FC6 updates-testing kernel based on 22.214.171.124
(kernel-2.6.19-1.2895.fc6) seems to have a broken IPv6 stateless
address autoconf. The previous, successfully tested version was based
on 126.96.36.199 (kernel-2.6.18-1.2860.fc6).
The problem appears to be that the host is not joining the all-nodes
multicast address (ff02::1) [netstat -g -n], so the IP layer isn't
processing the received router advertisements even though you can see
them with tcpdump. The solicited node mcast address is joined though.
Loopback interface doesn't join ff02::1 either.
linux-2.6-xen.patch at least seems to be making some modifications to
IPv6 ND code, but I didn't look closely enough if that's the culprit.
If it matters, the network adapter is Realtek 8139.
Pekka Savola "You each name yourselves king, yet the
Netcore Oy kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings
On Sun, Jan 28, 2007 at 02:45:25PM -0800, Greg Kroah-Hartman wrote:
> On Sun, Jan 28, 2007 at 04:47:43PM -0500, Jesse Keating wrote:
> > On Sunday 28 January 2007 15:02, Dave Jones wrote:
> > > What broke?
> > I think Bill said it kept reordering network devices. Missing symlinks and
> > such.
> That kernel option should have nothing to do with network devices for
> the 2.6.19 kernel release, so I would be very supprised if it changed
For the kernel in question, this was 2.6.20rc5.
(I don't think the option even existed back in .19 ?)
Whenever i try to run pungi (0.2.1-1), either straight from command line,
or following the directions on 'running pungi in mock', i get the
Traceback (most recent call last):
File "/usr/bin/pungi", line 166, in ?
File "/usr/bin/pungi", line 95, in main
File "/usr/lib/python2.4/site-packages/pypungi/gather.py", line 194, in
os.link(local, os.path.join(pkgdir, os.path.basename(remote)))
OSError: [Errno 17] File exists
Anyone have any idea how to get past this hurdle?
-- Chris Chabot
* Fri Jan 26 2007 Jeremy Katz <katzj(a)redhat.com> - 188.8.131.52-2
- rebuild against older pciutils
* Fri Jan 26 2007 Jeremy Katz <katzj(a)redhat.com> - 184.108.40.206-1
- Fix pkgorder
- Give indication of city being pointed at for timezone (clumens, #219417)
* Fri Jan 26 2007 Chip Coldwell <coldwell(a)redhat.com> - 22.0.93-5
- remove Tetris to avoid trademark problems (Ville Skyttä #224627)
* Thu Jan 25 2007 Chip Coldwell <coldwell(a)redhat.com> - 22.0.93-4
- fixup loaddefs.el dependencies (Dan Nicolaescu #176171)
- add BuildRequires: automake (changes to Makefile.in)
* Fri Jan 26 2007 Jesse Keating <jkeating(a)redhat.com> - 6.90-2
- Core? What Core?
* Fri Jan 26 2007 Bill Nottingham <notting(a)redhat.com>
- turn on CONFIG_SYSFS_DEPRECATED so that things actually work. *sigh*
* Thu Jan 25 2007 Daniel P. Berrange <berrange(a)redhat.com> - 0.100.0-3.fc7
- Make back-compat with old APIs for Cobbler sanity
* Thu Jan 25 2007 Daniel P. Berrange <berrange(a)redhat.com> - 0.100.0-2.fc7
- Fix errors with NFS mount based installs
* Thu Jan 25 2007 Dan Walsh <dwalsh(a)redhat.com> 2.5.2-2
- Fix squid cachemgr labeling
* Thu Jan 25 2007 Dan Walsh <dwalsh(a)redhat.com> 2.5.2-1
- Add ability to generate webadm_t policy
- Lots of new interfaces for httpd
- Allow sshd to login as unconfined_t
Broken deps for s390
systemtap - 0.5.10-1.fc7.s390 requires kernel >= 0:2.6.9-11
systemtap-runtime - 0.5.10-1.fc7.s390 requires kernel >= 0:2.6.9-11
I sent this message to fedora-list(a)redhat.com but i had no reaction so i
try also the devel one because i find this is problem we have to solve
for the core and extras merge, no ?
