linux-kernel - Re: [PATCH 15/20] ACPICA: executer/exsystem: Warn about sleeps greater than 10 ms

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: <fd14c5e5-9635-7ef5-0bf4-438dfaea5361@molgen.mpg.de>
Date:   Sun, 22 May 2022 01:28:47 +0200
From:   Paul Menzel <pmenzel@...gen.mpg.de>
To:     "Rafael J. Wysocki" <rafael@...nel.org>
Cc:     "Rafael J. Wysocki" <rjw@...ysocki.net>,
        Bob Moore <robert.moore@...el.com>,
        Linux ACPI <linux-acpi@...r.kernel.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 15/20] ACPICA: executer/exsystem: Warn about sleeps
 greater than 10 ms

Dear Rafael,


Am 21.05.22 um 18:11 schrieb Rafael J. Wysocki:
> On Mon, Apr 11, 2022 at 9:04 PM Rafael J. Wysocki <rjw@...ysocki.net> wrote:
>>
>> From: Paul Menzel <pmenzel@...gen.mpg.de>
>>
>> ACPICA commit 2a0d1d475e7ea1c815bee1e0692d81db9a7c909c
>>
>> Quick boottime is important, so warn about sleeps greater than 10 ms.
>>
>> Distribution Linux kernels reach initrd in 350 ms, so excessive delays
>> should be called out. 10 ms is chosen randomly, but three of such delays
>> would already make up ten percent of the boottime.
>>
>> Link: https://github.com/acpica/acpica/commit/2a0d1d47
>> Signed-off-by: Bob Moore <robert.moore@...el.com>
>> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@...el.com>
> 
> I have decided to revert this, because it spams logs with unuseful
> warnings on many production systems.

Thank you for the information. Can you please give an example?

> Power management AML uses sleeps above 10 ms for PCI spec compliance
> and polling is not useful in those cases.

Can you please tell me what delays are used? Maybe we can increase the 
threshold to the one required by the PCI spec?


Kind regards,

Paul


> I will submit an analogous revert for upstream ACPICA.
> 
>> ---
>>   exsystem.c |   10 ++++++++++
>>   1 file changed, 10 insertions(+)
>>
>> diff -Nurp linux.before_name/drivers/acpi/acpica/exsystem.c linux.after_name/drivers/acpi/acpica/exsystem.c
>> --- linux.before_name/drivers/acpi/acpica/exsystem.c    2022-04-01 18:26:54.958046947 +0200
>> +++ linux.after_name/drivers/acpi/acpica/exsystem.c     2022-04-01 18:26:51.760086923 +0200
>> @@ -170,6 +170,16 @@ acpi_status acpi_ex_system_do_sleep(u64
>>          acpi_ex_exit_interpreter();
>>
>>          /*
>> +        * Warn users about excessive sleep times, so ASL code can be improved to
>> +        * use polling or similar techniques.
>> +        */
>> +       if (how_long_ms > 10) {
>> +               ACPI_WARNING((AE_INFO,
>> +                             "Firmware issue: Excessive sleep time (%llu ms > 10 ms) in ACPI Control Method",
>> +                             how_long_us));
>> +       }
>> +
>> +       /*
>>           * For compatibility with other ACPI implementations and to prevent
>>           * accidental deep sleeps, limit the sleep time to something reasonable.
>>           */
>>
>>
>>