linux-kernel - Re: [PATCH] semaphore: Add might_sleep() to down

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]

Date:   Mon, 9 Aug 2021 22:33:18 +0800
From:   Xiaoming Ni <nixiaoming@...wei.com>
To:     Waiman Long <llong@...hat.com>, <linux-kernel@...r.kernel.org>,
        <peterz@...radead.org>, <mingo@...hat.com>, <will@...nel.org>,
        <boqun.feng@...il.com>
CC:     <wangle6@...wei.com>, <xiaoqian9@...wei.com>, <shaolexi@...wei.com>
Subject: Re: [PATCH] semaphore: Add might_sleep() to down_*() family

On 2021/8/9 20:52, Waiman Long wrote:
> On 8/8/21 11:51 PM, Xiaoming Ni wrote:
>> On 2021/8/9 11:01, Waiman Long wrote:
>>>
>>> I think it is simpler to just put a "might_sleep()" in 
>>> __down_common() which is the function where sleep can actually happen.
>>>
>>
>> If the actual atomic context hibernation occurs, the corresponding 
>> alarm log is generated in __schedule_bug().
>>     __schedule()
>>         --> schedule_debug()
>>             --> __schedule_bug()
>>
>> However, "might_sleep()" indicates the possibility of sleep, so that 
>> code writers can identify and fix the problem as soon as possible, but 
>> does not trigger atomic context sleep.
>>
>> Is it better to put "might_sleep()" in each down API entry than 
>> __down_common() to help identify potential code problems? 
> 
> Putting "might_sleep()" in each down_*() functions mean that whenever we 
> add a new API function, we have to remember to add "might_sleep()". If
> we put it in down_common(), it will work for any newly added API 
> function in the future even though I doubt we will add any.
> 
If the code enters down_common(), it is not "might" sleep but "will" 
sleep, and an alarm is printed in __schedule_bug() later.

"might_sleep()" is used to check potential problems, and 
"_schedule_bug()" is used to check actual faults.

So, I still think we should add "might_sleep()" to each down_*() 
function to alert code owner to potential problems early.

Thanks
Xiaoming Ni