I'm tired to have problems with rt2500 driver at each new fedora kernel
release because the versionning seems to not follow standards :
What do you think about this ?
Someone have similar problems with other modules ?
> So, what it the rationale for /etc/init.d not being in the default
> $PATH, for root at least?
Very very bad idea. Imagine only this: look at /sbin/iptables and
/etc/rc.d/init.d/iptables (or /etc/init.d/iptables, since /etc/init.d is a
symlink to /etc/rc.d/init.d). You want to run iptables binary (elf) and you
launch iptables service, that's call /sbin/iptables to initialize IP tables
with default values (using /etc/sysconfig/iptables for all default values you
I have also this question in the past, but I've found myself that this idea is
wrong, so I create symlinks to all the service I need naming with 'rc-' before
their names (/etc/rc.d/init.d/network become, as symlink, rc-network).
ing. Traian Gheorghe ONCIU
S.C. Dalila SoftWare SRL (Fedora(a)DalilaSoft.Ro)
begin 666 S.C. Dalila SoftWare SRL (DalilaSoft).vcf
M1D%8.BLT(# S-#$@-#(P(#$T-(a)T*0412.U=/4DLZ.RLT(# W,C,@-3,U(#,Q
M+B V,#M#;VYS=&%N_F$[4F]MXFYI83LY,# T.3<[4F]MXFYI80T*3$%"14P[
M5BP@87 N(#8P/3!$/3!!0V\]#0IN<W1A;CU&16$L(%)O;3U%,FYI82 Y,# T
I created an initial skeleton list of packages for the Fedora Serevr
spin . The list was generated by looking at all the packages ('server
packages') in current Core and Extras Rawhide that own a file
in /etc/rc.d/init.d or /etc/init.d.
The list is a first attempt at getting a package list together and needs
a lot more love from everybody; once this list is complete we'll look at
adding supporting packages like system-config-* and sendmail-cf.
I classified all the server packages in one of three groups: Maybe,
Include, and Exclude. Things landed in Exclude mainly because they
either were game servers or because they seemed only useful on a
desktop. Packages in Maybe seem like they should be excluded, but I am
Anybody interested in Fedora Server should have a look at the package
list and let me know if they think a package should be classified
differently; for the Maybe group, I'd appreciate comments on whether the
packages should be included or not (e.g., should FSS support dynamic DNS
and include the various clients for dyndns.org etc. ?)
Infrastructure, Extras, and the Board have had sporadic talks about
changing over from cvs to another version control system and what that
entails. Part of the discussions during the fedora-summit showed that
we can reimplement dist-cvs in another VCS (for instance, dist-hg) with
minimal change in workflow but there are some higher level questions
about what things we want to make easy that we should answer first. The
SCM SIG _ is responsible for driving this discussion and as part of
that, we're soliciting input on what our high level goals are.
Here's an initial list of questions that need to be answered. We will
be discussing this and likely making some decisions at FudCon so we're
hoping we can get some good input of these goals now to make sure we
take into account all the goals that people want the new system to
VCS Redesign Plan_
We need to answer these overarching goals before we can decide how to
redesign the VCS.
== Goals that the VCS can Help Achieve ==
Decide on goals and prioritize them so we can evaluate what
architectures implement these features the best.
* Work closer with upstream
* Work closer with downstream
* Make it easy to manage large numbers of patches
* Keep the barrier of entry for new contributors low
== High Level Architecture ==
There are some highlevel design decisions that will have an effect on
whichever underlying VCS we use. The architecture will enable us to
more effectively achieve our goals.
* Exploded trees
* Patches & spec plus lookaside cache
== Workflow for the VCS ==
* jcollie has started a page that abstracts our current workflow with
== What VCS Makes it Easiest to Implement our Goals ==
After we answer the above questions we can answer this.
_ : This information is recorded on the wiki at: