Re: [arch-general] Conky autostart problem

[Ralf Mardorf]

On Fri, 1 May 2015 21:36:30 -0500, Francis Gerund wrote:
>So, do I need to create a "conky.desktop" file, and if so, how?

"Add a ~/.config/autostart/conky.desktop:

[Desktop Entry]
Encoding=UTF-8
Version=0.9.4
Type=Application
Name=conky
Comment=
Exec=conky -d
StartupNotify=false
Terminal=false
Hidden=false" - https://wiki.archlinux.org/index.php/Conky

"Integrate with Gnome 3

Some have experienced problems with conky showing up under Gnome 3.

Add these lines to ~/.conkyrc:

own_window yes
own_window_type conky
own_window_transparent yes
own_window_hints undecorated,below,sticky,skip_taskbar,skip_pager

If you still experience problems with transparency. You could add these
lines.

own_window_argb_visual yes
own_window_argb_value 255" -
https://wiki.archlinux.org/index.php/Conky#Integrate_with_Gnome_3

This might not solve your issue. I'm neither a GNOME, nor a conky user.
$ pacman -Qi conky
error: package 'conky' was not found

However, if you are used to GUIs and GNOME shouldn't provide GUIs were
you can add autostart.desktop files to ~/.config/autostart/ that aren't
listed, use the tool of another DE following freedesktop.org thingies,
e.g. xfce4-session-settings .

[arch-general] Conky autostart problem

[Francis Gerund]

Hello!

Using Arch with Gnome, I can start and run Conky manually.  But I can't get
it to autostart.

As stated in the Arch wiki Gnome article, Tweak tool will add autostarts,
but there is no entry for Conky in the list of application programs
accessed by using the " + " button.

It also says to "copy the relevant .desktop file from
/usr/share/applications/ to ~/.config/autostart" . . .
Well, even after re-installing Conky, ~/.config/autostart is empty, and
/usr/share/applications has no entry for Conky.  Also,
/usr/share/applications does not seem to have anything like a
"conky.desktop" file.

So, do I need to create a "conky.desktop" file, and if so, how?

P.S. I can't seem to find a good answer using a search engine.  And I
really don't want to (and should not have to) try downgrading (and then
upgrading again) Gnome, just to try something obsolete referred to as
"gnome-session-properties".

Re: [arch-general] at times when booting system-logind fails and system hangs

[Javier Vasquez]

> On Fri, May 1, 2015 at 4:11 AM, Mark Lee  wrote:
> ...
>
> This is unrelated, but does Arch Linux respect rootwait?

Linux does, :-)  I don't need on this laptop, but I have Arch also on
a USB disk, and for that one, at times linux doesn't find root, so I
found it just needed to wait for it to become available...

> I don't see any issues in there that'd cause your systemd issues. Did you try
> reinstalling systemd?

No I haven't.  Though I wouldn't suppose such re-install (if no new
version implied) would help.  I'll upgrade Arch to see.  The other
option is to tried the systemd patches, or the patched systemd offered
by Christian.  I haven't done that either...

> Regards,
> Mark

Thanks,

-- 
Javier

Re: [arch-general] PKGBUILD inheritance

[Mark Lee]

On Friday, May 01, 2015 01:27:58 PM Roman Rader wrote:
> I want to be in sync with official kernel config.
> Script is good idea but universal PKGBUILD based on core/linux will keep me
> from updating the package manually every time kernel is updated.

If you want to be in sync with the official arch kernel config, use git rebase 
in the PKGBUILD.

I'm not sure if it's a wise idea to have  PKGBUILD that's updating against 
kernel releases but not updating its pkgver variable. It would mean different 
under the hood kernels, and the pkgver variable only reflects the static 
changes you make on top of the kernel. In other words: 
-1.0 /= -1.0 
since it's pulling in the kernel config branch.

Regards,
Mark

signature.asc
Description: This is a digitally signed message part.

[arch-general] run boinc with optirun but no GPU found

[Stephen Zhang]

Hi all,

Today I installed boinc, and I'd like to run it with `optirun`.

