linux-kernel - Re: [PATCH v2] arch: powerpc: platforms: 85xx: Add missing of_node_put in sgy

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

Message-ID: <87o7ysb2ot.fsf@mpe.ellerman.id.au>
Date:   Fri, 17 Jun 2022 09:37:06 +1000
From:   Michael Ellerman <mpe@...erman.id.au>
To:     Christophe JAILLET <christophe.jaillet@...adoo.fr>,
        Liang He <windhl@....com>, oss@...error.net, paulus@...ba.org,
        christophe.leroy@...roup.eu, nixiaoming@...wei.com
Cc:     linuxppc-dev@...ts.ozlabs.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] arch: powerpc: platforms: 85xx: Add missing
 of_node_put in sgy_cts1000.c

Christophe JAILLET <christophe.jaillet@...adoo.fr> writes:
> Le 16/06/2022 à 17:19, Liang He a écrit :
>> In gpio_halt_probe(), of_find_matching_node() will return a node pointer with
>> refcount incremented. We should use of_node_put() in each fail path or when it
>> is not used anymore.
>> 
>> Signed-off-by: Liang He <windhl@....com>
>> ---
>>   changelog:
>> 
>>   v2: use goto-label patch style advised by Christophe.
>>   v1: add of_node_put() before each exit.
>> 
>>   arch/powerpc/platforms/85xx/sgy_cts1000.c | 27 +++++++++++++++--------
>>   1 file changed, 18 insertions(+), 9 deletions(-)
>> 
>> diff --git a/arch/powerpc/platforms/85xx/sgy_cts1000.c b/arch/powerpc/platforms/85xx/sgy_cts1000.c
>> index 98ae64075193..e280f963d88c 100644
>> --- a/arch/powerpc/platforms/85xx/sgy_cts1000.c
>> +++ b/arch/powerpc/platforms/85xx/sgy_cts1000.c
>> @@ -73,6 +73,7 @@ static int gpio_halt_probe(struct platform_device *pdev)
...
>> @@ -122,8 +127,12 @@ static int gpio_halt_probe(struct platform_device *pdev)
>>   
>>   	printk(KERN_INFO "gpio-halt: registered GPIO %d (%d trigger, %d"
>>   	       " irq).\n", gpio, trigger, irq);
>> +	ret = 0;
>>   
>> -	return 0;
>> +err_put:
>> +	of_node_put(halt_node);
>> +	halt_node = NULL;
>
> Hi,
> so now we set 'halt_node' to NULL even in the normal case.
> This is really spurious.
>
> Look at gpio_halt_cb(), but I think that this is just wrong and badly 
> breaks this driver.

I agree, thanks for reviewing.

I think the cleanest solution is to use a local variable for the node in
the body of gpio_halt_probe(), and only assign to halt_node once all the
checks have passed.

So something like:

        struct device_node *child_node;

	child_node = of_find_matching_node(node, child_match);
        ...

	printk(KERN_INFO "gpio-halt: registered GPIO %d (%d trigger, %d"
	       " irq).\n", gpio, trigger, irq);
        ret = 0;
        halt_node = of_node_get(child_node);

out_put:
        of_node_put(child_node);
        
	return ret;
}


cheers