Hi,
On 2020-11-02 09:16:26 +0000, Jinbao Chen wrote:
> We directly call the heap function VM_ALL_VISIBLE in the
> IndexOnlyNext function. This is not in line with the design idea of
> table am. If the new storage type needs to implement index only
> scan, he must hack the IndexOnlyNext function.
Yea, it's something we should improve. Have you checked if this has
performance impact for heap? Should we also consider planning costs?
> So this patch add a new table am 'tid_visible' to test visibility
> of tid. So that index only scan can completely use table AM.
As far as I can tell you have not acually attached the patch.
Greetings,
Andres Freund