linux-kernel - Re: [PATCH] ACPI: processor: idle: Fix resource rollback in acpi_processor_power

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

Message-ID: <6a35291a-32e8-461e-a0e5-405b7b5d24ce@huawei.com>
Date: Thu, 3 Jul 2025 14:23:44 +0800
From: "lihuisong (C)" <lihuisong@...wei.com>
To: "Rafael J. Wysocki" <rafael@...nel.org>
CC: <lenb@...nel.org>, <linux-acpi@...r.kernel.org>,
	<linux-kernel@...r.kernel.org>, <linuxarm@...wei.com>,
	<jonathan.cameron@...wei.com>, <zhanjie9@...ilicon.com>,
	<zhenglifeng1@...wei.com>, <yubowen8@...wei.com>, <liuyonglong@...wei.com>
Subject: Re: [PATCH] ACPI: processor: idle: Fix resource rollback in
 acpi_processor_power_init

Hi,

Thanks for your review.


在 2025/7/3 1:42, Rafael J. Wysocki 写道:
> On Thu, Jun 19, 2025 at 8:13 AM Huisong Li <lihuisong@...wei.com> wrote:
>> There are two resource rollback issues in acpi_processor_power_init:
>> 1> Do not unregister acpi_idle_driver when do kzalloc failed.
>> 2> Do not free cpuidle device memory when register cpuidle device failed.
>>
>> Signed-off-by: Huisong Li <lihuisong@...wei.com>
>> ---
>>   drivers/acpi/processor_idle.c | 24 +++++++++++++++++-------
>>   1 file changed, 17 insertions(+), 7 deletions(-)
>>
>> diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c
>> index 2c2dc559e0f8..3548ab9dac9e 100644
>> --- a/drivers/acpi/processor_idle.c
>> +++ b/drivers/acpi/processor_idle.c
>> @@ -1392,8 +1392,10 @@ int acpi_processor_power_init(struct acpi_processor *pr)
>>                  }
>>
>>                  dev = kzalloc(sizeof(*dev), GFP_KERNEL);
>> -               if (!dev)
>> -                       return -ENOMEM;
>> +               if (!dev) {
>> +                       retval = -ENOMEM;
>> +                       goto unregister_driver;
> No, unregistering the driver here is pointless.
I don't quite understand why here is pointless. Can you explain it?
>
>> +               }
>>                  per_cpu(acpi_cpuidle_device, pr->id) = dev;
>>
>>                  acpi_processor_setup_cpuidle_dev(pr, dev);
>> @@ -1402,14 +1404,22 @@ int acpi_processor_power_init(struct acpi_processor *pr)
>>                   * must already be registered before registering device
>>                   */
>>                  retval = cpuidle_register_device(dev);
>> -               if (retval) {
>> -                       if (acpi_processor_registered == 0)
>> -                               cpuidle_unregister_driver(&acpi_idle_driver);
> Pretty much the same as here.
>
> It would be good to clean up dev here though.
It is ok if above is pointless.
>
>> -                       return retval;
>> -               }
>> +               if (retval)
>> +                       goto free_cpuidle_device;
>> +
>>                  acpi_processor_registered++;
>>          }
>>          return 0;
>> +
>> +free_cpuidle_device:
>> +       per_cpu(acpi_cpuidle_device, pr->id) = NULL;
>> +       kfree(dev);
>> +
>> +unregister_driver:
>> +       if (acpi_processor_registered == 0)
>> +               cpuidle_unregister_driver(&acpi_idle_driver);
>> +
>> +       return retval;
>>   }
>>
>>   int acpi_processor_power_exit(struct acpi_processor *pr)
>> --
>> 2.33.0
>>