help with pungi error: no attribute 'F8_Bootloader'
by Doug Chapman
Anybody know what this means? This is from pungi on my ia64 rawhide
build. I didn't see anything too odd in the log file before it hit this
error. Since this is all quite experimental I am not sure if this is
something I have set up wrong or possibly I have tripped over an anacona
regression?
Pungi.Pungi.INFO:
Running /usr/lib/anaconda-runtime/pkgorder /tmp/test/8.1/ia64/os ia64
Packages
Pungi.Pungi.ERROR: Got an error from /usr/lib/anaconda-runtime/pkgorder
Pungi.Pungi.ERROR: Traceback (most recent call last):
File "/usr/lib/anaconda-runtime/pkgorder", line 45, in <module>
from yuminstall import YumSorter
File "/usr/lib/anaconda/yuminstall.py", line 42, in <module>
from backend import AnacondaBackend
File "/usr/lib/anaconda/backend.py", line 31, in <module>
import kickstart
File "/usr/lib/anaconda/kickstart.py", line 149, in <module>
class Bootloader(commands.bootloader.F8_Bootloader):
AttributeError: 'module' object has no attribute 'F8_Bootloader'
Here are the key rpms that I have installed:
pungi-1.2.6-1.fc9
anaconda-11.4.0.21-1
anaconda-runtime-11.4.0.21-1
yum-3.2.8-2.fc9
thanks,
- Doug
16 years, 3 months
output of buildinstall from pungi?
by Doug Chapman
I am still hacking away at my ia64 builds. Occasionally pungi will fail
because something went wrong when it runs buildinstall (such as a
package missing). The only way I have found to see what went wrong is
to re-run the buildinstall command manually.
I don't see it in the pungi man page but is there an option to not hide
this output? Is it being logged someplace? If this functionality does
not exist perhaps I could see about implementing it (I am by no means a
python programmer but no time like the present to learn).
Also, I notice it passes --quiet to several other programs it uses.
Might be good to have a debug option to see output from those also when
things are not going well.
thanks,
- Doug
16 years, 3 months
Oddly hanging mock build
by Jason L Tibbitts III
I'm not sure what to make of this so I haven't yet filed a bug. But I
noticed that one of spot's jobs was taking way longer than it should:
http://koji.fedoraproject.org/koji/buildinfo?buildID=31678
and so I grabbed the srpm and tried to build it on my local builder.
It hung at the same place that spot's job was hung.
Basically mock installs dependencies, then calls rpmbuild and stops:
mock.util: Installed: ccache.x86_64 0:2.4-11.fc8 perl-DBI.x86_64 0:1.601-1.fc9 sqlite-devel.x86_64 0:3.5.4-2.fc9
mock.util: Dependency Installed: perl-ExtUtils-Embed.x86_64 0:1.26-31.fc9 perl-ExtUtils-MakeMaker.x86_64 0:6.30-31.fc9 perl-Test-Harness.x86_64 0:2.56-31.fc9 perl-devel.x86_64 4:5.8.8-31.fc9
mock.Root.state: State Changed: build
mock.util: run cmd timeout(0): ccache -M 4G
mock.util: Ran setarch 'x86_64'
mock.util: chroot /mock/fedora-development-x86_64/root/
mock.Root.build: ENTER do("bash -l -c 'rpmbuild -bb --target x86_64 --nodeps //builddir/build/SPECS/perl-DBD-SQLite.spec'", '/mock/fedora-development-x86_64/root/', 0, True, 0, <mock.uid.uidManager object at 0x90e210>, 7225, 493, 'x86_64', logger=<mock.trace_decorator.getLog object at 0x90e990>)
mock.Root.build: run cmd timeout(0): bash -l -c 'rpmbuild -bb --target x86_64 --nodeps //builddir/build/SPECS/perl-DBD-SQLite.spec'
mock.util: Ran setarch 'x86_64'
mock.util: chroot /mock/fedora-development-x86_64/root/
mock.util: elevate privs to run chroot
mock.util: back to other privs
mock.util: about to drop privs
There's no further output.
- J<
16 years, 3 months
[Patch] add BUILD_FLAGS support to make scratch-build
by Till Maas
Hiyas,
The scratch build target does not respect the BUILD_FLAGS variable,
that is repected for make build. This patch changes this and makes it possible
to use, e.g.:
make scratch-build BUILD_FLAGS=--nowait
Regards,
Till
16 years, 3 months
[PATCH] koji create repo when building only for a subarch
by Dennis Gilmore
The attached patch has koji build repos when you are building only for a
non-canonical arch. this would allow someone to shadowbuild fedora and build
for i686 for instance. they would get a 32 bit repo. I dont see any issues
with this unless you try to build for multiple sub-arches but that is no
different to what we have now. with this patch you will not get ppc repos
for instance if you are only building i386 but happen to have ppc packages
in the tree
Dennis
16 years, 3 months
[PATCH] Revert to getfile urls if the task is not successful (closed).
by Jesse Keating
This is needed because unless the task is closed, the log files won't
be in the packages/ tree, and the email will reference 404 urls.
