Re: [PATCH 01/14] parport: return value of attach and parport_register_driver

From: Sudip Mukherjee <sudipm.mukherjee@gmail.com>
To: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Arnd Bergmann <arnd@arndb.de>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Jean Delvare <jdelvare@suse.de>, Wolfram Sang <wsa@the-dreams.de>,
	Rodolfo Giometti <giometti@enneenne.com>,
	"James E.J. Bottomley" <JBottomley@parallels.com>,
	Mark Brown <broonie@kernel.org>, Willy Tarreau <willy@meta-x.org>,
	Jaroslav Kysela <perex@perex.cz>, Takashi Iwai <tiwai@suse.de>,
	devel@driverdev.osuosl.org, alsa-devel@alsa-project.org,
	linux-scsi@vger.kernel.org, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org, linux-spi@vger.kernel.org,
	linux-i2c@vger.kernel.org
Subject: Re: [PATCH 01/14] parport: return value of attach and parport_register_driver
Date: Wed, 8 Apr 2015 17:20:10 +0530	[thread overview]
Message-ID: <20150408115010.GA11153@sudip-PC> (raw)
In-Reply-To: <20150408113832.GH10964@mwanda>

On Wed, Apr 08, 2015 at 02:38:32PM +0300, Dan Carpenter wrote:
> 1) We can't apply this patch on its own so this way of breaking up the
> patches doesn't work.
yes, if the first patch is reverted for any reason all the others need
to be reverted also. so then everything in one single patch?
> 
> 2) I was thinking that all the ->attach() calls would have to succeed or
> we would bail.  Having some of them succeed and some fail doesn't seem
> like it will simplify the driver code very much.  But I can also see
> your point.  Hm...
to clarify my point more here: any system might have more than one
parallel port but the module might decide to use just one. so in that
case attach will return 0 for the port that it wishes to use, for others
it will be a error code. So in parport_register_driver if we get error
codes in all the attach calls then we know that attach has definitely
failed, but atleast one 0 means one attach call has succeeded, which
will happen in case of staging/panel, net/plip...

> 
> Minor comment:  No need to preserve the error code if there are lots
> which we miss.  We may as well hard code an error code.  But that's a
> minor thing.  Does this actually simplify the driver code?  That's the
> more important thing.

i don't think this will simplify the driver code, but atleast now
parport_register_driver() will not report success when we have actually
failed. And as a result module_init will also fail which is supposed to
be the actual behviour.

regards
sudip

> 
> regards,
> dan carpenter
> 