Re: [PATCH] regmap: silence GCC warning

[Paul Bolle <pebolle@tiscali.nl>]

On Sat, 2012-10-06 at 09:53 +0100, Mark Brown wrote:
> On Fri, Oct 05, 2012 at 06:20:44PM -0400, Valdis.Kletnieks@vt.edu wrote:
> > On Wed, 03 Oct 2012 09:23:36 +0200, Paul Bolle said:
> > > That is another way to silence GCC here.
> 
> > That's probably a preferable approach - that way, if a bogus val_count gets
> > passed in, the caller will be informed of the fact.  Which is a lot better than
> > just papering over the warning.
> 
> As I hinted earlier if someone were to send me a patch...

0) I was hoping to do that. But in the mean time I also filed a bug
report for GCC (at Fedora's bugzilla) [0].

1) In that report (after actually closing it) Jakub Jelinek pointed at
the type mismatch between regmap_volatile_range()'s 'num' (unsigned int)
and its callers (both use size_t, both through 'val_count'). And,
indeed, changing 'num' to size_t also makes this warning go away.

This might explain why you didn't see a warning on 32 bit arm (if that
is what you running while looking at my patch).

2) I hope to send in a second path shortly, changing 'num' to size_t. My
main doubt is whether its problematic that the loop index in
regmap_volatile_range() uses unsigned int too. If 'num' would exceed
UINT_MAX, that loop would never finish, wouldn't it? But is that
actually possible? Are there machines with that many registers?


Paul Bolle

0) https://bugzilla.redhat.com/show_bug.cgi?id=862620