Re: pg11.5: ExecHashJoinNewBatch: glibc detected...double free or corruption (!prev) - Mailing list pgsql-hackers

From Tom Lane
Subject Re: pg11.5: ExecHashJoinNewBatch: glibc detected...double free or corruption (!prev)
Date
Msg-id 2475.1566837924@sss.pgh.pa.us
Whole thread Raw
In response to Re: pg11.5: ExecHashJoinNewBatch: glibc detected...double free orcorruption (!prev)  (Thomas Munro <thomas.munro@gmail.com>)
Responses Re: pg11.5: ExecHashJoinNewBatch: glibc detected...double free orcorruption (!prev)
List pgsql-hackers
Thomas Munro <thomas.munro@gmail.com> writes:
> Of course it's entirely possible that we have a bug here and I'm very
> keen to find it, but I can't help noticing the common factor here is
> that they're all running ancient RHEL 6.x releases, except Merlin who
> didn't say.  Merlin?

Hm, Justin said he had
glibc-2.12-1.192.el6.x86_64
which is a little bit behind the times but not *that* ancient.

For reference, attached is the rpm --changelog output on an up-to-date
RHEL6 box for glibc running back to 2.12-1.192.  Unfortunately a lot
of the referenced bugzilla entries aren't public, but the one-liner
descriptions don't seem to suggest that they found any heap overruns
that would be interesting for us.

However ... there is some pretty interesting info at
https://bugzilla.redhat.com/show_bug.cgi?id=1338673
suggesting that compiling with a late-model gcc against older RHEL6
headers could result in bad code.  I wonder whether the reporters'
servers were built using such a configuration.  (Although the linkage,
if any, to this report still wouldn't be very clear.)

            regards, tom lane


* Mon Apr 01 2019 Florian Weimer <fweimer@redhat.com> - 2.12-1.212.3
- Remove abort() warning in manual (#1577437)

* Mon Apr 01 2019 Florian Weimer <fweimer@redhat.com> - 2.12-1.212.2
- ja_JP: Add new Japanese Era name (#1555930)

* Mon Apr 01 2019 Florian Weimer <fweimer@redhat.com> - 2.12-1.212.1
- regex: Fix false match in trailing characters (#1668169)

* Fri Nov 17 2017 Patsy Franklin <pfrankli@redhat.com> - 2.12-1.212
- CVE-2017-15670: glob: Fix one-byte overflow with GLOB_TILDE (#1504810)
- CVE-2017-15804: glob: Fix buffer overflow in GLOB_TILDE unescaping (#1504810)

* Mon Jun 19 2017 Florian Weimer <fweimer@redhat.com> - 2.12-1.211
- Avoid large allocas in the dynamic linker (#1452717)

* Wed Mar 29 2017 Carlos O'Donell <carlos@redhat.com> - 2.12-1.210
- Fix thread cancellation issues for setmntent() and others (#1437147).

* Wed Jan 25 2017 Florian Weimer <fweimer@redhat.com> - 2.12-1.209
- Fix AF_INET6 getaddrinfo with nscd (#1416496)

* Tue Oct 18 2016 Carlos O'Donell <carlos@redhat.com> - 2.12-1.208
- Update tests for struct sockaddr_storage changes (#1338673)

* Mon Oct 17 2016 Martin Sebor <msebor@redhat.com> - 2.12-1.207
- Use FL_CLOEXEC in internal calls to fopen (#1012343).

* Mon Oct 17 2016 Carlos O'Donell <carlos@redhat.com> - 2.12-1.206
- Fix CVE-2015-8779 glibc: Unbounded stack allocation in catopen function
  (#1358015).

* Mon Oct 17 2016 DJ Delorie <dj@redhat.com> - 2.12-1.205
- Make padding in struct sockaddr_storage explicit (#1338673)

* Thu Oct 13 2016 Carlos O'Donell <carlos@redhat.com> - 2.12-1.204
- Fix detection of Intel FMA hardware (#1384281).

* Tue Oct 11 2016 Carlos O'Donell <carlos@redhat.com> - 2.12-1.203
- Add support for el_GR@euro, ur_IN, and wal_ET locales (#1101858).

* Tue Oct 11 2016 Patsy Franklin <pfrankli@redhat.com> - 2.12-1.202
- Change malloc/tst-malloc-thread-exit.c to use fewer threads and
  avoid timeout (#1318380).

* Tue Oct 11 2016 Patsy Franklin <pfrankli@redhat.com> - 2.12-1.201
- df can fail on some systems (#1307029).

* Wed Sep 21 2016 DJ Delorie <dj@redhat.com> - 2.12-1.200
- Log uname, cpuinfo, meminfo during build (#1307029).

* Mon Sep 12 2016 DJ Delorie <dj@redhat.com> - 2.12-1.199
- Draw graphs for heap and stack only if MAXSIZE_HEAP and MAXSIZE_STACK
  are non-zero (#1331304).

* Mon Sep 12 2016 DJ Delorie <dj@redhat.com> - 2.12-1.198
- Avoid unneeded calls to __check_pf in getadddrinfo (#1270950)

* Mon Sep 12 2016 Martin Sebor <msebor@redhat.com> - 2.12-1.197
- Fix CVE-2015-8778 glibc: Integer overflow in hcreate and hcreate_r
  (#1358013).

* Mon Sep 12 2016 Martin Sebor <msebor@redhat.com> - 2.12-1.196
- Fix CVE-2015-8776 glibc: Segmentation fault caused by passing
  out-of-range data to strftime() (#1358011).

* Mon Sep 12 2016 Florian Weimer <fweimer@redhat.com> - 2.12-1.195
- tzdata-update: Ignore umask setting (#1373646)

* Thu Sep 08 2016 Florian Weimer <fweimer@redhat.com> - 2.12-1.194
- CVE-2014-9761: Fix unbounded stack allocation in nan* (#1358014)

* Thu Feb 04 2016 Florian Weimer <fweimer@redhat.com> - 2.12-1.193
- Avoid using uninitialized data in getaddrinfo (#1223095)

* Thu Jan 28 2016 Carlos O'Donell <carlos@redhat.com> - 2.12-1.192
- Update fix for CVE-2015-7547 (#1296029).



pgsql-hackers by date:

Previous
From: Anastasia Lubennikova
Date:
Subject: Re: Optimize single tuple fetch from nbtree index
Next
From: Justin Pryzby
Date:
Subject: Re: pg11.5: ExecHashJoinNewBatch: glibc detected...double free orcorruption (!prev)