Re: Crash report for some ICU-52 (debian8) COLLATE and work_mem values

From: "Daniel Verite" <daniel(at)manitou-mail(dot)org>
To: "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: "Peter Geoghegan" <pg(at)bowt(dot)ie>,"PostgreSQL mailing lists" <pgsql-bugs(at)postgresql(dot)org>,"Peter Eisentraut" <peter(dot)eisentraut(at)2ndquadrant(dot)com>
Subject: Re: Crash report for some ICU-52 (debian8) COLLATE and work_mem values
Date: 2017-08-07 13:52:33
Message-ID: b9379009-61d6-4fec-b7e9-786f3da7b2d3@manitou-mail.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs pgsql-hackers

Tom Lane wrote:

> I went to http://www.icu-project.org/ and downloaded icu4c-52_1-src.tgz.
> All the file dates therein seem to be 2013-10-04.
>
> Debian, for one, is evidently not trying very hard in that direction,
> since not only are the bugs still there but the line numbers I saw in
> my backtraces agreed with Daniel's, indicating they've not changed
> much of anything at all in ucol.cpp.

They have 2 small patches in ucol.cpp (diff attached),
but the last backtraces I've sent were against upstream, not
Debian, got from the same source as you, so they wouldn't differ
in the line numbers.
Anyway the behavior with segfaulting was identical to Debian's.

Speaking of upstream vs Debian, for the library as a whole there are
quite a few security patches that are not in upstream:

$ apt-get source libicu-dev
[...]
dpkg-source: info: extracting icu in icu-52.1
dpkg-source: info: unpacking icu_52.1.orig.tar.gz
dpkg-source: info: unpacking icu_52.1-8+deb8u5.debian.tar.xz
dpkg-source: info: applying icudata-stdlibs.patch
dpkg-source: info: applying gennorm2-man.patch
dpkg-source: info: applying icuinfo-man.patch
dpkg-source: info: applying malayalam-rendering.patch
dpkg-source: info: applying indic-ccmp.patch
dpkg-source: info: applying mlym-crash.patch
dpkg-source: info: applying two-digit-year-test.patch
dpkg-source: info: applying icu-config.patch
dpkg-source: info: applying CVE-2014-6585.patch
dpkg-source: info: applying CVE-2014-6591.patch
dpkg-source: info: applying CVE-2014-7923+7926.patch
dpkg-source: info: applying CVE-2014-7940.patch
dpkg-source: info: applying CVE-2014-9654.patch
dpkg-source: info: applying CVE-2014-8146.patch
dpkg-source: info: applying CVE-2014-8147.patch
dpkg-source: info: applying CVE-2015-4760.patch
dpkg-source: info: applying CVE-2014-6585+.patch
dpkg-source: info: applying CVE-2015-1270.patch
dpkg-source: info: applying CVE-2014-9911.patch
dpkg-source: info: applying CVE-2015-2632.patch
dpkg-source: info: applying CVE-2015-4844.patch
dpkg-source: info: applying CVE-2016-0494.patch
dpkg-source: info: applying CVE-2016-6293.patch
dpkg-source: info: applying CVE-2016-7415.patch
dpkg-source: info: applying CVE-2017-7867_CVE-2017-7868.patch

Independantly of the bug discussed in this thread, what is puzzling
to me is why upstream does not integrate any of these fixes.
Here's their policy about maintenance releases:

http://site.icu-project.org/processes/maintenance-releases

"When a critical problem is found in ICU libraries, we try to fix the
problem in the latest development stream first. If there is a demand
for the fix in a past release, an ICU project developer may escalate
the fix to be integrated in the release to the ICU project management
committee. Once the committee approved to merge the fix into back
level stream, the developer can merge the bug fix back to the past
release suggested by the committee. This merge activity must be
tracked by maintenance release place holder tickets and the developer
should provide original ticket number and description as the response
in each maintenance ticket. These fixes are automatically included in
a future ICU maintenance release."

Best regards,
--
Daniel Vérité
PostgreSQL-powered mailer: http://www.manitou-mail.org
Twitter: @DanielVerite

Attachment Content-Type Size
diff-debian-upstream-icu52.1.diff.gz application/octet-stream 23.2 KB

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message abo 2017-08-07 14:18:04 BUG #14772: psql autocommit does not work
Previous Message Tom Lane 2017-08-07 13:19:26 Re: Error of insert Foreign table with postgres_fdw

Browse pgsql-hackers by date

  From Date Subject
Next Message Fabrízio de Royes Mello 2017-08-07 15:14:52 Re: Patch: Add --no-comments to skip COMMENTs with pg_dump
Previous Message Amit Kapila 2017-08-07 13:49:34 Re: free space % calculation in pgstathashindex