On Thu, Jan 16, 2014 at 8:22 PM, Andreas Tille <andreas@an3as.eu> wrote:
Hi Jorge,
in case you are not (yet) subscribed to
debian-med-packaging@lists.alioth.debian.org
(alternatively you can subscribe the package snp-sites in the package
tracking system) I'm forwarding this bug report.
Kind regards
Andreas.
----- Forwarded message from "Aaron M. Ucko" <ucko@debian.org> -----
Date: Thu, 16 Jan 2014 13:50:46 -0500
From: "Aaron M. Ucko" <ucko@debian.org>
To: Debian Bug Tracking System <submit@bugs.debian.org>
Subject: Bug#735596: snp-sites: FTBFS: FAIL: run-all-tests
X-Debian-PR-Message: report 735596
X-Debian-PR-Package: src:snp-sites
X-Debian-PR-Keywords:
X-Debian-PR-Source: snp-sites
Source: snp-sites
Version: 1.3-1
Severity: important
Justification: fails to build from source
The builds of snp-sites for several architectures (most 32-bit ones,
plus s390x) failed:
FAIL: run-all-tests
make[5]: Entering directory `/«PKGBUILDDIR»/src'
make[5]: Nothing to be done for `all'.
make[5]: Leaving directory `/«PKGBUILDDIR»/src'
============================================================================
Testsuite summary for snp-sites 1
============================================================================
# TOTAL: 1
# PASS: 0
# SKIP: 0
# XFAIL: 0
# FAIL: 1
# XPASS: 0
# ERROR: 0
============================================================================
See src/test-suite.log
I was able to reproduce the problem on i386, and found that the
underlying error (which would have been helpful to print
automatically) was
../tests/check-snp-sites.c:167:F:alignment_file:valid_initial_reference_sequence:0: Assertion 'strcmp(actual_reference_sequence,expected_reference_sequence) == 0' failed
Running the program under a debugger, with CK_FORK=no in the
environment, revealed trailing garbage in actual_reference_sequence;
initializing the variable to an empty string (automatically NUL-padded
to the array's length) cleared the error up.
Could you please take care of it?