MySQL Bugs: #108789: OverflowBitsetTest.ZeroInitialize failure on x86

Bug #108789	OverflowBitsetTest.ZeroInitialize failure on x86
Submitted:	15 Oct 2022 8:09	Modified:	8 Nov 2022 15:00
Reporter:	Georgi Sotirov	Email Updates:
Status:	Unsupported	Impact on me:	None
Category:	MySQL Server: Tests	Severity:	S3 (Non-critical)
Version:	8.0.31	OS:	Linux (Slackware)
Assigned to:		CPU Architecture:	x86

Description:
While trying to build recently released MySQL 8.0.31 on Slackware Linux 15.0 x86 I'm getting the following test error:

$ cat /usr/src/tmp/mysql-8.0.31/build/Testing/Temporary/LastTest.log | sed -e '1,5049d' | head -10
[ RUN      ] OverflowBitsetTest.ZeroInitialize
/usr/src/tmp/mysql-8.0.31/unittest/gunit/overflow_bitset-t.cc:37: Failure
Expected equality of these values:
  31
  s.capacity()
    Which is: 63
[  FAILED  ] OverflowBitsetTest.ZeroInitialize (0 ms)

I never had this test failing before, but recently there was an issue with another test (see #106721).

Please, let me know what other information is necessary, so I could provide it.

How to repeat:
Download generic architecture independent source archive from https://dev.mysql.com/get/Downloads/MySQL-8.0/mysql-8.0.31.tar.gz and try to build the server on Slackware 15.0 (x86) with GCC 11.2.0 and the following configuration:

cmake -DCMAKE_BUILD_TYPE=Release \
      -DCMAKE_CXX_COMPILER=g++ \
      -DCMAKE_CXX_FLAGS="-O3" \
      -DCMAKE_C_COMPILER=gcc \
      -DCMAKE_C_FLAGS="-O3" \
      -DCMAKE_INSTALL_PREFIX=/usr \
      -DCMAKE_MAKE_PROGRAM=gmake \
      -DDOWNLOAD_BOOST=ON \
      -DENABLED_PROFILING=ON \
      -DFEATURE_SET="community" \
      -DINNODB_COMPILER_HINTS=ON \
      -DINSTALL_BINDIR=bin \
      -DINSTALL_DOCDIR=doc/mysql-8.0.31 \
      -DINSTALL_DOCREADMEDIR=doc/mysql-8.0.31 \
      -DINSTALL_INCLUDEDIR=include/mysql \
      -DINSTALL_LAYOUT=STANDALONE \
      -DINSTALL_LIBDIR=lib \
      -DINSTALL_MANDIR=man \
      -DINSTALL_MYSQLDATADIR=lib/mysql \
      -DINSTALL_MYSQLSHAREDIR=share/mysql \
      -DINSTALL_PLUGINDIR=lib/mysql/plugin \
      -DINSTALL_SBINDIR=sbin \
      -DINSTALL_SHAREDIR=share/mysql \
      -DINSTALL_SUPPORTFILESDIR=/etc/mysql \
      -DMYSQL_DATADIR=/var/lib/mysql \
      -DSYSCONFDIR=/etc/mysql \
      -DWITH_ARCHIVE_STORAGE_ENGINE=ON \
      -DWITH_AUTHENTICATION_LDAP=ON \
      -DWITH_BLACKHOLE_STORAGE_ENGINE=ON \
      -DWITH_BOOST=/tmp \
      -DWITH_CURL=system \
      -DWITH_EDITLINE=system \
      -DWITH_EXAMPLE_STORAGE_ENGINE=ON \
      -DWITH_EXTRA_CHARSETS=all \
      -DWITH_FEDERATED_STORAGE_ENGINE=ON \
      -DWITH_ICU=system \
      -DWITH_INNODB_MEMCACHED=ON \
      -DWITH_LIBEVENT=system \
      -DWITH_LZ4=system \
      -DWITH_NUMA=ON \
      -DWITH_PROTOBUF=system \
      -DWITH_RAPIDJSON=bundled \
      -DWITH_ROUTER=OFF \
      -DWITH_SSL=system \
      -DWITH_UNIT_TESTS=ON \
      -DWITH_ZLIB=system \
      -DWITH_ZSTD=system

Hi Mr. Sotirov,

Thank you for your bug report.

However, there is no way in which we could check your findings. Furthermore, Slackware is not one of the supported Linux platforms, so we do not have means to build and test properly our binaries.

On our download pages for the Community Server, you can find Linux versions that we fully support. For other versions, we offer a source code , so that you can build MySQL, but we can not control whether your build is faultless or not. It is also out of control whether all installed libraries are of the exact version and release that we require. We do not keep those data on libraries, we just keep version and releases of the Linux distributions that we support.

Unsupported.

I'm not asking you to reproduce anything. I'm only reporting you a problem encountered when building MySQL from source, which I'm pretty much sure could be reproduced on other Linux distributions as I'm not doing anything specific related to Slackware Linux - I'm only building the MySQL server from source with the configuration I provided. And I'm building the server on Slackware Linux for the past 17 years...

Have you checked #106721 at all? I was the same case - a test failing and it was fixed in next release. I expect the same here or otherwise please, explain why the test fails. Of course let me know if any additional information is necessary so I could provide it.

Hi Mr. Sotirov,

You wrote that you are not asking us to reproduce anything.

But, this is a forum for the  reproducible bugs. If we can not reproduce it, then we can not process it, at all.

Also, no, we have not encountered any of the problems that you are reporting on the Linux distributions that we support.

Unsupported.

Sure, "unsupported", but you have not reproduced the problem I reported in #106721 either, right? And still you were able to identify the problem _internally_ and fix the issue with next release, right? So, what is the difference now?

Again, please, let me know what other information I could provide? How could I find more about the problem with this particular test in release 8.0.31?