On 10/07/2016 03:19 PM, Zefir Kurtisi wrote:
> On 10/07/2016 02:15 PM, Alexandru Ardelean wrote:
>> On Fri, Oct 7, 2016 at 3:09 PM, Felix Fietkau <n...@nbd.name> wrote:
>>> Instead of introducing yet another timer, wouldn't it also be possible
>>> to close this race window by registering the event handler before
>>> attempting the lookup?
>>> - Felix
>> I've also seen this race.
>> I tried something like this:
>> But never got around to doing much testing to see if the race goes
>> away completely.
>> So, I never pushed it upstream.
>> @Zefir, maybe you could try it ?
> Hi Alex,
> your assumption is right, that's the root cause for the random timeouts.
> Unfortunately, it is hard to provide a positive proof, since for me the effect
> went away when I added some logging in between.
> My patch made it disappear, but of course what Felix suggests and you already
> implemented is the better approach. I'll take your commit instead and test it.
> From looking at the changes it should do, but to get some confidence it will
> some time.
After some days of testing, it seems like the patch fixes the race.
Alex, feel free to add me to the 'Tested-by' list if you are going to make a
request for that commit.
Lede-dev mailing list