Skip to contentSkip to conversation

Yocto Project Status WW09'20

DISCUSSION [email protected] 3 participants 6 messages Mar 3, 2020 – Mar 7, 2020

lore
Share & tools
lore
#0 lore raw

Current Dev Position: YP 3.1 M3 - At Feature Freeze, build pending

Next Deadline: YP 3.1 M3 build date 2/24/2020

Next Team Meetings:

* Bug Triage meeting Thursday Mar. 5th at 7:30am PDT (
<https://zoom.us/j/454367603> https://zoom.us/j/454367603)
* Monthly Project Meeting Tuesday Mar. 3rd at 8am PDT (
<https://zoom.us/j/990892712> https://zoom.us/j/990892712)
* Weekly Engineering Sync Tuesday Mar. 10th at 8am PDT (
<https://zoom.us/j/990892712> https://zoom.us/j/990892712)
* Twitch - See http://www.twitch.tv/letoatreidesthe2nd

Key Status/Updates:

* YP 2.7.3 and 3.0.2 have been released.
* YP 3.1 has been announced as an LTS release:
<https://www.yoctoproject.org/yocto-project-long-term-support-announced/>
https://www.yoctoproject.org/yocto-project-long-term-support-announced/
* We are now past M3 feature freeze. A number of key issues have been
addressed, thanks to everyone who contributed:

* psplash systemd race issues fixed (Thanks Scott)
* make mips issue root caused and glibc fix pulled in (Thanks
Victor/Khem)
* ltp upgrade resolved (Thanks Petr)
* util-linux update resolved (Thanks Pierre-Jean)
* Uninative tumbleweed issue resolved (Thanks Michael)
* libgpg-error upgrade (Thanks Alex)

* There were many other fixes and upgrades, thanks all!
* The remaining issues we need to look at addressing before we build
M3 are:

* bind upgrade (needs libuv maintainer fix and resend from Armin)
* coreutils ptest blocked on libmodule-build-perl reproducibility
issue
* Locked signature selftest failure
* SystemExit() intermittent selftest failure
* pseudo sysroot high priority bug
* gcc buildtools tarball solution for centos7 support

* Recipe upgrades are now much less likely to be accepted unless there
is a pressing need, ideally these should be held until 3.1 is released and development for 3.2 is started. We do not have a tracking branch for 3.2 as we'd like to focus people on 3.1 right now.
* Bug metrics are not good with a rise in WDD when typically it should
be falling for this point in a release cycle.

YP 3.1 Milestone Dates:

* YP 3.1 M3 build date 2/24/2020
* YP 3.1 M3 release date 3/6/2020
* YP 3.1 M4 build date 3/30/2020
* YP 3.1 M4 release date 4/24/2020

Planned upcoming dot releases:

* YP 3.0.3 build date 5/4/2020
* YP 3.0.3 release date 5/15/2020
* YP 2.7.4 build date 5/18/2020
* YP 2.7.4 release date 5/29/2020

Tracking Metrics:

* WDD 2720 (last week 2665) (
<https://wiki.yoctoproject.org/charts/combo.html>
https://wiki.yoctoproject.org/charts/combo.html)
* Poky Patch Metrics

* Total patches found: 1346 (last week 1356)
* Patches in the Pending State: 539 (40%) [last week 543 (40%)]

The Yocto Project's technical governance is through its Technical Steering Committee, more information is available at:

<https://wiki.yoctoproject.org/wiki/TSC>
https://wiki.yoctoproject.org/wiki/TSC

The Status reports are now stored on the wiki at:
<https://wiki.yoctoproject.org/wiki/Weekly_Status>
https://wiki.yoctoproject.org/wiki/Weekly_Status

[If anyone has suggestions for other information you'd like to see on this weekly status update, let us know!]

Thanks,

Stephen K. Jolley

Yocto Project Program Manager

* Cell: (208) 244-4460

* Email: [email protected]
<mailto:[email protected]>

#1 lore raw
Quoted 2 lines · On Tue, 3 Mar 2020 at 16:59, <[email protected]> wrote:

coreutils ptest blocked on libmodule-build-perl reproducibility issue

I sent a patch for this.

