Re: [PATCH 2/6] firewire: isochronous and asynchronous I/O

From: Stefan Richter <stefanr@s5r6.in-berlin.de>
To: "Kristian Høgsberg" <krh@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>,
	linux-kernel@vger.kernel.org,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux1394-devel <linux1394-devel@lists.sourceforge.net>
Subject: Re: [PATCH 2/6] firewire: isochronous and asynchronous I/O
Date: Thu, 03 May 2007 10:54:34 +0200	[thread overview]
Message-ID: <4639A34A.5030904@s5r6.in-berlin.de> (raw)
In-Reply-To: <46392800.4090408@redhat.com>

Kristian Høgsberg wrote:
> Christoph Hellwig wrote:
[...]
>>> +	retval = fw_core_add_descriptor(&model_id_descriptor);
>>> +	BUG_ON(retval < 0);
>> 
>> These kinds of bug checks look wrong.  Either the operations
>> can't fail in which case they should not return an error value
>> or you should handle them properly.
> 
> The fw_core_add_descriptor() checks that the descriptor block it's passed is 
> internally consistent and is used for blocks passed in from userspace too.  In 
> these two cases, the blocks are static const arrays in the driver and if 
> fw_core_add_descriptor returns < 0 it's a bug in the driver.

When you submitted it, I too wondered whether these BUG_ONs are correct
(they are) and whether they are worth having (not so sure).

A plus of a smaller stack is that we need less assertions to guard its
integrity. :-)
-- 
Stefan Richter
-=====-=-=== -=-= ---==
http://arcgraph.de/sr/