On Mon, Dec 01, 2014 at 03:08:51PM -0800, Andrew Morton wrote:
> On Thu, 27 Nov 2014 14:13:12 +0900 Joonsoo Kim <iamjoonsoo.kim@xxxxxxx> wrote:
>
> > @@ -1190,7 +1190,7 @@ static int do_mmap_private(struct vm_area_struct *vma,
> > kdebug("try to alloc exact %lu pages", total);
> > base = alloc_pages_exact(len, GFP_KERNEL);
> > } else {
> > - base = __get_free_pages(GFP_KERNEL, order);
> > + base = (void *)__get_free_pages(GFP_KERNEL, order);
> > }
>
> __get_free_pages() is so irritating. I'm counting 268 calls, at least
> 172 of which have to typecast the return value.
>
> static inline void *
> someone_think_of_a_name_for_this(gfp_t gfp_mask, unsigned int order)
> {
> return (void *)__get_free_pages(gfp, order);
> }
>
Hello,
I think that changing return type of __get_free_pages() is better than
introducing new interface. With it, we only need to fix 268 - 172 = 96
callsites. And, get_zeroed_page() should also be fixed. Almost every
caller of this function do typecast. :)
I'm not familiar with this kind of massive change so it'd be better to
be done it by another developer.
Thanks.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>