linux-kernel - Re: [PATCH][next] io_uring: Fix memory leak on error return path.

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <0aa1bcfe-9e45-ac54-1292-43caad8b9b06@canonical.com>
Date:   Thu, 29 Apr 2021 11:36:51 +0100
From:   Colin Ian King <colin.king@...onical.com>
To:     Pavel Begunkov <asml.silence@...il.com>,
        Jens Axboe <axboe@...nel.dk>, io-uring@...r.kernel.org
Cc:     kernel-janitors@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH][next] io_uring: Fix memory leak on error return path.

On 29/04/2021 11:32, Pavel Begunkov wrote:
> On 4/29/21 11:26 AM, Colin King wrote:
>> From: Colin Ian King <colin.king@...onical.com>
>>
>> Currently the -EINVAL error return path is leaking memory allocated
>> to data. Fix this by kfree'ing data before the return.
>>
>> Addresses-Coverity: ("Resource leak")
>> Fixes: c3a40789f6ba ("io_uring: allow empty slots for reg buffers")
>> Signed-off-by: Colin Ian King <colin.king@...onical.com>
>> ---
>>  fs/io_uring.c | 4 +++-
>>  1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/fs/io_uring.c b/fs/io_uring.c
>> index 47c2f126f885..beeb477e4f6a 100644
>> --- a/fs/io_uring.c
>> +++ b/fs/io_uring.c
>> @@ -8417,8 +8417,10 @@ static int io_sqe_buffers_register(struct io_ring_ctx *ctx, void __user *arg,
>>  		ret = io_buffer_validate(&iov);
>>  		if (ret)
>>  			break;
>> -		if (!iov.iov_base && tag)
>> +		if (!iov.iov_base && tag) {> +			kfree(data);
>>  			return -EINVAL;
>> +		}
> 
> Buggy indeed, should have been:
> 
> ret = -EINVAL;
> break;
Ah, thanks.

> 
> Colin, can you resend with the change?

Will do in a moment or so.
> 
>>  
>>  		ret = io_sqe_buffer_register(ctx, &iov, &ctx->user_bufs[i],
>>  					     &last_hpage);
>>
>