vpa1977 | Hi, if someone has a bit of time would it be possible to take a look at jsurf-alggeo patch here https://code.launchpad.net/~vpa1977/ubuntu/+source/jsurf-alggeo/+git/jsurf-alggeo/+merge/439875 This package autopkgtest blocks openjdk from time to time and the patch solves race condition in the failing test. Thank you!!!!!! | 02:37 |
---|---|---|
=== sem2peie- is now known as sem2peie | ||
=== SuperL4g is now known as SuperLag | ||
ahasenack | sergiodj: hi, you did openldap uploads in the past | 14:51 |
ahasenack | I'm struggling with this exim4 tls error in dep8 | 14:51 |
ahasenack | in this case, for kinetic | 14:51 |
ahasenack | but the history of that test is bad | 14:51 |
ahasenack | https://autopkgtest.ubuntu.com/packages/e/exim4/kinetic/ppc64el | 14:51 |
ahasenack | lots of retries, most fail, but a migration-reference/0 passed in the first attempt | 14:51 |
ahasenack | I have a ppc64el vm, and that test runs just fine there | 14:51 |
ahasenack | have you investigated this further at some point in the past, for your openldap srus, or just hammered the retry button? | 14:52 |
sergiodj | ahasenack: hey. I haven't investigated this problem yet; I was just hammering the retry button, as you say. this seems to be related to openssl3 and not openldap, though. | 14:55 |
ahasenack | oh, I take it back, it fails with the updated libldap | 14:56 |
ahasenack | fascinating | 14:56 |
ahasenack | just on ppc64el | 14:56 |
sergiodj | hm, really? | 14:58 |
sergiodj | so it may be something related to openldap after all | 14:59 |
ahasenack | yeah, but this is basically an openldap rebuild as far as exim4 is concerned | 14:59 |
ahasenack | it's the gcc flag change for the sha2 module, which is not even loaded | 14:59 |
ahasenack | I mean, slapd is not even installed (which is what has that module) | 15:00 |
ahasenack | so it's something else | 15:00 |
sergiodj | this is strange. this openssl error reminds me of those EOF problems we sometimes encounter | 15:02 |
sergiodj | ahasenack: can you give me access to the ppc64el machine? | 15:33 |
sergiodj | I'd like to take a look at the problem | 15:33 |
sergiodj | ahasenack: I downgraded libldap to 2.5.14+dfsg-0ubuntu0.22.10.1 and could still reproduce the problem | 15:51 |
sergiodj | ahasenack: I found some interesting stuff. I'll ping you when I get back from lunch | 16:46 |
ahasenack | sergiodj: did you restart exim4 after that downgrade? | 16:49 |
ahasenack | gnutls assert | 17:45 |
ahasenack | unknown error code | 17:46 |
ahasenack | helpful as ever | 17:46 |
ahasenack | oh my, what aberration is this: libgnutls-openssl27 :) | 17:48 |
ahasenack | "This package contains the runtime library of the GnuTLS OpenSSL wrapper" | 17:48 |
tobhe | if you like the worst of both worlds | 17:49 |
ahasenack | so with the previous gnutls, the test passes | 18:04 |
ahasenack | gnutls had a security update | 18:04 |
ahasenack | I wonder if the exim4 tests passed when that security update came out | 18:04 |
ahasenack | ...and now the test is passing | 18:17 |
* ahasenack scratches head | 18:17 | |
vorlon | ahasenack: libgnutls-openssl27> omg is something actually using that? in the archive? | 18:40 |
vorlon | we should really kill it, it's incomplete and useless | 18:40 |
vorlon | also it's the only part of libgnutls that was licensed GPL instead of LGPL, out of spite | 18:40 |
vorlon | (which is why nobody cared about making it good) | 18:40 |
ahasenack | vorlon: yeah, in kinetic at least, where I'm troubleshooting something | 18:43 |
ahasenack | $ dpkg -s libgnutls28-dev | grep Depends | 18:43 |
ahasenack | Depends: libc6-dev | libc-dev, libgnutls-dane0 (= 3.7.7-2ubuntu2.1), libgnutls-openssl27 (= 3.7.7-2ubuntu2.1), libgnutls30 (= 3.7.7-2ubuntu2.1), libgnutlsxx30 (= 3.7.7-2ubuntu2.1), libidn2-dev, libp11-kit-dev (>= 0.23.10), libtasn1-6-dev, nettle-dev (>= 3.6) | 18:43 |
tobhe | just wait until openssl sunsets 1.1 in september and everyone is forced to switch to gnutls-openssl ;) | 18:45 |
mdeslaur | *lolsnort* | 18:46 |
liushuyu | Hi there, I need some help with retrying the autopkgtests, the links follow: | 20:07 |
liushuyu | https://autopkgtest.ubuntu.com/request.cgi?release=kinetic&arch=amd64&package=ycmd&trigger=llvm-toolchain-15/1%3A15.0.7-0ubuntu0.22.10.1 | 20:07 |
liushuyu | https://autopkgtest.ubuntu.com/request.cgi?release=kinetic&arch=i386&package=dkms&trigger=llvm-toolchain-15/1%3A15.0.7-0ubuntu0.22.10.1 | 20:08 |
liushuyu | -- end of links. Thank you! | 20:08 |
rbasak | How is podiff not packaged? https://man.gnu.org.ua/manpage/?1+podiff | 20:22 |
rbasak | https://github.com/amandasaurus/podiff might be easier | 20:25 |
enr0n | bdmurray: do these denylisted tests actually block migration? Or is this something the update-excuses page just doesn't know about? https://ubuntu-archive-team.ubuntu.com/proposed-migration/focal/update_excuses.html#systemd | 21:06 |
vorlon | enr0n: they do block migration barring an SRU team override | 21:13 |
vorlon | (this is a think I don't like about autopkgtest blacklist entries, they fall out of sync with the hints files) | 21:13 |
vorlon | enr0n: I'd be more worried about the non-blacklisted kernel autopkgtest regressions listed, tho | 21:14 |
enr0n | vorlon: thanks for the info. And yeah I'm trying to sift through the noise | 21:16 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!