Re: Inquiry about the removal of flag O_NONBLOCK on /dev/random

["Guozihua (Scott)" <guozihua@huawei.com>]

On 2022/9/7 21:03, Jason A. Donenfeld wrote:
> On Tue, Sep 06, 2022 at 12:16:56PM +0200, Jason A. Donenfeld wrote:
>> On Thu, Jul 28, 2022 at 10:25 AM Guozihua (Scott) <guozihua@huawei.com> wrote:
>>>
>>> On 2022/7/26 19:33, Guozihua (Scott) wrote:
>>>> On 2022/7/26 19:08, Jason A. Donenfeld wrote:
>>>>> Hi,
>>>>>
>>>>> On Tue, Jul 26, 2022 at 03:43:31PM +0800, Guozihua (Scott) wrote:
>>>>>> Thanks for all the comments on this inquiry. Does the community has any
>>>>>> channel to publishes changes like these? And will the man pages get
>>>>>> updated? If so, are there any time frame?
>>>>>
>>>>> I was under the impression you were ultimately okay with the status quo.
>>>>> Have I misunderstood you?
>>>>>
>>>>> Thanks,
>>>>> Jason
>>>>> .
>>>>
>>>> Hi Jason.
>>>>
>>>> To clarify, I does not have any issue with this change. I asked here
>>>> only because I would like some background knowledge on this flag, to
>>>> ensure I am on the same page as the community regarding this flag and
>>>> the change. And it seems that I understands it correctly.
>>>>
>>>> However I do think it's a good idea to update the document soon to avoid
>>>> any misunderstanding in the future.
>>>>
>>>
>>> Our colleague suggests that we should inform users clearly about the
>>> change on the flag by returning -EINVAL when /dev/random gets this flag
>>> during boot process. Otherwise programs might silently block for a long
>>> time, causing other issues. Do you think this is a good way to prevent
>>> similar issues on this flag?
>>
>> I still don't really understand what you want. First you said this was
>> a problem and we should reintroduce the old behavior. Then you said no
>> big deal and the docs just needed to be updated. Now you're saying
>> this is a problem and we should reintroduce the old behavior?
>>
>> I'm just a bit lost on where we were in the conversation.
>>
>> Also, could you let me know whether this is affecting real things for
>> Huawei, or if this is just something you happened to notice but
>> doesn't have any practical impact?
> 
> Just following up on this again...
> .

Hi Jason,

Thank you for the timely respond and your patient. And sorry for the 
confusion.

First of all, what we think is that this change (removing O_NONBLOCK) is 
reasonable. However, this do cause issue during the test on one of our 
product which uses O_NONBLOCK flag the way I presented earlier in the 
Linux 4.4 era. Thus our colleague suggests that returning -EINVAL when 
this flag is received would be a good way to indicate this change.

For example:


-- 
Best
GUO Zihua

-- 
Best
GUO Zihua