Fedora 8 blocker update - testers needed (still!)

[Will Woods]

As promised, here's an update on the current Fedora 8 Blocker list. 

As a note: the deadline for blocker fixes is Tuesday, October 29. If we
still have unfixed critical bugs Tuesday, the release will have to be
delayed.

But I don't think that'll happen. Things are looking OK. We currently
have 14 bugs on the blocker list, which I'll talk about in detail below.
About half of them have fixes that are waiting for confirmation, and a
couple of them could be dropped if necessary because they only affect
particular, uncommon hardware.

What we need help testing is:

1) ALSA - 1.0.15 was released last week, and we decided it was a
low-risk update that would fix audio problems. for a lot of people. You
probably saw mail from Chuck Ebbert about this already.

2) Intel wireless - John Linville, our fearless (and handsome) wireless
maintainer, just added a fix for iwlwifi that should help some Intel
wireless users. 

Both of the above patches are in kernel-2.6.23.1-37.fc8, which should
land in rawhide tomorrow. You can get it from koji in the meantime:
http://koji.fedoraproject.org/koji/buildinfo?buildID=22511

3) dmraid, especially Intel
See bug 349161 below. Something is causing mkinitrd to mess up the /init
script, which leaves the reporter's system unbootable. We can't
reproduce it at all. 

If you've got dmraid (aka BIOS RAID aka fakeraid) hardware, ESPECIALLY
the one mentioned in this bug (ICH9R chipset + Intel Matrix Raid),
please try to install rawhide and let us know if it works or not.

Here's the full blocker list, in detail:

=== Miscellaneous bugs from last time ===

* Bug 208607: ASSIGNED - kernel maintainers
- mkinitrd: iSCSI root requires crc32c module

I believe we plan to fix this by building crc32c into the kernel for
release (which would make this a non-blocker). We'll hopefully have a
smarter fix for F9 that will actually close the bug.


* Bug 242766: ASSIGNED - alan
- [pata_via via82c596a] hard drive not found during install attempt

All testing with pata_via seems to indicate that it works for most
people. Alan's debugging efforts indicate that this is a drive-specific
failure on a machine with a buggy BIOS. Fun. This will be dropped from
the blocker list, since it works under normal circumstances.

(By the way- thanks to everyone who answered the call for testing here.
You guys rock!)


* Bug 335161: ASSIGNED - triad (Linus Walleij)
- multilib file conflicts in gnokii

Fix was incomplete, needs a rebuild. This prevents people from being
able to yum upgrade from F8Test3 to F8 final. 


* Bug 330931: MODIFIED - katzj
- liveinst fails - sg devices confuse anaconda?

This was waiting on some info from the reporter, which has been
provided. It involves some uncommon hardware so this bug may be dropped
from the blocker list if it can't be fixed easily.


=== NetworkManager bugs ===

* Bug 311511: NEW - dcbw
- x86_64 iwl3945 NM crash after Keyring loop Deny

Comments indicate that the crash is fixed, but it still loops on bad
passwords. 


* Bug 323371: NEEDINFO - dcbw
- NM 0.7 doesn't support WPA/WPA2 Enterprise (certificate-based
authentication)

Reporter seems to indicate that this is mostly fixed, but some testing
on a fresh install is needed to confirm the fix.


* Bug 353091: MODIFIED - dcbw
- NetworkManager-0.7.0-0.3.svn3020.fc8 crashes when trying to switch to
wireless Network

nm-applet crasher. A fix will be in the next build, and then this can be
retested and (hopefully) closed. 


=== Added since Wednesday ===

* Bug 349161: NEW - pjones
- Missing entries in the init script of the initrd image makes the
system unbootable on dmraid

This is a mysterious one. mkinitrd is creating a bad /init script for a
dmraid (aka fakeraid, aka BIOS RAID) raid0 set. But we can't reproduce
it on any of the dmraid machines we've tried.

It's a fixable problem, and I don't think the hardware is very common,
and we can't reproduce the bug ourselves, so.. if push comes to shove
this bug might get dropped from the blocker list and fixed post-F8.


* Bug 352511: MODIFIED - dwalsh
- return of an avc on useradd with livecd user

Should be fixed in selinux-policy -33 - we'll confirm as soon as Jeremy
makes some Live images.


* Bug 353571: MODIFIED - dwalsh
- /usr/bin/crontab blocked by selinux


Apparently fixed in selinux-policy-3.0.8-36; waiting for confirmation
from the reporter.

* Bug 353961: NEW - twoerner
- lokkit doesn't write iptables config with --nostart

This was reported earlier this morning - no response from the maintainer
yet. Technically I don't think this breaks any of the ReleaseCriteria,
so it could be dropped. But proper firewall function is important for
security so we want to keep an eye on this.


* Bug 354531: ASSIGNED - dwmw2
- script failure

This is a simple scriptlet problem with bluez-utils, but it keeps the
'bluetooth' service from being installed properly. Improved Bluetooth
support is one of the major features of F8, so we want this to work
right. The maintainer is aware of the problem and working to fix it.


* Bug 354571: MODIFIED - anaconda maintainers
- live install fails when there's a pre-existing install

Fixed in anaconda CVS - the next set of Live images should allow us to
retest and close this.


* Bug 354781: NEW - skvidal
- yum-skip-broken changes yum's dependency solver even when it's not
active

This is a bug with the popular yum-skip-broken plugin that could cause
it to make upgrades from F7 fail. Upgrade failures are bad. 



And there you have it. Most of those bugs already have fixes pending,
and none of the rest are hair-on-fire critical bugs, so I think we're in
good shape. If you can help test the stuff I mentioned above, or you
have found some huge problem that we've missed, please let us know.

Thanks for helping make Fedora more awesome!

-w
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://listman.redhat.com/archives/fedora-devel-list/attachments/20071026/a25cad5c/attachment.sig>