Re: segmentation fault when cassert enabled

On Tue, Dec 17, 2019 at 10:09 AM Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
>
> On Mon, Dec 16, 2019 at 9:16 PM Jehan-Guillaume de Rorthais
> <jgdr(at)dalibo(dot)com> wrote:
> >
> > On Mon, 16 Dec 2019 13:27:43 +0100
> > Peter Eisentraut <peter(dot)eisentraut(at)2ndquadrant(dot)com> wrote:
> >
> > > On 2019-12-16 11:11, Amit Kapila wrote:
> > > > I agree that this is a timing issue. I also don't see a way to write
> > > > a reproducible test for this. However, I could reproduce it via
> > > > debugger consistently by following the below steps. I have updated a
> > > > few comments and commit messages in the attached patch.
> > > >
> > > > Peter E., Petr J or anyone else, do you have comments or objections on
> > > > this patch? If none, then I am planning to commit (and backpatch)
> > > > this patch in a few days time.
> > >
> > > The patch seems fine to me. Writing a test seems hard. Let's skip it.
> > >
>
> Okay.
>
> > > The commit message has a duplicate "building"/"built" in the first sentence.
> >
> > I think the sentence is quite long. I had to re-read it to get it.
> >
> > What about:
> >
> > This patch allows building the local relmap cache for a subscribed relation
> > after processing pending invalidation messages and potential relcache
> > updates.
> >
>
> Attached patch with updated commit message based on suggestions. I am
> planning to commit this tomorrow unless there are more comments.
>

While testing the patch on back versions, I found that the patch does
not apply on PG 11 & PG 10 branch. Attached patch has the changes for
PG 11 & PG 10 branch. Only difference in the patch was that table_open
needed to be changed to heap_open. I have verified the patch on back
branches and found it to be working. For PG 12 & current the previous
patch that Amit need to be used, I'm not reattaching here.

Regards,
Vignesh
EnterpriseDB: http://www.enterprisedb.com