Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1308819 > unrolled thread

fsl/fman: Clarification for implementation details in dtsec_config()

Started bySF Markus Elfring <elfring@users.sourceforge.net>
First post2016-01-13 22:30 +0100
Last post2016-01-14 23:00 +0100
Articles 6 — 4 participants

Back to article view | Back to linux.kernel

Contents

  fsl/fman: Clarification for implementation details in dtsec_config() SF Markus Elfring <elfring@users.sourceforge.net> - 2016-01-13 22:30 +0100
    Re: fsl/fman: Clarification for implementation details in  dtsec_config() Dan Carpenter <dan.carpenter@oracle.com> - 2016-01-13 22:40 +0100
      Re: fsl/fman: Clarification for implementation details in  dtsec_config() Andrew Lunn <andrew@lunn.ch> - 2016-01-13 23:10 +0100
        [PATCH] fsl/fman: Delete one function call "put_device" in  dtsec_config() SF Markus Elfring <elfring@users.sourceforge.net> - 2016-01-14 16:10 +0100
          Re: [PATCH] fsl/fman: Delete one function call "put_device" in  dtsec_config() Andrew Lunn <andrew@lunn.ch> - 2016-01-14 17:30 +0100
          Re: [PATCH] fsl/fman: Delete one function call "put_device" in  dtsec_config() David Miller <davem@davemloft.net> - 2016-01-14 23:00 +0100

#1308819 — fsl/fman: Clarification for implementation details in dtsec_config()

FromSF Markus Elfring <elfring@users.sourceforge.net>
Date2016-01-13 22:30 +0100
Subjectfsl/fman: Clarification for implementation details in dtsec_config()
Message-ID<qQB9o-5uC-1@gated-at.bofh.it>
Hello,

The Coccinelle semantic patch "deref_null.cocci" pointed out a potential
problem in the implementation of the function "dtsec_config" (lines 1434-1439):
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/tree/drivers/net/ethernet/freescale/fman/fman_dtsec.c?id=e5a03bfd873c29eb786655ef2e95e53ed242b404#n1434
…
	dtsec->tbiphy = of_phy_find_device(params->internal_phy_node);
	if (!dtsec->tbiphy) {
		pr_err("of_phy_find_device (TBI PHY) failed\n");
		put_device(&dtsec->tbiphy->mdio.dev);
		goto err_dtsec_drv_param;
	}
…

If the pointer "dtsec->tbiphy" is NULL, then it will especially matter
if the shown address computation will succeed.
Should the call of the function "put_device" be removed,
or was another argument intended?

Regards,
Markus

[toc] | [next] | [standalone]

#1308833 — Re: fsl/fman: Clarification for implementation details in dtsec_config()

FromDan Carpenter <dan.carpenter@oracle.com>
Date2016-01-13 22:40 +0100
SubjectRe: fsl/fman: Clarification for implementation details in dtsec_config()
Message-ID<qQBj5-5yt-31@gated-at.bofh.it>
In reply to#1308819
I think you can remove the put_device() line.

regards,
dan carpenter

[toc] | [prev] | [next] | [standalone]

#1308856 — Re: fsl/fman: Clarification for implementation details in dtsec_config()

FromAndrew Lunn <andrew@lunn.ch>
Date2016-01-13 23:10 +0100
SubjectRe: fsl/fman: Clarification for implementation details in dtsec_config()
Message-ID<qQBM6-5Z4-21@gated-at.bofh.it>
In reply to#1308833
On Thu, Jan 14, 2016 at 12:35:37AM +0300, Dan Carpenter wrote:
> I think you can remove the put_device() line.

Agreed.

of_phy_find_device() only does a get_device() when it finds a matching
device.

	Andrew

[toc] | [prev] | [next] | [standalone]

#1309369 — [PATCH] fsl/fman: Delete one function call "put_device" in dtsec_config()

FromSF Markus Elfring <elfring@users.sourceforge.net>
Date2016-01-14 16:10 +0100
Subject[PATCH] fsl/fman: Delete one function call "put_device" in dtsec_config()
Message-ID<qQRHd-nE-47@gated-at.bofh.it>
In reply to#1308856
From: Markus Elfring <elfring@users.sourceforge.net>
Date: Thu, 14 Jan 2016 15:46:28 +0100

The Coccinelle semantic patch script "deref_null.cocci" pointed a problem
out in the implementation of the function "dtsec_config".

