users October 2017

users@lists.fedoraproject.org

105 participants
91 discussions

suspends shutdowns system instead of suspending
by Sudhir Khanger 21 Jul '18

21 Jul '18

Hi, A new bug has cropped up on my system Dell Precision 5510 running Fedora 25 Plasma Desktop. Anytime I suspend my device it shutdowns and obviously I lose all my unsaved work. I have tried all the kernel currently installed on my system 4.10.10-200.fc25.x86_64, 4.10.9, and 4.10.8. I have also tried downgrading systemd to systemd-231-10.fc25.x86_64. I continue to have the problem. sysinfo - https://paste.fedoraproject.org/paste/9nY1-YH4IKXt4wJhyIbMGV5M1UNdIGYhyRLiv… dmesg - https://paste.fedoraproject.org/paste/WK2NAqy3PRxrR6Fze2baIl5M1UNdIGYhyRLiv… What I can do to troubleshoot the problem?

8 17

F26 update: gnome-terminal doesn't launch; locale issues
by Mike Zingale 07 May '18

07 May '18

I just did an update from F25 to F26 via dnf. After it boots, I can no longer start gnome terminal (nothing shows). If I ssh into the machine and try it remotely, I see the following: $ gnome-terminal (process:3586): Gtk-WARNING **: Locale not supported by C library. Using the fallback 'C' locale. Error constructing proxy for org.gnome.Terminal:/org/gnome/Terminal/Factory0: Error calling StartServiceByName for org.gnome.Terminal: Timeout was reached So it looks like there is a locale issue in my update. Does anyone know how to fix this?

2 2

how to correctly log without journald (and its blobs)?
by Franta Hanzlík 09 Apr '18

09 Apr '18

In previous Fedora distros was (IMHO) right way to log only to syslog, without journald and its (for me) unwanted annoying binary logs, by using configuration as: *) Set systemd log target to syslog: systemd.log_target=syslog (syslog-or-kmsg) ---- on kernel cmdline systemd.default_standard_output=syslog --/ and LogTarget=syslog ---- in systemd.conf DefaultStandardOutput=syslog --/ *) configuring rsyslog to listen on /dev/log unix socket: $SystemLogSocketName /dev/log ------ in rsyslog.conf $ModLoad imuxsock ----/ $OmitLocalLogging off ---/ $AddUnixListenSocket /run/systemd/journal/syslog --/(legacy directives) $AddUnixListenSocket /run/systemd/journal/socket -/ $AddUnixListenSocket /run/systemd/journal/stdout / (not sure when all last three directives needed) *) prevent to run journald: systemctl mask systemd-journald.service (and maybe mask systemd-journal-flush.service) And it works fine. But in actual Fedora 24 in systemd man page values syslog-or-kmsg and syslog are missing in LogTarget option. As systemd/journald man pages say hardly anything about exact mean of appropriate configuration for this purpose, then please when someone more knowledgeable can advise: - where to direct the systemd output? Maybe to kmsg an then read it in rsyslog via imklog? (rsyslog should have also imkmsg module, but is not in rsyslog-8.12.0-3.fc24) - what about /run/log/journal/.../system.journal ? On my test F24 system it have open rsyslogd, abrt-dump-journal-oops and abrt-dump-journal-xorg (no need for these last two) processes. Uses it also something else? - what about /run/systemd/journal/{dev-log,socket,stdout} unix sockets? Should rsyslogd listen on them? - /dev/log seems be now symlink to /run/systemd/journal/dev-log socket (as defined by systemd-journald-dev-log.socket) Should it be left (because now something sends to /run/systemd/journal/dev-log), or can be /run/systemd/journal/dev-log removed? Thanks in advance for some clarification about this. Franta Hanzlik

5 14

how to (re-?)construct grub menu?
by William 22 Feb '18

22 Feb '18

Good afternoon, I found out the rather hard way what caused those boot failure problems that were the subject of a threat starting last month. After my weekly "dnf upgrade" last Thursday, I checked my root e-mail, and found the hard drive had 8 bad sectors, then 16 bad sectors. A smartctl short test showed several parameters "pre-fail", and the others "old-age". I obtained a new drive. My Cray Research friend gave me a USB stick with "Clonezilla-Live-Version" on it. That duplicated the old hard drive onto the new one, but apparently not 100%. After the old drive was removed and the new drive installed, I could not boot. The system was being put into a grub shell(?). (It displayed the "grub> " prompt.) After beating out heads against the wall (or the phone), my Cray research friend gave me another USB stick with "Boot-Repair-Disk" on it. That rebuilt a grub file, but not correctly. But at least I got a grub menu when booting, and I was able to edit the shell that the grub menu entry would run. I had to take the "efi" off the commands "linuxefi" and "initrdefi". That allows me to boot up Fedora. But the menu is missing an entry for windows, and I still have to edit that script every time. I want the grub menu to offer the three most recent patches of Fedora, the most recent Fedora rescue shell, and windows-7, in that order. (This is a dual-boot system.) And I want the shells launched by the menu entries to be correct. How do I get that accomplished - the correct Fedora-25 way? If it matters, the motherboard uses UEFI bios. Thank-you in advance. Bill.

