Re: pg_upgrade test script creates port conflicts in parallel testing

Andrew Dunstan <andrew@dunslane.net> writes:
> On 01/03/2013 12:58 PM, Tom Lane wrote:
>> Does anyone have an objection to fixing the pg_upgrade test script
>> to suppress the TCP socket?

> Should be OK. We can't do that on Windows, though, so please make it 
> conditional so we don't break Mingw buildfarm members. The test script 
> already contains a few Windows variants.

I'm planning to do it like this:
testhost=`uname -s`
+case $testhost in
+    MINGW*)    LISTEN_ADDRESSES="localhost" ;;
+    *)        LISTEN_ADDRESSES="" ;;
+esac
+
+POSTMASTER_OPTS="-F -c listen_addresses=$LISTEN_ADDRESSES"
+temp_root=$PWD/tmp_check

which matches the existing Windows-specific switches.
        regards, tom lane