Re: tx_timeout and timer serialisation

[kuznet@xxxxxxxxxxxxx]

Hello!

> I posted a patch to the async del_timer() behaviour onto linux-kernel
> yesterday.
> 
> I think we need to change the default behaviour of del_timer() to be
> async, and then run around and fix the resulting deadlocks. If we accept
> anything less, we have tens or even hundreds of subtle bugs.
> 
> I'd be very interested in your opinions.

Seems, I like this. 8)

Only I still do not undertsnad how it works. 8)
I am lost in these patches... Could you prepare one large
patch doing all of the described things?

Alexey