Ask Your Question

BH2017's profile - activity

2018-03-02 10:25:21 +0000 received badge  Notable Question (source)
2018-03-01 08:29:10 +0000 commented question Is it possible to just use a LTS kernel on Fedora?

I was hoping there was a LTS kernel provided by Fedora somewhere I hadn't looked. But if not maybe from a Copr, or some other third party repo. Or a nice automated build tool for creating your own custom kernel whenever a new kernel in the LTS series comes out. But alas it seems I'm out of luck.

2018-03-01 07:05:43 +0000 commented answer Is it possible to just use a LTS kernel on Fedora?

Wouldn't these regression tests only work if the system could actually successfully boot on the defective kernel? I've read the info at the link you gave, followed one of the links there (namely https://fedoraproject.org/wiki/Kernel... ) and so far it seems like I actually have to be able to boot the problem kernel and be able to access the command-line to run these tests. When this problem I'm having with the 4.15.x kernels occurs the system doesn't boot to a system where I can actually open a terminal, not even an agetty terminal (i.e. those opened with Ctrl + Alt + Fn).

2018-03-01 05:35:47 +0000 commented answer Is it possible to just use a LTS kernel on Fedora?

Oh and all I wanted to use the LTS of was the kernel. Everything else bleeding-edge. This is possible on the bleeding-edge distros Arch, Gentoo, Manjaro and Void. I was hoping a LTS kernel was available from somewhere like it is on these distros.

2018-03-01 05:29:03 +0000 commented answer Is it possible to just use a LTS kernel on Fedora?

Plus of course if it weren't the 4.15.x kernel and systemd having a problem this shouldn't affect other systemd-using bleeding-edge distros.

2018-03-01 05:05:39 +0000 commented answer Is it possible to just use a LTS kernel on Fedora?

The systemd services just time out. Most of my fstab is left as it was from the Anaconda installer, only change I made was adding:

/dev/sdb1    /data    ext4    defaults 0 1

to it. My fstab is: http://paste.debian.net/1012510/ . Frankly I doubt my fstab is relevant as how would it be a problem with the 4.15.x series kernels, but not the 4.14.x series kernel? SELinux labelling is fine, when it's not almost all services fail, in this case it's just the disk-related services that are failing. If SELinux labelling was off I wouldn't be able to boot with any kernel.

2018-02-28 06:03:24 +0000 asked a question Is it possible to just use a LTS kernel on Fedora?

I do like Fedora, but for some bizarre reason the 4.15.x series kernel seems to not be working well with systemd for me. Several disk-related services (those that mount file systems based on the contents of /etc/fstab) fail and the system fails to boot. This isn't exclusive to Fedora either, for me Debian unstable, openSUSE Tumbleweed and Arch Linux face the same problem with the 4.15.x kernel, despite working fine on earlier kernels. Oddly systemd-free bleeding edge distros like Gentoo and Void do not seem to have this problem.

What I do know is that the 4.14.x and earlier series kernels work fine on Fedora. In fact I'm typing this from F27 running on the 4.14.18 kernel, so is there a way I can use the 4.14.x series kernels, while still getting bug fix and security updates (i.e. I don't want to just stay on 4.14.18 while new patches come out for the 4.14.x series) until this issue is resolved?

EDIT: To be clear, however, I only want the LTS of the kernel, everything else the standard Fedora versions. I thought this'd be obvious, but clearly I expect people to think more like me than they do.

2018-01-16 23:41:08 +0000 marked best answer Broadcom wireless issues with Fedora 27 wl.ko exists but not detected by modprobe

I just installed Fedora 27 and installed broadcom-wl from the rpmfusion non-free repository. I then upgraded my system's packages with sudo dnf update -y (as broadcom-wl was built for a newer kernel than the one I had freshly after installing F27) and rebooted. When I logged in I noticed something bizarre there was no WiFi icon in the menu at the top-right hand side of the screen (where Battery, volume, bluetooth, etc. options appear). Running nmtui (after installing NetworkManager-tui) also does not show any WiFi option either. I verified that broadcom-wl was installed with sudo dnf install -y broadcom-wl and it was. I then ran sudo modprobe -a wl and it returned:

modprobe: WARNING: Module wl not found in directory /lib/modules/4.13.11-300.fc27.x86_64


find /lib/modules/4.13.11-300.fc27.x86_64 -name "wl.ko"



so clearly something is up with the Broadcom wireless as the module exists but it's not detected by modprobe. Any ideas what the problem is?

2018-01-14 23:20:56 +0000 received badge  Popular Question (source)
2018-01-11 07:19:14 +0000 marked best answer How do I overcome these DNF and RPM errors?

