Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> Perhaps we should just fix this one in master and consider
> back-patching it if and when we get some plausibly related bug
> reports.
I'm not completely clear on what one would do to be vulnerable to
hitting the bug, or what the impact of hitting it would be. Tom
said:
> The potential consequences are hugely worse than that, though:
> with a bit more time between the two operations, it'd be possible
> for someone else to reclaim the dead tuple and replace it with
> something else. As long as the TID is live when we get to it,
> heap_update will blindly replace it, whether or not it has
> anything to do with the tuple we think we're replacing.
What would a worst-case bug report based on hitting that look like?
-Kevin