Re: [PATCH] libata-sff: Don't call bmdma_stop on non DMA capable controllers

From: Mark Lord <lkml@rtr.ca>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: David Woodhouse <dwmw2@infradead.org>,
	Alan <alan@lxorguk.ukuu.org.uk>, Jeff Garzik <jgarzik@pobox.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] libata-sff: Don't call bmdma_stop on non DMA capable controllers
Date: Fri, 26 Jan 2007 10:29:34 -0500	[thread overview]
Message-ID: <45BA1E5E.6050405@rtr.ca> (raw)
In-Reply-To: <Pine.LNX.4.64.0701251832040.25027@woody.linux-foundation.org>

> (Yeah, mmap() has taught some people about MAP_FAIL, but that's pretty 
> unusual too.

It's so unusual, that I regularly find buggy code checking for NULL
instead of MAP_FAILED.  Just about every program I look at has it wrong
in at least one place.

Ditto for the result value from wait() -- lots of code fails to use
the fancy macros required for correct interpretation of it.

> The quality of our drivers is low;

I think that's a bit harsh.  But yes, compared to core parts
of the kernel perhaps (eg. page dirty mappings: remember the recent bugfix?).

The real deal is that using zero/NULL makes it *easier* to get the code right.
And we certainly have tons more people working on drivers than anything else,
so making it *easier* for those coders to reduce the bug count is a Good Thing(tm).

Cheers