gitsm and own-mirror lead to unhappiness

Patrick Venture venture at google.com
Wed Feb 13 03:37:05 AEDT 2019


On Tue, Feb 12, 2019 at 7:53 AM Andrew Geissler <geissonator at gmail.com> wrote:
>
> On Tue, Feb 12, 2019 at 9:21 AM Patrick Venture <venture at google.com> wrote:
> >
> > Has anyone else run into this issue?
> >
> > openbmcbmc works fine when building zaius as phosphor:
> >
> > PREMIRRORS="bzr://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n cvs://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n git://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n gitsm://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n hg://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n osc://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n p4://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n svn://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n"
> >
> > when using a mirror via the own-mirror class it has (and fails):
> >
> > PREMIRRORS=" cvs://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n svn://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n git://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n gitsm://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n hg://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n bzr://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n p4://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n osc://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n https?\$://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n ftp://.*/.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n npm://.*/?.*
> > file:///google/data/ro/teams/gbmc/mirror/sources/ \n bzr://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n cvs://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n git://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n gitsm://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n hg://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n osc://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n p4://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n svn://.*/.*
> > http://downloads.yoctoproject.org/mirror/sources/ \n"
> >
> > Here's the error:
> >
> > ERROR: dbus-broker-16+gitAUTOINC+fc874afa09-r0 do_unpack: Fetcher
> > failure: Fetch command export PSEUDO_DISABLED=1; export
> > P4CONFIG=".p4config"; export
> > DBUS_SESSION_BUS_ADDRESS="unix:abstract=/tmp/dbus-ikQVQidiXM,guid=017a369eba1bf971869460b55c475ed9";
> > export SSH_AGENT_PID="64102"; export
> > SSH_AUTH_SOCK="/tmp/ssh-lw8WN8I9mmbS/agent.63979"; export
> > PATH="/usr/local/google/git/workspaces/building_rebase_failure/build/tmp/work/armv5e-openbmc-linux-gnueabi/dbus-broker/16+gitAUTOINC+fc874afa09-r0/recipe-sysroot-native/usr/bin/python3-native:/usr/local/google/git/workspaces/building_rebase_failure/scripts:/usr/local/google/git/workspaces/building_rebase_failure/build/tmp/work/armv5e-openbmc-linux-gnueabi/dbus-broker/16+gitAUTOINC+fc874afa09-r0/recipe-sysroot-native/usr/bin/arm-openbmc-linux-gnueabi:/usr/local/google/git/workspaces/building_rebase_failure/build/tmp/work/armv5e-openbmc-linux-gnueabi/dbus-broker/16+gitAUTOINC+fc874afa09-r0/recipe-sysroot/usr/bin/crossscripts:/usr/local/google/git/workspaces/building_rebase_failure/build/tmp/work/armv5e-openbmc-linux-gnueabi/dbus-broker/16+gitAUTOINC+fc874afa09-r0/recipe-sysroot-native/usr/sbin:/usr/local/google/git/workspaces/building_rebase_failure/build/tmp/work/armv5e-openbmc-linux-gnueabi/dbus-broker/16+gitAUTOINC+fc874afa09-r0/recipe-sysroot-native/usr/bin:/usr/local/google/git/workspaces/building_rebase_failure/build/tmp/work/armv5e-openbmc-linux-gnueabi/dbus-broker/16+gitAUTOINC+fc874afa09-r0/recipe-sysroot-native/sbin:/usr/local/google/git/workspaces/building_rebase_failure/build/tmp/work/armv5e-openbmc-linux-gnueabi/dbus-broker/16+gitAUTOINC+fc874afa09-r0/recipe-sysroot-native/bin:/usr/local/google/git/workspaces/building_rebase_failure/poky/bitbake/bin:/usr/local/google/git/workspaces/building_rebase_failure/build/tmp/hosttools";
> > export HOME="/usr/local/google/home/venture"; cp -fpLR
> > /usr/local/google/git/workspaces/building_rebase_failure/build/downloads/git2/github.com.bus1.dbus-broker/modules/subprojects/c-dvar
> > /usr/local/google/git/workspaces/building_rebase_failure/build/tmp/work/armv5e-openbmc-linux-gnueabi/dbus-broker/16+gitAUTOINC+fc874afa09-r0/git/.git/modules/subprojects/c-dvar
> > failed with exit code 1, output:
> > cp: cannot stat
> > '/usr/local/google/git/workspaces/building_rebase_failure/build/downloads/git2/github.com.bus1.dbus-broker/modules/subprojects/c-dvar/modules/subprojects/c-utf8':
> > No such file or directory
> >
> > ERROR: dbus-broker-16+gitAUTOINC+fc874afa09-r0 do_unpack: Function
> > failed: base_do_unpack
> > ERROR: Logfile of failure stored in:
> > /usr/local/google/git/workspaces/building_rebase_failure/build/tmp/work/armv5e-openbmc-linux-gnueabi/dbus-broker/16+gitAUTOINC+fc874afa09-r0/temp/log.do_unpack.86989
> > ERROR: Task (/usr/local/google/git/workspaces/building_rebase_failure/meta-openembedded/meta-oe/recipes-core/dbus/dbus-broker_git.bb:do_unpack)
> > failed with exit code '1'
> > NOTE: Tasks Summary: Attempted 1317 tasks of which 0 didn't need to be
> > rerun and 1 failed.
>
> Yeah, we've seen this on a few of our builder machines and some
> internal machines - https://github.com/openbmc/openbmc/issues/3480
>
> Sometimes rebuilding a few times fixes it, but sometimes it's a hard
> failure (builder4 is offline due to this).

Interesting. So, our failures are not identical but symptomatically
similar -- unable to get submodules.  It's frustrating to say the
least.  I'll try doing the rebuilds you suggested.  If it builds once,
perhaps and i have the tarball built, it'll work for the future --
since it's a host-specific issue for you, it makes me think perhaps
it's a git version issue or related to some package on which it
depends varying between systems.  However, I imagine the build systems
are identical except for architecture.  We're seeing the failure
building on x86-64.

>
> Andrew
>
> >
> >
> >
> > Patrick


More information about the openbmc mailing list