Re: TAP output format in pg_regress

> On 22 Mar 2022, at 00:49, Andres Freund <andres(at)anarazel(dot)de> wrote:

> This is failing with segmentation faults on cfbot:
> https://cirrus-ci.com/task/4879227009892352?logs=test_world#L21
>
> Looks like something around isolationtester is broken?

It could be triggered by plpgsql tests as well, and was (as usual) a silly
mistake easily fixed when found. The attached survices repeated check-world
for me.

> Seems like we might not have energy to push this forward in the next two
> weeks, so maybe the CF entry should be marked returned or moved for now?

Since there is little use for this without the Meson branch, it should target
the same version as that patch. I'll move the patch to the next CF for now.

As we discussed off-list I extended this patchset with an attempt to minimize
noise as per 20220221164736(dot)rq3ornzjdkmwk2wo(at)alap3(dot)anarazel(dot)de, but it's not
yet done. The attached has a rough draft of adding a --verbose option which
gives the current output (and potentially more for debugging), but which by
default is off producing minimal noise.

--
Daniel Gustafsson https://vmware.com/