USB issues in fedora-27 ("can't get device qualifier: Resource
temporarily unavailable") not happening when booted into CentOS-6.2, same hardware
by Tom Hodder
Hi All,
I've been having a bunch of trouble with the USB ports on my desktop, in
particular the webcam doesn't work and there are lots of errors in the
output <https://paste.fedoraproject.org/paste/qYxQ2rH-tlsE8pJqWz442w>
of *"lsusb
-v"* when I try to debug. (I get the "can't get device qualifier: Resource
temporarily unavailable" errors, whether the webcam is plugged in or not,
which seems to eliminate the webcam itself as the source of the problem)
On the positive side, a usb keyboard, mouse and USB sticks seem to work ok.
The hardware is a Supermicro SuperWorkstation 7037A-i
<https://www.supermicro.com/products/system/tower/7037/sys-7037a-i.cfm>,
with an X9dai motherboard
<https://www.supermicro.com/products/motherboard/xeon/c600/x9dai.cfm>.
I assumed the problem was hardware related and raised a ticket with
supermicro. They asked me to test the machine using a supported OS, and the
nearest to fedora-27 was CentOS 6.2 (compatibility matrix
<https://www.supermicro.com/support/resources/OS/C602_listing2.cfm>). So
today I booted into Centos-6.2. To my surprise there were no errors in
the "lsusb
-v" output <https://paste.fedoraproject.org/paste/yVFlRYGi9m4ZCt9PZPF3Gg>,
no errors in dmesg or syslog, and the webcam worked fine.
This seems to suggest it's not a hardware problem, but some difference in
the drivers. I am pretty sure the usb/webcam used to work, but that's going
back a few versions of fedora, and probably 3/4 years.
Any ideas on what might be causing this?
Many Thanks,
Tom
Various debugging output;
*Fedora-27*
lsusb -v
https://paste.fedoraproject.org/paste/qYxQ2rH-tlsE8pJqWz442w
dmidecode
https://paste.fedoraproject.org/paste/gW9D10O0fy6-p58I-YgJOQ
dmesg
https://paste.fedoraproject.org/paste/hnVj1hyjDpyYtiInxPGZ3Q
lsmod
https://paste.fedoraproject.org/paste/bZTVW-7P8mD19cLO6nogww
*CentOS-6.2*
lsusb -v
https://paste.fedoraproject.org/paste/yVFlRYGi9m4ZCt9PZPF3Gg
dmesg output
https://paste.fedoraproject.org/paste/TTyqm9zDrf1m4kID9rl09Q
dmidecode
https://paste.fedoraproject.org/paste/FfafH1-KdNzy5IA17oD9mg
lsmod
https://paste.fedoraproject.org/paste/~orePKeYsJRlBRmP1P2vvw
Plugging in the webcam produces this;
Mar 30 02:19:24 pc03.config kernel: usb 2-1.7: new high-speed USB device
number 8 using ehci-pci
Mar 30 02:19:25 pc03.config kernel: usb 2-1.7: New USB device found,
idVendor=045e, idProduct=0294
Mar 30 02:19:25 pc03.config kernel: usb 2-1.7: New USB device strings:
Mfr=1, Product=2, SerialNumber=3
Mar 30 02:19:25 pc03.config kernel: usb 2-1.7: Product: Video Camera
Mar 30 02:19:25 pc03.config kernel: usb 2-1.7: Manufacturer: Microsoft
Mar 30 02:19:25 pc03.config kernel: usb 2-1.7: SerialNumber:
000F330333475838
Mar 30 02:19:25 pc03.config kernel: uvcvideo: Found UVC 1.00 device Video
Camera (045e:0294)
Mar 30 02:19:26 pc03.config kernel: uvcvideo 2-1.7:1.0: Entity type for
entity Extension 3 was not initialized!
Mar 30 02:19:26 pc03.config kernel: uvcvideo 2-1.7:1.0: Entity type for
entity Processing 2 was not initialized!
Mar 30 02:19:26 pc03.config kernel: uvcvideo 2-1.7:1.0: Entity type for
entity Camera 1 was not initialized!
...
Mar 30 02:19:19 pc03.config kernel: uvcvideo: Failed to set UVC probe
control : -71 (exp. 26).
Mar 30 02:19:19 pc03.config kernel: uvcvideo: Failed to set UVC probe
control : -71 (exp. 26).