Alex

#2 lore raw
Quoted 6 lines · Alexander Kanavin
Quoted 3 lines · On Tue, 3 Mar 2020 at 16:59, <[email protected]> wrote:

coreutils ptest blocked on libmodule-build-perl reproducibility issue

I sent a patch for this.

Thanks!

Just FYI I think there may also be a couple of other packages coreutils pulls in and they may also have reproducibility issues (gettext-ptest, glibc-locale-* and procps*).

Cheers,

Richard

#3 lore raw
Quoted 5 lines · Richard Purdie

Just FYI I think there may also be a couple of other packages coreutils pulls in and they may also have reproducibility issues (gettext-ptest, glibc-locale-* and procps*).

Patch for gettext also sent :) glibc-locale and procps I could not reproduce :-/

Alex

#4 lore raw
Quoted 10 lines · Alexander Kanavin
Quoted 5 lines · Richard Purdie

Just FYI I think there may also be a couple of other packages coreutils
pulls in and they may also have reproducibility issues (gettext- ptest,
glibc-locale-* and procps*).

Patch for gettext also sent :) glibc-locale and procps I could not reproduce :-/

Thanks for that fix, it helps!

FWIW, what I'm seeing with glibc-locale is that these files "disappear" from the packages in one of my builds:

./usr/share/locale/vi/LC_MESSAGES/libc.mo
./usr/share/locale/ko/LC_MESSAGES/libc.mo
./usr/share/locale/en_GB/LC_MESSAGES/libc.mo
./usr/share/locale/ca/LC_MESSAGES/libc.mo
./usr/share/locale/fr/LC_MESSAGES/libc.mo
./usr/share/locale/pl/LC_MESSAGES/libc.mo
./usr/share/locale/id/LC_MESSAGES/libc.mo
./usr/share/locale/hr/LC_MESSAGES/libc.mo
./usr/share/locale/uk/LC_MESSAGES/libc.mo
./usr/share/locale/ja/LC_MESSAGES/libc.mo
./usr/share/locale/gl/LC_MESSAGES/libc.mo
./usr/share/locale/es/LC_MESSAGES/libc.mo
./usr/share/locale/el/LC_MESSAGES/libc.mo
./usr/share/locale/pt_BR/LC_MESSAGES/libc.mo
./usr/share/locale/sl/LC_MESSAGES/libc.mo
./usr/share/locale/lt/LC_MESSAGES/libc.mo
./usr/share/locale/rw/LC_MESSAGES/libc.mo
./usr/share/locale/sv/LC_MESSAGES/libc.mo
./usr/share/locale/cs/LC_MESSAGES/libc.mo
./usr/share/locale/nl/LC_MESSAGES/libc.mo
./usr/share/locale/zh_CN/LC_MESSAGES/libc.mo
./usr/share/locale/ia/LC_MESSAGES/libc.mo
./usr/share/locale/fi/LC_MESSAGES/libc.mo
./usr/share/locale/be/LC_MESSAGES/libc.mo
./usr/share/locale/ru/LC_MESSAGES/libc.mo
./usr/share/locale/nb/LC_MESSAGES/libc.mo
./usr/share/locale/zh_TW/LC_MESSAGES/libc.mo
./usr/share/locale/tr/LC_MESSAGES/libc.mo
./usr/share/locale/de/LC_MESSAGES/libc.mo
./usr/share/locale/da/LC_MESSAGES/libc.mo
./usr/share/locale/pt/LC_MESSAGES/libc.mo
./usr/share/locale/hu/LC_MESSAGES/libc.mo
./usr/share/locale/eo/LC_MESSAGES/libc.mo
./usr/share/locale/it/LC_MESSAGES/libc.mo
./usr/share/locale/bg/LC_MESSAGES/libc.mo
./usr/share/locale/sk/LC_MESSAGES/libc.mo

I can conform they're not in the glibc stashed-locale sstate object so they disappear somewhere before that in the glibc recipe itself.

