netdev
[Top] [All Lists]

Re: tx_timeout and timer serialisation

To: "David S. Miller" <davem@xxxxxxxxxx>
Subject: Re: tx_timeout and timer serialisation
From: Donald Becker <becker@xxxxxxxxx>
Date: Sat, 20 May 2000 03:25:53 -0400 (EDT)
Cc: netdev@xxxxxxxxxxx
In-reply-to: <200005200524.WAA18905@xxxxxxxxxxxxxxx>
Sender: owner-netdev@xxxxxxxxxxx
On Fri, 19 May 2000, David S. Miller wrote:

> I just noticed this thread, and has del_timer_sync been mentioned yet?
> That is what should be used to make sure the timer is done in 2.3.x,
> unless something else prevents it's usage (locking conflict).

Yes, it was used as an example of badness.
 "Something broke with an interface.  It should have been obvious that it
 broke, if for no other reason than a new function had to be created that
 did what the old function used to do."

The new interface/semantics should have been named  del_timer_async(), with
del_timer() being the synchronous version. 
Given the questionable semantics of the  _async() version, I doubt that
there would have been much of a demand to use it.

Backwards compatibility is pointlessly made more painful by such a change.
It's almost as bad as changing a interface function's argument count without
changing the name, or changing the element order of a structure that is
commonly statically initialized.

Donald Becker                           becker@xxxxxxxxx
Scyld Computing Corporation
410 Severn Ave. Suite 210
Annapolis MD 21403



<Prev in Thread] Current Thread [Next in Thread>