11 60

Failure with Intel wireless on Dell XPS 17
by Go Canes 09 Jan '18

09 Jan '18

After a dnf update today, I was unable to bring up my normal wireless connection on a Dell XPS 17 laptop. I was able to get a connection via my cellphone, and did a "dnf downgrade iwl*-firmware NetworkManager*", and after a reboot I am back online. From info in /var/log/messages, it looks like the relevant firmware packages are iwl7260-firmware-1:25.30.13.0-76.fc25.noarch which was downgraded to iwl7260-firmware-1:25.30.13.0-68.fc25.noarch. I'm happy to open a bugzilla if there isn't one on this issue, but it seems to me I should have more info than what I posted above, and so am looking for advice/guidance from the list. Thanks in advance!

2 2

since upgrading to F26, 'ps' will wedge the entire OS
by Lonni J Friedman 09 Dec '17

09 Dec '17

Greetings, I've been running Fedora since the Fedora Core days. About 11 days ago, I upgraded my x86_64 system from 25 to 26. Prior to upgrading to 26, I was running 25 for over 6 months without any stability issues. Since the upgrade to 26, something inspecting the process list is wedging the OS. The problem isn't even obvious until I happen to run 'ps' and the command hangs indefinitely, and can't be killed. However, at that point, the load avg takes off, and just keeps climbing over time. The system is still mostly responsive, but clearly its very unhappy about something. I'm able to check 'dmesg' as well as /var/log/messages while its in the bad state, and there's nothing new in either, and definitely nothing bad, unusual or suspicious. Unfortunately, I can't even shutdown/reboot gracefully at this point, as anything & everything that wants to inspect the process list also hangs, and so the system never fully shuts down (I waited a full hour the first time hoping that eventually it would proceed). I have to hit RESET and then (thankfully) the system boots up just fine again. The first hang happened about 28 hours after upgrading to 26. Things were fine after that for just over 9 days of uptime when the same thing happened all over again. I've tried googling to see if there's any one else running into this, and didn't find much. This was the closest thing, but its over a year ago: https://forums.fedoraforum.org/archive/index.php/t-310268.html At this point, I'm wondering whether anyone: 0) recognizes this as a known bug 1) has any ideas how to debug it further thanks!

3 5

Converting a Fedora installation from BIOS to UEFI
by Patrick O'Callaghan 08 Nov '17

08 Nov '17

I have a new external drive with Windows 10, cloned from my working QEMU/KVM setup under F26. I want the option of continuing to use it as a VM, or occasionally dual-booting directly into Windows. The VM is set up to use UEFI but my host is running under BIOS, so I need to change one of them to match the other. I prefer to change the host to UEFI as that is the preferred method for running VMs. So how do I go about it? If I simply boot using UEFI I get the EFI menu with a bunch of block devices and no \EFI directory, which I think means that I need to do something specific in Fedora. Is this some magic incantation with Grub? The Grub man pages are not clear. poc

3 5

This morning's update & bluetooth
by Wells, Roger K. 06 Nov '17

06 Nov '17

Computer is Thinkpad x260 Uname -a: Linux rwells-x260 4.13.4-200.fc26.x86_64 #1 SMP Thu Sep 28 20:46:39 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux Just updated from 4.12.14 I was using two Bluetooth devices, mouse & speaker, when I started the update. After the update neither would connect. Rebooting the previous kernel, both connect again Same for Wayland & Xorg (no surprise) -- Roger Wells, P.E. leidos 221 Third St Newport, RI 02840 401-847-4210 (voice) 401-849-1585 (fax) roger.k.wells(a)leidos.com

5 7

hybrid graphics laptop and nouveau
by AV 02 Nov '17

02 Nov '17

Hello, This concerns an Asus Zenbook with hybrid graphics (Intel/Nvidia). 1) There are 2 ways to deactivate the nouveau driver: by adding 'modprobe.nouveau=0' to the kernel cmd line OR by adding 'modprobe.blacklist=nouveau rd.blacklist=nouveau'. Which is to be preferred? 2) I do the above because there is as yet no Nvidia driver that gives a hybrid solution like under Windows and the Intel driver is more than enough for my needs. However when using a solution as described in 1) the Nvidia chip will still drain power. Is there anyway to deactivate the chip? (short of removing it from the motherboard if possible). AV

2 3

strange screen tearing effect
by Tim 01 Nov '17

01 Nov '17

Since updating to Fedora 26, I get a strange screen tearing effect on any moving image (scrolling webpages, watching live video, anything that moves). The effect is unnoticeable on static displayes. The effect is like this diagram: http://imgbox.com/TqED2C76 There's a portion at the top of the screen related to one moment in time, with a dividing line going across the screen, related to another moment. The really strange thing is that there's a diagonal portion to this division. The division slowly rolls down the screen, so it can be seen that the screen refresh rate is different from the source material (whatever it may be). According to driconf, which offers options I know nothing about, and offers nothing I see related to setting any vertical sync options, the graphics set is using: NV98 (nouveau) The display configurator (I'm using Mate) gives me 60 or 75 Hz refresh rates, and they will change the rate that the dividing line rolls down the screen, but not eliminate it. I'm using a LCD screen connected through a DVI cable (both ends, no conversions). Mate's display configurator says its DVI-I How can I get rid of this? -- [tim@localhost ~]$ uname -rsvp Linux 4.13.5-200.fc26.x86_64 #1 SMP Thu Oct 5 16:53:13 UTC 2017 x86_64 Boilerplate: All mail to my mailbox is automatically deleted. There is no point trying to privately email me, I only get to see the messages posted to the mailing list. Using Windows software is like coating all your handtools with sewage.

3 5

← Newer
1
2
3
4
5
6
7
8
9
10
Older →