Quite where/why I don't know as yet.

For procps, the timestamps in the output are different, that is the only difference. That means something is indeterminate with the source epoch for that recipe. Again, why, I don't know. The two stamps are:

2019-12-08·04:06:03
2019-12-08·04:05:49

This latter one could potentially be a bug in hashequiv (cc Joshua). It may need that stamps patch I have fixed up to resolve it. The fix it needs is to make it apply only to certain tasks.

Cheers,

Richard

#5 lore raw
Quoted 55 lines · Richard Purdie
Quoted 10 lines · Alexander Kanavin
Quoted 5 lines · Richard Purdie

Just FYI I think there may also be a couple of other packages coreutils
pulls in and they may also have reproducibility issues (gettext- ptest,
glibc-locale-* and procps*).

Patch for gettext also sent :) glibc-locale and procps I could not reproduce :-/

Thanks for that fix, it helps!

FWIW, what I'm seeing with glibc-locale is that these files "disappear"
from the packages in one of my builds:

./usr/share/locale/vi/LC_MESSAGES/libc.mo
./usr/share/locale/ko/LC_MESSAGES/libc.mo
./usr/share/locale/en_GB/LC_MESSAGES/libc.mo
./usr/share/locale/ca/LC_MESSAGES/libc.mo
./usr/share/locale/fr/LC_MESSAGES/libc.mo
./usr/share/locale/pl/LC_MESSAGES/libc.mo
./usr/share/locale/id/LC_MESSAGES/libc.mo
./usr/share/locale/hr/LC_MESSAGES/libc.mo
./usr/share/locale/uk/LC_MESSAGES/libc.mo
./usr/share/locale/ja/LC_MESSAGES/libc.mo
./usr/share/locale/gl/LC_MESSAGES/libc.mo
./usr/share/locale/es/LC_MESSAGES/libc.mo
./usr/share/locale/el/LC_MESSAGES/libc.mo
./usr/share/locale/pt_BR/LC_MESSAGES/libc.mo
./usr/share/locale/sl/LC_MESSAGES/libc.mo
./usr/share/locale/lt/LC_MESSAGES/libc.mo
./usr/share/locale/rw/LC_MESSAGES/libc.mo
./usr/share/locale/sv/LC_MESSAGES/libc.mo
./usr/share/locale/cs/LC_MESSAGES/libc.mo
./usr/share/locale/nl/LC_MESSAGES/libc.mo
./usr/share/locale/zh_CN/LC_MESSAGES/libc.mo
./usr/share/locale/ia/LC_MESSAGES/libc.mo
./usr/share/locale/fi/LC_MESSAGES/libc.mo
./usr/share/locale/be/LC_MESSAGES/libc.mo
./usr/share/locale/ru/LC_MESSAGES/libc.mo
./usr/share/locale/nb/LC_MESSAGES/libc.mo
./usr/share/locale/zh_TW/LC_MESSAGES/libc.mo
./usr/share/locale/tr/LC_MESSAGES/libc.mo
./usr/share/locale/de/LC_MESSAGES/libc.mo
./usr/share/locale/da/LC_MESSAGES/libc.mo
./usr/share/locale/pt/LC_MESSAGES/libc.mo
./usr/share/locale/hu/LC_MESSAGES/libc.mo
./usr/share/locale/eo/LC_MESSAGES/libc.mo
./usr/share/locale/it/LC_MESSAGES/libc.mo
./usr/share/locale/bg/LC_MESSAGES/libc.mo
./usr/share/locale/sk/LC_MESSAGES/libc.mo

Quoted 5 lines

I can conform they're not in the glibc stashed-locale sstate object so
they disappear somewhere before that in the glibc recipe itself.

Quite where/why I don't know as yet.

Just to save anyone digging, the difference is in configure, one has msgfmt in the sysroot, one does not. It is there later in the build so its probably task dependencies not being right.

Cheers,

Richard