Test failures on s390x

[opoplawski]

I'm working on updating hdf5 in Fedora to 1.10.5, and I'm getting test failures with matio 1.5.14 only on s390x:

621: Read 2D Unicode character array                 FAILED (mat5_compressed_read_le.at:182)
Read/Write Compressed MAT version 5
2825: Copy variables                                  skipped (mat5_compressed_readwrite.at:28)
2826: Copy variables                                  FAILED (mat5_compressed_readwrite.at:38)
2827: Delete variables                                FAILED (mat5_compressed_readwrite.at:46)
2828: Copy large struct                               skipped (mat5_compressed_readwrite.at:149)
2829: Copy large struct                               ok

Full s390x build log: https://kojipkgs.fedoraproject.org//work/tasks/7129/33587129/build.log

[tbeu]

Probably duplicate of #108.

[opoplawski]

More details:

## ---------------------- ##
## Detailed failed tests. ##
## ---------------------- ##
#                             -*- compilation -*-
621. mat5_compressed_read_le.at:180: testing Read 2D Unicode character array ...
./mat5_compressed_read_le.at:182: cp $srcdir/results/read-var95-5.out expout
         $builddir/test_mat readvar $srcdir/datasets/matio_test_cases_compressed_le.mat var95
--- /dev/null   2019-03-19 02:23:16.979272966 +0000
+++ /builddir/build/BUILD/matio-1.5.14/test/testsuite.dir/at-groups/621/stderr  2019-03-19 02:24:28.489272966 +0000
@@ -0,0 +1,2 @@
+corrupted size vs. prev_size
+/builddir/build/BUILD/matio-1.5.14/test/testsuite.dir/at-groups/621/test-source: line 17: 46408 Aborted                 (core dumped) $builddir/test_mat readvar $srcdir/datasets/matio_test_cases_compressed_le.mat var95
--- expout  2019-03-19 02:24:27.539272966 +0000
+++ /builddir/build/BUILD/matio-1.5.14/test/testsuite.dir/at-groups/621/stdout  2019-03-19 02:24:27.539272966 +0000
@@ -1,9 +0,0 @@
-      Name: var95
-      Rank: 2
-Dimensions: 2 x 4
-Class Type: Character Array
- Data Type: Unicode UTF-16 Encoded Character Data
-{
-بجلی
-điện
-}
./mat5_compressed_read_le.at:182: exit code was 134, expected 0
621. mat5_compressed_read_le.at:180: 621. Read 2D Unicode character array (mat5_compressed_read_le.at:180): FAILED (mat5_compressed_read_le.at:182)
#                             -*- compilation -*-
2826. mat5_compressed_readwrite.at:34: testing Copy variables ...
./mat5_compressed_readwrite.at:38: $builddir/test_mat -v 5 -z copy $srcdir/datasets/matio_test_cases_compressed_le.mat
--- /dev/null   2019-03-19 02:23:16.979272966 +0000
+++ /builddir/build/BUILD/matio-1.5.14/test/testsuite.dir/at-groups/2826/stderr 2019-03-19 02:26:55.349282966 +0000
@@ -0,0 +1,2 @@
+corrupted size vs. prev_size
+/builddir/build/BUILD/matio-1.5.14/test/testsuite.dir/at-groups/2826/test-source: line 19: 52122 Aborted                 (core dumped) $builddir/test_mat -v 5 -z copy $srcdir/datasets/matio_test_cases_compressed_le.mat
./mat5_compressed_readwrite.at:38: exit code was 134, expected 0
2826. mat5_compressed_readwrite.at:34: 2826. Copy variables (mat5_compressed_readwrite.at:34): FAILED (mat5_compressed_readwrite.at:38)
#                             -*- compilation -*-
2827. mat5_compressed_readwrite.at:43: testing Delete variables ...
./mat5_compressed_readwrite.at:46: $builddir/test_mat -v 5 -z copy $srcdir/datasets/matio_test_cases_compressed_le.mat
--- /dev/null   2019-03-19 02:23:16.979272966 +0000
+++ /builddir/build/BUILD/matio-1.5.14/test/testsuite.dir/at-groups/2827/stderr 2019-03-19 02:26:56.379282966 +0000
@@ -0,0 +1,2 @@
+corrupted size vs. prev_size
+/builddir/build/BUILD/matio-1.5.14/test/testsuite.dir/at-groups/2827/test-source: line 16: 52148 Aborted                 (core dumped) $builddir/test_mat -v 5 -z copy $srcdir/datasets/matio_test_cases_compressed_le.mat
./mat5_compressed_readwrite.at:46: exit code was 134, expected 0
2827. mat5_compressed_readwrite.at:43: 2827. Delete variables (mat5_compressed_readwrite.at:43): FAILED (mat5_compressed_readwrite.at:46)

Full testsuite.log can be seen here: https://kojipkgs.fedoraproject.org//work/tasks/5100/33625100/build.log

[tbeu]

@opoplawski Can you please confirm that 02625a0e394eeb8bf3ea61641f73de907296a2c4 fixes the issue? Thanks.

[tbeu]

@opoplawski Waiting for your final approval here to release a new version replacing the unfortunate v1.5.14.

[tbeu]

Hm, not sure how long to wait to finally get a confirmation.

[opoplawski]

Sorry for the delay - very busy here. 1.5.15 builds and tests fine on s390x.