Re: [PATCH] rbd: enforce parent overlap

[Josh Durgin <josh.durgin@inktank.com>]

On 04/22/2013 01:43 PM, Alex Elder wrote:
> On 04/22/2013 01:34 PM, Josh Durgin wrote:
>>> +     * We need to zero anything beyond the parent overlap
>>> +     * boundary.  Since rbd_img_obj_request_read_callback()
>>> +     * will zero anything beyond the end of a short read, an
>>> +     * easy way to do this is to pretend the data from the
>>> +     * parent came up short--ending at the overlap boundary.
>>> +     */
>
> Sorry, I missed this one.
>
>>> +    rbd_assert(obj_request->img_offset < U64_MAX - obj_request->length);
>>> +    obj_end = obj_request->img_offset + obj_request->length;
>>> +    rbd_dev = obj_request->img_request->rbd_dev;
>>> +    if (obj_end > rbd_dev->parent_overlap) {
>>
>> Shouldn't this be >=, since the overlap is a size?
>
> Does the overlap define the maximum byte offste included in
> the overlap, or does it define the first offset not included?
>
> If it's the former, then I agree with you (and it's not
> what I thought).

It's the latter. Now I see that obj_end is the first offset not 
included, so it's correct in your patch.

> 					-Alex
>
>>> +        u64 xferred = 0;
>>> +
>>> +        if (obj_request->img_offset < rbd_dev->parent_overlap)
>>> +            xferred = rbd_dev->parent_overlap -
>>> +                    obj_request->img_offset;
>>> +
>>> +        obj_request->xferred = min(img_request->xferred, xferred);
>>> +    } else {