I actually have the same patch sitting on my branch, reported by Dan. I
have been testing this code before sending it out.
Reviewed-by: Liam R. Howlett <Liam.Howlett@xxxxxxxxxx>
* Yang Yingliang <yangyingliang@xxxxxxxxxx> [220623 09:32]:
> Return error code when munmap_sidetree() fails in do_mas_align_munmap().
>
> Fixes: 81f5504dfb36 ("mm/mmap: change do_mas_align_munmap() to avoid preallocations for sidetree")
> Reported-by: Hulk Robot <hulkci@xxxxxxxxxx>
> Signed-off-by: Yang Yingliang <yangyingliang@xxxxxxxxxx>
> ---
> mm/mmap.c | 6 ++++--
> 1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/mm/mmap.c b/mm/mmap.c
> index 1bdf56804048..bb978061a790 100644
> --- a/mm/mmap.c
> +++ b/mm/mmap.c
> @@ -2427,7 +2427,8 @@ do_mas_align_munmap(struct ma_state *mas, struct vm_area_struct *vma,
>
> mas_set(mas, end);
> split = mas_prev(mas, 0);
> - if (munmap_sidetree(split, &mas_detach))
> + error = munmap_sidetree(split, &mas_detach);
> + if (error)
> goto munmap_sidetree_failed;
>
> count++;
> @@ -2435,7 +2436,8 @@ do_mas_align_munmap(struct ma_state *mas, struct vm_area_struct *vma,
> vma = split;
> break;
> }
> - if (munmap_sidetree(next, &mas_detach))
> + error = munmap_sidetree(next, &mas_detach);
> + if (error)
> goto munmap_sidetree_failed;
>
> count++;
> --
> 2.25.1
>
