Re: Shave a few instructions from child-process startup sequence

On 1/20/14, 8:08 PM, Alvaro Herrera wrote:
> Peter Eisentraut escribió:
>> src/backend/postmaster/postmaster.c:2255: indent with spaces.
>> + else
>> src/backend/postmaster/postmaster.c:2267: indent with spaces.
>> + break;
>
> I just checked the Jenkins page for this patch:
> http://pgci.eisentraut.org/jenkins/job/postgresql_commitfest_world/243/
> just to make sure I can figure out what it means. You reported it as
> "build unstable" in the commitfest entry:
> https://commitfest.postgresql.org/action/patch_view?id=1277
> However, looking at Jenkins, I couldn't figure out what the problem is.

In this case, it was the whitespace violation. (Yeah, I'm constantly
debating with myself whether it's worth reporting that, but at the
moment I'm still on the side of the fence that wants to make people
submit clean patches.)

In general, it's sometimes a bit hard to find out what caused the build
to fail. Jenkins can detect and report that for standard tools (e.g.,
compiler warnings, JUnit test results), but not for our custom test
tools. Another issue is that the build is running with make -k, so the
issue could be somewhere in the middle of the build log. I'm exploring
new plugins to improve that, as it's a significant problem.