[LEDE-DEV] [PATCH] uloop: use a waker for notifying sigchld and loop cancel events

[Xinxing Hu]

Hi Mats,

I think you should try the latest one. Felix made a mistake in this
commit and waker_pipe was not initialized correctly.

Best Regards,
Xinxing


> Date: Wed, 15 Jun 2016 15:07:25 +0200
> From: Mats Karrman <mats.dev.list at gmail.com>
> To: Yousong Zhou <yszhou4tech at gmail.com>, Felix Fietkau <nbd at nbd.name>
> Cc: lede-dev at lists.infradead.org
> Subject: Re: [LEDE-DEV] [PATCH] uloop: use a waker for notifying
>         sigchld and loop cancel events
> Message-ID: <5761530D.6010509 at gmail.com>
> Content-Type: text/plain; charset=windows-1252; format=flowed
>
>
> On 2016-06-15 13:20, Yousong Zhou wrote:
>> On 15 June 2016 at 17:58, Felix Fietkau <nbd at nbd.name> wrote:
>>> On 2016-06-09 04:20, Yousong Zhou wrote:
>>>> Fix a race condition when do_sigchld, uloop_cancelled were set just
>>>> before epoll_wait(timeout=-1), resulting the loop stuck in the syscall
>>>> without noticing the events just happened
>>>>
>>>> Signed-off-by: Yousong Zhou <yszhou4tech at gmail.com>
>>> I found a few bugs and inconsistencies in that patch, and pushed a
>>> fixed version here:
>>> https://git.lede-project.org/?p=project/libubox.git;a=commitdiff;h=4e3a47a4cb438866bc4b9cb2f5d16226ffb48502
>>>
>>> Please review and test it some more before I push an update to LEDE.
>>>
>>> Thanks,
>>>
>>> - Felix
>> Ack from me.
>>
>> The patch looks better.  Thanks
>>
>>                  yousong
> Tried this version together with procd-2016-03-09
> (b12bb150ed38a4409bef5127c77b060ee616b860)
> but I still see the same behavior/error.
>
> BR // Mats
>
>
>
> ------------------------------
>

-- 
Sincerely
Xinxing Hu