Re: sysfs file creation result nightmare (WAS radeonfb: Fix sysfs_create_bin_file warnings)

[Andrew Morton <akpm@osdl.org>]

On Sat, 9 Dec 2006 23:34:19 +0100
Olivier Galibert <galibert@pobox.com> wrote:

> On Sat, Dec 09, 2006 at 01:58:29PM -0800, Andrew Morton wrote:
> > On Sat, 9 Dec 2006 22:44:53 +0100
> > Olivier Galibert <galibert@pobox.com> wrote:
> > > Hmmm, I don't understand.  Which is the bug, having a sysfs file
> > > creation fail or going on if it happens?
> > 
> > Probably the former, probably the latter.
> > 
> > There may be situations in which we want do to "create this sysfs file if
> > it doesn't already exist", but I'm not aware of any such.
> > 
> > Generally speaking, if sysfs file creation went wrong, it's due to a bug. 
> > The result is that the driver isn't working as intended: tunables or
> > instrumentation which it is designed to make available are not present.  We
> > want to know about that bug asap so we can get it fixed.
> 
> Hmmm, then why don't you just drop the return value from the creation
> function and BUG() in there is something went wrong.  That would allow
> for better error messages too.

And (ultimately) make the function return void.

Yes, that's probably a valid approach - we've discussed it before but nobody has
taken it further.