Re: [PATCH 6.12.y v2] mm/mmap: fix __mmap_region() error handling in rare merge failure case

[Lorenzo Stoakes <lorenzo.stoakes@oracle.com>]

On Tue, Nov 19, 2024 at 12:59:45PM -0500, Liam R. Howlett wrote:
> From: "Liam R. Howlett" <Liam.Howlett@Oracle.com>
>
> The mmap_region() function tries to install a new vma, which requires a
> pre-allocation for the maple tree write due to the complex locking
> scenarios involved.
>
> Recent efforts to simplify the error recovery required the relocation of
> the preallocation of the maple tree nodes (via vma_iter_prealloc()
> calling mas_preallocate()) higher in the function.
>
> The relocation of the preallocation meant that, if there was a file
> associated with the vma and the driver call (mmap_file()) modified the
> vma flags, then a new merge of the new vma with existing vmas is
> attempted.
>
> During the attempt to merge the existing vma with the new vma, the vma
> iterator is used - the same iterator that would be used for the next
> write attempt to the tree.  In the event of needing a further allocation
> and if the new allocations fails, the vma iterator (and contained maple
> state) will cleaned up, including freeing all previous allocations and
> will be reset internally.
>
> Upon returning to the __mmap_region() function, the error is available
> in the vma_merge_struct and can be used to detect the -ENOMEM status.
>
> Hitting an -ENOMEM scenario after the driver callback leaves the system
> in a state that undoing the mapping is worse than continuing by dipping
> into the reserve.
>
> A preallocation should be performed in the case of an -ENOMEM and the
> allocations were lost during the failure scenario.  The __GFP_NOFAIL
> flag is used in the allocation to ensure the allocation succeeds after
> implicitly telling the driver that the mapping was happening.
>
> The range is already set in the vma_iter_store() call below, so it is
> not necessary and is dropped.
>
> Reported-by: syzbot+bc6bfc25a68b7a020ee1@syzkaller.appspotmail.com
> Closes: https://syzkaller.appspot.com/x/log.txt?x=17b0ace8580000
> Fixes: 5de195060b2e2 ("mm: resolve faulty mmap_region() error path behaviour")
> Signed-off-by: Liam R. Howlett <Liam.Howlett@Oracle.com>
> Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> Cc: Vlastimil Babka <vbabka@suse.cz>
> Cc: Jann Horn <jannh@google.com>
> Cc: <stable@vger.kernel.org>

Looks good to me:

Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

I mean ideally we'd not have to handle this scenario, but 6.13 resolves
this in a different way, and since this will never nearly happen (perhaps
actually never in reality), I think having an operation that will nearly
always be a no-op beats out alternative solutions.

> ---
>  mm/mmap.c | 13 ++++++++++++-
>  1 file changed, 12 insertions(+), 1 deletion(-)
>
> Changes since v1:
>  - Don't bail out and force the allocation when the merge failure is
>    -ENOMEM - Thanks Lorenzo
>
> diff --git a/mm/mmap.c b/mm/mmap.c
> index 79d541f1502b2..4f6e566d52faa 100644
> --- a/mm/mmap.c
> +++ b/mm/mmap.c
> @@ -1491,7 +1491,18 @@ static unsigned long __mmap_region(struct file *file, unsigned long addr,
>  				vm_flags = vma->vm_flags;
>  				goto file_expanded;
>  			}
> -			vma_iter_config(&vmi, addr, end);
> +
> +			/*
> +			 * In the unlikely even that more memory was needed, but
> +			 * not available for the vma merge, the vma iterator
> +			 * will have no memory reserved for the write we told
> +			 * the driver was happening.  To keep up the ruse,
> +			 * ensure the allocation for the store succeeds.
> +			 */
> +			if (vmg_nomem(&vmg)) {
> +				mas_preallocate(&vmi.mas, vma,
> +						GFP_KERNEL|__GFP_NOFAIL);
> +			}
>  		}
>
>  		vm_flags = vma->vm_flags;
> --
> 2.43.0
>