Test suite fails on ibm 390x

[nirgal]

Hi

There seems to be a issue in rpn2 test suite on 390x architecture.
Here's the test-suite.log

=========================================
   rrdtool 1.5.4: tests/test-suite.log
=========================================

# TOTAL: 20
# PASS:  19
# SKIP:  0
# XFAIL: 0
# FAIL:  1
# XPASS: 0
# ERROR: 0

.. contents:: :depth: 2

FAIL: rpn2
==========

OK: create
OK: update
--- -   2015-08-13 17:48:47.758470590 +0200
+++ /home/nirgal/rrdtool/tests/rpn2.output      2015-08-13 17:41:40.000000000 +0200
@@ -20,19 +20,19 @@
     [ 1.0000000e-01, 1.4400000e+03, 0.0000000e+00, 1.4400000e+03, 0.0000000e+00, 1.4400000e+03, 0.0000000e+00, 1.4400000e+03, 0.0000000e+00 ],
     [ 1.0000000e-01, 2.1600000e+03, 0.0000000e+00, 2.1600000e+03, 0.0000000e+00, 2.1600000e+03, 0.0000000e+00, 2.1600000e+03, 0.0000000e+00 ],
     [ 1.0000000e-01, 2.8800000e+03, 0.0000000e+00, 2.8800000e+03, 0.0000000e+00, 2.8800000e+03, 0.0000000e+00, 2.8800000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 7.2000000e+02, 1.0000000e+00, 3.6000000e+03, 0.0000000e+00, 3.6000000e+03, 0.0000000e+00, 3.6000000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 1.4400000e+03, 0.0000000e+00, 4.3200000e+03, 0.0000000e+00, 4.3200000e+03, 0.0000000e+00, 4.3200000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 2.1600000e+03, 0.0000000e+00, 5.0400000e+03, 0.0000000e+00, 5.0400000e+03, 0.0000000e+00, 5.0400000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 2.8800000e+03, 0.0000000e+00, 5.7600000e+03, 0.0000000e+00, 5.7600000e+03, 0.0000000e+00, 5.7600000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 3.6000000e+03, 0.0000000e+00, 6.4800000e+03, 0.0000000e+00, 6.4800000e+03, 0.0000000e+00, 6.4800000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 4.3200000e+03, 0.0000000e+00, 7.2000000e+03, 0.0000000e+00, 7.2000000e+03, 0.0000000e+00, 7.2000000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 5.0400000e+03, 0.0000000e+00, 7.9200000e+03, 0.0000000e+00, 7.9200000e+03, 0.0000000e+00, 7.9200000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 5.7600000e+03, 0.0000000e+00, 8.6400000e+03, 0.0000000e+00, 8.6400000e+03, 0.0000000e+00, 8.6400000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 6.4800000e+03, 0.0000000e+00, 9.3600000e+03, 0.0000000e+00, 9.3600000e+03, 0.0000000e+00, 9.3600000e+03, 0.0000000e+00 ],
-    [ 1.0000000e-01, 7.2000000e+03, 0.0000000e+00, 1.0080000e+04, 0.0000000e+00, 1.0080000e+04, 0.0000000e+00, 1.0080000e+04, 0.0000000e+00 ],
-    [ 1.0000000e-01, 7.9200000e+03, 0.0000000e+00, 1.0800000e+04, 0.0000000e+00, 1.0800000e+04, 0.0000000e+00, 1.0800000e+04, 0.0000000e+00 ],
-    [ 1.0000000e-01, 8.6400000e+03, 0.0000000e+00, 1.1520000e+04, 0.0000000e+00, 1.1520000e+04, 0.0000000e+00, 1.1520000e+04, 0.0000000e+00 ],
-    [ 1.0000000e-01, 7.2000000e+02, 1.0000000e+00, 7.2000000e+02, 1.0000000e+00, 7.2000000e+02, 1.0000000e+00, 7.2000000e+02, 1.0000000e+00 ],
