off-by: Mikko Perttunen <mperttunen@...dia.com>
>>>> ---
>>>> v2:
>>>> * Adjusted commit message.
>>>> * Patch 2/2 dropped for now since it is more controversial,
>>>>    and this patch is more critical.
>>>>
>>>>   drivers/thermal/tegra/tegra-bpmp-thermal.c | 9 ++++++++-
>>>>   1 file changed, 8 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/drivers/thermal/tegra/tegra-bpmp-thermal.c 
>>>> b/drivers/thermal/tegra/tegra-bpmp-thermal.c
>>>> index f5fd4018f72f..4ffc3bb3bf35 100644
>>>> --- a/drivers/thermal/tegra/tegra-bpmp-thermal.c
>>>> +++ b/drivers/thermal/tegra/tegra-bpmp-thermal.c
>>>> @@ -52,6 +52,8 @@ static int __tegra_bpmp_thermal_get_temp(struct 
>>>> tegra_bpmp_thermal_zone *zone,
>>>>       err = tegra_bpmp_transfer(zone->tegra->bpmp, &msg);
>>>>       if (err)
>>>>           return err;
>>>> +    if (msg.rx.ret == -BPMP_EFAULT)
>>>> +        return -EAGAIN;
>>>>       if (msg.rx.ret)
>>>>           return -EINVAL;
>>>> @@ -259,7 +261,12 @@ static int tegra_bpmp_thermal_probe(struct 
>>>> platform_device *pdev)
>>>>           zone->tegra = tegra;
>>>>           err = __tegra_bpmp_thermal_get_temp(zone, &temp);
>>>> -        if (err < 0) {
>>>> +
>>>> +        /*
>>>> +         * Sensors in powergated domains may temporarily fail to be 
>>>> read
>>>> +         * (-EAGAIN), but will become accessible when the domain is 
>>>> powered on.
>>>> +         */
>>>> +        if (err < 0 && err != -EAGAIN) {
>>>>               devm_kfree(&pdev->dev, zone);
>>>>               continue;
>>>>           }
>>>
>>
>