[lkml]   [2020]   [Jan]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH -next v2] mm/hotplug: silence a lockdep splat with printk()
On Wed 15-01-20 07:35:45, Qian Cai wrote:
> > On Jan 15, 2020, at 4:22 AM, Michal Hocko <> wrote:
> >
> > Is this correct? CMA pages cannot be comound?
> I never saw a CMA page also a compound page. Also, in
> alloc_contig_range(), it calls migrate_pages()—>
> unmap_and_move_huge_page() which will free the old huge page, so I
> think that will clear pageCompound.

This sounds like the implementation detail. I do not think you want to
depend on that. Anyway, see my other response. If you rebase on top of
Vlastimil's patch you do not really have to care.
Michal Hocko

 \ /
  Last update: 2020-01-15 15:46    [W:0.058 / U:3.416 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site