Re: [PATCH] fs/dax: Fix pmd vs pte conflict detection

[Matthew Wilcox]

On Sat, Oct 19, 2019 at 09:26:19AM -0700, Dan Williams wrote:
> Check for NULL entries before checking the entry order, otherwise NULL
> is misinterpreted as a present pte conflict. The 'order' check needs to
> happen before the locked check as an unlocked entry at the wrong order
> must fallback to lookup the correct order.
> 
> Reported-by: Jeff Smits <jeff.smits@xxxxxxxxx>
> Reported-by: Doug Nelson <doug.nelson@xxxxxxxxx>
> Cc: <stable@xxxxxxxxxxxxxxx>
> Fixes: 23c84eb78375 ("dax: Fix missed wakeup with PMD faults")
> Cc: Jan Kara <jack@xxxxxxx>
> Cc: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx>
> Signed-off-by: Dan Williams <dan.j.williams@xxxxxxxxx>
> ---
>  fs/dax.c |    5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/dax.c b/fs/dax.c
> index a71881e77204..08160011d94c 100644
> --- a/fs/dax.c
> +++ b/fs/dax.c
> @@ -221,10 +221,11 @@ static void *get_unlocked_entry(struct xa_state *xas, unsigned int order)
>  
>  	for (;;) {
>  		entry = xas_find_conflict(xas);
> +		if (!entry || WARN_ON_ONCE(!xa_is_value(entry)))
> +			return entry;
>  		if (dax_entry_order(entry) < order)
>  			return XA_RETRY_ENTRY;
> -		if (!entry || WARN_ON_ONCE(!xa_is_value(entry)) ||
> -				!dax_is_locked(entry))
> +		if (!dax_is_locked(entry))
>  			return entry;

Yes, I think this works.  Should we also add:

 static unsigned int dax_entry_order(void *entry)
 {
+	BUG_ON(!xa_is_value(entry));
 	if (xa_to_value(entry) & DAX_PMD)
 		return PMD_ORDER;
 	return 0;
 }

which would have caught this logic error before it caused a performance
regression?