Re: enable certain TAP tests for MSVC builds

From: Andrew Dunstan <andrew(at)dunslane(dot)net>
To: Noah Misch <noah(at)leadboat(dot)com>
Cc: PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: enable certain TAP tests for MSVC builds
Date: 2021-12-07 20:40:29
Message-ID: 9461cfff-5685-b46b-ac2e-17a67c59c836@dunslane.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers


On 12/5/21 18:32, Noah Misch wrote:
> On Sun, Dec 05, 2021 at 06:00:08PM -0500, Andrew Dunstan wrote:
>> On 12/5/21 14:47, Noah Misch wrote:
>>> On Sun, Dec 05, 2021 at 11:57:31AM -0500, Andrew Dunstan wrote:
>>>> Certain TAP tests rely on settings that the Make files provide for them.
>>>> However vcregress.pl doesn't provide those settings. This patch remedies
>>>> that, and I propose to apply it shortly (when we have a fix for the SSL
>>>> tests that I will write about separately) and backpatch it appropriately.
>>>> --- a/src/tools/msvc/vcregress.pl
>>>> +++ b/src/tools/msvc/vcregress.pl
>>>> @@ -59,6 +59,12 @@ copy("$Config/autoinc/autoinc.dll", "src/test/regress");
>>>> copy("$Config/regress/regress.dll", "src/test/regress");
>>>> copy("$Config/dummy_seclabel/dummy_seclabel.dll", "src/test/regress");
>>>>
>>>> +# Settings used by TAP tests
>>>> +$ENV{with_ssl} = $config->{openssl} ? 'openssl' : 'no';
>>>> +$ENV{with_ldap} = $config->{ldap} ? 'yes' : 'no';
>>>> +$ENV{with_icu} = $config->{icu} ? 'yes' : 'no';
>>>> +$ENV{with_gssapi} = $config->{gss} ? 'yes' : 'no';
>>> That's appropriate. There are more variables to cover:
>>>
>>> $ git grep -n ^export ':(glob)**/Makefile'
>>> src/bin/pg_basebackup/Makefile:22:export LZ4
>>> src/bin/pg_basebackup/Makefile:23:export TAR
>>> src/bin/pg_basebackup/Makefile:27:export GZIP_PROGRAM=$(GZIP)
>>> src/bin/psql/Makefile:20:export with_readline
>>> src/test/kerberos/Makefile:16:export with_gssapi with_krb_srvnam
>>> src/test/ldap/Makefile:16:export with_ldap
>>> src/test/modules/ssl_passphrase_callback/Makefile:3:export with_ssl
>>> src/test/recovery/Makefile:20:export REGRESS_SHLIB
>>> src/test/ssl/Makefile:18:export with_ssl
>>> src/test/subscription/Makefile:18:export with_icu
>> LZ4/TAR/GZIP_PROGAM: It's not clear what these should be set to. The TAP
>> tests skip tests that use them if they are not set.
> Could add config.pl entries for those. Preventing those skips on Windows may
> or may not be worth making config.pl readers think about them.
>
>> with_readline: we don't build with readline on Windows, period. I guess
>> we could just set it to "no".
>> with_krb_srvnam: the default is "postgres", we could just set it to that
>> I guess.
> Works for me.

All done.

cheers

andrew

--
Andrew Dunstan
EDB: https://www.enterprisedb.com

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Peter Geoghegan 2021-12-07 20:43:57 Re: Why doesn't pgstat_report_analyze() focus on not-all-visible-page dead tuple counts, specifically?
Previous Message Robert Haas 2021-12-07 20:27:36 Re: Why doesn't pgstat_report_analyze() focus on not-all-visible-page dead tuple counts, specifically?