Re: [PATCH repost 12/16] parisc/uaccess: fix sparse errors

[James Bottomley]

On Wed, 2014-12-31 at 20:38 +0200, Michael S. Tsirkin wrote:
> On Wed, Dec 31, 2014 at 09:17:20AM -0800, James Bottomley wrote:
[...]
> > OK, parisc developers still being dense, but this does look like an
> > abuse of the bitwise type.
> 
> To give you another example:
> 
> 	__le16 __user *p;
> 	__le16 foo;
> 	int rc = get_user(v, p);
> 
> really should be fine, ATM this gives a warning.

OK, I think I've figured it out.  You're saying that casting __gu_val to
a bitwise annotated type is an automatic sparse failure because it has
to be a long in our assembly code to receive the load/store as a
register.  However, this is required for sparse to do the correct lvalue
type = rvalue type check in the assignment to x.  We were all thinking
the __force just killed these sparse type checks.

In that case, I think parisc is fine with this.

James


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/