Re: [PATCH v2] printk: Relocate wake_klogd check close to the end of console_unlock()

[Petr Mladek]

On Fri 2018-02-09 12:28:30, Sergey Senozhatsky wrote:
> On (02/08/18 17:48), Petr Mladek wrote:
> > Note that we need Tejun's patch for-4.16 because it fixes a potential
> > race introduced by the console waiter logic.
> 
> I understand it. I think I put my Reviewed-by on v1, maybe not...
> Here it is
> 
> Reviewed-by: Sergey Senozhatsky <sergey.senozhatsky@xxxxxxxxx>

You actually provided Reviewed-by for v1. But I was not sure if
you would agree with the updated commit log and pushing this for-4.16.

OK, I am going to push v2 into for-4.16-rc3 branch. I will push it
to Linus once I am back from vacation. If nobody is against it.


> But can we may be come up with a more general solution? Like the one
> that we discussed above. I've seen some user space klogd lost messages
> reports recently. I assume it was partly caused by that very late
> wakeup that we have.

Sure. A better solution makes sense. I am just not 100% sure about
the exact variant. But it is more or less about details that we
need to settle on.

Best Regards,
Petr