-    [ 1.0000000e-01, 1.4400000e+03, 0.0000000e+00, 1.4400000e+03, 0.0000000e+00, 1.4400000e+03, 0.0000000e+00, 1.4400000e+03, 0.0000000e+00 ]
+    [ 1.0000000e-01, 7.2000000e+02, 1.0000000e+00, 7.2000000e+02, 1.0000000e+00, 3.6000000e+03, 0.0000000e+00, 3.6000000e+03, 0.0000000e+00 ],
+    [ 1.0000000e-01, 1.4400000e+03, 0.0000000e+00, 1.4400000e+03, 0.0000000e+00, 4.3200000e+03, 0.0000000e+00, 4.3200000e+03, 0.0000000e+00 ],
+    [ 1.0000000e-01, 2.1600000e+03, 0.0000000e+00, 2.1600000e+03, 0.0000000e+00, 5.0400000e+03, 0.0000000e+00, 5.0400000e+03, 0.0000000e+00 ],
+    [ 1.0000000e-01, 2.8800000e+03, 0.0000000e+00, 2.8800000e+03, 0.0000000e+00, 5.7600000e+03, 0.0000000e+00, 5.7600000e+03, 0.0000000e+00 ],
+    [ 1.0000000e-01, 3.6000000e+03, 0.0000000e+00, 3.6000000e+03, 0.0000000e+00, 6.4800000e+03, 0.0000000e+00, 6.4800000e+03, 0.0000000e+00 ],
+    [ 1.0000000e-01, 4.3200000e+03, 0.0000000e+00, 4.3200000e+03, 0.0000000e+00, 7.2000000e+03, 0.0000000e+00, 7.2000000e+03, 0.0000000e+00 ],
+    [ 1.0000000e-01, 5.0400000e+03, 0.0000000e+00, 5.0400000e+03, 0.0000000e+00, 7.9200000e+03, 0.0000000e+00, 7.9200000e+03, 0.0000000e+00 ],
+    [ 1.0000000e-01, 5.7600000e+03, 0.0000000e+00, 5.7600000e+03, 0.0000000e+00, 8.6400000e+03, 0.0000000e+00, 8.6400000e+03, 0.0000000e+00 ],
+    [ 1.0000000e-01, 6.4800000e+03, 0.0000000e+00, 6.4800000e+03, 0.0000000e+00, 9.3600000e+03, 0.0000000e+00, 9.3600000e+03, 0.0000000e+00 ],
+    [ 1.0000000e-01, 7.2000000e+03, 0.0000000e+00, 7.2000000e+03, 0.0000000e+00, 1.0080000e+04, 0.0000000e+00, 1.0080000e+04, 0.0000000e+00 ],
+    [ 1.0000000e-01, 7.9200000e+03, 0.0000000e+00, 7.9200000e+03, 0.0000000e+00, 1.0800000e+04, 0.0000000e+00, 1.0800000e+04, 0.0000000e+00 ],
+    [ 1.0000000e-01, 8.6400000e+03, 0.0000000e+00, 8.6400000e+03, 0.0000000e+00, 1.1520000e+04, 0.0000000e+00, 1.1520000e+04, 0.0000000e+00 ],
+    [ 1.0000000e-01, 7.2000000e+02, 1.0000000e+00, 9.3600000e+03, 0.0000000e+00, 7.2000000e+02, 1.0000000e+00, 7.2000000e+02, 1.0000000e+00 ],
+    [ 1.0000000e-01, 1.4400000e+03, 0.0000000e+00, 1.0080000e+04, 0.0000000e+00, 1.4400000e+03, 0.0000000e+00, 1.4400000e+03, 0.0000000e+00 ]
   ]
 }
FAILED: (rc=1) xport
FAIL rpn2 (exit status: 1)

I've spent some time on that issue, but I'm stuck for now. Maybe you'll have a good idea?

Looks like another first day of the week issue?

The build log with detected configuration is there:
https://buildd.debian.org/status/fetch.php?pkg=rrdtool&arch=s390x&ver=1.5.4-3&stamp=1439466415

[nirgal]

I got it! In rrd_rpncalc.c:static int find_first_weekday(void)
week_1stday_l == 85775350214164480, so we fall back to hard coded value.
Note that 85775350214164480 == 19971130 << 32

[oetiker]

hmm so the question is why is 390x lacking the weekday information ? maybe the locales package not installed ?

[nirgal]

I have a patch ready. Please stand by :)

It was tricky not to add warnings on platforms where long is only 32 bits ...

[c72578]

Here is the link to the PR fixing this issue: #662
And the merged commit: 750a713