linux-kernel - Re: [PATCH] of: dynamic: fix memory leak related to properties of __of_node

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: <CAL_JsqLaOYL64SatOv-xvq0AqXTKm1BUtienFfx_qwZ8TiN+xw@mail.gmail.com>
Date:   Tue, 17 Oct 2017 09:03:44 -0500
From:   Rob Herring <robh+dt@...nel.org>
To:     Frank Rowand <frowand.list@...il.com>
Cc:     alawang <alan.1.wang@...ia-sbell.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "devicetree@...r.kernel.org" <devicetree@...r.kernel.org>
Subject: Re: [PATCH] of: dynamic: fix memory leak related to properties of __of_node_dup

On Fri, Oct 13, 2017 at 7:59 PM, Frank Rowand <frowand.list@...il.com> wrote:
> On 10/13/17 13:47, Rob Herring wrote:
>> On Thu, Oct 12, 2017 at 8:42 PM, alawang <alan.1.wang@...ia-sbell.com> wrote:
>>> It is possible a node was dynamically allocated but without any
>>> property. The properies will be got from devices and added to the
>>> node when devices got connected.
>>> When release this node, all properties of which had been moved to
>>> deadprops.
>>> In this case, the properties in the deadprops list are never
>>> deallocated.
>>
>> This is not new, right? Do you actually hit this case? If so, how?
>>
>>>
>>> Signed-off-by: alawang <alan.1.wang@...ia-sbell.com>
>>> ---
>>>  drivers/of/dynamic.c | 4 ++++
>>>  1 file changed, 4 insertions(+)
>>>
>>> diff --git a/drivers/of/dynamic.c b/drivers/of/dynamic.c
>>> index 301b6db..465d43b 100644
>>> --- a/drivers/of/dynamic.c
>>> +++ b/drivers/of/dynamic.c
>>> @@ -335,6 +335,10 @@ void of_node_release(struct kobject *kobj)
>>>         if (!of_node_check_flag(node, OF_DYNAMIC))
>>>                 return;
>>>
>>> +       if (!prop) {
>>> +               prop = node->deadprops;
>>> +               node->deadprops = NULL;
>>> +       }
>>>         while (prop) {
>>>                 struct property *next = prop->next;
>>>                 kfree(prop->name);
>>> --
>>> 2.6.2
>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe devicetree" in
>>> the body of a message to majordomo@...r.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
>
> The code looks right to me.  The patch description is confusing to me.
>
> If a node with no properties is dynamically added, then a property is
> dynamically added to the node, then the property is dynamically removed,
> the result will be node->properties == NULL and node->deadprops != NULL.

What if you add 2 properties and then remove only 1. This would leave
both properties and deadprops !NULL. I think the while loop needs to
be a separate function so we iterate over both lists.

Rob