Re: Teach pg_upgrade test to honor NO_TEMP_INSTALL

On 5/24/19 11:22 AM, Andres Freund wrote:
> Hi,
>
> On 2019-05-22 13:08:41 -0700, Andres Freund wrote:
>> On 2019-05-22 16:04:34 -0400, Andrew Dunstan wrote:
>>> If I disable install, the buildfarm fails the upgrade check even when
>>> not using NO_TEMP_INSTALL.
>>>
>>>
>>> excerpts from the log:
>>> sh: /home/pgl/npgl/pg_head/bfroot/HEAD/inst/bin/psql: No such file or
>>> directory
>> That's the issue I was talking to Tom about above. Need to
>> unconditionally have
>> ....
> Andrew, after the latest set of changes, the reversed order should now
> work reliably?
>

With the latest changes I don't get the above failure:

andrew(at)emma:pg_head (master)$ ~/bf/client-code/run_build.pl 
--skip-steps=install
master:HEAD          [19:21:45] creating vpath build dir
/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build ...
master:HEAD          [19:21:45] running configure ...
master:HEAD          [19:21:52] running make ...
master:HEAD          [19:23:42] running make check ...
master:HEAD          [19:24:37] running make contrib ...
master:HEAD          [19:24:45] running make testmodules ...
master:HEAD          [19:24:45] checking pg_upgrade
master:HEAD          [19:26:41] checking test-decoding
master:HEAD          [19:26:59] running make ecpg check ...
master:HEAD          [19:27:25] OK

cheers

andrew

--
Andrew Dunstan https://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services