[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <336d516d-a256-fd70-1c08-8c42880a2821@gmail.com>
Date: Mon, 21 Nov 2016 08:58:55 +0100
From: Quentin Lambert <lambert.quentin@...il.com>
To: Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
linux-kernel@...r.kernel.org, kernel-janitors@...r.kernel.org
Subject: Re: [PATCH] stm class: Add a missing call to put_device
On 11/21/2016 08:32 AM, Alexander Shishkin wrote:
> Quentin Lambert <lambert.quentin@...il.com> writes:
>
>> Most error branches following the call to class_find_device contain
>> a call to put_device. This patch add calls to put_device where
>> they are missing.
>>
>> This issue was found with Hector.
>>
>> Signed-off-by: Quentin Lambert <lambert.quentin@...il.com>
>>
>> ---
>> drivers/hwtracing/stm/core.c | 4 +++-
>> 1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> --- a/drivers/hwtracing/stm/core.c
>> +++ b/drivers/hwtracing/stm/core.c
>> @@ -368,8 +368,10 @@ static int stm_char_open(struct inode *i
>> return -ENODEV;
>>
>> stmf = kzalloc(sizeof(*stmf), GFP_KERNEL);
>> - if (!stmf)
>> + if (!stmf) {
>> + put_device(dev);
>> return -ENOMEM;
>> + }
> There is a goto label at the bottom of this function which is supposed
> to deal with this. See the fix that we already have [1] for this issue.
>
> [1] https://git.kernel.org/cgit/linux/kernel/git/ash/stm.git/commit/?h=stm-for-greg-20161118&id=a0ebf519b8a2666438d999c62995618c710573e5
>
> Regards,
> --
> alex
Your fix is better, you are right, I did not use the goto label because
of the free.
Thanks for your feedback though.
Regards,
Quentin
Powered by blists - more mailing lists