On Wed, Nov 10, 2021 at 11:20 AM Andres Freund <andres@anarazel.de> wrote:
> I hit a crash once in 13 with a slightly evolved version of the test (many
> connections creating / dropping the partitions as in the original scenario,
> using :client_id to target different tables). It's possible that my
> instrumentation was the cause of that. Unfortunately it took quite a few hours
> to hit the problem in 13...
Have you thought about the case where a transaction does a HOT update
of the same row twice, and then aborts?
I'm asking because I notice that the fragile "We need this primarily
to handle aborted HOT updates" precheck for
HeapTupleHeaderIsHeapOnly() doesn't just check if the heap-only tuple
is DEAD before deciding to mark it LP_UNUSED. It also checks
HeapTupleHeaderIsHotUpdated() against the target tuple -- that's
another condition of the tuple being marked unused. Of course, whether
or not a given tuple is considered HeapTupleHeaderIsHotUpdated() can
change from true to false when an updater concurrently aborts. Could
that have race conditions?
In other words: what if the aforementioned "aborted HOT updates"
precheck code doesn't deal with a DEAD tuple, imagining that it's not
a relevant tuple, while at the same time the later HOT-chain-chasing
code *also* doesn't get to the tuple? What if they each assume that
the other will/has taken care of it, due to a race?
So far we've been worried about cases where these two code paths
clobber each other -- that's what we've actually seen. We should also
at least consider the possibility that we have the opposite problem,
which is what this really is.
--
Peter Geoghegan