This is incomplete so please hold off on this patch for now.
On Mon, Jun 3, 2024 at 5:47 PM Leah Rumancik <leah.rumancik@xxxxxxxxx> wrote:
>
> The original backport didn't move the code to link the vma into the MT
> and also the code to increment the map_count causing ~15 xfstests
> (including ext4/303 generic/051 generic/054 generic/069) to hard fail
> on some platforms. This patch resolves test failures.
>
> Fixes: 0c42f7e039ab ("fork: defer linking file vma until vma is fully initialized")
> Signed-off-by: Leah Rumancik <leah.rumancik@xxxxxxxxx>
> ---
> kernel/fork.c | 17 +++++++++--------
> 1 file changed, 9 insertions(+), 8 deletions(-)
>
> diff --git a/kernel/fork.c b/kernel/fork.c
> index 7e9a5919299b..3b44960b1385 100644
> --- a/kernel/fork.c
> +++ b/kernel/fork.c
> @@ -668,6 +668,15 @@ static __latent_entropy int dup_mmap(struct mm_struct *mm,
> if (is_vm_hugetlb_page(tmp))
> hugetlb_dup_vma_private(tmp);
>
> + /* Link the vma into the MT */
> + mas.index = tmp->vm_start;
> + mas.last = tmp->vm_end - 1;
> + mas_store(&mas, tmp);
> + if (mas_is_err(&mas))
> + goto fail_nomem_mas_store;
> +
> + mm->map_count++;
> +
> if (tmp->vm_ops && tmp->vm_ops->open)
> tmp->vm_ops->open(tmp);
>
> @@ -687,14 +696,6 @@ static __latent_entropy int dup_mmap(struct mm_struct *mm,
> i_mmap_unlock_write(mapping);
> }
>
> - /* Link the vma into the MT */
> - mas.index = tmp->vm_start;
> - mas.last = tmp->vm_end - 1;
> - mas_store(&mas, tmp);
> - if (mas_is_err(&mas))
> - goto fail_nomem_mas_store;
> -
> - mm->map_count++;
> if (!(tmp->vm_flags & VM_WIPEONFORK))
> retval = copy_page_range(tmp, mpnt);
>
> --
> 2.45.1.288.g0e0cd299f1-goog
>