I have an installation of Fedora 22 on a removable hard disk drive (HDD) (that I always run via USB link to the same computer, if you're wondering) that used to work fine until recently (keeping in mind while it's about a week ago, it is recent to me as I only use this OS once a week or fortnight or so) when I received errors described in this question. These errors appear to be gone now but now I am getting the following errors from running sudo dnf install terminator:

error: db5 error(-30973) from dbenv->open: BDB0087 DB_RUNRECOVERY: Fatal error, run database recovery
error: cannot open Packages index using db5 -  (-30973)
error: cannot open Packages database in /var/lib/rpm
warning: /var/cache/dnf/i386/22/fedora/packages/terminator-0.97-6.fc21.noarch.rpm: Header V3 RSA/SHA256 Signature, key ID 8e1431d5: NOKEY
error: db5 error(-30973) from dbenv->open: BDB0087 DB_RUNRECOVERY: Fatal error, run database recovery
error: cannot open Packages index using db5 -  (-30973)
error: cannot open Packages database in /var/lib/rpm
error: db5 error(-30973) from dbenv->open: BDB0087 DB_RUNRECOVERY: Fatal error, run database recovery
error: cannot open Packages index using db5 -  (-30973)
error: cannot open Packages database in /var/lib/rpm
Traceback (most recent call last):
  File "/bin/dnf", line 36, in <module>
    main.user_main(sys.argv[1:], exit_code=True)
  File "/usr/lib/python2.7/site-packages/dnf/cli/main.py", line 185, in user_main
    errcode = main(args)
  File "/usr/lib/python2.7/site-packages/dnf/cli/main.py", line 84, in main
    return _main(base, args)
  File "/usr/lib/python2.7/site-packages/dnf/cli/main.py", line 141, in _main
    ret = resolving(cli, base)
  File "/usr/lib/python2.7/site-packages/dnf/cli/main.py", line 160, in resolving
  File "/usr/lib/python2.7/site-packages/dnf/cli/cli.py", line 216, in do_transaction
  File "/usr/lib/python2.7/site-packages/dnf/cli/cli.py", line 249, in gpgsigcheck
    self.getKeyForPackage(po, fn)
  File "/usr/lib/python2.7/site-packages/dnf/base.py", line 1742, in getKeyForPackage
    if misc.keyInstalled(ts, info.rpm_id, info.timestamp) >= 0:
  File "/usr/lib/python2.7/site-packages/dnf/yum/misc.py", line 285, in keyInstalled
    for hdr in ts.dbMatch('name', 'gpg-pubkey'):
  File "/usr/lib/python2.7/site-packages/dnf/rpm/transaction.py", line 58, in dbMatch
    mi = self.ts.dbMatch(*args, **kwds)
_rpm.error: rpmdb open failed

Now these are just lines of output from this command that appeared to be errors to me (although they did all appear in exactly that order in the terminal, with no missing lines if you're wondering) the full output is VERY LONG and is here. If the output of the commands free -h and uname -r are still relevant (as was mentioned in the linked question) here they are:

free -h:

              total        used        free      shared  buff/cache   available
Mem:           3.7G        755M        2.2G        134M        820M        2.8G
Swap:          3.0G          0B        3.0G

uname -r:

2018-01-08 17:32:38 +0000 asked a question Guake: it starts, but pressing F12 doesn't reveal it all the time

I'm a fan of drop-down terminals but under GNOME on Fedora 27 (regularly updated with dnf update -y --refresh) guake starts when added to applications started on start-up, and when launched manually but a lot of the time pressing F12 does not reveal it. It is very irritating as only sometimes does F12 reveal it and the whole purpose of drop-down terminals is that you can use your keyboard alone to interact with it. The same problem seems to affect Yakuake too.

This is especially the case when all applications have been minimized, or when the only applications opened are not maximized. When an application is open and maximized (e.g. my web browser) F12 seems to work much more reliably.

EDIT: openSUSE Tumbleweed doesn't seem to have this problem. Guake starts even when all programs are minimized.

2018-01-05 12:11:16 +0000 asked a question Is there a way to replace the "Applications" text in Applications Menu with an icon?

I would like to replace the text "Applications" in GNOME's Application Menu extension with the Fedora logo. Specifically the "Applications" shown in the bottom left-hand corner of this screenshot (and yes I have significantly customized my GNOME 3 session):

image description

I have tried editing /usr/share/gnome-shell/extensions/app... to try and make this happen (and yes I am aware that with the next update of gnome-shell-extension-apps-menu my edits will likely be discarded), but alas it has failed (with no visible change). Specifically what I have tried is changing:

        this._label = new St.Label({  text: _("Applications"),
                                 y_expand: true,
                                 y_align: Clutter.ActorAlign.CENTER });


        this._label = new St.Label({ style_class: 'overview-icon',
                               //  text: _("Applications"),
                                 y_expand: true,
                                 y_align: Clutter.ActorAlign.CENTER });

As I thought this must dictate the look of the "Applications" button. The Fedora logo is defined as the .show-apps .overview-icon CSS style in the United Fedora Dark theme I am using (if you'd like to check whether I'm right you can get the theme here). While the full file is here. To test whether this works I have tried logging in and out of my GNOME session, after I saw no change after that I decided to reboot for good measure, still no change.

