Fix a possible memory leak in __stmmac_open when stmmac_init_phy fails.
It's also needed to free everything allocated by stmmac_setup_dma_desc
and not just the dma_conf struct.

Correctly call free_dma_desc_resources on the new dma_conf passed to
__stmmac_open on error.

Reported-by: Jose Abreu <Jose.Abreu@synopsys.com>
Fixes: ba39b344e924 ("net: ethernet: stmicro: stmmac: generate stmmac dma conf before open")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Cc: stable@vger.kernel.org
---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index fa07b0d50b46..0966ab86fde2 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -3877,10 +3877,10 @@ static int __stmmac_open(struct net_device *dev,
 
 	stmmac_hw_teardown(dev);
 init_error:
-	free_dma_desc_resources(priv, &priv->dma_conf);
 	phylink_disconnect_phy(priv->phylink);
 init_phy_error:
 	pm_runtime_put(priv->device);
+	free_dma_desc_resources(priv, dma_conf);
 	return ret;
 }
 
-- 
2.40.1

On Wed, Jun 14, 2023 at 09:32:41AM +0200, Christian Marangi wrote:
> Fix a possible memory leak in __stmmac_open when stmmac_init_phy fails.
> It's also needed to free everything allocated by stmmac_setup_dma_desc
> and not just the dma_conf struct.
> 
> Correctly call free_dma_desc_resources on the new dma_conf passed to
> __stmmac_open on error.
> 
> Reported-by: Jose Abreu <Jose.Abreu@synopsys.com>
> Fixes: ba39b344e924 ("net: ethernet: stmicro: stmmac: generate stmmac dma conf before open")
> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> Cc: stable@vger.kernel.org
> ---
>  drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> index fa07b0d50b46..0966ab86fde2 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> @@ -3877,10 +3877,10 @@ static int __stmmac_open(struct net_device *dev,
>  
>  	stmmac_hw_teardown(dev);
>  init_error:
> -	free_dma_desc_resources(priv, &priv->dma_conf);
>  	phylink_disconnect_phy(priv->phylink);
>  init_phy_error:
>  	pm_runtime_put(priv->device);
> +	free_dma_desc_resources(priv, dma_conf);

Hi Christian,

Are these resources allocated by the caller?
If so, perhaps it would be clearer if a symmetric approach
was taken and the caller handled freeing them on error.

>  	return ret;
>  }
>  
> -- 
> 2.40.1
> 
>

On Wed, Jun 14, 2023 at 04:55:10PM +0200, Simon Horman wrote:
> On Wed, Jun 14, 2023 at 09:32:41AM +0200, Christian Marangi wrote:
> > Fix a possible memory leak in __stmmac_open when stmmac_init_phy fails.
> > It's also needed to free everything allocated by stmmac_setup_dma_desc
> > and not just the dma_conf struct.
> > 
> > Correctly call free_dma_desc_resources on the new dma_conf passed to
> > __stmmac_open on error.
> > 
> > Reported-by: Jose Abreu <Jose.Abreu@synopsys.com>
> > Fixes: ba39b344e924 ("net: ethernet: stmicro: stmmac: generate stmmac dma conf before open")
> > Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> > Cc: stable@vger.kernel.org
> > ---
> >  drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> > index fa07b0d50b46..0966ab86fde2 100644
> > --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> > +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> > @@ -3877,10 +3877,10 @@ static int __stmmac_open(struct net_device *dev,
> >  
> >  	stmmac_hw_teardown(dev);
> >  init_error:
> > -	free_dma_desc_resources(priv, &priv->dma_conf);
> >  	phylink_disconnect_phy(priv->phylink);
> >  init_phy_error:
> >  	pm_runtime_put(priv->device);
> > +	free_dma_desc_resources(priv, dma_conf);
> 
> Hi Christian,
> 
> Are these resources allocated by the caller?
> If so, perhaps it would be clearer if a symmetric approach
> was taken and the caller handled freeing them on error.
>

Yes, they are. Handling in the caller would require some additional
delta to this and some duplicated code but if preferred I can implement
it. I can provide a v2 shortly if it's ok and you prefer this
implementation.

> >  	return ret;
> >  }
> >  
> > -- 
> > 2.40.1
> > 
> > 

-- 
	Ansuel

On Wed, Jun 14, 2023 at 10:18:20AM +0200, Christian Marangi wrote:
> On Wed, Jun 14, 2023 at 04:55:10PM +0200, Simon Horman wrote:
> > On Wed, Jun 14, 2023 at 09:32:41AM +0200, Christian Marangi wrote:
> > > Fix a possible memory leak in __stmmac_open when stmmac_init_phy fails.
> > > It's also needed to free everything allocated by stmmac_setup_dma_desc
> > > and not just the dma_conf struct.
> > > 
> > > Correctly call free_dma_desc_resources on the new dma_conf passed to
> > > __stmmac_open on error.
> > > 
> > > Reported-by: Jose Abreu <Jose.Abreu@synopsys.com>
> > > Fixes: ba39b344e924 ("net: ethernet: stmicro: stmmac: generate stmmac dma conf before open")
> > > Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> > > Cc: stable@vger.kernel.org
> > > ---
> > >  drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 2 +-
> > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > 
> > > diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> > > index fa07b0d50b46..0966ab86fde2 100644
> > > --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> > > +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> > > @@ -3877,10 +3877,10 @@ static int __stmmac_open(struct net_device *dev,
> > >  
> > >  	stmmac_hw_teardown(dev);
> > >  init_error:
> > > -	free_dma_desc_resources(priv, &priv->dma_conf);
> > >  	phylink_disconnect_phy(priv->phylink);
> > >  init_phy_error:
> > >  	pm_runtime_put(priv->device);
> > > +	free_dma_desc_resources(priv, dma_conf);
> > 
> > Hi Christian,
> > 
> > Are these resources allocated by the caller?
> > If so, perhaps it would be clearer if a symmetric approach
> > was taken and the caller handled freeing them on error.
> >
> 
> Yes, they are. Handling in the caller would require some additional
> delta to this and some duplicated code but if preferred I can implement
> it. I can provide a v2 shortly if it's ok and you prefer this
> implementation.

Thanks, I think that would be best,
unless the new delta is enormous for some reason.

-- 
pw-bot: cr