Re: Race conditions in logical decoding

Alvaro Herrera <alvherre(at)kurilemu(dot)de> wrote:

> On 2026-Sep-17, Antonin Houska wrote:
>
> > > + for (int i = 0; i < nrunning; i++)
> > > + {
> > > + TransactionId running_xid = running->xids[i];
> > > +
> > > + if (bsearch(&running_xid, snap->xip, snap->xcnt,
> > > + sizeof(TransactionId), xidComparator) != NULL)
> > > + XactLockTableWait(running_xid, NULL, NULL, XLTW_None);
> > > + }
> > > + }
> >
> > I don't understand why you check all transactions in procarray, instead of
> > only those in snap->xip.
>
> Hmm, but he does: for all the transactions that are running, only those
> that are found by bsearch() in the snap->xip array are waited for. Is
> that not what we want?
>
> I guess we could do it the other way around: iterate for each item on
> snap->xip and search for those in running->xids. Is that what you
> suggest?
>
> We don't know offhand which array is largest; it would be better to
> iterate on the smaller one and bsearch the largest. (Or maybe if both
> are sorted, scan them simultaneously.) I don't find any reference to
> say that running_xid is sorted.

Maybe I miss the point, but what's wrong about modifying the existing loop
that inverts the meaning of the ->xip array

/*
* snapbuild.c builds transactions in an "inverted" manner, which means it
* stores committed transactions in ->xip, not ones in progress. Build a
* classical snapshot by marking all non-committed transactions as
* in-progress. This can be expensive.
*/
for (xid = snap->xmin; NormalTransactionIdPrecedes(xid, snap->xmax);)
{
...
}

by calling XactLockTableWait() for each XID we find in the array (i.e. each
committed transaction)?

> I don't understand these two paragraphs:
>
> * A subtransaction is covered by its top-level transaction, which is in
> * snap->xip as well, or was purged from it because it is below xmin and
> * thus finished long ago.

Me neither. AFAIU SnapBuildCommitTxn() adds both top-level transaction and
subtransactions to the builder's array of committed transaction.

> * Historic snapshots do not need this: between xmin and xmax they rely on
> * xip alone, and transactions below xmin had left the procarray by the
> * time the xl_running_xacts record that set xmin was written.

I think this is related to the note that HeapTupleSatisfiesHistoricMVCC() does
not really use CLOG in the 3rd paragraph in [1].

[1] https://www.postgresql.org/message-id/CAHWVJhHXyLtS-8mdL9WhEWfsERb%3DFN7JdPD0GYAXgTmCnqbYGw%40mail.gmail.com

--
Antonin Houska
Web: https://www.cybertec-postgresql.com