2017-12-31 16:00:57 +0000 received badge  Notable Question (source)
2017-12-31 16:00:57 +0000 received badge  Popular Question (source)
2017-12-25 13:19:41 +0000 asked a question Is it possible to install Fedora from a chroot in a way analogous to Arch Linux?

I would like to build a Fedora system from the ground up and I'm wondering if there's a lesser known chroot method for installing Fedora. If it is possible I think the most interesting part of it would be how to install a base system into which one can chroot. The only way I can come up with is getting the base file system from a Fedora netinstall ISO image.

2017-12-24 18:51:56 +0000 commented answer fedora on ZFS

None of these links give methods of installing Fedora on ZFS. They say how to install the ZFS tools on an existing Fedora installation but not how to install Fedora to an existing ZFS file system, for example. Mentioning how one may go about this sounds like a useful addition to your answer, although the way the question was phrased seems to suggest the opposite was wanted. Nonetheless sounds like a useful addition to your answer, it might even be what originalmax wanted.

2017-12-21 03:13:06 +0000 marked best answer Is it possible to get DNF to install packages listed in a text file?

I am writing a shell script for installing Moksha (the E17 fork used by Bodhi Linux) on Fedora 23 and it would help if I could get DNF to install packages listed in a plain text file (.txt).

Such a text file might be formatted like this, for example:

.  .   .

with each package name on a different line, if it makes any difference to DNF.

2017-12-21 03:13:01 +0000 received badge  Nice Question (source)
2017-12-20 09:32:57 +0000 commented answer Boot has become slow due to service failures

For some reason now the service is running right, I missed a few NVIDIA packages and I uninstalled them.

2017-12-20 06:42:36 +0000 answered a question Are there any major differences between fedora and red hat?

From what I understand Red Hat Enterprise Linux (RHEL) is based on Fedora with the main difference being it's designed for servers and other circumstances where stability is far more important than having the latest software. So its software is fairly outdated, like it still uses GCC 4.8.5 (from ca. mid 2015) when 7.2.0 is out and has been out for over a month. The reason for older software is that its undergoes more testing before entering their repositories.

It also isn't free of financial charge to the end-user like Fedora, releases are supported for longer and paid support is available from Red Hat.

2017-12-20 06:20:02 +0000 edited question Boot has become slow due to service failures

When I first installed Fedora 27 its boot wasn't blazing fast but it was acceptably fast at least. Now it's intolerably slow, at least partly because a dev-disk-by service times out (after 1 & 1/2 minutes). It seems to correspond to something I dare not remove from my /etc/fstab file: my swap partition. The precise service is:


My /etc/fstab file is here:

# /etc/fstab
# Created by anaconda on Sun Dec 17 01:24:32 2017
# Accessible filesystems, by reference, are maintained under '/dev/disk'
# See man pages fstab(5), findfs(8), mount(8) and/or blkid(8) for more info
UUID=c42b821c-e07d-4322-8e47-ff1d76a0fd75 /                       btrfs   subvol=root     0 0
UUID=b7e401ea-a7e5-4ed9-94c1-acf8f8ba5b83 /boot                   ext4    defaults        1 2
UUID=D796-7F56          /boot/efi               vfat    umask=0077,shortname=winnt 0 0
UUID=0a85c259-e28b-4fb7-aa9e-a5894701c6b4 swap                    swap    defaults        0 0
/dev/sdb1               /data        ext4        defaults  0 1
#/dev/sda3               /arch        btrfs       defaults  0 1
#/dev/sdb1               /arch        ext4        defaults  0 1

as you can see the UUID seems to line up with my swap partition. I don't really want to disable swap. So is there a way to fix this issue?

I also see kernel module load service failing when I start up. I think the kernels fail due to the fact I use the Broadcom DKMS (akmod-wl) package in RPMFusion nonfree as I need it to connect to the WiFi (my Broadcom chip is BCM4352).

If you're wondering how "unacceptable" the boot time is well systemd-analyze time gives 3 minutes and 12.5 seconds.

EDIT: I uninstalled akmods-wl (using broadcom-wl, as I think it's pre-built against the latest kernel, right?), NVIDIA and bbswitch kernel modules and still systemd-modules-load.service is failing. Any ideas?

2017-12-20 06:19:29 +0000 commented question Boot has become slow due to service failures

@sideburns Ya I know it's not systemd's fault but I wasn't sure if I could really call the device issue a service problem per se. But I suppose if you's call it one I'll rephrase the title.