[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <f6d9b092-20e8-436e-9307-2c24cb0ba3a5@arm.com>
Date: Fri, 1 Dec 2023 14:02:38 +0000
From: Lukasz Luba <lukasz.luba@....com>
To: Greg KH <gregkh@...uxfoundation.org>
Cc: linux-kernel@...r.kernel.org, linux-pm@...r.kernel.org,
daniel.lezcano@...aro.org, rafael@...nel.org,
stable@...r.kernel.org
Subject: Re: [PATCH] powercap: DTPM: Fix the missing cpufreq_cpu_put() calls
Hi Greg,
On 12/1/23 12:44, Greg KH wrote:
> On Fri, Dec 01, 2023 at 12:32:05PM +0000, Lukasz Luba wrote:
>> The policy returned by cpufreq_cpu_get() has to be released with
>> the help of cpufreq_cpu_put() to balance its kobject reference counter
>> properly.
>>
>> Add the missing calls to cpufreq_cpu_put() in the code.
>>
>> Fixes: 0aea2e4ec2a2 ("powercap/dtpm_cpu: Reset per_cpu variable in the release function")
>> Fixes: 0e8f68d7f048 ("powercap/drivers/dtpm: Add CPU energy model based support")
>> Cc: <stable@...r.kernel.org> # v5.10+
>
> But the Fixes: tags are for commits that are only in 5.12 and newer, how
> can this be relevant for 5.10?
My apologies, you're right. Somehow I checked that this dtpm_cpu.c
was introduced in v5.10. It was in v5.12 indeed. I messed that up.
Also, the code in that v5.12 had different implementation and there was
a function cpuhp_dtpm_cpu_offline() which had the cpufreq_cpu_get().
I can craft for that v5.12 special extra patch fix addressing it and
send directly to stable list. Would that make sense?
So this patch would only be applicable for v5.16+ AFAICS.
Regards,
Lukasz
Powered by blists - more mailing lists