On Wed, Sep 21, 2016 at 10:13:58AM -0400, Lance Richardson wrote:
> Sparse emits a large number of warnings for the linux kernel source
> tree of the form:
> ./arch/x86/include/asm/uaccess.h:735:18: \
> warning: incorrect type in argument 1 (different modifiers)
> ./arch/x86/include/asm/uaccess.h:735:18: expected void *<noident>
> ./arch/x86/include/asm/uaccess.h:735:18: got void const *from
>
> Fix by making the first parameter to __builtin_object_size()
> type "const void *" instead of "void *", which is consistent with GCC
> behavior (the prototype for this builtin in GCC documentation is evidently
> incorrect).
>
> Signed-off-by: Lance Richardson <lrichard@xxxxxxxxxx>
> ---
> lib.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lib.c b/lib.c
> index aa2af68..122dd6f 100644
> --- a/lib.c
> +++ b/lib.c
> @@ -917,7 +917,7 @@ void declare_builtin_functions(void)
> add_pre_buffer("extern long double __builtin_nanl(const char *);\n");
>
> /* And some __FORTIFY_SOURCE ones.. */
> - add_pre_buffer ("extern __SIZE_TYPE__ __builtin_object_size(void *, int);\n");
> + add_pre_buffer ("extern __SIZE_TYPE__ __builtin_object_size(const void *, int);\n");
> add_pre_buffer ("extern void * __builtin___memcpy_chk(void *, const void *, __SIZE_TYPE__, __SIZE_TYPE__);\n");
> add_pre_buffer ("extern void * __builtin___memmove_chk(void *, const void *, __SIZE_TYPE__, __SIZE_TYPE__);\n");
> add_pre_buffer ("extern void * __builtin___mempcpy_chk(void *, const void *, __SIZE_TYPE__, __SIZE_TYPE__);\n");
> --
Acked-by: Luc Van Oostenryck <luc.vanoostenryck@xxxxxxxxx>
Chris, can you please take this patch (or Johannes' one).
When compiling the kernel, there is more than 2000 false warnings
only for the uaccess.h case described here above.
Luc
--
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
