[PATCH v1 0/8] Fix set_huge_pte_at() panic on arm64

[Ryan Roberts]

On 21/09/2023 17:30, Andrew Morton wrote:
> On Thu, 21 Sep 2023 17:19:59 +0100 Ryan Roberts <ryan.roberts at arm.com> wrote:
> 
>> Hi All,
>>
>> This series fixes a bug in arm64's implementation of set_huge_pte_at(), which
>> can result in an unprivileged user causing a kernel panic. The problem was
>> triggered when running the new uffd poison mm selftest for HUGETLB memory. This
>> test (and the uffd poison feature) was merged for v6.6-rc1. However, upon
>> inspection there are multiple other pre-existing paths that can trigger this
>> bug.
>>
>> Ideally, I'd like to get this fix in for v6.6 if possible? And I guess it should
>> be backported too, given there are call sites where this can theoretically
>> happen that pre-date v6.6-rc1 (I've cc'ed stable at vger.kernel.org).
> 
> This gets you a naggygram from Greg.  The way to request a backport is
> to add cc:stable to all the changelogs.  I'll make that change to my copy.

Ahh, sorry about that... I just got the same moan from the kernel test robot too.

> 
> 
>> Ryan Roberts (8):
>>   parisc: hugetlb: Convert set_huge_pte_at() to take vma
>>   powerpc: hugetlb: Convert set_huge_pte_at() to take vma
>>   riscv: hugetlb: Convert set_huge_pte_at() to take vma
>>   s390: hugetlb: Convert set_huge_pte_at() to take vma
>>   sparc: hugetlb: Convert set_huge_pte_at() to take vma
>>   mm: hugetlb: Convert set_huge_pte_at() to take vma
>>   arm64: hugetlb: Convert set_huge_pte_at() to take vma
>>   arm64: hugetlb: Fix set_huge_pte_at() to work with all swap entries
>>
>>  arch/arm64/include/asm/hugetlb.h              |  2 +-
>>  arch/arm64/mm/hugetlbpage.c                   | 22 ++++----------
>>  arch/parisc/include/asm/hugetlb.h             |  2 +-
>>  arch/parisc/mm/hugetlbpage.c                  |  4 +--
>>  .../include/asm/nohash/32/hugetlb-8xx.h       |  3 +-
>>  arch/powerpc/mm/book3s64/hugetlbpage.c        |  2 +-
>>  arch/powerpc/mm/book3s64/radix_hugetlbpage.c  |  2 +-
>>  arch/powerpc/mm/nohash/8xx.c                  |  2 +-
>>  arch/powerpc/mm/pgtable.c                     |  7 ++++-
>>  arch/riscv/include/asm/hugetlb.h              |  2 +-
>>  arch/riscv/mm/hugetlbpage.c                   |  3 +-
>>  arch/s390/include/asm/hugetlb.h               |  8 +++--
>>  arch/s390/mm/hugetlbpage.c                    |  8 ++++-
>>  arch/sparc/include/asm/hugetlb.h              |  8 +++--
>>  arch/sparc/mm/hugetlbpage.c                   |  8 ++++-
>>  include/asm-generic/hugetlb.h                 |  6 ++--
>>  include/linux/hugetlb.h                       |  6 ++--
>>  mm/damon/vaddr.c                              |  2 +-
>>  mm/hugetlb.c                                  | 30 +++++++++----------
>>  mm/migrate.c                                  |  2 +-
>>  mm/rmap.c                                     | 10 +++----
>>  mm/vmalloc.c                                  |  5 +++-
>>  22 files changed, 80 insertions(+), 64 deletions(-)
> 
> Looks scary but it's actually a fairly modest patchset.  It could
> easily be all rolled into a single patch for ease of backporting. 
> Maybe Greg has an opinion?

Yes, I thought about doing that; or perhaps 2 patches - one for the interface
change across all arches and core code, and one for the actual bug fix?

But I thought the arch people might prefer to see exactly what's going on in
each arch. Let me know the preference and I can repost if necessary.

>