On Wed, Nov 22, 2017 at 10:28:48PM +0100, Luc Van Oostenryck wrote:
> > - root->gfp_mask &= (1 << ROOT_TAG_SHIFT) - 1;
> > + root->gfp_mask &= (__force gfp_t)((1 << ROOT_TAG_SHIFT) - 1);
>
> IMO, it would be better to define something for that in radix-tree.h,
> like it has been done for ROOT_IS_IDR.
If we were keeping the radix tree around, I'd agree, but the point of
the rest of this patch set is replacing it ;-) I should probably have
just dropped this patch, to be honest.
From 1584807521595998052@xxx Wed Nov 22 22:37:09 +0000 2017
X-GM-THRID: 1584803315806607151
X-Gmail-Labels: Inbox,Category Forums,HistoricalUnread
On Wed, Nov 22, 2017 at 02:24:02PM -0800, Matthew Wilcox wrote:
> On Wed, Nov 22, 2017 at 10:28:48PM +0100, Luc Van Oostenryck wrote:
> > > - root->gfp_mask &= (1 << ROOT_TAG_SHIFT) - 1;
> > > + root->gfp_mask &= (__force gfp_t)((1 << ROOT_TAG_SHIFT) - 1);
> >
> > IMO, it would be better to define something for that in radix-tree.h,
> > like it has been done for ROOT_IS_IDR.
>
> If we were keeping the radix tree around, I'd agree, but the point of
> the rest of this patch set is replacing it ;-) I should probably have
> just dropped this patch, to be honest.
Ah OK, sure.
I confess I didn't saw the whole series, just this patch.
-- Luc
From 1584803632374546509@xxx Wed Nov 22 21:35:20 +0000 2017
X-GM-THRID: 1584803315806607151
X-Gmail-Labels: Inbox,Category Forums,HistoricalUnread