Re: [PATCH] cryptoloop

From: Jari Ruusu <jari.ruusu@pp.inet.fi>
To: Chris Friesen <cfriesen@nortelnetworks.com>
Cc: Andrew Morton <akpm@osdl.org>,
	Andries.Brouwer@cwi.nl, akpm@digeo.com,
	linux-kernel@vger.kernel.org, torvalds@osdl.org
Subject: Re: [PATCH] cryptoloop
Date: Fri, 04 Jul 2003 10:43:10 +0300	[thread overview]
Message-ID: <3F05300E.AA26A021@pp.inet.fi> (raw)
In-Reply-To: 3F046A30.6080509@nortelnetworks.com

Chris Friesen wrote:
> Jari Ruusu wrote:
> > Because loop-AES attempts to be compatible with structures in loop.h by not
> > modifying loop.h at all. This is what the "no kernel sources patched or
> > replaced" means. Breakage in loop.h breaks loop-AES, and I have to clean the
> > mess.
> 
> We're in a development stream.  It is kind of expected that in-kernel
> APIs may change if the developers feel it will lead to some improvement.
> 
> This sucks for people that are trying to track those APIs with
> out-of-kernel patches, but its a fact of life.

I know. I already have to deal with API breakages.

Changing transfer function prototype may be a tiny speed improvement for one
implementation that happens to use unoptimal API, but at same time be tiny
speed degration to other implementations that use more saner APIs. I am
unhappy with that change, because I happen to maintain four such transfers
that would be subject to tiny speed degration.

Regards,
Jari Ruusu <jari.ruusu@pp.inet.fi>