Mar 30 02:19:19 pc03.config kernel: uvcvideo: Failed to set UVC probe
control : -71 (exp. 26).
Mar 30 02:19:19 pc03.config kernel: uvcvideo: Failed to set UVC probe
control : -71 (exp. 26).
5 years, 12 months
easiest way to encrypt existing home dir?
by Neal Becker
I'm running f27 with everything on one drive. /home is 1 btrfs subvolume
and / is another btrfs subvolume.
I want to move my /home to encryption. One possibility seems to be to use
ecryptfs. I tried creating a test user, and the following:
https://cloud-ninja.org/2014/04/05/fedora-encrypting-your-home-directory/
(following "easy way" instructions).
This didn't do anything AFAICT, the test user's files still appear to be
unencrypted, and I don't see any ecryptfs mount.
Any suggestions? Is ecryptfs the way to go here or something else?
Thanks,
Neal
5 years, 12 months
rpm says package isn't installed but it is, and related question
by Chris Murphy
[chris@f27h ~]$ rpm -q kernel-core
kernel-core-4.16.0-0.rc6.git2.1.fc29.x86_64
kernel-core-4.15.12-300.fc27.x86_64
kernel-core-4.16.0-0.rc7.git1.1.fc29.x86_64
kernel-core-4.15.14-300.fc27.x86_64
[chris@f27h ~]$ sudo rpm -q -i kernel-core-4.15.14-300.fc27.x86_64.rpm
Name : kernel-core
Version : 4.15.14
Release : 300.fc27
Architecture: x86_64
Install Date: (not installed)
Group : System Environment/Kernel
Size : 60650235
License : GPLv2 and Redistributable, no modification permitted
Signature : (none)
Source RPM : kernel-4.15.14-300.fc27.src.rpm
Build Date : Thu 29 Mar 2018 11:02:25 AM MDT
Build Host : bkernel02.phx2.fedoraproject.org
Relocations : (not relocatable)
Packager : Fedora Project
Vendor : Fedora Project
URL : http://www.kernel.org/
Bug URL : https://bugz.fedoraproject.org/kernel
Summary : The Linux kernel
Description :
The kernel package contains the Linux kernel (vmlinuz), the core of any
Linux operating system. The kernel handles the basic functions
of the operating system: memory allocation, process allocation, device
input and output, etc.
[chris@f27h Downloads]$
This specific rpm file I'm getting info on, was installed yesterday.
But this query says it's not installed. In fact all RPMs I've
downloaded and installed, when -qi is run on them, say they're not
installed. Wha wha waht?
Also, is there any way to know whether and what kind of compression is
used in an RPM? I'm guessing the Size value of 60M above is installed
size. The package size is 24M so there must be some kind of
compression being used. But 'man rpm | grep -i compress' yields no
results.
--
Chris Murphy
5 years, 12 months
`plymouth show-splash` is not working
by Basix
Hello, I'm running f27 with kernel 4.15.12-301.fc27.x86_64.
I tried to run `plymouth show-splash` as root, but it dosen't displayed
anything and just returned 1. I tried `sudo plymouth --show-splash` and
others but it isn't working. Any solutions?
5 years, 12 months
Re: Node.js 6.14.1, 8.11.1 and 9.10.1 will *not* be built for Fedora
or EPEL
by Stephen Gallagher
On Fri, Mar 30, 2018 at 8:41 AM Robert Van Voorhees <rcvanvo(a)gmail.com>
wrote:
> While on the topic of Node.js versions, do we plan on moving Fedora 28 to
> Node 10, since their release dates coincide, or will Node 10 be pushed
> until Fedora 29?
>
> For those not aware, Node 10 is the next LTS, however it's LTS support
> does not start until October even though the release is slated for April:
> https://github.com/nodejs/Release
>
>
My plan is for Fedora 28 traditional RPM repositories to ship Node.js 8
because I don't like changing defaults after Beta release. However, as
Fedora 28 is now the first release with support for Fedora Modularity, you
may have noticed that it is also providing a non-default module for Node.js
6.x and 9.x that one can switch to. So what I intend to do is package up
Node.js 10.x as quickly as possible and make that a non-default module in
F28 and the default for F29.
