linux-kernel - Re: [PATCH] comedi: drivers: pcl730: Fix potential memory leak in pcl730

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: <2f4dc5e8-1b2c-e071-a363-d3f509cc4ccc@huawei.com>
Date:   Tue, 22 Nov 2022 20:24:09 +0800
From:   shangxiaojing <shangxiaojing@...wei.com>
To:     Ian Abbott <abbotti@....co.uk>, <hsweeten@...ionengravers.com>,
        <gregkh@...uxfoundation.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] comedi: drivers: pcl730: Fix potential memory leak in
 pcl730_attach()



On 2022/11/22 20:18, Ian Abbott wrote:
> On 22/11/2022 12:04, Shang XiaoJing wrote:
>> pcl730_attach() calls comedi_request_region() and won't release the
>> resource allocated by alloc_resource() when pcl730_attach() failed 
>> latter.
>> Add release_region() to prevent memory leak.
>>
>> Fixes: 6f9aa29b47f6 ("staging: comedi: pcl730: use 
>> comedi_request_region()")
>> Signed-off-by: Shang XiaoJing <shangxiaojing@...wei.com>
>> ---
>>   drivers/comedi/drivers/pcl730.c | 8 +++++++-
>>   1 file changed, 7 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/comedi/drivers/pcl730.c 
>> b/drivers/comedi/drivers/pcl730.c
>> index d2733cd5383d..c463de8a14e1 100644
>> --- a/drivers/comedi/drivers/pcl730.c
>> +++ b/drivers/comedi/drivers/pcl730.c
>> @@ -274,8 +274,14 @@ static int pcl730_attach(struct comedi_device *dev,
>>           return ret;
>>       ret = comedi_alloc_subdevices(dev, board->n_subdevs);
>> -    if (ret)
>> +    if (ret) {
>> +        if (dev->iobase && dev->iolen) {
>> +            release_region(dev->iobase, dev->iolen);
>> +            dev->iobase = 0;
>> +            dev->iolen = 0;
>> +        }
>>           return ret;
>> +    }
>>       subdev = 0;
> 
> This is not needed.  If the 'attach' handler pcl730_attach() returns an 
> error, the 'detach' handler comedi_legacy_detach() will be called to 
> clean up the allocated resources.  All the comedi drivers work that way. 
> (A lot of them have an 'auto_attach' handler instead of an 'attach' 
> handler, but the error handling is basically the same.)
> 

ok, thanks for the advice!

Thanks,
-- 
Shang XiaoJing