Re: bin/55976 (usr.bin/xlint/lint1/t_integration:all_messages test case fails)

[Roland Illig <roland.illig%gmx.de@localhost>]

The following reply was made to PR bin/55976; it has been noted by GNATS.

From: Roland Illig <roland.illig%gmx.de@localhost>
To: Andreas Gustafsson <gson%NetBSD.org@localhost>
Cc: gnats-bugs%netbsd.org@localhost
Subject: Re: bin/55976 (usr.bin/xlint/lint1/t_integration:all_messages test
 case fails)
Date: Thu, 4 Feb 2021 23:41:31 +0100

 On 04.02.2021 20:55, Andreas Gustafsson wrote:
 > rillig%NetBSD.org@localhost wrote:
 >> State-Changed-From-To: open->closed
 >> State-Changed-By: rillig%NetBSD.org@localhost
 >> State-Changed-When: Thu, 04 Feb 2021 19:21:49 +0000
 >> State-Changed-Why:
 >> This was fixed on msg_324.c 1.4 from 2021-01-06 09:23:04.
 >
 > Yes, but it was broken again on 2021-01-08 21:25:03:
 >
 >    http://releng.netbsd.org/b5reports/i386/commits-2021.01.html#2021.01.=
 08.21.25.03
 >
 > and still failed today:
 >
 >    http://releng.netbsd.org/b5reports/i386/2021/2021.02.04.15.08.44/test=
 .html#usr.bin_xlint_lint1_t_integration_all_messages
 
 Given that the test is called "all_messages" and contains 338 sub-tests,
 it's no surprise that some of them fail from time to time since lint
 behaves differently on 32-bit and 64-bit platforms.  It's not the same
 sub-tests that fail though.  The PR is about message 324, while
 currently messages 56 and 259 fail.
 
 I am fixing these tests as soon as I notice them, from my point of view
 there is no need to file a PR for that.  For 259 I'm unsure how to fix
 it for all platforms at once, therefore I only committed some
 documentation that the test fails on i386.
 
 In my pre-commit hook, I verify that the tests run fine on x86_64, to
 prevent me from committing complete crap.  I should probably add i386 to
 that as well.  From lint's point of view, there is no difference between
 i386 and sparc, they are both ILP32 platforms.  Therefore I think its
 unnecessary to test sparc or sparc64 as well.
 
 Roland