Re: [PATCH 2/2] tty: Fix lockless tty buffer race

[Peter Hurley <peter@hurleysoftware.com>]

On 05/02/2014 10:56 AM, Peter Hurley wrote:
> Commit 6a20dbd6caa2358716136144bf524331d70b1e03,
> "tty: Fix race condition between __tty_buffer_request_room and flush_to_ldisc"
> correctly identifies an unsafe race condition between
> __tty_buffer_request_room() and flush_to_ldisc(), where the consumer
> flush_to_ldisc() prematurely advances the head before consuming the
> last of the data committed. For example:
>
>             CPU 0                     |            CPU 1
> __tty_buffer_request_room            | flush_to_ldisc
>    ...                                |   ...
>                                       |   count = head->commit - head->read
>    n = tty_buffer_alloc()             |
>    b->commit = b->used                |
>    b->next = n                        |
>                                       |   if (!count)                /* T */
>                                       |     if (head->next == NULL)  /* F */
>                                       |     buf->head = head->next
>
> In this case, buf->head has been advanced but head->commit may have
> been updated with a new value.
>
> Instead of reintroducing an unnecessary lock, fix the race locklessly.
> Read the commit-next pair in the reverse order of writing, which guarantees
> the commit value read is the latest value written if the head is
> advancing.
>
> Reported-by: Manfred Schlaegl <manfred.schlaegl@gmx.at>
> Cc: <stable@vger.kernel.org> # 3.12.x+

The patch submitted by Manfred notes the commits which introduced the
race [1], but attributes those commits to the 3.11 cycle. Those commits
were merged in the 3.12 cycle.

Regards,
Peter Hurley

[1] commits e8437d7ecbc50198705331449367d401ebb3181f,
"tty: Make driver-side flip buffers lockless", and
e9975fdec0138f1b2a85b9624e41660abd9865d4,
"tty: Ensure single-threaded flip buffer consumer with mutex"