So, you'll be able to use it in Fedora 28 by doing `dnf module install
nodejs:10` if you want to. Also available as a module will be 6.x and 9.x
if you really want or need them, though they won't carry updates past
upstream's EOL, which is in April 2019 for 6.x and June 2018 for 9.x).
6 years
Node.js 6.14.1, 8.11.1 and 9.10.1 will *not* be built for Fedora or EPEL
by Stephen Gallagher
Just a general notice to anyone using Node.js in Fedora. We're not going to
be rebuilding for the releases in the Subject because there's no need. The
X.Y.1 releases were created because of a compilation mistake made by the
upstream binary release process and contain no additional commits. Since
Fedora and EPEL rebuild all of the sources in our own build environment,
there is no need to issue an update at this time.
6 years
Sharing USB Serial data with 2 processes ?
by linux guy
Hi.
I've got a USB serial device that supplies data to a process via
/dev/ttyUSB0.
I need to use the same data for a 2nd, independent process. I have code
for both processes and can make changes, but I would prefer not to.
Is there a way that both processes can use the same serial USB data stream ?
Thanks.
PS: Is there a better place to ask this question ?
6 years
HYPERKITTY does not show messages parts being addressed in replies.
by home user
Good afternoon,
In the recent thread "alternative to skype", one member commented
"Don't quote the entire email. In fact, don't quote anything at all when you reply. On k-9 mail there's a little x button (in the default config) that I can click to get rid of all the quotes text."
Another responded on March 25:
"No, do quote the part you are commenting on. It makes no sense to omit this. Of course it's all there in the archives but making your readers open a browser just to get the context of what you are talking about is poor practice. I note that posts via HyperKitty seem to do this but don't know if that's the way it works by default. If so it's a bug and should be fixed."
I myself have experienced what the second member said, and I agree with him. List members respond to my posts in parts, but I have trouble knowing what specific part of my post the member is responding to.
I agree. I submitted fedora-infrastructure issue #6802 "HYPERKITTY does not show what posting text is being addressed in a reply." to address this. It can be viewed here:
"https://pagure.io/fedora-infrastructure/issue/6802".
Bill.
6 years
Killing only selected Chrome windows from the CLI ?
by Philip Rhoades
People,
This command exits the chrome process tree gracefully, in all window
managers:
pkill --oldest chrome
but it is not what I want to do - when Chrome goes mad and the whole
system starts grinding to a halt (I know from experience it is Chrome) I
want to try and work out what specific Chrome tab or window is the
problem. Chrome has its own task manager:
SHIFT ESC
but you can't kill a window with all its tabs from there.
Using:
ps aux | grep "beta\/chrome"
seems to indicate individual tabs?
Using:
xprop _NET_WM_PID
and clicking on the different windows always give the same id.
I wrote the script below using wmctrl and xkill to allow me to kill the
Chrome windows in reverse order of creation but killing the first Chrome
window kills all Chrome windows. The xkill man page says the "program
is very dangerous" and:
-id resource
This option specifies the X identifier for the resource whose creator is
to be aborted.
- so it looks like aborting "creator" gets rid of everything.
Any suggestions about fixing my script to selectively kill single Chrome
windows?
Thanks,
Phil.
#!/bin/bash
wins=`wmctrl -l | sort -r`
IFS=$'\n' # bash 4
readarray -t winsarr <<< "$wins"
for win in "${winsarr[@]}"
do
IFS=' ' read id junk1 junk2 name <<< $win
echo "$id $name"
echo -n "Kill?: "
read junk
if [ "$junk" == "Y" ]; then
xkill -frame -id "$id"
fi
done
--
Philip Rhoades
PO Box 896
Cowra NSW 2794
Australia
E-mail: phil(a)pricom.com.au
6 years
F27 boots to grub> after (failed?) dnf update ?
by linux guy
I ran DNF update via ssh on my x86_64 FC27 server the other day. For some
reason it didn't finish and when I rebooted it, it boots to a grub prompt.
"Minimal BASH-like line editing..." grub>
So...
grub> set root=(lvm/fedora/root)
grub> linuxefi (hd0,gpt2)/linuz-4.14.13-300.fc27.x86_64
grub>initrdefi (hd0,gpt2)/initramfs-4.14.13-300.fc27.x86_64.img
grub>boot
This gives me a switch root error and stops at the emergency mode prompt #.
How do I get my server to boot again ?
Thanks
6 years