Bug#942609: pspp: build with -O2 to avoid misbuild with CFLAGS=-O3

[Ben Pfaff]

On Fri, Oct 18, 2019 at 4:27 PM Steve Langasek
<steve.langasek at canonical.com> wrote:
>
> Package: pspp
> Version: 1.2.0-3
> Severity: minor
> Tags: patch
> User: ubuntu-devel at lists.ubuntu.com
> Usertags: origin-ubuntu focal ubuntu-patch
>
> Dear maintainers,
>
> In Ubuntu, pspp was failing to build on ppc64el because the Ubuntu ppc64el
> port, unlike Debian's, uses -O3 as the default optimization level; and when
> built this way, at least on ppc64el the package fails one of its tests:
>
> [...]
> ## ------------------------ ##
> Failed tests:
> GNU PSPP 1.2.0 test suite test groups:
>
>  NUM: FILE-NAME:LINE     TEST-GROUP-NAME
>       KEYWORDS
>
>  126: sys-file-reader.at:2951 zero or one variable in mrset
>       sack synthetic system file negative multiple response
>
> ## ---------------------- ##
> ## Detailed failed tests. ##
> ## ---------------------- ##
>
> #                             -*- compilation -*-
> 126. sys-file-reader.at:2951: testing zero or one variable in mrset ...
> ./sys-file-reader.at:2970: sack --$variant sys-file.sack > sys-file.sav
> ./sys-file-reader.at:2973: pspp -O format=csv sys-file.sps
> --- -   2019-01-16 11:24:48.338741568 +0000
> +++ /<<PKGBUILDDIR>>/tests/testsuite.dir/at-groups/126/stdout   2019-01-16 11:24:48.335475863 +0000
> @@ -1,3 +1,5 @@
> +warning: `sys-file.sav' near offset 0xd8: Missing new-line parsing variable names at offset 22 in MRSETS record.

Thank you very much for the note and the patch.

Do you have any idea whether this is likely to be a previously unknown bug
in PSPP (as opposed to a bug somewhere else, such as in GCC)? I feel like
I ought to invest some time into trying to find a PSPP bug here, but on the
other hand I don't want to chase a ghost.