GNU bug report logs - #19609
coreutils tests on illumos

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 19609 in the body.
You can then email your comments to 19609 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Alexander Pyhalov <alp <at> rsu.ru>
To: bug-coreutils <at> gnu.org
Subject: coreutils tests on illumos
Date: Thu, 15 Jan 2015 21:05:29 +0300

Hello.
I was looking at coreutils test suite on illumos.
There are currently about 5 failing tests.

First are 2 df tests. They assume that struct mntent *getmntent (FILE 
*fp) is used by df. But it's not true on SystemV. Instead, int getmntent 
(FILE *fp, struct mnttab *mp) is used. Patches to correct tests on 
illumos is here:
https://github.com/OpenIndiana/oi-userland/blob/oi/hipster/components/coreutils/patches/tests_df_no-mtab-status.sh.patch
https://github.com/OpenIndiana/oi-userland/blob/oi/hipster/components/coreutils/patches/tests_df_skip-duplicates.sh.patch

3 remaining failing tests are from glib-tests directory.

First one is test-locale.c. It assumes that LC_GLOBAL_LOCALE is a 
constant. On illumos it's a function, so compilation fails. Fix is here: 
https://github.com/OpenIndiana/oi-userland/blob/oi/hipster/components/coreutils/patches/gnulib-tests_test-locale.c.patch

Second is test-getlogin. It seems to be a bit strange: it compares $USER 
and getlogin(). Under "su" they don't match, so it can occasionally fail.

Third (test-getlogin) is a bit tricky. It uses 
gl_locale_name_thread_unsafe, which simply doesn't know about 
illumos/Solaris and returns NULL, so the test fails. I don't have a fix 
for this one.

-- 
Best regards,
Alexander Pyhalov,
system administrator of Southern Federal University IT department

Message #8 received at 19609 <at> debbugs.gnu.org (full text, mbox):

From: Pádraig Brady <P <at> draigBrady.com>
To: Alexander Pyhalov <alp <at> rsu.ru>, 19609 <at> debbugs.gnu.org
Subject: Re: bug#19609: coreutils tests on illumos
Date: Mon, 19 Jan 2015 13:52:34 +0000

tag 19609 notabug
close 19609
stop

On 15/01/15 18:05, Alexander Pyhalov wrote:
> Hello.
> I was looking at coreutils test suite on illumos.
> There are currently about 5 failing tests.
> 
> First are 2 df tests. They assume that struct mntent *getmntent (FILE 
> *fp) is used by df. But it's not true on SystemV. Instead, int getmntent 
> (FILE *fp, struct mnttab *mp) is used. Patches to correct tests on 
> illumos is here:
> https://github.com/OpenIndiana/oi-userland/blob/oi/hipster/components/coreutils/patches/tests_df_no-mtab-status.sh.patch
> https://github.com/OpenIndiana/oi-userland/blob/oi/hipster/components/coreutils/patches/tests_df_skip-duplicates.sh.patch

The above are already addressed in v8.23 with:
http://git.sv.gnu.org/gitweb/?p=coreutils.git;a=commitdiff;h=39e2a4cf

> 3 remaining failing tests are from glib-tests directory.
> 
> First one is test-locale.c. It assumes that LC_GLOBAL_LOCALE is a 
> constant. On illumos it's a function, so compilation fails. Fix is here: 
> https://github.com/OpenIndiana/oi-userland/blob/oi/hipster/components/coreutils/patches/gnulib-tests_test-locale.c.patch
> 
> Second is test-getlogin. It seems to be a bit strange: it compares $USER 
> and getlogin(). Under "su" they don't match, so it can occasionally fail.

There have been a few changes to that test in gnulib since
coreutils 8.21 was released, so I'm hopeful that's addressed too.

> Third (test-getlogin) is a bit tricky.

I presume you mean test-localename.c

> It uses gl_locale_name_thread_unsafe, which simply doesn't know about 
> illumos/Solaris and returns NULL, so the test fails. I don't have a fix 
> for this one.

I'll follow up on bug-gnulib <at> gnu.org

thanks,
Pádraig

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.