Hi, On 2019-04-08 14:05:02 -0700, Peter Geoghegan wrote: > However, now that I take a closer look I also notice that there is > core code that calls PageGetFreeSpace() when it probably shouldn't, > either. For example, what business does heap_xlog_visible() have > calling PageGetFreeSpace()?
I'm not sure I understand what the problem is. We got to get the information for the fsm from somewhere? Are you arguing we should instead have it included as an explicit xlog record payload? Or just that it should use PageGetExactFreeSpace()? I assume the former based on your "what business" language, but that'd not make terribly much sense to me. I don't think precision terribly matters here... Greetings, Andres Freund