Re: [PATCH 6.18.y] kho: init alloc tags when restoring pages from reserved memory

[Sasha Levin <sashal@kernel.org>]

On Wed, Feb 04, 2026 at 06:06:16PM +0100, Greg KH wrote:
>On Wed, Feb 04, 2026 at 08:46:35AM -0800, Suren Baghdasaryan wrote:
>> On Wed, Feb 4, 2026 at 1:59 AM Greg KH <gregkh@linuxfoundation.org> wrote:
>> >
>> > On Tue, Feb 03, 2026 at 07:26:54PM -0500, Sasha Levin wrote:
>> > > From: Ran Xiaokai <ran.xiaokai@zte.com.cn>
>> > >
>> > > [ Upstream commit e86436ad0ad2a9aaf88802d69b68f02cbd1f04a9 ]
>> > >
>> > > Memblock pages (including reserved memory) should have their allocation
>> > > tags initialized to CODETAG_EMPTY via clear_page_tag_ref() before being
>> > > released to the page allocator.  When kho restores pages through
>> > > kho_restore_page(), missing this call causes mismatched
>> > > allocation/deallocation tracking and below warning message:
>> > >
>> > > alloc_tag was not set
>> > > WARNING: include/linux/alloc_tag.h:164 at ___free_pages+0xb8/0x260, CPU#1: swapper/0/1
>> > > RIP: 0010:___free_pages+0xb8/0x260
>> > >  kho_restore_vmalloc+0x187/0x2e0
>> > >  kho_test_init+0x3c4/0xa30
>> > >  do_one_initcall+0x62/0x2b0
>> > >  kernel_init_freeable+0x25b/0x480
>> > >  kernel_init+0x1a/0x1c0
>> > >  ret_from_fork+0x2d1/0x360
>> > >
>> > > Add missing clear_page_tag_ref() annotation in kho_restore_page() to
>> > > fix this.
>> > >
>> > > Link: https://lkml.kernel.org/r/20260122132740.176468-1-ranxiaokai627@163.com
>> > > Fixes: fc33e4b44b27 ("kexec: enable KHO support for memory preservation")
>> > > Signed-off-by: Ran Xiaokai <ran.xiaokai@zte.com.cn>
>> > > Reviewed-by: Pratyush Yadav <pratyush@kernel.org>
>> > > Reviewed-by: Pasha Tatashin <pasha.tatashin@soleen.com>
>> > > Reviewed-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
>> > > Cc: Alexander Graf <graf@amazon.com>
>> > > Cc: Kent Overstreet <kent.overstreet@linux.dev>
>> > > Cc: Suren Baghdasaryan <surenb@google.com>
>> > > Cc: <stable@vger.kernel.org>
>> > > Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
>> > > Signed-off-by: Sasha Levin <sashal@kernel.org>
>> > > ---
>> > >  kernel/kexec_handover.c | 8 ++++++++
>> > >  1 file changed, 8 insertions(+)
>> > >
>> > > diff --git a/kernel/kexec_handover.c b/kernel/kexec_handover.c
>> > > index 03d12e27189fc..db08c1a2e1f80 100644
>> > > --- a/kernel/kexec_handover.c
>> > > +++ b/kernel/kexec_handover.c
>> > > @@ -260,6 +260,14 @@ static struct page *kho_restore_page(phys_addr_t phys)
>> > >       if (info.order > 0)
>> > >               prep_compound_page(page, info.order);
>> > >
>> > > +     /* Always mark headpage's codetag as empty to avoid accounting mismatch */
>> > > +     clear_page_tag_ref(page);
>> > > +     if (!is_folio) {
>> > > +             /* Also do that for the non-compound tail pages */
>> > > +             for (unsigned int i = 1; i < nr_pages; i++)
>> > > +                     clear_page_tag_ref(page + i);
>> > > +     }
>> > > +
>> >
>> > Breaks the build :(
>>
>> Which config? I built both defconfig and CONFIG_MEM_ALLOC_PROFILING=y,
>> they didn't fail. Could you please send me your failing config?
>
>is_folio is not defined in this function, how are you even building this
>file?

This is a stupid issue I've hit last time too: allmodconfig sets
CONFIG_DEFERRED_STRUCT_PAGE_INIT=y which disables KHO.

I'll figure out a patch that still builds KHO if CONFIG_COMPILE_TEST is set,
unless someone beats be to it...

-- 
Thanks,
Sasha