- -- I was under the assumption that only float8 and datetime (possibly) would
- -- be broken right now...but it seems others (point/polygon and create
- -- function, and a couple of others) are broken too.

=============================== OUTPUT OF BUILD EFFORT ++++++++++++++++++++++++

Platform: Sparc Solaris 2.5 / gcc 2.7.2.1
Version: 6.1 970428

postgres$ mkdir 970428 ; cd 970428
postgres$ gzcat ../970428.tar.gz | tar xvf -
postgres$ cd src
postgres$ ./configure --prefix=/opt/postgres --with-template=sparc_solaris-gcc
postgres$ make
postgres$ make install
postgres$ initdb
postgres$ TZ=PST8PDT7,M04.01.00,M10.05.03 postmaster -S
postgres$ cd test/regress
postgres$ make all runtest
postgres$ cat regress.out


=============== Notes... =================
postmaster must already be running for the regression tests to succeed.
The non-GEQ optimizer will give more consistant results than will the
GEQ optimizer. See the regression testing README for more details.
The time zone might need to be set to PST/PDT for the date and time data
types to pass the regression tests; to do this type
setenv TZ PST8PDT7,M04.01.00,M10.05.03
before starting the postmaster.

=============== destroying old regression database... =================
=============== creating new regression database... =================
=============== running regression queries... =================
boolean .. ok
box .. ok
char .. ok
char16 .. ok
char2 .. ok
char4 .. ok
char8 .. ok
float4 .. ok
float8 .. failed
int2 .. ok
int4 .. ok
oid .. ok
oidint2 .. ok
oidint4 .. ok
oidname .. ok
point .. failed
polygon .. failed
text .. ok
timespan .. ok
datetime .. failed
reltime .. failed
abstime .. failed
tinterval .. failed
create_function_1 .. failed
create_type .. ok
create_table .. ok
create_function_2 .. ok
copy .. ok
create_misc .. ok
create_aggregate .. ok
create_operator .. ok
create_view .. ok
create_index .. ok
sanity_check .. failed
select .. ok
select_into .. ok
select_distinct .. ok
select_distinct_on .. ok
transactions .. diff: expected/transactions.out: No such file or directory
ok
portals .. diff: expected/portals.out: No such file or directory
ok
errors .. ok
misc .. failed
random .. diff: expected/random.out: No such file or directory
ok
arrays .. diff: expected/arrays.out: No such file or directory
ok
btree_index .. diff: expected/btree_index.out: No such file or directory
ok
hash_index .. diff: expected/hash_index.out: No such file or directory
ok
aggregates .. diff: expected/aggregates.out: No such file or directory
ok
select_views .. diff: expected/select_views.out: No such file or directory
ok
alter_table .. diff: expected/alter_table.out: No such file or directory
ok
purge .. diff: expected/purge.out: No such file or directory
ok
portals_p2 .. diff: expected/portals_p2.out: No such file or directory
ok

------------------------------

Search Discussions

  • The Hermit Hacker at Apr 29, 1997 at 2:20 am

    On Mon, 28 Apr 1997 adrian@waltham.harvard.net wrote:

    -- I was under the assumption that only float8 and datetime (possibly) would
    -- be broken right now...but it seems others (point/polygon and create
    -- function, and a couple of others) are broken too.
    *Most* of the failed results, if you look at the diffs, are a result
    of ordering problems, and not true bugs...

    The final 'failures' you see are those tests with no current
    'expected' files, which Thomas is working on...


    Marc G. Fournier
    Systems Administrator @ hub.org
    primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org

    ------------------------------
  • Vadim B. Mikheev at Apr 29, 1997 at 3:20 am

    The Hermit Hacker wrote:
    On Mon, 28 Apr 1997 adrian@waltham.harvard.net wrote:

    -- I was under the assumption that only float8 and datetime (possibly) would
    -- be broken right now...but it seems others (point/polygon and create
    -- function, and a couple of others) are broken too.
    *Most* of the failed results, if you look at the diffs, are a result
    of ordering problems, and not true bugs...
    ^^^^^^^^
    I have patches to turn GEQO ON/OFF via SET var TO... and want to add
    them in 1-2 hours - may it help ?

    Vadim

    ------------------------------
  • The Hermit Hacker at Apr 29, 1997 at 8:18 am

    On Tue, 29 Apr 1997, Vadim B. Mikheev wrote:

    The Hermit Hacker wrote:
    On Mon, 28 Apr 1997 adrian@waltham.harvard.net wrote:

    -- I was under the assumption that only float8 and datetime (possibly) would
    -- be broken right now...but it seems others (point/polygon and create
    -- function, and a couple of others) are broken too.
    *Most* of the failed results, if you look at the diffs, are a result
    of ordering problems, and not true bugs...
    ^^^^^^^^
    I have patches to turn GEQO ON/OFF via SET var TO... and want to add
    them in 1-2 hours - may it help ?
    Yes!! Those would help immensely :) But, we need to test the
    GEQO code also. How does one test the GEQO code though?

    Marc G. Fournier
    Systems Administrator @ hub.org
    primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org

    ------------------------------
  • Thomas Lockhart at Apr 29, 1997 at 1:43 pm

    ... But, we need to test the GEQO code also.
    How does one test the GEQO code though?
    The results are the same as without GEQO, just (sometimes) in
    a different order. If the other tests pass, then we can turn on
    GEQ and do diffs on _sorted_ output against _sorted_ expected.
    If we/you can do the sorts on the fly then your "ok/failed"
    info will be accurate and the results/geq.out will not be
    jumbled and can be poked through to analyse problems.

    We should probably put in some regression tests for cases suited
    to GEQ, such as joins of 10 tables, etc. I'm ignoring new
    regression tests until we have the new regression structure
    firmed up.

    Will try committing updates to regression this morning...

    - Tom

    ------------------------------

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
grouppgsql-hackers @
categoriespostgresql
postedApr 29, '97 at 2:07a
activeApr 29, '97 at 1:43p
posts5
users4
websitepostgresql.org...
irc#postgresql

People

Translate

site design / logo © 2022 Grokbase