[PATCH 1/1] pci: host: pci-layerscape: add missing of_node_put after calling of_parse_phandle

[PATCH 1/1] pci: host: pci-layerscape: add missing of_node_put after calling of_parse_phandle

[Peter Chen]

of_node_put needs to be called when the device node which is got
from of_parse_phandle has finished using.

Cc: Minghuan Lian <minghuan.Lian@freescale.com>
Cc: Mingkai Hu <mingkai.hu@freescale.com>
Cc: Roy Zang <tie-fei.zang@freescale.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>
---
 drivers/pci/host/pci-layerscape.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/pci/host/pci-layerscape.c b/drivers/pci/host/pci-layerscape.c
index 114ba81..573b996 100644
--- a/drivers/pci/host/pci-layerscape.c
+++ b/drivers/pci/host/pci-layerscape.c
@@ -173,6 +173,8 @@ static int ls_pcie_msi_host_init(struct pcie_port *pp,
 		return -EINVAL;
 	}
 
+	of_node_put(msi_node);
+
 	return 0;
 }
 
-- 
1.9.1

Re: [PATCH 1/1] pci: host: pci-layerscape: add missing of_node_put after calling of_parse_phandle

[Bjorn Helgaas]

Hi Peter,

On Fri, Aug 12, 2016 at 09:34:41AM +0800, Peter Chen wrote:
> of_node_put needs to be called when the device node which is got
> from of_parse_phandle has finished using.
> 
> Cc: Minghuan Lian <minghuan.Lian@freescale.com>
> Cc: Mingkai Hu <mingkai.hu@freescale.com>
> Cc: Roy Zang <tie-fei.zang@freescale.com>
> Signed-off-by: Peter Chen <peter.chen@nxp.com>
> ---
>  drivers/pci/host/pci-layerscape.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/pci/host/pci-layerscape.c b/drivers/pci/host/pci-layerscape.c
> index 114ba81..573b996 100644
> --- a/drivers/pci/host/pci-layerscape.c
> +++ b/drivers/pci/host/pci-layerscape.c
> @@ -173,6 +173,8 @@ static int ls_pcie_msi_host_init(struct pcie_port *pp,
>  		return -EINVAL;
>  	}
>  
> +	of_node_put(msi_node);
> +

Can you please look for and fix similar errors in other drivers in
drivers/pci/host/*?

For example:

  advk_pcie_probe() and iproc_pcie_msi_enable() call
  of_parse_phandle() but don't call of_node_put() in failure paths.

  dra7xx_pcie_init_irq_domain(), nwl_pcie_init_irq_domain(), and
  xilinx_pcie_init_irq_domain() call of_get_next_child() but don't
  call of_node_put() in failure paths.

  ks_pcie_get_irq_controller_info() calls of_find_node_by_name().

I think there may be others, e.g., the pci_host_bridge_msi_domain()
path calls of_parse_phandle(), but I'm not sure of_node_put() is
called on failure paths.

When we find bugs like this, it's nice to fix one occurrence, but it's
really great if we can squash similar bugs nearby so the bug isn't
copied into new drivers.

>  	return 0;
>  }
>  
> -- 
> 1.9.1
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-pci" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 1/1] pci: host: pci-layerscape: add missing of_node_put after calling of_parse_phandle

[Peter Chen]

=20
>
>On Fri, Aug 12, 2016 at 09:34:41AM +0800, Peter Chen wrote:
>> of_node_put needs to be called when the device node which is got from
>> of_parse_phandle has finished using.
>>
>> Cc: Minghuan Lian <minghuan.Lian@freescale.com>
>> Cc: Mingkai Hu <mingkai.hu@freescale.com>
>> Cc: Roy Zang <tie-fei.zang@freescale.com>
>> Signed-off-by: Peter Chen <peter.chen@nxp.com>
>> ---
>>  drivers/pci/host/pci-layerscape.c | 2 ++
>>  1 file changed, 2 insertions(+)
>>
>> diff --git a/drivers/pci/host/pci-layerscape.c
>> b/drivers/pci/host/pci-layerscape.c
>> index 114ba81..573b996 100644
>> --- a/drivers/pci/host/pci-layerscape.c
>> +++ b/drivers/pci/host/pci-layerscape.c
>> @@ -173,6 +173,8 @@ static int ls_pcie_msi_host_init(struct pcie_port *p=
p,
>>  		return -EINVAL;
>>  	}
>>
>> +	of_node_put(msi_node);
>> +
>
>Can you please look for and fix similar errors in other drivers in drivers=
/pci/host/*?
>
>For example:
>
>  advk_pcie_probe() and iproc_pcie_msi_enable() call
>  of_parse_phandle() but don't call of_node_put() in failure paths.
>
>  dra7xx_pcie_init_irq_domain(), nwl_pcie_init_irq_domain(), and
>  xilinx_pcie_init_irq_domain() call of_get_next_child() but don't
>  call of_node_put() in failure paths.
>
>  ks_pcie_get_irq_controller_info() calls of_find_node_by_name().
>

Would you agree that I fix the issues for drivers that call of_parse_phandl=
e or of_get_next_child
or of_find_node_by_name? I can grep the symbols, and check if of_node_put i=
s called properly.
I will group all fixed within one patch set.

>I think there may be others, e.g., the pci_host_bridge_msi_domain() path c=
alls
>of_parse_phandle(), but I'm not sure of_node_put() is called on failure pa=
ths.
>

I find pci_host_bridge_msi_domain does not call of_parse_phandle directly, =
only of_msi_get_domain calls of_parse_phandle, and
of_node_put is called properly.

Peter


>When we find bugs like this, it's nice to fix one occurrence, but it's rea=
lly great if we
>can squash similar bugs nearby so the bug isn't copied into new drivers.
>
>>  	return 0;
>>  }
>>
>> --
>> 1.9.1
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-pci"
>> in the body of a message to majordomo@vger.kernel.org More majordomo
>> info at  http://vger.kernel.org/majordomo-info.html