users September 2016

users@lists.fedoraproject.org

136 participants
109 discussions

Cleaning up after an aborted upgrade
by CLOSE Dave 08 Sep '16

08 Sep '16

I have a machine where I tried to do a system-upgrade from F23 to F24. The download portion worked fine but after the reboot, the upgrade process aborted. I'd like to know why, but first I need to clean up the mess so I can try it again. Right now I have 4782 total packages installed including 1490 fc24 packages. I haven't checked all of them but it appears that all are paired with the same fc23 package. Trying to repeat the download step, I get this result. # dnf --allowerasing -y system-upgrade download --releasever 24 Last metadata expiration check: 0:23:15 ago on Tue Sep 6 15:53:47 2016. Dependencies resolved. Error: The operation would result in removing the following protected packages: systemd. Evidently, there are two systemd packages installed: # rpm -q systemd systemd-222-14.fc23.x86_64 systemd-229-13.fc24.x86_64 No new kernel was installed and the machine is running fine as F23. But I'd like to get it upgraded... I tried using "dnf history rollback" but that didn't work, reporting "A transaction cannot be undone". Since dnf won't erase the extra stuff, I can probably do it with something like, "rpm erase --force $(rpm -qa|grep fc24)" but that seems iffy. Better suggestions are very welcome. -- Dave Close

4 10

*Still* having grief trying to stick a samsung SSD into ASUS laptop running fedora 23
by Robert P. J. Day 08 Sep '16

08 Sep '16

(this is almost certainly a kernel issue but here's hoping someone can give me some guidance.) once upon a time, i whined pathetically about problems sticking a samsung 950 PRO M.2 SSD into my ASUS G752VL-DH71 gaming laptop currently running fedora: http://osdir.com/ml/fedora-users/2016-07/msg00410.html eventually, i tripped over this ASUS forum page: https://rog.asus.com/forum/archive/index.php/t-79590.html where one gets the distinct impression that that particular samsung SSD is not compatible with my laptop, but one is assured that what does work is the samsung SM951, which i picked up, with still no luck, still getting early error message: ata1.00: failed to enable AA (error_mask=0x4) ... etc etc ... this is the SSD: https://www.amazon.com/Samsung-SM951-512GB-NVMe-MZVPV512HDGL-00000/dp/B015C… and i'm running out of ideas; do i need to roll my own kernel to add support for this thing? i would have thought that if i don't have support, the SSD would just be ignored, and not choke the system. i'm about to do more testing this weekend, but i sure would like some advice. rday -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

2 1

Issue with ftp making connection but not list?
by Michael D. Setzer II 08 Sep '16

08 Sep '16

Use ftp to transfer files, but just had issues today in which connection is made and login works fine, but doing a ls or trying to download a file fails? Tried with passive on and off? Did use it just the other day. We did have some power issues, since two of the Island main generators are down, and they have had 3 hour rolling outages?? I can do sftp connections, and they seem to work just fine, but anonymous doesn't work with that. Have tried both using ncftp and regular ftp. Have systems in my classroom that are on a separate network sitting side by side, and then can not connect. Can connect with ftp server on same machine using IP, and it works? Not at the site at moment, but will try to power cycle switches and computers to see. I can ncftp to servers on campus from off campus, but from inside campus to same machines doesn't work? Don't know if MIS might have made some changes?? +----------------------------------------------------------+ Michael D. Setzer II - Computer Science Instructor Guam Community College Computer Center mailto:mikes@kuentos.guam.net mailto:msetzerii@gmail.com Guam - Where America's Day Begins G4L Disk Imaging Project maintainer http://sourceforge.net/projects/g4l/ +----------------------------------------------------------+ http://setiathome.berkeley.edu (Original) Number of Seti Units Returned: 19,471 Processing time: 32 years, 290 days, 12 hours, 58 minutes (Total Hours: 287,489) BOINC@HOME CREDITS ABC 16613838.513356 | EINSTEIN 111590261.788695 ROSETTA 47988082.134319 | SETI 91238390.558019

