[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#1033811: marked as done (unblock: mariadb/1:10.11.2-2)



Your message dated Thu, 20 Apr 2023 12:02:01 +0100
with message-id <CAM8zJQs4G_g7b7q_uVDt15R4mBydLoWThUjeUnDn1UXSbmhhLw@mail.gmail.com>
and subject line Re: Bug#1033811: unblock: mariadb/1:10.11.2-2
has caused the Debian Bug report #1033811,
regarding unblock: mariadb/1:10.11.2-2
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 owner@bugs.debian.org
immediately.)


-- 
1033811: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1033811
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems
--- Begin Message ---
Package: release.debian.org
Severity: normal
User: release.debian.org@packages.debian.org
Usertags: unblock
Control: affects -1 + src:mariadb

Please unblock package 'mariadb' so that the bug fixes in the Debian
revision 2 can be included in the release of Debian 12 "Bookworm".

The package is currently 7/20 days old, but even after 20 days it will
not automatically migrate because it is listed as a key package.

Also, it has autopkgtest failure for armel/armhf. The failure is safe
to ignore because it is from a new test that had a typo[1] on these
two platforms. Autopkgtests fully pass on all other platforms. I
decided to not do a re-upload just for this typo for now, it would not
be frugal use of builders and Debian infra.


This Debian revision has been carefully crafted to only include bug
fixes, test improvements and translations (following the release
policy[2]). There are no risky functional changes, and the package has
not had any new bugs reported for several weeks and confidence that
this is the best version for Bookworm is high.

The changes can be reviewed by reading individual commits (each change
is well documented)[3]. Below summary from changelog[4]:

mariadb (1:10.11.2-2) unstable; urgency=medium
 .
   [ Otto Kekäläinen ]
   * SUMMARY: This version has a lot of bug fixes, quality fixes, documentation
     and translation updates and it is tailored for the Debian 12 "Bookworm"
     release and all potentially functional changes have been left out and
     are pending review and merge post-Bookworm release at
     https://salsa.debian.org/mariadb-team/mariadb-server/-/merge_requests,
     where contributions both as reviews and new submissions are welcome!
   * Update NEWS to summarize what is new in MariaDB 10.11 (compared to 10.6)
   * Update and activate configuration tracing in autopkgtests to ensure that
     for the Debian 12 "Bookworm" release cycle no upstream server config change
     would slip in unnoticed
   * Enable mariadb-plugin-rocksdb for riscv64 (fixes autopkgtests for riscv64)
   * Sync downstream (where applicable) with upstream 10.11 debian/* contents
     so that using diff/meld to compare changes are easier
   * Add important upstream 10.11.2+ fixes as packes:
     * Stack overflow in pinbox allocator (PR#2541)
     * Upgrades from MySQL 5.7 to MariaDB 10.11 (MDEV-30483) (Closes: #866751)
     * Misc compiler warnings in upstream build
     * Incomplete stack traces if MariaDB crashes
     * Make mariadbd emit warnings if mariadb-upgrade was not run
     * Binlog failures due to 'character_set_client' (MDEV-30824)
   * Prevent mariadb-test-run from using native I/O on ppc64el and s390x due to
     Linux kernel bug (Related: #1031656)
   * Add patch to better diagnose potential failures in main.order_by_innodb
   * Add patch to fix cross-compilation failure on uca-dump (Closes: #1029165)
   * Update mariadb-test-run skip test lists to not run tests that are known
     to be broken or unstable, and that have been reported upstream
   * Limit check of running mysqld/mariadbd to system users (Closes: #1032047)
   * Make error more helpful in case server restart fails (Related: #1033234)
   * Update Lintian overrides after rigorous review of all Lintian issues
     * Remove incorrect Multi-Arch definitions
     * Fix man pages syntax issues (Closes: #1032861)
     * Fix spelling in MariaDB and components (Closes: #1032860)
     * Refresh patches metadata
     * Update upstream signing key
     * Fix dependency of obsolete libncurses5-dev
 .
   [ Ekaterine Papava ]
   * Add Georgian translation (error messages)
 .
   [ Tuukka Pasanen ]
   * Update README files to correct versions



[1] https://salsa.debian.org/mariadb-team/mariadb-server/-/merge_requests/42
[2] https://release.debian.org/testing/freeze_policy.html
[3] https://salsa.debian.org/mariadb-team/mariadb-server/-/commits/debian/1%2510.11.2-2
[4] https://tracker.debian.org/news/1427834/accepted-mariadb-110112-2-source-into-unstable/

--- End Message ---
--- Begin Message ---
Hi Otto

On Sun, 2 Apr 2023 at 04:54, Otto Kekäläinen <otto@debian.org> wrote:
> This Debian revision has been carefully crafted to only include bug
> fixes, test improvements and translations (following the release
> policy[2]).

I see many changes in your 1:10.11.2-2 upload that do not follow
policy, see Appropriate changes during Hard and Full Freeze [1].
Policy refers to targeted fixes for release critical bugs and fixes
for severity: important bugs, while your upload contains many
unrelated packaging changes.

> Also, it has autopkgtest failure for armel/armhf. The failure is safe
> to ignore because it is from a new test that had a typo[1] on these
> two platforms. Autopkgtests fully pass on all other platforms. I
> decided to not do a re-upload just for this typo for now, it would not
> be frugal use of builders and Debian infra.

Allowing the autopkgtest to regress in testing would prevent us from
detecting real regressions triggered by other packages.  You don't
need to spare the buildds and autopkgtest infrastructure, they exist
to be used.

I appreciate the hard work you've put into this upload, which would
have been welcome in testing before the hard freeze, but sorry, now is
just too late.  If you'd like to get some of these fixes into
bookworm, I suggest preparing a new 1:10.11.2-3 upload containing only
targeted fixes for release critical and important bugs, and fixing the
armel/armhf autopkgtests.

Regards
Graham


[1] https://release.debian.org/testing/freeze_policy.html#appropriate

--- End Message ---

Reply to: