Re: [PATCH 00/16] mm: prepare for converting vm->vm_flags to 64-bit

[Konstantin Khlebnikov <khlebnikov@openvz.org>]

Konstantin Khlebnikov wrote:
> Andrew Morton wrote:
>> On Thu, 22 Mar 2012 21:28:11 +0000
>> Al Viro<viro@ZenIV.linux.org.uk>   wrote:
>>
>>> On Thu, Mar 22, 2012 at 02:26:47PM -0700, Andrew Morton wrote:
>>>> It would be nice to find some way of triggering compiler warnings or
>>>> sparse warnings if someone mixes a 32-bit type with a vm_flags_t.  Any
>>>> thoughts on this?
>>>>
>>>> (Maybe that's what __nocast does, but Documentation/sparse.txt doesn't
>>>> describe it)
>>>
>>> Use __bitwise for that - check how gfp_t is handled.
>>
>> So what does __nocast do?
>
> Actually it forbid any non-forced casts, but its implementation in sparse seems buggy:
> __nocast generates some strange false positives. For example it sometimes forgot about
> type attributes in function arguments, I saw this for vm_flags argument in ksm_madvise().
> I can reproduce this bug, if somebody interested.
>

1.c:

# define __nocast	__attribute__((nocast))

typedef long __nocast long_t;

static void f(long_t *x)
{
}

static void g(void)
{
	long_t x;

	f(&x);
}

$ sparse 1.c
1.c:13:12: warning: incorrect type in argument 1 (different modifiers)
1.c:13:12:    expected int [nocast] [usertype] *x
1.c:13:12:    got int *<noident>
1.c:13:12: warning: implicit cast to nocast type

Is this ok?

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>