[PATCH 04/18] drivers/ipc: bufp: fix read-write, write-write preemption cases

[Philippe Gerum]

On 2/25/19 7:17 PM, Jan Kiszka wrote:
> On 05.02.19 12:20, Philippe Gerum via Xenomai wrote:
>> On 2/4/19 7:55 PM, Jan Kiszka wrote:
>>>> -             * Release the lock while copying the data to
>>>> -             * keep latency low.
>>>> +             * We have to drop the lock while reading in
>>>> +             * data, but we can't rollback on bad read
>>>> +             * from user because some other thread might
>>>> +             * have populated the memory ahead of our
>>>> +             * write slot already: bluntly clear the
>>>> +             * unavailable bytes on copy error.
>>>>                 */
>>>>                cobalt_atomic_leave(s);
>>>> -            ret = xnbufd_copy_to_kmem(rsk->bufmem + wroff, bufd, n);
>>>> -            if (ret < 0)
>>>> -                return ret;
>>>> +            xret = xnbufd_copy_to_kmem(rsk->bufmem + wroff, bufd, n);
>>>>                cobalt_atomic_enter(s);
>>>> -            /*
>>>> -             * In case we were preempted while copying the
>>>> -             * message, we have to write the whole thing
>>>> -             * again.
>>>> -             */
>>>> -            if (rsk->wrtoken != wrtoken) {
>>>> -                xnbufd_reset(bufd);
>>>> -                goto redo;
>>>> +            if (xret < 0) {
>>>> +                memset(rsk->bufmem + wroff + n - xret, 0, xret);
>>>
>>> This looks fishy, to the compiler and also to me.
>>>
>> Oh, well. Paste&copy blunder, the fix originates from a distinct
>> co-kernel implementation. I'll resubmit.
>>
> 
> Did I get that update already? I don't think so.
> 

You may want to check my local dev. tree. I generally don't send every
trivial change to a reviewed patch to the mailing list, but I'm
certainly maintaining a branch for you to pick from there:

https://lab.xenomai.org/xenomai-rpm.git/commit/?h=for-upstream/next&id=99dc7076ee757c2c82ffa6c58930388974808c09

-- 
Philippe.