linux-kernel - Re: [PATCH] usb: phy: Initialize struct usb_phy list

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: <8837bf19-dd98-40f9-b2ff-192a511c357c@tecnico.ulisboa.pt>
Date: Fri, 21 Nov 2025 14:55:35 +0000
From: Diogo Ivo <diogo.ivo@...nico.ulisboa.pt>
To: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Cc: Felipe Balbi <felipe.balbi@...ux.intel.com>,
 Baolin Wang <baolin.wang@...ux.alibaba.com>, linux-usb@...r.kernel.org,
 linux-kernel@...r.kernel.org, stable@...r.kernel.org
Subject: Re: [PATCH] usb: phy: Initialize struct usb_phy list_head



On 11/21/25 14:09, Greg Kroah-Hartman wrote:
> On Thu, Nov 13, 2025 at 02:59:06PM +0000, Diogo Ivo wrote:
>> When executing usb_add_phy() and usb_add_phy_dev() it is possible that
>> usb_add_extcon() fails (for example with -EPROBE_DEFER), in which case
>> the usb_phy does not get added to phy_list via
>> list_add_tail(&x->head, phy_list).
>>
>> Then, when the driver that tried to add the phy receives the error
>> propagated from usb_add_extcon() and calls into usb_remove_phy() to
>> undo the partial registration there will be an unconditional call to
>> list_del(&x->head) which is notinitialized and leads to a NULL pointer
>> dereference.
>>
>> Fix this by initializing x->head before usb_add_extcon() has a chance to
>> fail.
>>
>> Fixes: 7d21114dc6a2d53 ("usb: phy: Introduce one extcon device into usb phy")
>> Cc: stable@...r.kernel.org
>> Signed-off-by: Diogo Ivo <diogo.ivo@...nico.ulisboa.pt>
>> ---
>>   drivers/usb/phy/phy.c | 4 ++++
>>   1 file changed, 4 insertions(+)
>>
>> diff --git a/drivers/usb/phy/phy.c b/drivers/usb/phy/phy.c
>> index e1435bc59662..5a9b9353f343 100644
>> --- a/drivers/usb/phy/phy.c
>> +++ b/drivers/usb/phy/phy.c
>> @@ -646,6 +646,8 @@ int usb_add_phy(struct usb_phy *x, enum usb_phy_type type)
>>   		return -EINVAL;
>>   	}
>>   
>> +	INIT_LIST_HEAD(&x->head);
>> +
>>   	usb_charger_init(x);
>>   	ret = usb_add_extcon(x);
>>   	if (ret)
>> @@ -696,6 +698,8 @@ int usb_add_phy_dev(struct usb_phy *x)
>>   		return -EINVAL;
>>   	}
>>   
>> +	INIT_LIST_HEAD(&x->head);
>> +
>>   	usb_charger_init(x);
>>   	ret = usb_add_extcon(x);
>>   	if (ret)
>>
> 
> Shouldn't you be also removing an existing call to INIT_LIST_HEAD()
> somewhere?  This is not "moving" the code, it is adding it.

 From my understanding that's exactly the problem, currently there is no
call to INIT_LIST_HEAD() anywhere on these code paths, meaning that if
we do not reach the point of calling list_add_tail() at the end of
usb_add_phy() and usb_phy_add_dev() then x->head will remain uninitialized
and fault when running usb_remove_phy().

Best regards,
Diogo

> thanks,
> 
> greg k-h