Re: [PATCH 5/6] xen/gnttab: Fold error handling in _cache_flush()

[Jan Beulich <jbeulich@xxxxxxxx>]

On 03.11.2026 16:29, Andrew Cooper wrote:
> No functional change.
> 
> Signed-off-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>

Acked-by: Jan Beulich <jbeulich@xxxxxxxx>

> @@ -3683,13 +3682,15 @@ static int _cache_flush(gnttab_cache_flush_t *cflush, 
> grant_ref_t *cur_ref)
>      else
>          ret = 0;
>  
> +    unmap_domain_page(v);
> +
>      if ( d != owner )
>      {
>          active_entry_release(act);

I was first meaning to complain about the unnecessary new lock nesting here,
but the earlier map_domain_page() also is issued with the active entry lock
held. You move things to a more consistent pattern this way, which is good.
Yet at the same time avoiding that nesting might also be a good idea.
Something for a later change, I guess.

Jan

> +    gnt_unlock:
>          grant_read_unlock(owner->grant_table);
>      }
>  
> -    unmap_domain_page(v);
>      put_page(page);
>  
>      return ret;