[Bug 1468733] New: Switch to erlang-lager/lager instead of basho'
s lager and update to >= 3.4.2
by bugzilla＠redhat.com
https://bugzilla.redhat.com/show_bug.cgi?id=1468733
Bug ID: 1468733
Summary: Switch to erlang-lager/lager instead of basho's lager
and update to >= 3.4.2
Product: Fedora
Version: rawhide
Component: erlang-lager
Severity: medium
Assignee: lemenkov(a)gmail.com
Reporter: randy(a)electronsweatshop.com
QA Contact: extras-qa(a)fedoraproject.org
CC: erlang(a)lists.fedoraproject.org, lemenkov(a)gmail.com,
randy(a)electronsweatshop.com
Description of problem:
The latest ejabberd requires lager >= 3.4.2 but Fedora has 3.2.4. I was curious
why the-new-hotness hadn't filed a bug against erlang-lager and I found that
erlang-lager is using basho's lager[0] instead of erlang-lager[1]. The
upstream's latest release is 3.5.1. The basho downstream hasn't made a release
since October.
I propose that we switch to the upstream lager for our source and update to the
latest release.
Additional info:
[0] https://github.com/basho/lager
[1] https://github.com/erlang-lager/lager/
--
You are receiving this mail because:
You are on the CC list for the bug.
6 years, 8 months
[Bug 1470583] New: Elixir built with wrong endianness
by bugzilla＠redhat.com
https://bugzilla.redhat.com/show_bug.cgi?id=1470583
Bug ID: 1470583
Summary: Elixir built with wrong endianness
Product: Fedora
Version: 26
Component: elixir
Severity: urgent
Assignee: relrod(a)redhat.com
Reporter: dagolav(a)prestegarden.com
QA Contact: extras-qa(a)fedoraproject.org
CC: erlang(a)lists.fedoraproject.org,
jochen(a)herr-schmitt.de, lemenkov(a)gmail.com,
martin(a)laptop.org, puiterwijk(a)redhat.com,
relrod(a)redhat.com
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML,
like Gecko) Chrome/61.0.3153.4 Safari/537.36
Build Identifier:
Upon starting iex, elixirc, mix or other Elixir software the software crashes
with the following message:
warning: Elixir is running in a system with a different endianness than the one
its source code was compiled in. Please make sure Elixir and all source files
were compiled in a machine with the same endianness as the current one: little
Reproducible: Always
Steps to Reproduce:
1. Install elixir on fedora 26 x86_64 from official repos
2. Run elixirc --version
Actual Results:
Software crashes with warning messange and crash dump
Expected Results:
Software running, in this case returning version information.
Package info:
Name : elixir
Version : 1.4.2
Release : 1.fc26
Running Linux 4.11.9-300.fc26.x86_64 #1 SMP Wed Jul 5 16:21:56 UTC 2017 x86_64
x86_64 x86_64 GNU/Linux
Worth noting: I observe that the fc26 package is built using a PPC build host
buildvm-ppc64-04.ppc.fedoraproject.org while the fc25 package is built using an
x86_64 host buildvm-27.phx2.fedoraproject.org
--
You are receiving this mail because:
You are on the CC list for the bug.
6 years, 8 months
Erlang noarch libs in /usr/share
by Randy Barlow
Hello!
I finally have a patch for the erlang spec file that lets the VM look
for dependencies in /usr/share/erlang/lib[0]. I tested it by moving one
of ejabberd's dependencies there, and the server does start up.
I've also written a patch for the macros to put libraries in that
location[1].
There is one problem I've not solved yet - rebar does not seem to find
dependencies when running eunit, as I reported in a comment in [0]. I'm
not sure why but it seems that Rebar does not use the Erlang VM's
search path? Does anybody know how we can convince Rebar to find
dependencies in /usr/share?
[0] https://bugzilla.redhat.com/show_bug.cgi?id=1476614
[1] https://github.com/lemenkov/erlang-rpm-macros/pull/4/
6 years, 9 months
[Bug 1470657] New: Couchdb installation broken.
by bugzilla＠redhat.com
https://bugzilla.redhat.com/show_bug.cgi?id=1470657
Bug ID: 1470657
Summary: Couchdb installation broken.
Product: Fedora
Version: 26
Component: couchdb
Severity: high
Assignee: lemenkov(a)gmail.com
Reporter: slavikvin(a)quintagroup.com
QA Contact: extras-qa(a)fedoraproject.org
CC: erlang(a)lists.fedoraproject.org, lemenkov(a)gmail.com,
wtogami(a)gmail.com
Description of problem:
Couchdb installation on fresh fedora 26 installation is broken due to
js-1.8.5-27.fc26.x86_64.rpm(dependency) was patched for 48-bit VA
(https://fedoraproject.org//wiki/Changes/aarch64-48bitVA).
Version-Release number of selected component (if applicable):
couchdb-1.6.1-18.fc26.x86_64.rpm
js-1.8.5-27.fc26.x86_64.rpm
How reproducible:
Steps to Reproduce:
1. Install Fedora 26;
2. dnf install couchdb;
3. sudo systemctl start couchdb;
4. run verification http://localhost:5984/_utils/verify_install.html
Actual results:
It says "Running..." freezes and unable to finish verification.
Expected results:
Must complete test, and say "Your installation looks fine. Time to Relax."
Additional info:
In log i say many repeatable messages: " [error] [<0.221.0>] OS Process Error
<0.2722.0> :: {os_process_error, {exit_status,1}}"
--
You are receiving this mail because:
You are on the CC list for the bug.
6 years, 9 months
[Bug 1433985] New: CVE-2016-10253 erlang:
Heap-buffer overflow via regular expressions
by bugzilla＠redhat.com
https://bugzilla.redhat.com/show_bug.cgi?id=1433985
Bug ID: 1433985
Summary: CVE-2016-10253 erlang: Heap-buffer overflow via
regular expressions
Product: Security Response
Component: vulnerability
Keywords: Security
Severity: medium
Priority: medium
Assignee: security-response-team(a)redhat.com
Reporter: anemec(a)redhat.com
CC: aortega(a)redhat.com, apevec(a)redhat.com,
ayoung(a)redhat.com, chrisw(a)redhat.com,
cvsbot-xmlrpc(a)redhat.com,
erlang(a)lists.fedoraproject.org, jeckersb(a)redhat.com,
jjoyce(a)redhat.com, jschluet(a)redhat.com,
kbasil(a)redhat.com, lemenkov(a)gmail.com, lhh(a)redhat.com,
lpeer(a)redhat.com, markmc(a)redhat.com,
plemenko(a)redhat.com, rbryant(a)redhat.com,
rhbugs(a)n-dimensional.de, rjones(a)redhat.com,
sclewis(a)redhat.com, s(a)shk.io, tdecacqu(a)redhat.com
An issue was discovered in Erlang/OTP Erlang's generation of compiled regular
expressions is vulnerable to a heap overflow. Regular expressions using a
malformed extpattern can indirectly specify an offset that is used as an array
index. This ordinal permits arbitrary regions within the erts_alloc arena to be
both read and written to.
References:
https://github.com/erlang/otp/pull/1108
--
You are receiving this mail because:
You are on the CC list for the bug.
6 years, 9 months
