[PATCH v3 1/2] net: page_pool: check page pool ethtool stats

[PATCH v3 1/2] net: page_pool: check page pool ethtool stats

[Sven Auhagen]

If the page_pool variable is null while passing it to
the page_pool_get_stats function we receive a kernel error.

Check if the page_pool variable is at least valid.

Signed-off-by: Sven Auhagen <sven.auhagen@voleatech.de>
Reported-by: Paulo Da Silva <Paulo.DaSilva@kyberna.com>
---
 net/core/page_pool.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/core/page_pool.c b/net/core/page_pool.c
index 2396c99bedea..4c5dca6b4a16 100644
--- a/net/core/page_pool.c
+++ b/net/core/page_pool.c
@@ -65,6 +65,9 @@ bool page_pool_get_stats(struct page_pool *pool,
 	if (!stats)
 		return false;
 
+	if (!pool)
+		return false;
+
 	/* The caller is responsible to initialize stats. */
 	stats->alloc_stats.fast += pool->alloc_stats.fast;
 	stats->alloc_stats.slow += pool->alloc_stats.slow;
-- 
2.42.0

Re: [PATCH v3 1/2] net: page_pool: check page pool ethtool stats

[Leon Romanovsky]

On Sun, Oct 15, 2023 at 02:37:27PM +0200, Sven Auhagen wrote:
> If the page_pool variable is null while passing it to
> the page_pool_get_stats function we receive a kernel error.
> 
> Check if the page_pool variable is at least valid.
> 
> Signed-off-by: Sven Auhagen <sven.auhagen@voleatech.de>
> Reported-by: Paulo Da Silva <Paulo.DaSilva@kyberna.com>
> ---
>  net/core/page_pool.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/net/core/page_pool.c b/net/core/page_pool.c
> index 2396c99bedea..4c5dca6b4a16 100644
> --- a/net/core/page_pool.c
> +++ b/net/core/page_pool.c
> @@ -65,6 +65,9 @@ bool page_pool_get_stats(struct page_pool *pool,
>  	if (!stats)
>  		return false;
>  
> +	if (!pool)
> +		return false;
> +

I would argue that both pool and stats shouldn't be NULL and must be
checked by caller. This API call named get-stats-from-pool.

Thanks

>  	/* The caller is responsible to initialize stats. */
>  	stats->alloc_stats.fast += pool->alloc_stats.fast;
>  	stats->alloc_stats.slow += pool->alloc_stats.slow;
> -- 
> 2.42.0
> 
> 

Re: [PATCH v3 1/2] net: page_pool: check page pool ethtool stats

[Sven Auhagen]

On Sun, Oct 15, 2023 at 08:27:10PM +0300, Leon Romanovsky wrote:
> On Sun, Oct 15, 2023 at 02:37:27PM +0200, Sven Auhagen wrote:
> > If the page_pool variable is null while passing it to
> > the page_pool_get_stats function we receive a kernel error.
> > 
> > Check if the page_pool variable is at least valid.
> > 
> > Signed-off-by: Sven Auhagen <sven.auhagen@voleatech.de>
> > Reported-by: Paulo Da Silva <Paulo.DaSilva@kyberna.com>
> > ---
> >  net/core/page_pool.c | 3 +++
> >  1 file changed, 3 insertions(+)
> > 
> > diff --git a/net/core/page_pool.c b/net/core/page_pool.c
> > index 2396c99bedea..4c5dca6b4a16 100644
> > --- a/net/core/page_pool.c
> > +++ b/net/core/page_pool.c
> > @@ -65,6 +65,9 @@ bool page_pool_get_stats(struct page_pool *pool,
> >  	if (!stats)
> >  		return false;
> >  
> > +	if (!pool)
> > +		return false;
> > +
> 
> I would argue that both pool and stats shouldn't be NULL and must be
> checked by caller. This API call named get-stats-from-pool.
> 
> Thanks

I can do it either way, I only need to know which version to send.
Can someone let me know how to proceed with the patch?

Best
Sven

> 
> >  	/* The caller is responsible to initialize stats. */
> >  	stats->alloc_stats.fast += pool->alloc_stats.fast;
> >  	stats->alloc_stats.slow += pool->alloc_stats.slow;
> > -- 
> > 2.42.0
> > 
> > 

Re: [PATCH v3 1/2] net: page_pool: check page pool ethtool stats

[Yunsheng Lin]

On 2023/10/25 10:12, Sven Auhagen wrote:
> On Sun, Oct 15, 2023 at 08:27:10PM +0300, Leon Romanovsky wrote:
>> On Sun, Oct 15, 2023 at 02:37:27PM +0200, Sven Auhagen wrote:
>>> If the page_pool variable is null while passing it to
>>> the page_pool_get_stats function we receive a kernel error.
>>>
>>> Check if the page_pool variable is at least valid.
>>>
>>> Signed-off-by: Sven Auhagen <sven.auhagen@voleatech.de>
>>> Reported-by: Paulo Da Silva <Paulo.DaSilva@kyberna.com>
>>> ---
>>>  net/core/page_pool.c | 3 +++
>>>  1 file changed, 3 insertions(+)
>>>
>>> diff --git a/net/core/page_pool.c b/net/core/page_pool.c
>>> index 2396c99bedea..4c5dca6b4a16 100644
>>> --- a/net/core/page_pool.c
>>> +++ b/net/core/page_pool.c
>>> @@ -65,6 +65,9 @@ bool page_pool_get_stats(struct page_pool *pool,
>>>  	if (!stats)
>>>  		return false;
>>>  
>>> +	if (!pool)
>>> +		return false;
>>> +
>>
>> I would argue that both pool and stats shouldn't be NULL and must be
>> checked by caller. This API call named get-stats-from-pool.
>>
>> Thanks
> 
> I can do it either way, I only need to know which version to send.
> Can someone let me know how to proceed with the patch?

From the message in v1, it seems more appropriate to check it in
the driver, because:
1. if the driver only support page pool, then the check is not
   really necessary.
2. if the driver support both page_pool API and alloc_pages()
   API, then a check is already needed in the driver to switch
   between those two API, it might be better to reuse that check.

"First the page pool is only available if the bm is not used.
The page pool is also not allocated when the port is stopped.
It can also be not allocated in case of errors."

> 
> Best
> Sven
> 
>>
>>>  	/* The caller is responsible to initialize stats. */
>>>  	stats->alloc_stats.fast += pool->alloc_stats.fast;
>>>  	stats->alloc_stats.slow += pool->alloc_stats.slow;
>>> -- 
>>> 2.42.0
>>>
>>>
> 
> .
>