A null pointer was assigned to the data structure member "tbiphy" of the
variable "dtsec" if a matching device was not found.
A call of the function "put_device" was unnecessary then because
a previous call of the function "get_device" was not triggered.
Thus remove the function call "put_device" after the printing of the
desired error message.

Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
---
 drivers/net/ethernet/freescale/fman/fman_dtsec.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/net/ethernet/freescale/fman/fman_dtsec.c b/drivers/net/ethernet/freescale/fman/fman_dtsec.c
index 6b1261c..7c92eb8 100644
--- a/drivers/net/ethernet/freescale/fman/fman_dtsec.c
+++ b/drivers/net/ethernet/freescale/fman/fman_dtsec.c
@@ -1434,7 +1434,6 @@ struct fman_mac *dtsec_config(struct fman_mac_params *params)
 	dtsec->tbiphy = of_phy_find_device(params->internal_phy_node);
 	if (!dtsec->tbiphy) {
 		pr_err("of_phy_find_device (TBI PHY) failed\n");
-		put_device(&dtsec->tbiphy->mdio.dev);
 		goto err_dtsec_drv_param;
 	}
 
-- 
2.6.3

[toc] | [prev] | [next] | [standalone]

#1309460 — Re: [PATCH] fsl/fman: Delete one function call "put_device" in dtsec_config()

FromAndrew Lunn <andrew@lunn.ch>
Date2016-01-14 17:30 +0100
SubjectRe: [PATCH] fsl/fman: Delete one function call "put_device" in dtsec_config()
Message-ID<qQSWC-197-23@gated-at.bofh.it>
In reply to#1309369
On Thu, Jan 14, 2016 at 04:00:46PM +0100, SF Markus Elfring wrote:
> From: Markus Elfring <elfring@users.sourceforge.net>
> Date: Thu, 14 Jan 2016 15:46:28 +0100
> 
> The Coccinelle semantic patch script "deref_null.cocci" pointed a problem
> out in the implementation of the function "dtsec_config".
> 
> A null pointer was assigned to the data structure member "tbiphy" of the
> variable "dtsec" if a matching device was not found.
> A call of the function "put_device" was unnecessary then because
> a previous call of the function "get_device" was not triggered.
> Thus remove the function call "put_device" after the printing of the
> desired error message.
> 
> Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>

Fixes: 57ba4c9b56d8 ("fsl/fman: Add FMan MAC support")
Reviewd-by: Andrew Lunn <andrew@lunn.ch>

	    Andrew
> ---
>  drivers/net/ethernet/freescale/fman/fman_dtsec.c | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/freescale/fman/fman_dtsec.c b/drivers/net/ethernet/freescale/fman/fman_dtsec.c
> index 6b1261c..7c92eb8 100644
> --- a/drivers/net/ethernet/freescale/fman/fman_dtsec.c
> +++ b/drivers/net/ethernet/freescale/fman/fman_dtsec.c
> @@ -1434,7 +1434,6 @@ struct fman_mac *dtsec_config(struct fman_mac_params *params)
>  	dtsec->tbiphy = of_phy_find_device(params->internal_phy_node);
>  	if (!dtsec->tbiphy) {
>  		pr_err("of_phy_find_device (TBI PHY) failed\n");
> -		put_device(&dtsec->tbiphy->mdio.dev);
>  		goto err_dtsec_drv_param;
>  	}
>  
> -- 
> 2.6.3
>

[toc] | [prev] | [next] | [standalone]

#1309686 — Re: [PATCH] fsl/fman: Delete one function call "put_device" in dtsec_config()

FromDavid Miller <davem@davemloft.net>
Date2016-01-14 23:00 +0100
SubjectRe: [PATCH] fsl/fman: Delete one function call "put_device" in dtsec_config()
Message-ID<qQY5Y-4Io-13@gated-at.bofh.it>
In reply to#1309369
From: SF Markus Elfring <elfring@users.sourceforge.net>
Date: Thu, 14 Jan 2016 16:00:46 +0100

> From: Markus Elfring <elfring@users.sourceforge.net>
> Date: Thu, 14 Jan 2016 15:46:28 +0100
> 
> The Coccinelle semantic patch script "deref_null.cocci" pointed a problem
> out in the implementation of the function "dtsec_config".
> 
> A null pointer was assigned to the data structure member "tbiphy" of the
> variable "dtsec" if a matching device was not found.
> A call of the function "put_device" was unnecessary then because
> a previous call of the function "get_device" was not triggered.
> Thus remove the function call "put_device" after the printing of the
> desired error message.
> 
> Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>

Applied, thanks.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web