linux-kernel - Re: [PATCH] staging: unisys: Properly test debugfs_create

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20220329174639.GC3293@kadam>
Date:   Tue, 29 Mar 2022 20:46:39 +0300
From:   Dan Carpenter <dan.carpenter@...cle.com>
To:     "Fabio M. De Francesco" <fmdefrancesco@...il.com>
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        David Kershner <david.kershner@...sys.com>,
        sparmaintainer@...sys.com, linux-staging@...ts.linux.dev,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] staging: unisys: Properly test debugfs_create_dir()
 return values

On Tue, Mar 29, 2022 at 07:00:49PM +0200, Fabio M. De Francesco wrote:
> On martedì 22 marzo 2022 09:38:58 CEST Fabio M. De Francesco wrote:
> > debugfs_create_dir() returns a pointers to a dentry objects. On failures
> > it returns errors. Currently the values returned to visornic_probe()
> > seem to be tested for being equal to NULL in case of failures.
> > 
> > Properly test with "if (IS_ERR())" and then assign the correct error 
> > value to the "err" variable.
> > 
> > Signed-off-by: Fabio M. De Francesco <fmdefrancesco@...il.com>
> > ---
> >  drivers/staging/unisys/visornic/visornic_main.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/staging/unisys/visornic/visornic_main.c b/drivers/staging/unisys/visornic/visornic_main.c
> > index 643432458105..58d03f3d3173 100644
> > --- a/drivers/staging/unisys/visornic/visornic_main.c
> > +++ b/drivers/staging/unisys/visornic/visornic_main.c
> > @@ -1922,11 +1922,11 @@ static int visornic_probe(struct visor_device *dev)
> >  	/* create debug/sysfs directories */
> >  	devdata->eth_debugfs_dir = debugfs_create_dir(netdev->name,
> >  						      visornic_debugfs_dir);
> > -	if (!devdata->eth_debugfs_dir) {
> > +	if (IS_ERR(devdata->eth_debugfs_dir)) {
> >  		dev_err(&dev->device,
> >  			"%s debugfs_create_dir %s failed\n",
> >  			__func__, netdev->name);
> > -		err = -ENOMEM;
> > +		err = PTR_ERR(devdata->eth_debugfs_dir);
> >  		goto cleanup_register_netdev;
> >  	}
> >  
> > -- 
> > 2.34.1
> > 
> Hi Greg, Dan,
> 
> Now I have time to rework this patch but, if I'm not misunderstanding, you 
> asked for two contrasting works to do here...
> 
> Dan wrote that "[in] this case you can delete the whole devdata->eth_debugfs_dir 
> and the related code.".
> 
> Greg wrote that "We really shouldn't be checking this value at all.  There's 
> no reason to check the return value of a debugfs_* call. Can you fix up the code to
> do that instead?".
> 
> I'm confused because they look like two incompatible requests. What should I do?

Greg is saying delete the tests and the error handling.  But I am
saying delete the tests, the error handling, the devdata->eth_debugfs_dir
related code and the call to debugfs_create_dir().

There is no conflict.  Delete it all.

regards,
dan carpenter