linux-kernel - Re: [PATCH v1 2/9] ACPI: processor: idle: Return failure if entry method is not buffer or integer type

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: <a577bfd3-a4b8-4d06-b95b-a78b2916b8a4@huawei.com>
Date: Fri, 24 Oct 2025 17:25:35 +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>, <Sudeep.Holla@....com>,
	<linuxarm@...wei.com>, <jonathan.cameron@...wei.com>,
	<zhanjie9@...ilicon.com>, <zhenglifeng1@...wei.com>, <yubowen8@...wei.com>
Subject: Re: [PATCH v1 2/9] ACPI: processor: idle: Return failure if entry
 method is not buffer or integer type


在 2025/10/23 18:07, Rafael J. Wysocki 写道:
> On Thu, Oct 23, 2025 at 11:25 AM lihuisong (C) <lihuisong@...wei.com> wrote:
>>
>> 在 2025/10/22 3:34, Rafael J. Wysocki 写道:
>>> On Mon, Sep 29, 2025 at 11:38 AM Huisong Li <lihuisong@...wei.com> wrote:
>>>> According to ACPI spec, entry method in LPI sub-package must be buffer
>>>> or integer. However, acpi_processor_evaluate_lpi() regeards it as success
>>>> and treat it as an effective LPI state.
>>> Is that the case?  AFAICS, it just gets to the next state in this case
>>> and what's wrong with that?
>> The flatten_lpi_states() would consider the state with illegal entry
>> method sub-package as a valid one
>> if the flag of this state is enabled(ACPI_LPI_STATE_FLAGS_ENABLED is set).
>> And then cpuidle governor would use it because the caller of
>> acpi_processor_ffh_lpi_probe() also don't see the return value.
> So the problem appears to be that lpi_state increments in every step
> of the loop, but it should only increment if the given state is valid.
Yes,
So set the flag of the state with illegal entry method sub-package to 
zero so that this invalid LPI state will be skiped in 
flatten_lpi_states(), ok?