Bug#1025492: marked as done (openjdk-17: Upstream commit f71b21b causes segfault in hdf5 java test suite on arches i386 and mips64el)

Debian Bug Tracking System Sun, 18 Dec 2022 15:03:23 -0800

Your message dated Sun, 18 Dec 2022 23:58:42 +0100
with message-id <[email protected]>
and subject line Re: openjdk-17: Upstream commit f71b21b causes segfault in 
hdf5 java test suite on arches i386 and mips64el
has caused the Debian Bug report #1025492,
regarding openjdk-17: Upstream commit f71b21b causes segfault in hdf5 java test 
suite on arches i386 and mips64el
to be marked as done.


This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [email protected]
immediately.)


-- 
1025492: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1025492
Debian Bug Tracking System
Contact [email protected] with problems

--- Begin Message ---

Package: openjdk-17
Version: 17~14-1
Severity: serious
Tags: upstream
Justification: makes unrelated software on the system break

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Hi,

A few weeks ago default-jdk was switched from openjdk-11 to openjdk-17.
Afterward hdf5 has been suffering FTBFS on arches i386 [1] and mips64el [2]
caused by a segfault in one of its java test case:

junit.sh: line 926: 3107321 Aborted                 ( $RUNSERIAL $JAVAEXE 
$JAVAEXEFLAGS -Xmx1024M -Dorg.slf4j.simpleLogger.defaultLog=trace 
-Djava.library.path=$BLDLIBDIR -cp $CLASSPATH -ea org.junit.runner.JUnitCore 
test.TestH5Arw > JUnit-TestH5Arw.ext )
**FAILED**    JUnit-TestH5Arw
    Expected result differs from actual result
    *** JUnit-TestH5Arw.txt     2022-11-26 00:28:33.024585625 +0000
    --- JUnit-TestH5Arw.out     2022-11-26 00:28:42.316468486 +0000
    ***************
    *** 7,13 ****
      .testH5Aread_32bit_floats
      .testH5Aread_16bit_ints
      
    ! Time:  XXXX
    ! 
    ! OK (7 tests)
    ! 
    --- 7,27 ----
      .testH5Aread_32bit_floats
      .testH5Aread_16bit_ints
      
    ! #
    ! # A fatal error has been detected by the Java Runtime Environment:
    ! #
    ! #  SIGSEGV (0xb) at pc=0xf74070c5, pid=3107321, tid=3107322
    ! #
    ! # JRE version: OpenJDK Runtime Environment (version (number)) (build 
17.0.5+8-Debian-2)
    ! # Java VM: OpenJDK Server VM (version (number))
    ! # Problematic frame:
    ! # V  [libjvm.so+0x65e0c5]
    ! #
    ! # No core dump will be written. Core dumps have been disabled. To enable 
core dumping, try "ulimit -c unlimited" before starting Java again
    ! #
    ! # An error report file with more information is saved as:
    ! # /<<BUILDDIR>>/hdf5-version (number)
    ! #
    ! # If you would like to submit a bug report, please visit:
    ! #   https://bugs.debian.org/openjdk-17
    ! #

[1] 
https://buildd.debian.org/status/fetch.php?pkg=hdf5&arch=i386&ver=1.10.8%2Brepack-3&stamp=1669422533&raw=0
[2] 
https://buildd.debian.org/status/fetch.php?pkg=hdf5&arch=mips64el&ver=1.10.8%2Brepack-3&stamp=1669424344&raw=0

Thanks to snapshot.debian.or I was able to spot that openjdk-17 17~14-1
was the first release with this issue in Debian.

I then ran a bisect against the openjdk upstream repo and found out that
upstream commit f71b21b [3]  was the culprit. Reverting this commit on
current openjdk-17 source (17.0.5+8-2) makes the hdf5 java test suite
successful.

[3] https://github.com/openjdk/jdk/commit/f71b21b

I don't know what to do from here. Any chance to release openjdk-17 with a patch
reverting the faulty commit?

Setting this bug as serious since it causes an unrelated package to FTBFS.

Thanks,
_g.