---
builder/kojid | 5 ++++-
1 files changed, 4 insertions(+), 1 deletions(-)
diff --git a/builder/kojid b/builder/kojid
index f4b7509..c03af74 100755
--- a/builder/kojid
+++ b/builder/kojid
@@ -2118,7 +2118,10 @@ Build Info: %(weburl)s/buildinfo?buildID=%(build_id)i\r
if filetype == 'rpms':
output += " %s\r\n" % '/'.join([options.pkgurl, build['name'], build['version'], build['release'], task['build_arch'], file])
elif filetype == 'logs':
- output += " %s\r\n" % '/'.join([options.pkgurl, build['name'], build['version'], build['release'], 'data', 'logs', task['build_arch'], file])
+ if tasks[task_state] != 'closed':
+ output += " %s/getfile?taskID=%s&name=%s\r\n" % (weburl, task['id'], file)
+ else:
+ output += " %s\r\n" % '/'.join([options.pkgurl, build['name'], build['version'], build['release'], 'data', 'logs', task['build_arch'], file])
elif task[filetype] == 'misc':
output += " %s/getfile?taskID=%s&name=%s\r\n" % (weburl, task['id'], file)
output += "\r\n"
--
1.5.3.8
--
Jesse Keating
Fedora -- All my bits are free, are yours?
16 years, 3 months
mock 0.9.5 tmpfs plugin failure
by Matt Domsch
When I enabled the tmpfs plugin in mock 0.9.5 today, I get the
appended traceback, I believe during clean. With tmpfs plugin
disabled (the default), I get no such failure.
--
Matt Domsch
Linux Technology Strategist, Dell Office of the CTO
linux.dell.com & www.dell.com/linux
ERROR: [Errno 16] Device or resource busy: '/var/lib/mock/fedora-development-i386-GeoIP-1.4.3-1.fc8.src.rpm/root'
Traceback (most recent call last):
File "/usr/sbin/mock", line 601, in <module>
main(retParams)
File "/usr/sbin/mock", line 546, in main
do_rebuild(config_opts, chroot, args)
File "<peak.util.decorators.rewrap wrapping __main__.do_rebuild at 0x008CE050>", line 3, in do_rebuild
def do_rebuild(config_opts, chroot, srpms): return __decorated(config_opts, chroot, srpms)
File "/usr/lib/python2.5/site-packages/mock/trace_decorator.py", line 70, in trace
result = func(*args, **kw)
File "/usr/sbin/mock", line 359, in do_rebuild
chroot.clean()
File "<peak.util.decorators.rewrap wrapping mock.backend.clean at 0x008A7320>", line 3, in clean
def clean(self): return __decorated(self)
File "/usr/lib/python2.5/site-packages/mock/trace_decorator.py", line 70, in trace
result = func(*args, **kw)
File "/usr/lib/python2.5/site-packages/mock/backend.py", line 126, in clean
mock.util.rmtree(self.basedir)
File "<peak.util.decorators.rewrap wrapping mock.util.rmtree at 0x008A67D0>", line 3, in rmtree
def rmtree(path, *args, **kargs): return __decorated(path, *args, **kargs)
File "/usr/lib/python2.5/site-packages/mock/trace_decorator.py", line 70, in trace
result = func(*args, **kw)
File "/usr/lib/python2.5/site-packages/mock/util.py", line 59, in rmtree
shutil.rmtree(path, *args, **kargs)
File "/usr/lib64/python2.5/shutil.py", line 169, in rmtree
rmtree(fullname, ignore_errors, onerror)
File "/usr/lib64/python2.5/shutil.py", line 178, in rmtree
onerror(os.rmdir, path, sys.exc_info())
File "/usr/lib64/python2.5/shutil.py", line 176, in rmtree
os.rmdir(path)
OSError: [Errno 16] Device or resource busy: '/var/lib/mock/fedora-development-i386-GeoIP-1.4.3-1.fc8.src.rpm/root'
16 years, 3 months
bash login shell invocation in mock
by Paul Howarth
mock now uses a login shell for the build phase, invoking bash as:
bash -l -c ...
It would be nice if this was changed to:
bash --login -c ...
(twi lines need changing in backend.py)
This is because the --login long option pre-dates the -l short option in
bash; the short version isn't understood in some really old
distributions (like RHL7, don't ask!) and its use precludes use of mock
in building packages for those releases. The long version seems to work
everywhere.
Paul.
16 years, 3 months
mock: enable gpgcheck for f8 config file
by Till Maas
Hiyas,
now that the groups repo is not used anymore in mock, imho the gpgcheck option
can be enabled by default and only be disabled for the local repo. It will
only need the required gpg-keys be included in the mock rpm and some more
lines in the config files. I will write a patch for this if you will apply
it.
Regards,
Till
16 years, 3 months