Re: [PATCH v2] devpts: allow mounting with uid/gid of uint32_t

[Andrew Morton <akpm@linux-foundation.org>]

On Wed, 19 Aug 2015 09:24:31 +0200 Dongsu Park <dpark@posteo.net> wrote:

> > 	unsigned long uidl;
> > 
> > 	rc = kstrtoul(uidstr, 0, &uidl);
> > 	uidval = uidl;
> 
> That's a good point. I'll do it.
> 
> > > +			if (rc)
> > >  				return -EINVAL;
> > 
> > I don't get it.  From my reading, kstrtouint->parse_integer() returns
> > "number of characters parsed or -E".  So this code won't work.  But
> > presumably it *does* work, so why?
> 
> It's probably because kstrtouint() returns just 0 on success.
> That's what functions in the call chain of kstrtouint() -> kstrtoull() ->
> _kstrtoull() -> _parse_integer() are actually doing.
> _parse_integer() actually returns rv, i.e. number of characters parsed.
> But after that, if there's no error, _kstrtoull() simply returns 0.

whoa, wait, I was looking at the -mm tree which changes kstrtouint():

static inline int __must_check kstrtouint(const char *s, unsigned int base, unsigned int *res)
{
	return parse_integer(s, base | PARSE_INTEGER_NEWLINE, res);
}

and

 * Return number of characters parsed or -E.
 ...
 */
#define parse_integer(s, base, val)	\



Alexey, doesn't this mean that code which does

	if (kstrtouint(...))
		return -EFOO;

will break?  Is it intended that parse_integer-convert-*.patch will fix
every callsite in the kernel?  If so, how do we know there haven't been
concurrent additions in -next which need review/conversion?  Let alone
out-of-tree things...