- -- System Information:
Debian Release: buster/sid
Architecture: amd64 (x86_64)

Kernel: Linux 5.10.0-19-amd64 (SMP w/12 CPU threads)
Kernel taint flags: TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=fr_FR.UTF-8, LC_CTYPE=fr_FR.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled

-----BEGIN PGP SIGNATURE-----

iQEzBAEBCgAdFiEEoJObzArDE05WtIyR7+hsbH/+z4MFAmOORD8ACgkQ7+hsbH/+
z4MFRQf+LG9hnV03M1QsxDPg7mHeDQBIzQEaBl6jTOjIgnQ0n7qwlxk+hKdhi4pw
5oKrDWN3XxzxpwDcP+PH8/7JBbtp1b6m+xFFTe12fogj3/So7/hsJRoImZFajbO3
VKEeCyV8K1d11T13nJdXXvcFtQ1ergeApI5ClY6JIsT499Lj0r6tTZNvblGXfDMp
qq2MHnNFM4AoRcXY+PGhnYJY4InmvL6Cg/1gUnWul45n63WrFE+R19MnLbnR+rLW
K1XAUDKG0jKxhkbOeZ6B80sYldza+vhuAsilga9Y6I1NludMLzR/91+u9GApBYRd
mUiUpIeWB3X/iwgPCkpS4v9DLvR+qg==
=KzlR
-----END PGP SIGNATURE-----

--- End Message ---

--- Begin Message ---
Gilles Filippini a écrit le 05/12/2022 à 22:57 :
On Mon, 05 Dec 2022 20:19:37 +0100 Gilles Filippini <[email protected]> wrote:
Package: openjdk-17
Version: 17~14-1
Severity: serious
Tags: upstream
Justification: makes unrelated software on the system break

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Hi,

A few weeks ago default-jdk was switched from openjdk-11 to openjdk-17.
Afterward hdf5 has been suffering FTBFS on arches i386 [1] and mips64el [2]
caused by a segfault in one of its java test case:
junit.sh: line 926: 3107321 Aborted ( $RUNSERIAL $JAVAEXE $JAVAEXEFLAGS -Xmx1024M -Dorg.slf4j.simpleLogger.defaultLog=trace -Djava.library.path=$BLDLIBDIR -cp $CLASSPATH -ea org.junit.runner.JUnitCore test.TestH5Arw > JUnit-TestH5Arw.ext )
**FAILED**    JUnit-TestH5Arw
    Expected result differs from actual result
    *** JUnit-TestH5Arw.txt    2022-11-26 00:28:33.024585625 +0000
    --- JUnit-TestH5Arw.out    2022-11-26 00:28:42.316468486 +0000
    ***************
    *** 7,13 ****
      .testH5Aread_32bit_floats
      .testH5Aread_16bit_ints
    ! Time:  XXXX
    !     ! OK (7 tests)
    !     --- 7,27 ----
      .testH5Aread_32bit_floats
      .testH5Aread_16bit_ints
    ! #
    ! # A fatal error has been detected by the Java Runtime Environment:
    ! #
    ! #  SIGSEGV (0xb) at pc=0xf74070c5, pid=3107321, tid=3107322
    ! #
! # JRE version: OpenJDK Runtime Environment (version (number)) (build 17.0.5+8-Debian-2)
    ! # Java VM: OpenJDK Server VM (version (number))
    ! # Problematic frame:
    ! # V  [libjvm.so+0x65e0c5]
    ! #
! # No core dump will be written. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
    ! #
    ! # An error report file with more information is saved as:
    ! # /<<BUILDDIR>>/hdf5-version (number)
    ! #
    ! # If you would like to submit a bug report, please visit:
    ! #   https://bugs.debian.org/openjdk-17
    ! #
[1] https://buildd.debian.org/status/fetch.php?pkg=hdf5&arch=i386&ver=1.10.8%2Brepack-3&stamp=1669422533&raw=0 [2] https://buildd.debian.org/status/fetch.php?pkg=hdf5&arch=mips64el&ver=1.10.8%2Brepack-3&stamp=1669424344&raw=0
Closing this issue as it appears the test failure was due to an array overflow in the test itself.
Best,
_g.
OpenPGP_signature
Description: OpenPGP digital signature

--- End Message ---

Bug#1025492: marked as done (openjdk-17: Upstream commit f71b21b causes segfault in hdf5 java test suite on arches i386 and mips64el)

Reply via email to