4 11

CUDA on F24?
by Hiisi 07 Sep '16

07 Sep '16

Hi! Have a fresh install of F24. The machine has Nvidia GTX 960 video card. I would like to enable GPU rendering in blender. Have installed nvidia drivers from rpmfusion repo and CUDA toolkit from nvidia. Can select now CUDA as rendering option in blender settings. However trying to render anything fails. Initially nvcc couldn't compile using gcc-6.1.1. After installing compat-gcc I have gcc-3.4. Added softlinks to it: sudo ln -s /usr/bin/gcc34 /usr/local/cuda-7.5/bin/gcc And same for g++ Now CUDA's kernel compilation fails with the following error: /usr/lib/gcc/x86_64-redhat-linux/3.4.6/../../../../include/c++/3.4.6/bits/cpp_type_traits.h (334): error: expected an identifier 1 error detected in the compilation of "/tmp/tmpxft_00001fa8_00000000-7_kernel.cpp1.ii". As I understand the issue is still related to some uncompatible versions of C++. Maybe not the compiler itself but std standard. Is there workaround? Anyone using successfully CUDA under F24? There must be thousands of you! Please give me some hint. -- Hiisi. Registered Linux User #487982. Be counted at: https://linuxcounter.net/ -- Spandex is a privilege, not a right.

2 2

Re: unable to start rc-local.service[SOLVED]
by R. G. Newbury 07 Sep '16

07 Sep '16

On 09/05/2016 05:07 AM, oe Zeff <joe(a)zeff.us> wrote: > Subject: Re: unable to start rc-local.service[SOLVED] > To: Community support for Fedora users <users(a)lists.fedoraproject.org> > Message-ID: <57CC87EA.3090508(a)zeff.us> > Content-Type: text/plain; charset=utf-8; format=flowed > > On 09/04/2016 01:34 PM, François Patte wrote: >> > Just unwanted spaces before and after #!/bin/bash >> > >> > The script works when executed for itself, but it seems that systemd is >> > quite supercilious about that spaces and very avaricious to give some >> > hints about the error encountered! >> > >> > Thank you for helping. > Interesting. I have some blank lines in mine, to improve readability, > but not there. I'll have to keep that in mind for future reference. > It's rather like mount insisting that every line in /etc/fstab end in a > newline, even the last one, but not giving a clear error message if you > forget it. > > And, it's occurred to me that there's a way to get what's needed here > done that's easier because it doesn't try to make the same change during > every boot. Alas, I've gotten the impression that my advice isn't > wanted in this thread, so I'll leave it as an exercise for the reader. Lennart Poettering has a hate for rc.local. He discusses that somewhere in his systemd blog. For whatever reason, he dislikes it and thinks that no-one should use it, or be able to use it. So you must sacrifice the correct animal, at the exactly correct time, for it to work. And of course, just as with pulseaudio, systemd does not give meaningful or useful error messages. Since most items in rc.local are run-one-time items, the answer is to NOT USE SYSTEMD. Create your rc.local file, and chmod it u+x, and then add the following to your (root) crontab: @reboot /etc/rc.d/rc.local @reboot in a cron does exactly what it implies: it runs, once at boot time. See 'man 5 crontab' Problem solved without systemd. -- Geoff

4 3

Getting luks dependency problems on boot
by Martín Marqués 07 Sep '16

07 Sep '16

Hi all, I have my laptop with F23 with the root partition encrypted with luks, working great for about a year already. I had some pending upgrades which I applied today (kernel, glibc and systemd are the ones that made me open my eyes wide). The thing is that system starts booting, asks for the password to decrypt (actually decrypts the partition) but then fails and falls back to a rescue shell. Reading the logs there (/run/initramfs/rdsosreport.txt) it seems there is a Dependency problem somewhere: Dependency failed for Cryptography Setup for luks-..... Dependency failed for Encrypted Volumes I'm able to mount the root partition in the rescue shell and see it's contents, but I can't get the system to boot correctly. I've already tried using one of the old kernels (have always 3 around). Any thoughts on how I can get passed this? -- Martín Marqués select 'martin.marques' || '@' || 'gmail.com' DBA, Programador, Administrador

3 5

Dolphin SMB read only?
by Mark Haney 07 Sep '16

07 Sep '16

I'm not sure if this is a Samba client issue, or an issue in Dolphin, so I'm posting to the KDE and General users lists. Is anyone else connecting to a Windows share and not being able to write to it from Dolphin? I've got a handful of Windows servers that I regularly download items to my Fedora laptop and copy over via Dolphin. Up until the F24 upgrade that is. Originally I thought it was just a bug that would be addressed shortly after F24 was released. Now, though, I can still connect to the shares, but I cannot copy/paste to them. None of them. Is there something simple I'm missing? FYI, it's not a permissions issue as I can login to those shares from a Win10 VM I have on my laptop and write to those same shares without trouble. The Windows servers, other than being patched are the same versions (ie no upgrades recently), so I'm at a loss as to the problem. Dolphin version: dolphin-libs-16.04.3-1.fc24.x86_64 dolphin-16.04.3-1.fc24.x86_64 dolphin4-libs-16.04.3-1.fc24.x86_64 Samba version: samba-common-libs-4.4.5-1.fc24.x86_64 samba-client-libs-4.4.5-1.fc24.x86_64 samba-client-4.4.5-1.fc24.x86_64 samba-common-4.4.5-1.fc24.noarch -- Mark Haney ::: Senior Systems Engineer *VIF* *International Education* P.O. Box 3566 ::: Chapel Hill, N.C. 27515 ::: USA 919-265-5006 office Global learning for all. www.viflearn.com Find VIF on Facebook <http://facebook.com/VIFInternationalEducation> | Twitter <https://twitter.com/vifglobaled> | LinkedIn <http://www.linkedin.com/company/vif-international-education> Recognized as a ‘Best for the World’ <http://bestfortheworld.bcorporation.net/> B Corp!

