Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH] PCI/hotplug: Remove unneeded of_node_put() in pnv_php
To:     Nathan Lynch <nathanl@linux.ibm.com>,
        Wan Jiabing <wanjiabing@vivo.com>
Cc:     kael_w@yeah.net, Michael Ellerman <mpe@ellerman.id.au>,
        Benjamin Herrenschmidt <benh@kernel.crashing.org>,
        Paul Mackerras <paulus@samba.org>,
        Bjorn Helgaas <bhelgaas@google.com>,
        linuxppc-dev@lists.ozlabs.org, linux-pci@vger.kernel.org,
        linux-kernel@vger.kernel.org, ajd@linux.ibm.com
References: <20211020094604.2106-1-wanjiabing@vivo.com>
 <87tuhcx6v6.fsf@linux.ibm.com>
From:   Tyrel Datwyler <tyreld@linux.ibm.com>
Message-ID: <0622b8ea-7587-2e6b-5558-fae1847b14b9@linux.ibm.com>
Date:   Wed, 20 Oct 2021 09:57:23 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.13.0
MIME-Version: 1.0
In-Reply-To: <87tuhcx6v6.fsf@linux.ibm.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 10/20/21 4:39 AM, Nathan Lynch wrote:
> Wan Jiabing <wanjiabing@vivo.com> writes:
>> Fix following coccicheck warning:
>> ./drivers/pci/hotplug/pnv_php.c:161:2-13: ERROR: probable double put.
>>
>> Device node iterators put the previous value of the index variable, so
>> an explicit put causes a double put.
> 
> I suppose Coccinelle doesn't take into account that this code is
> detaching and freeing the nodes.
> 
> 
>> diff --git a/drivers/pci/hotplug/pnv_php.c b/drivers/pci/hotplug/pnv_php.c
>> index f4c2e6e01be0..f3da4f95d73f 100644
>> --- a/drivers/pci/hotplug/pnv_php.c
>> +++ b/drivers/pci/hotplug/pnv_php.c
>> @@ -158,7 +158,6 @@ static void pnv_php_detach_device_nodes(struct device_node *parent)
>>  	for_each_child_of_node(parent, dn) {
>>  		pnv_php_detach_device_nodes(dn);
>>  
>> -		of_node_put(dn);
>>  		of_detach_node(dn);
>>  	}
> 
> The code might be improved by comments explaining how the bare
> of_node_put() corresponds to a "get" somewhere else in the driver, and
> how it doesn't render the ongoing traversal unsafe. It looks suspicious
> on first review, but I believe it's intentional and probably correct as
> written.
> 

This is a common usage pattern which if we put a comment about the pattern here
we need to do it every where. I suppose a better solution is to wrap this put in
a more descriptive function name like of_node_long_put() or something of the
sort the makes it obvious we are dropping a long held global scope reference.

-Tyrel