Re: [PATCH] hugetlb: Fix an error code in hugetlb_reserve_pages()

[Mike Kravetz <mike.kravetz@oracle.com>]

On 12/11/20 2:13 AM, Dan Carpenter wrote:
> Preserve the error code from region_add() instead of returning success.
> 
> Fixes: 0db9d74ed884 ("hugetlb: disable region_add file_region coalescing")
> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
> ---
> From static analysis.  Untested.

Thanks Dan.

> 
>  mm/hugetlb.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index 1f3bf1710b66..ac2e48b9f1d7 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -5113,6 +5113,7 @@ int hugetlb_reserve_pages(struct inode *inode,
>  
>  		if (unlikely(add < 0)) {
>  			hugetlb_acct_memory(h, -gbl_reserve);
> +			ret = add;
>  			goto out_put_pages;
>  		} else if (unlikely(chg > add)) {
>  			/*
> 


That error path is VERY unlikely to be taken, but is indeed incorrect.

When looking at this, I noticed that callers of hugetlb_reserve_pages only
check for 0 or !0.  This changed as the code evolved to add reservation
cgroup support.  The routine type can be changed to a bool and simplified
some.  I'll send that as a follow up patch not for stable.

Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
-- 
Mike Kravetz