Re: + mm-fix-huge-page-table-not-free-after-memory-unplug.patch added to mm-hotfixes-unstable branch

[Mike Rapoport <rppt@kernel.org>]

On Tue, Dec 23, 2025 at 09:27:52AM +0100, David Hildenbrand (Red Hat) wrote:
> On 12/23/25 02:15, Andrew Morton wrote:
>
> > From: Yuan Liu <yuan1.liu@intel.com>
> > Subject: mm: fix huge page table not free after memory unplug
> > Date: Sun, 21 Dec 2025 23:11:17 -0500
> > 
> > Newly plugged memory is marked as prot_sethuge via phys_pmd_init without
> > setting PG_head.  During memory unplug, free_hugepage_table frees the page
> > table as 2M, but pagetable_free handles it as 4K.
> > 
> > Link: https://lkml.kernel.org/r/20251222041117.44865-1-yuan1.liu@intel.com
> > Fixes: bf9e4e30f353 ("x86/mm: use pagetable_free()")
> > Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
> > Cc: Jason Gunthorpe <jgg@nvidia.com>
> > Cc: Baolu Lu <baolu.lu@linux.intel.com>
> > Cc: David Hildenbrand <david@redhat.com>
> > Cc: Mike Rapoport <rppt@kernel.org>
> > Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
> > ---
> > 
> >   arch/x86/mm/init_64.c |    2 +-
> >   1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > --- a/arch/x86/mm/init_64.c~mm-fix-huge-page-table-not-free-after-memory-unplug
> > +++ a/arch/x86/mm/init_64.c
> > @@ -1028,7 +1028,7 @@ static void __meminit free_pagetable(str
> >   		free_reserved_pages(page, nr_pages);
> >   #endif
> >   	} else {
> > -		pagetable_free(page_ptdesc(page));
> > +		__free_pages(page, order);
> 
> Hm, but this way we are skipping deconstructors, no?
> 
> In particular, clearing the kernel flag (ptdesc_clear_kernel()) and not
> handling ASYNC_KERNEL_PGTABLE_FREE correctly.

And it seems to be x86 material at any rate.

-- 
Sincerely yours,
Mike.