When I boot my system I can play music etc. using Rhythmnbox. I also can
use the volume control. After running Firefox, the pulse application is
gone (pulseaudio --check returns a 1) and I am not able to restart it.
But the audio from firefox which must not be using pulse still works
fine. Sorry, but I am forced to run an old version of the kernel
(version 3.1.0-7.fc16.x86_64) to get video to work in my Dell 1721
laptop with a Radeon video card.
The message in /var/log/messages is:
Apr 20 15:52:52 dfpc47 kernel: [20655807.089535] rhythmbox:
segfault at 0 ip 00000032888c46e1 sp 00007fff5cf30a00 error 4 in
Apr 20 15:52:54 dfpc47 abrtd: Directory 'ccpp-2014-04-20-15:52:52-28699'
Apr 20 15:52:54 dfpc47 abrt: Saved core dump of pid 28699
(/usr/bin/rhythmbox) to /var/spool/abrt/ccpp-2014-04-20-15:52:52-28699
Apr 20 15:52:54 dfpc47 abrtd: Size of '/var/spool/abrt' >= 1000 MB,
Apr 20 15:52:57 dfpc47 abrtd: New problem directory
Apr 20 15:53:42 dfpc47 pulseaudio: socket-server.c: bind():
Address already in use
Apr 20 15:53:42 dfpc47 pulseaudio: module.c: Failed to load
module "module-esound-protocol-unix" (argument: ""): initialization failed.
I find the segfault is rough to decode form the core dump. Is there
something special I need to set up to prevent firefox from killing pulse?
Please have mercy. It has been a couple of years since I posted.
Does anybody know what the corelation between fedora package version
numbers and mozilla version numbers?
e.g. While looking into CSS3 compatibility everybody refers to
firefox-3.x; but fedora firefox->about reports firefox 26.0.
Hard to imagine fedora is 23 major versions ahead of mozilla's major
A few months ago I installed Fedora 20 as dual boot on a new Dell
desktop - one with UEFI. But I got confused while installing it and
set the BIOS to "legacy boot mode" before the installation.
As a result, I have a PC which boots to Grub and Fedora works fine.
However I can't get to the Windows 8 installation and if I switch back
to UEFI boot mode, then the system won't boot at all. This works fine
for me most of the time, but occasionally I want to run Windows on the
system. So I'd like to fix my errors and end up with a real dual-boot
As I see it, my plan should be something like this:
1/ Back up everything important from the Linux partitions.
2/ Use the Windows restore CD to repair the Windows UEFI boot.
3/ Reinstall Fedora using UEFI boot instead of legacy.
I'm hoping that if I note the partitions and directories before I
start, then I'll be able to re-use the same partitions for the same
directories as I currently have. Only my system partition will have
changed and I'll need to reinstall some RPMs.
Does this sound like a reasonable plan? Is there anything I might be
able to do which will make it easier?
Oh, and the system has an SSD which it uses for booting Windows. Is
there any chance that I install Fedora so that it uses this (but
without disturbing the Windows boot files)?
Any advice would be much appreciated.
Dave Cross :: dave(a)dave.org.uk
Installed 3.13.10-200.fc20.x86_64 kernel
When shutting down from inside a session I get a stop job is running for
user session andit takes a long time to shut down, that is normal if I
shut down after exiting any session.
Linux Fedora F20 (Heisenbug)
on Fujitsu Lifebook A512
I installed Fedora 20 on /dev/sdb1 and after installation was completed
it didn't select /dev/sdb1 and bootup .
I have Windows on sda1 so i put a sdb in computer for Linux.
Is the command, grub2-install /dev/sdb1 the correct command for booting
Fedora 20 on sdb1 ?
In all of the hubbub of Red Hat Summit, I forgot to post this to the
mailing lists. Well, better late than never, I guess. Here's last
week's Five Things in Fedora This Week, reposted from
Fedora is a big project, and it’s hard to follow it all. This series
highlights interesting happenings in five different areas every week.
It isn’t comprehensive news coverage — just quick summaries with links
to each. Here are the five things for April 15th, 2014:
Live from Red Hat Summit
I'm posting this from the Fedora booth at Red Hat Summit in San
Francisco. The Summit is Red Hat's annual showcase conference, and of
course we have a good Fedora presence. Fedora Project Leader Robyn
Bergeron is on the scene; Ruth Suehle is setting up a demo on the big
TV behind us; and Tom Callaway has the Lulzbot up and making 3D prints.
