intermittent regression check problems

Hello,

yesterday I reported an almost total failure in make check. As it was
suggested, I installed and then did "make check" again.

using that, I get a lower number of failures, around 20. I Still think
they are too much. Also the number is not stable, everytime I run it I
get a different result.

I past here two. I don't know how to itnerpret them. Maybe some errors
are expected in the regressions.diff file?
--- 404,409 ----
***************
*** 849,855 ****
  ERROR:  column "........pg.dropped.1........" of relation "test"
does not exist
  copy test from stdin;
  ERROR:  extra data after last expected column
! CONTEXT:  COPY test, line 1: "10    11    12"
  select * from test;
   b | c
  ---+---
--- 838,844 ----
  ERROR:  column "........pg.dropped.1........" of relation "test"
does not exist
  copy test from stdin;
  ERROR:  extra data after last expected column
! lost synchronization with server, resetting connection
  select * from test;
   b | c
  ---+---

if I run make installcheck I get
============== dropping database "regression"         ==============
dropdb: could not connect to database postgres: FATAL:  role "multix"
does not exis

which is the user I am running from

Riccardo