Has anyone tried to get native Optimus running on a Fedora 19 box? When I attempt it, I get a blank black screen. I posted on the Nvidia forums, but no one has replied yet and I'm getting anxious ;-)
The general consensus is that the kernel must be 3.9 or higher, and I am running 3.9.9-302. The Nvidia readme<http://us.download.nvidia.com/XFree86/Linux-x86-ARM/319.32/README/randr14...> indicates CONFIG_DRM has to be enabled and the following driver interfaces present:
I have verified that CONFIG_DRM is enabled within the kernel as a module and it's loaded, but when I attempted to verify the list of module symbols, I only found two (the ones in bold).
The way I determined this information was through:
`cat /boot/config-3.9.9-302.fc19.x86_64 | grep CONFIG_DRM`
`cat /lib/modules/3.9.9-302.fc19.x86_64/modules.symbols | grep gem_prime`
Am I missing something or does it appear that F19 kernel is missing vital parts for native Nvidia Optimus support?
For those more interested, here is the post<https://devtalk.nvidia.com/default/topic/551709/linux/319-23-the-infamous...> on the nvidia forums.
On F19 with Mate desktop, sometimes after login 8 or 9 or even 10 copies
of Caja (Mate File Manger) are started up.
Most of the time not, but perhaps once every day or two this will happen.
the rest of the time it's completely normal.
I found these /var/log/messages entries at around the right time, but I'm
not certain whether they are from login or the boot a small bit earlier:
Jul 19 19:23:56 aspirebox dbus-daemon: dbus: [system] Rejected send me
ssage, 2 matched rules; type="method_call", sender=":1.52" (uid=1000 pid=1410 co
mm="caja -n ") interface="org.freedesktop.DBus.Properties" member="GetAll" error
name="(unset)" requested_reply="0" destination=":1.12" (uid=0 pid=638 comm="/us
r/sbin/console-kit-daemon --no-daemon ")
Jul 19 19:23:56 aspirebox dbus: [system] Rejected send message, 2 matched rules; type="method_call", sender=":1.52" (uid=1000 pid=1410 comm="caja -n ") interface="org.freedesktop.DBus.Properties" member="GetAll" error name="(unset)" requested_reply="0" destination=":1.12" (uid=0 pid=638 comm="/usr/sbin/console-kit-daemon --no-daemon ")
Anybody got any clues?
Under no circumstances will I ever purchase anything offered to me as
the result of an unsolicited e-mail message. Nor will I forward chain
letters, petitions, mass mailings, or virus warnings to large numbers
of others. This is my contribution to the survival of the online
--Roger Ebert, December, 1996
----------------------------- The Boulder Pledge -----------------------------
I have been having this weird problem; everytime I there is a
selinux-policy update, I get the following error:
libsepol.print_missing_requirements: hotplug's global requirements
were not met: bool init_systemd (No such file or directory).
libsemanage.semanage_link_sandbox: Link packages failed (No such file
FWIW, this machine was updated from F17 to F19 with yum. I tried
searching for the error above, but could not make sense of what I found.
Open source is the future. It sets us free.
sorry for breaking the thread. I was not subscribed up to now
and did not yet find an archive that exposes message ids.
It is not that bad with Blu-ray. At least with the backends.
(I am the developer of libburn and use BD-R and BD-RE media daily.)
But wodim (or "cdrecord") out of cdrkit should really be used
only for CD.
The slowliness of BD media comes from their Defect Management
which checkreads the written blocks while they are still present
in the drive's buffer. In case of poor read quality the blocks
get written again or redirected to the Spare Area where the data
get stored in replacement blocks.
This checkreading reduces write speed to at most half of the
nominal speed of drive and BD medium.
BD-R do Defect Management by default if they are formatted.
BD-RE can only be used formatted and do Defect Management by default.
Original cdrecord out of cdrtools was already discussed in the this
thread. To my knowledge it is fast with BD-R, because it does not
format them. It is not fast with BD-RE, because it uses SCSI command
WRITE(10) which does not allow to override the slow Defect Management.
growisofs out of dvd+rw-tools has a little bug with BD-R.
It is supposed to be fixed in
The tiny patch is shown in
Workaround is to explicitely format the BD-R before use by
or to use option
when giving an unformatted BD-R to growisofs.
growisofs formats BD-R by default and then uses slow Defect
Management. This can be prevented by option
growisofs has no means to prevent Defect Management on BD-RE media.
My own programs, libburn based cdrskin and xorriso, do not format
BD-R by default. So they are as fast as with cdrecord.
Other than with cdrecord it is possible to format BD-R on request
cdrskin -v dev=/dev/sr0 blank=format_if_needed
xorriso -outdev /dev/sr0 -format as_needed
which makes them capable of slow Defect Management.
Other than growisofs and cdrecord, cdrskin and xorriso offer the
opportunity to disable Defect Management on formatted BD-R and
cdrskin stream_recording=on ...cdrecord-like-burn-options...
xorriso -stream_recording on ...xorriso-burn-commands...
This is especially helpful with 2x BD-RE, which elsewise need
an awfully long time to write their capacity of 22.5 GiB.
Have a nice day :)
I recently noticed that I have been getting warnings
Too many objects for client ':1.15', client request failed.
on my server.
I must admit I didn't know I was running avahi on the server.
I looked at a couple of avahi howtos,
and it seemed that for almost any usage other packages are required,
in particular nss-mdns .
