linux-kernel - Re: [PATCH] staging: comedi: comedi_test: return error when comedi_test

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: <ca049de4-92f1-409c-6aa7-9c2ddc336638@mev.co.uk>
Date:   Tue, 10 Dec 2019 10:58:40 +0000
From:   Ian Abbott <abbotti@....co.uk>
To:     Chuhong Yuan <hslester96@...il.com>
Cc:     H Hartley Sweeten <hsweeten@...ionengravers.com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Cheah Kok Cheong <thrust73@...il.com>,
        devel@...verdev.osuosl.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] staging: comedi: comedi_test: return error when
 comedi_test_init fails

On 10/12/2019 09:14, Chuhong Yuan wrote:
> comedi_test_init() forgets to return error when it fails.
> Modify the return value to fix it.
> 
> Fixes: 9ff7400bd38c ("Staging: comedi: drivers: comedi_test: Add auto-configuration capability")
> Signed-off-by: Chuhong Yuan <hslester96@...il.com>
> ---
>   drivers/staging/comedi/drivers/comedi_test.c | 4 +++-
>   1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/staging/comedi/drivers/comedi_test.c b/drivers/staging/comedi/drivers/comedi_test.c
> index ef4c7c8a2b71..133ed99a89f1 100644
> --- a/drivers/staging/comedi/drivers/comedi_test.c
> +++ b/drivers/staging/comedi/drivers/comedi_test.c
> @@ -799,12 +799,14 @@ static int __init comedi_test_init(void)
>   	if (!config_mode) {
>   		ctcls = class_create(THIS_MODULE, CLASS_NAME);
>   		if (IS_ERR(ctcls)) {
> +			ret = PTR_ERR(ctcls);
>   			pr_warn("comedi_test: unable to create class\n");
>   			goto clean3;
>   		}
>   
>   		ctdev = device_create(ctcls, NULL, MKDEV(0, 0), NULL, DEV_NAME);
>   		if (IS_ERR(ctdev)) {
> +			ret = PTR_ERR(ctdev);
>   			pr_warn("comedi_test: unable to create device\n");
>   			goto clean2;
>   		}
> @@ -826,7 +828,7 @@ static int __init comedi_test_init(void)
>   clean3:
>   	ctcls = NULL;
>   
> -	return 0;
> +	return ret;
>   }
>   module_init(comedi_test_init);
>   
> 

I suspect the current behaviour was by design because the module is not 
completely broken if it fails to create the "comedi_test" device.  The 
module will just behave as if the "noauto" module parameter (which sets 
the 'config_mode' variable) was set to "true" if there is an error 
during class or device creation.  That's probably why the kernel log 
messages are written with 'pr_warn()' rather than 'pr_err()'.

As long as the core "comedi" module has been loaded with the 
"comedi_num_legacy_minors" module parameter set to a non-zero value, it 
would still be possible to configure one of the "legacy" comedi devices 
to use the "comedi_test" driver via the COMEDI_DEVCONFIG ioctl.

However, returning an error might be preferable here, especially since 
the comedi core module's "comedi_num_legacy_minors" parameter defaults 
to zero.

I'm happy with the change of behaviour, but would suggest changing the 
'pr_warn()' calls to 'pr_err()' to match this change.

-- 
-=( Ian Abbott <abbotti@....co.uk> || Web: www.mev.co.uk )=-
-=( MEV Ltd. is a company registered in England & Wales. )=-
-=( Registered number: 02862268.  Registered address:    )=-
-=( 15 West Park Road, Bramhall, STOCKPORT, SK7 3JZ, UK. )=-