On Sat, Mar 27, 2010 at 10:16:38AM +0100, Kamil Dudka wrote:
> On Wednesday 24 of March 2010 11:07:04 Christopher Li wrote:
> > That is just too much. Most of the warning is coming from enum or
> > operation. e.g.
> > .type = KW_SPECIFIER | KW_SHORT,
> > lookup_keyword(token->ident, NS_KEYWORD | NS_TYPEDEF);
>
> Looking once again, I don't think that's the case. The warnings do not come
> from the "enum or" operation, but they come from passing the integral result
> to enum variable (or arg). Instead of weaking my patch, we may improve the
> code of sparse and add explicit casts back to enum:
>
> .type = (enum keyword) (KW_SPECIFIER | KW_SHORT),
> lookup_keyword(token->ident, (enum namespace) (NS_KEYWORD | NS_TYPEDEF));
That looks wrong. .type doesn't contain a value of type "enum keyword",
it contains the bitwise or of such values, which won't represent a valid
enum value. Thus, .type should have an integral type, not an enum type.
The same goes for the second parameter of lookup_keyword.
> The whole problem can be narrowed down to a simple test-case:
>
> int main()
> {
> enum {
> A = 0x1,
> B = 0x2
> } val = A | B;
This code seems semantically wrong, as described above. val should only
have values 0x1 or 0x2, and you've assigned it 0x3, which doesn't
represent a valid value of its enum type.
> Here is what sparse gives:
> $ ./sparse enum.c
> enum.c:1:10: warning: non-ANSI function declaration of function 'main'
> enum.c:6:15: warning: conversion of
> enum.c:6:15: int to
> enum.c:6:15: int enum <noident>
>
>
> Here is what g++ gives:
> $ g++ enum.c
> enum.c: In function ‘int main()’:
> enum.c:6: error: invalid conversion from ‘int’ to ‘main()::<anonymous enum>’
Yup, both of these warnings seem correct. Don't fix them by casting,
fix them by declaring "val" with an appropriate integral type.
- Josh Triplett
--
To unsubscribe from this list: send the line "unsubscribe linux-sparse" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
