On Fri, Feb 11, 2022 at 03:21:43PM -0300, Alvaro Herrera wrote:
> On 2022-Jan-28, Justin Pryzby wrote:
> > Have you looked at code coverage ? I have an experimental patch to add that to
> > cirrus, and ran it with this patch; visible here:
> > https://cirrus-ci.com/task/6362512059793408
>
> Ah, thanks, this is useful. I think it is showing that the new code is
> generally well covered, but there are some exceptions, particularly
> ExecMergeMatched in some concurrent cases (TM_Deleted and
> TM_SelfModified after table_tuple_lock -- those code pages have
> user-facing errors but are not covered by any tests.)
>
> How does this work? I notice it is reporting for
> src/bin/pg_upgrade/relfilenode.c, but that file is not changed by this
> patch.
Because I put your patch on top of some other branch with the CI coverage (and
other stuff).
It tries to only show files changed by the branch being checked:
https://github.com/justinpryzby/postgres/commit/d668142040031915
But it has to figure out where the branch "starts". Which I did by looking at
"git diff --cherry-pick origin..."
Andres thinks that does the wrong thing if CI is run manually (not by CFBOT)
for patches against backbranches. I'm not sure git diff --cherry-pick is
widely known/used, but I think using that relative to master may be good
enough.
Ongoing discussion here.
https://www.postgresql.org/message-id/20220203035827.GG23027%40telsasoft.com
--
Justin
