* Michael Ellerman <mpe@xxxxxxxxxxxxxx> [240812 04:26]:
> Add an optional close() callback to struct vm_special_mapping. It will
> be used, by powerpc at least, to handle unmapping of the VDSO.
>
> Although support for unmapping the VDSO was initially added
> for CRIU[1], it is not desirable to guard that support behind
> CONFIG_CHECKPOINT_RESTORE.
>
> There are other known users of unmapping the VDSO which are not related
> to CRIU, eg. Valgrind [2] and void-ship [3].
>
> The powerpc arch_unmap() hook has been in place for ~9 years, with no
> ifdef, so there may be other unknown users that have come to rely on
> unmapping the VDSO. Even if the code was behind an ifdef, major distros
> enable CHECKPOINT_RESTORE so users may not realise unmapping the VDSO
> depends on that configuration option.
>
> It's also undesirable to have such core mm behaviour behind a relatively
> obscure CONFIG option.
>
> Longer term the unmap behaviour should be standardised across
> architectures, however that is complicated by the fact the VDSO pointer
> is stored differently across architectures. There was a previous attempt
> to unify that handling [4], which could be revived.
>
> See [5] for further discussion.
Thank you for tackling this issue, it's much improved.
For the whole series:
Reviewed-by: Liam R. Howlett <Liam.Howlett@xxxxxxxxxx>
>
> [1]: commit 83d3f0e90c6c ("powerpc/mm: tracking vDSO remap")
> [2]: https://sourceware.org/git/?p=valgrind.git;a=commit;h=3a004915a2cbdcdebafc1612427576bf3321eef5
> [3]: https://github.com/insanitybit/void-ship
> [4]: https://lore.kernel.org/lkml/20210611180242.711399-17-dima@xxxxxxxxxx/
> [5]: https://lore.kernel.org/linuxppc-dev/shiq5v3jrmyi6ncwke7wgl76ojysgbhrchsk32q4lbx2hadqqc@kzyy2igem256
>
> Suggested-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
> Signed-off-by: Michael Ellerman <mpe@xxxxxxxxxxxxxx>
> Reviewed-by: David Hildenbrand <david@xxxxxxxxxx>
> ---
> include/linux/mm_types.h | 3 +++
> mm/mmap.c | 6 ++++++
> 2 files changed, 9 insertions(+)
>
> v2:
> - Add some blank lines as requested.
> - Expand special_mapping_close() comment.
> - Add David's reviewed-by.
> - Expand change log to capture review discussion.
>
> diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
> index 485424979254..78bdfc59abe5 100644
> --- a/include/linux/mm_types.h
> +++ b/include/linux/mm_types.h
> @@ -1313,6 +1313,9 @@ struct vm_special_mapping {
>
> int (*mremap)(const struct vm_special_mapping *sm,
> struct vm_area_struct *new_vma);
> +
> + void (*close)(const struct vm_special_mapping *sm,
> + struct vm_area_struct *vma);
> };
>
> enum tlb_flush_reason {
> diff --git a/mm/mmap.c b/mm/mmap.c
> index d0dfc85b209b..af4dbf0d3bd4 100644
> --- a/mm/mmap.c
> +++ b/mm/mmap.c
> @@ -3620,10 +3620,16 @@ void vm_stat_account(struct mm_struct *mm, vm_flags_t flags, long npages)
> static vm_fault_t special_mapping_fault(struct vm_fault *vmf);
>
> /*
> + * Close hook, called for unmap() and on the old vma for mremap().
> + *
> * Having a close hook prevents vma merging regardless of flags.
> */
> static void special_mapping_close(struct vm_area_struct *vma)
> {
> + const struct vm_special_mapping *sm = vma->vm_private_data;
> +
> + if (sm->close)
> + sm->close(sm, vma);
> }
>
> static const char *special_mapping_name(struct vm_area_struct *vma)
> --
> 2.45.2
>