If you're at Summit, come by and say hello, ask questions, tell us
answers, and so on. (Except, don't, because, remember, this was
actually last week.)
Rawhide is Fedora's continuously updated development tree. Long ago, it
came with the scary warning that it would "eat babies" and other
horrible things, but that's been *gone* from the official line for
about five years now. In fact, if you're up for a little adventure (and
have a second computer to get work done if it goes wrong), we'd love
you to help test.
Fedora QA Community Monkey¹ Adam Williamson recently revamped the
Rawhide wiki page to be more useful and clear, and launched an
initiative for quality assurance volunteers to do ongoing validation
testing so any problems are caught long before release. If this sounds
interesting, join the Fedora QA team and help out.
I run Rawhide on my main desktop system but not my travel laptop; that
way, I can always get some work done even when there's a problem. But,
really, it's been mostly painless. I recommend keeping active on the
mailing lists and checking the Rawhide Watch and This Week in Rawhide
(1. That's Adam's semi-official title.)
Fedora Heartbleed Followup
Last week's main excitement — not just in Fedora! — was the
"Heartbleed" vulnerability in OpenSSL, referred to by the
industry-standard Common Vulnerabilities and Exposures identifier
`CVE-2014-0160`. (If you missed the official Fedora updates, see Update
on CVE-2014-0160, aka “Heartbleed” and Fedora Infrastructure
information on Openssl vulnerability.)
I think our response to this went fairly well overall, but there's
always something to be learned from these incidents and things we could
do better. There are two specific areas where I think we could improve.
First, our normal process of releasing updates requires a lot of data
crunching. I think it'd be useful to have a separate "urgent updates"
repository that would enable us to get updates to users more quickly
once they are ready to go. Second, it would be nice to have a standard
process for getting packagers, a communications team, security experts,
quality assurance, release engineers, and other responders "on the
scene" quickly — a sort of "Fedora bat-signal." Both of these ideas
need some development, of course.
Another important thing we can do in advance is develop more test cases
for key software. These can be followed as part of making sure updates
are good to be pushed to users, rather than relying on ad hoc
assessments ("works for me!"). This is an area where anyone can
contribute — you don't have to be a coder or a package maintainer. Pick
a package that's interesting to you and take a look at the instructions
for creating a test case.
Fedocal — the Fedora Project calendar
A couple of weeks ago, I mentioned that the Fedora wiki is weighed down
by too many different uses. One of those uses is scheduling: wiki pages
and wiki tables are used as editable calendars. If you've ever tried to
update one of these, you know that it can be painful. But we have a
better tool: Fedocal.
Fedora hackers Pierre-Yves Chibon (a.k.a. pingou) and Ralph Bean
(threebean) recently added location support, which means that IRC-based
meetings can use that for scheduling. In general, if you're scheduling
something in Fedora, this is the tool to use. Let's avoid further wiki
Although the North American and European events are replaced by Flock
to Fedora, our Fedora User and Developer Conferences are going strong
worldwide. Upcoming events include FUDCon Beijing 2014 and FUDCon
Managua 2014 in Nicaragua.
Also, speaking of Flock, don't forget to vote on session proposals.
Thanks this week to Ankur Sinha for feeding me notes from Fedora
Matthew Miller -- Fedora Project -- <mattdm(a)fedoraproject.org>
"Tepid change for the somewhat better!"
I want to report a freeze with latest fedora 3.13 kernel just on login
screen. The issue is not present on the initial kernel coming with
fedora 20 life CD.
The problem is that when I boot I can't switch to a text console after
the freeze. I don't have good access to the laptop because it's my
mother's so I'm wondering what would be the easiest way to collect
useful information and report it.
btw I think it is nouveau issue because I see some messages realated to
nouveau just before entering graphics mode.
Perhaps I can boot to the freeze, then reboot into older kernel, and
collect the log but my there is some risk the log to not be complete due
to unclean reboot. As well I'm not sure if plain log is enough. I would
like to file a good bug report from the first attempt to avoid the
hassle repeating the operation for small details.