If this is in fact required by avahi, it seems surprising
that it is not automatically installed at the same time.
(None of the examples of avahi usage actually worked on my system.)
But my real query is whether avahi (or zeroconf)
could actually offer me anything useful.
I'm running dhcpd on the server,
and I'm not looking for information about printers
or other users on my home network.
Also, is there any way of discovering who "client 1.15" is?
And is anything likely to go wrong if I simply "service avahi-daemon stop"?
e-mail: gayleard /at/ eircom.net
tel: +353-86-2336090, +353-1-2842366
School of Mathematics, Trinity College, Dublin 2, Ireland
Just want to check first if I am the only one having this problem with an
upgraded F19 setup.
Upgraded the system from F18 (and previously from 17-16-15) with fedup
--network. Before the upgrade I was running Gnome fine in the classical
style (dropdown app menu, no hotspots, workspace grid on lower bar etc.).
After the upgrade logged on with Gnome Classic selected -> worked fairly
ok, I have a grey top and bottom bar. Logged out and in again in order to
get some additional settings (the workspace for both dual displays). After
the login (making sure that I still have the Gnome classic selected) the
top and bottom bars are not the same.
Specifically the bottom bar (where the running apps should be displayed as
well as the workspace app) is missing altogether and the top bar is now
the new Gnome-styled, ie. has the hotspot and no app-menu at all.
This is reproducable, if I log out from this state, select the normal
gnome session, then I (again) get the same new Gnome desktop as described
above, and after an additional logout + login with the Gnome Classic
selected, then I end back in the actual classic for that particular
So it seems that the mode selection is not really working, and neither is
the selection remembered correctly.
Does this sound familiar to anyone?
Paul-Erik Törrönen "When science and the Bible differ,
poltsi(a)777-team.org science has obviously misinterpreted
+358 40 703 1231 its data"
http://poltsi.fi/ Henry Morris, father of 'modern creationism'
I have a Toshiba Satellite L850 Laptop for which the WiFi card appears
to be incompatible
with any drivers available for Linux. To try to get around this problem
I recently purchased
a USB WiFi device: An ASUS USB-N110. They (ASUS) provide drivers which
install. After a great deal of travail and fumbling around (the
instructions were both unclear
--- apparently written by someone for whom English was at best a *third*
and slightly misleading) I managed to get the device *partially* working.
The device appears to work immediately after boot-up but if I close the
lid of my laptop,
(which causes it to go into hibernation (???) mode) or if I log out and
log back in,
the WiFi connection gets lost. When I try to restore the connection it
seems to keep
trying to use the WiFi card that is built into the laptop rather than
the USB WiFi device.
(And the built-in device doesn't work; as I said, its drivers are
incompatible with Linux OSes.
Which is why I bought the USN-N10 in the first place.) The only way to
get the laptop to
make use of the USB WiFi device seems to be to shut down and restart.
This is unsatisfactory.
Is there any way to "disable" the (non-functional) built-in WiFi card so
that the laptop will
*not* try to use it but will rather go straight to the (functional) USB
device? I Googled around
a bit and found some instructions for disabling the built-in WiFi card
under Ubuntu, but
these instruction did not (as far as I could tell) mesh with the Fedora
system that I am
running. Can anyone tell me how to do what I want under Fedora?
If you reply, ***PLEASE*** be as clear and explicit as you possibly
can. I am not terribly
swift with OS matters. I know and understand ***some*** things, but
there are huge
lacunae in my knowledge.
I am running Fedora 17; output of "uname -a" is:
> Linux localhost.localdomain 3.3.4-5.fc17.x86_64 #1 SMP Mon May 7
> 17:29:34 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
Thanks for any help that anyone can give me.
> On 30/07/13 12:41, Steven Stern wrote:
> >/ If you're using the repo from virtualbox.org*, type
> />/ sudo /sbin/service vboxdrv setup
> />/ It will use DKMS to remove the virtual box drivers, recompile them for
> />/ the new kernel, and install them into the active kernel modules.
> />/ * This may apply to the Fedora repos but I'm not sure. I'm using the one
> />/ from virtualbox.org.Seehttps://www.virtualbox.org/wiki/Linux_Downloads <seehttps://www.virtualbox.org/wiki/Linux_Downloads>
> />/ and scroll down to RPM-based Linux distributions
> This is a yum install but I will look into that.
If you have DKMS installed it will automatically update the modules when
a new kernel
in installed. Works great for me.
From the doc referenced above.
*Note:*Users of Fedora/Mandriva might want to install the*dkms*package
(not available on openSUSE) to ensure that the VirtualBox host kernel
modules (/vboxdrv/,/vboxnetflt/and/vboxnetadp/) are properly updated if
the linux kernel version changes during the next kernel update.
Is there some way to create launchers on the Gnome 3 Classic Mode
panel, as was easily done in Gnome 2?
I'm trying really hard to like Gnome 3 Classic Mode, but having a very
difficult time of it, because it seems to only provide a superficial
resemblance to the Gnome 2 desktop, without actually providing
I'm using Linux 3.10.x kernel and I have AMD proprietary graphics driver running akmod. The 3.9.9 kernel series was working fine until the latest update came.
Now I can't log in using the new kernel, tthe screen goes black and it logs itself out. I can't even log back in.
The older kernel is working flawlessly though. Any solution?
Junayeed Ahnaf Nirjhor
Software Engineer @ Hulu
Twitter @ Nirjhor