Bug 701982 – sys-libs/pam QA: the address of 'binary_prompt' will always evaluate as 'true' [-Waddress]

• Product: Gentoo Linux
• Status: RESOLVED INVALID

[Description 7_9j~HTz 2019-12-04 17:36:54 UTC]

Created attachment 598470 [details]
sys conf

Though it is mentioned to report upstream I would rather not because the 1.3.0 release is dating back to 2016 and since then superseded by the 1.3.1 (May 2018) release which has not made into the Gentoo stable branch.

The issue reproduces on each attempt to get the package compiled.


[ebuild   R   ] sys-libs/pam-1.3.0-r2  USE="cracklib filecaps* nls (pie) (split-usr) -audit -berkdb -debug -nis (-selinux) -test -vim-syntax" ABI_X86="(64) -32 (-x32)"

 * QA Notice: Package triggers severe warnings which indicate that it
 *            may exhibit random runtime failures.
 * /var/tmp/portage/sys-libs/pam-1.3.0-r2/work/Linux-PAM-1.3.0/libpam_misc/misc_conv.c:329:19: warning: the address of 'binary_prompt' will always evaluate as 'true' [-Waddress]

 * Please do not file a Gentoo bug and instead report the above QA
 * issues directly to the upstream developers of this software.
 * Homepage: http://www.linux-pam.org/
 * ERROR: sys-libs/pam-1.3.0-r2::gentoo failed:
 *   install aborted due to severe warnings shown above
 *
 * Call stack:
 *   misc-functions.sh, line 586:  Called install_qa_check
 *   misc-functions.sh, line 132:  Called source 'install_symlink_html_docs'
 *      90gcc-warnings, line 167:  Called gcc_warn_check
 *      90gcc-warnings, line 161:  Called die
 * The specific snippet of code:
 *                            has stricter ${FEATURES} && \
 *                                    die "install aborted due to severe warnings shown above"
 *

[Comment 1 7_9j~HTz 2019-12-04 17:52:21 UTC]

1.3.1-r1 exhibits the same issue. It is not clear to me why being directed to upstream when PAM works on other distros?


* QA Notice: Package triggers severe warnings which indicate that it
 *            may exhibit random runtime failures.
 * /var/tmp/portage/sys-libs/pam-1.3.1-r1/work/linux-pam-1.3.1/libpam_misc/misc_conv.c:329:19: warning: the address of ▒binary_prompt▒ will always evaluate as ▒true▒ [-Waddress]

 * Please do not file a Gentoo bug and instead report the above QA
 * issues directly to the upstream developers of this software.
 * Homepage: https://github.com/linux-pam/linux-pam
 * ERROR: sys-libs/pam-1.3.1-r1::gentoo failed:
 *   install aborted due to severe warnings shown above
 *
 * Call stack:
 *   misc-functions.sh, line 586:  Called install_qa_check
 *   misc-functions.sh, line 132:  Called source 'install_symlink_html_docs'
 *      90gcc-warnings, line 167:  Called gcc_warn_check
 *      90gcc-warnings, line 161:  Called die
 * The specific snippet of code:
 *                            has stricter ${FEATURES} && \
 *                                    die "install aborted due to severe warnings shown above"

[Comment 2 7_9j~HTz 2019-12-04 17:59:08 UTC]

bug 701928 exhibits also QA issues and prints the same

 * Call stack:
 *     misc-functions.sh, line 586:  Called install_qa_check
 *     misc-functions.sh, line 132:  Called source 'install_symlink_html_docs'

[Comment 3 Brian Evans (RETIRED) 2019-12-04 18:27:52 UTC]

FWIW, your build only fails because you set FEATURES="stricter"  otherwise this is non-fatal

[Comment 4 7_9j~HTz 2019-12-04 18:31:20 UTC]

stricter is a legitimate setting and nothing wrong with deploying it, or? Plenty of packages are building with that setting.

[Comment 5 Brian Evans (RETIRED) 2019-12-04 19:02:01 UTC]

(In reply to 7_9j~HTz from comment #4)
> stricter is a legitimate setting and nothing wrong with deploying it, or?
> Plenty of packages are building with that setting.

"Plenty of packages" don't have "QA Warning"s either which stricter makes fatal.  If you remove that option, then this will install fine. This allows upstream to react to such an issue without impact on your system.

[Comment 6 7_9j~HTz 2019-12-04 19:24:17 UTC]

Will lodge an issue upstream. 

Is it possible to obsolete the option selectively with FEATURES="-stricter" for the package via /etc/portage/env - from the documentation it is not clear whether the syntax to obsolete an option from the global context is viable?

[Comment 7 Mikle Kolyada (RETIRED) 2019-12-04 20:18:37 UTC]

Not gentoo business.

[Comment 8 7_9j~HTz 2019-12-04 21:02:27 UTC]

selectively dropping FEATURES="-stricter" for the package via /etc/portage/env worked and the package compiled indeed.

Appreciate the pointer to the culprit because from the log output it is not apparent to uninitiated and neither the justification of the the deferral to upstream. Which would have been awkward to lodge the issue with not knowing what the log output implied.