1 0

During rescue startup boot process halts with "cannot open access to console, the root account is locked"
by Jobst Schmalenbach 07 Sep '16

07 Sep '16

Fedora-Live-Workstation-x86_64-23-10.iso USB based DVD Lenovo Yoga Description of problem: Originally I tried to boot to rescue mode on a Lenovo Yoga that has Ubuntu installed. I wanted to start rescue mode, then wipe the ubuntu install by formatting the partitions (/ and boot using mkfs.ext4) and replace Ubuntu with Fedora 23. So I inserted the Fedora 23 DVD, wait for the bios boot menu to appear, select the DVD drive with inserted Fedora 23 and boot to that. When the grub menu appears I select tab and change the kernel line to have "rescue" at the end, delete the "quiet" so I can see what is going on and add "vag=771". Just after the "Started D-Bus System Message Bus" it stops: "Cannot open access to console, the root account is locked. See sulogin(8) man page for more details. As this is "rescue mode" and the idea of rescue mode is to be able to access anything that is on the computer even with a messed up password it is tricky. How reproducible: I have tried this on three computers: 1: a beefy, selfmade workstation that has Fedora 22 installed on the local hard drive 2: a Dell XPS 13, that has Windows installed (but UEFI disabled) 3: Lenovo Yoga that has Ubuntu 16.04 installed (UEFI disabled) all of them have the same issue Additional info: It does not matter what setting UEFI boot is set to, I tried this with every possible setting in the BIOS that was given to me by the different manufacturers, I could not get pass the "root account is locked" Anybody any ideas, please. thanks Jobst

6 9

How do I beat systemd user daemons into submission?
by Tom Horsley 06 Sep '16

06 Sep '16

For the longest time, the systemd "user" daemon ran nothing other than something that called itself (sd-pam). Now I see a vast forest or utter crap being started by the systemd user daemon. Most of this is stuff I long ago eradicated, and now it is pulled it back in. Where the devil does systemd stash the info about what crap to start in the "user daemon" and how do I change it? 2080 ? Ss 0:00 /usr/lib/systemd/systemd --user 2083 ? S 0:00 \_ (sd-pam) 2111 ? Ss 0:00 \_ /usr/bin/dbus-daemon --session --address=systemd: --nofork --nopidfile --systemd-activation 2235 ? Ssl 0:00 \_ /usr/libexec/at-spi-bus-launcher 2240 ? S 0:00 | \_ /bin/dbus-daemon --config-file=/usr/share/defaults/at-spi2/accessibility.conf --nofork --print-address 3 2243 ? Sl 0:00 \_ /usr/libexec/at-spi2-registryd --use-gnome-session 2292 ? Ssl 0:00 \_ /usr/libexec/gvfsd 2297 ? Sl 0:00 \_ /usr/libexec/gvfsd-fuse /run/user/1000/gvfs -f -o big_writes 4646 ? S 0:00 \_ /usr/libexec/gconfd-2 6157 ? Sl 0:00 \_ /usr/bin/gnome-keyring-daemon --start --foreground -- components=secrets I utterly despise the rancid gnome virtual filesystems nonsense. I previous managed to make it go away, but now it is back. I hate the gnome-keyring-daemon, and it is back too (and comes back again when I kill it - no doubt courtesy of systemd). AAAAAAAGGGGGGGHHHHHHH!!!!!!!!

4 4

Question about export display
by Todor Petkov 06 Sep '16

06 Sep '16

Hello, is there a way to run a graphical program via ssh/X forward with low quality, i.e. 16 bit mode/800x600 etc? Sometimes I need to run firefox/chrome from a remote server, and I would like, if possible, to run them not like it's 1024x768 or higher. I searched google, but nothing comes so far. Thanks in advance.

5 8