The first time I run `optirun bonic_client --dir /var/lib/boinc`, it showed
me that `No usable GPUs found`. So, I installed `opencl-nvidia` package
from the repo. After `opencl-nvidia` installed, I restart the boinc with
optirun, it showed the same prompt. But after a suspend of my computer,
it worked fine! the prompt became like this:
,--
01-May-2015 11:36:53 [---] CUDA: NVIDIA GPU 0: GeForce GT 620M (driver version 
unknown, CUDA version 7.0, compute capability 2.1, 1024MB, 1009MB available, 
240 GFLOPS peak)
01-May-2015 11:36:53 [---] OpenCL: NVIDIA GPU 0: GeForce GT 620M (driver 
version 349.16, device version OpenCL 1.1 CUDA, 1024MB, 1009MB available, 240 
GFLOPS peak)
`--
And a project began to use GPU to caculate as well.

However, after I reboot my computer, the boinc can't find GPUs any more!
The prompt is always `No usable GPUs found`. I also tried to write a
simple CUDA program[1] to test if my CUDA device was ok, and the result was
the CUDA program worked fine, it printed 'Hello CUDA!' as expected.

My GPU model is NVIDIA GeForce GT 620M as mentioned above. My laptop
model is Acer E1-471G 32352G50Mnks.

I also have bbswitch installed, and `cat /proc/acpi/bbswitch` shows:
,--
:01:00.0 ON
`--

And here is the output of `lsmod |grep nvidia`:
,--
nvidia_uvm 73728  0
nvidia   8409088  1 nvidia_uvm
drm   282624  6 i915,drm_kms_helper,nvidia
i2c_core   53248  8 
drm,i915,i2c_i801,drm_kms_helper,i2c_algo_bit,nvidia,v4l2_common,videodev
`--

Anyone knows how I can run boinc with `optirun` correctly, or how I can
run boinc with my GPU?
Any reply will be appreciated!

[1] https://cfp.vim-cn.com/cbSR/c

-- 
Regards,
Stephen

Re: [arch-general] PKGBUILD inheritance

[Roman Rader]

I want to be in sync with official kernel config.
Script is good idea but universal PKGBUILD based on core/linux will keep me
from updating the package manually every time kernel is updated.

Пт, 1 трав. 2015 15:53 Maarten de Vries  пише:

​
>
> On 1 May 2015 at 14:13, Roman Rader  wrote:
>
> > Well yes, basically what I want to do is to download the PKGBUILD in the
> > prepare function and do makepkg inside my build function. I can then
> unpack
> > content of resulting package and put it to destination directory to pack
> it
> > again, but it's hacky and doesn't look nice. I'm looking for the proper
> way
> > to do it.
> >
>
> ​Indeed, that doesn't sound very nice. But running makepkg on the original
> package seems wasteful anyway​ if you're going to compile again in the
> build stage. It would seem more logical to use the same sources array plus
> a patch or script to modify some files and apply those in the prepare()
> function.
>
> I'm not certain what exactly you're trying to do though. If you only want
> to modify an upstream PKGBUILD it might be easier to just write a script to
> do what you do manually now.
>
> Regards,
> Maarten
>

Re: [arch-general] PKGBUILD inheritance

[Maarten de Vries]

​

On 1 May 2015 at 14:13, Roman Rader  wrote:

> Well yes, basically what I want to do is to download the PKGBUILD in the
> prepare function and do makepkg inside my build function. I can then unpack
> content of resulting package and put it to destination directory to pack it
> again, but it's hacky and doesn't look nice. I'm looking for the proper way
> to do it.
>

​Indeed, that doesn't sound very nice. But running makepkg on the original
package seems wasteful anyway​ if you're going to compile again in the
build stage. It would seem more logical to use the same sources array plus
a patch or script to modify some files and apply those in the prepare()
function.

I'm not certain what exactly you're trying to do though. If you only want
to modify an upstream PKGBUILD it might be easier to just write a script to
do what you do manually now.

Regards,
Maarten

Re: [arch-general] PKGBUILD inheritance

[Roman Rader]

Well yes, basically what I want to do is to download the PKGBUILD in the
prepare function and do makepkg inside my build function. I can then unpack
content of resulting package and put it to destination directory to pack it
again, but it's hacky and doesn't look nice. I'm looking for the proper way
to do it.

Thanks,
Roman

Пт, 1 трав. 2015 13:06 Maarten de Vries  пише:

> On 29 April 2015 at 22:48, Roman Rader  wrote:
>
> > Hi
> >
> > I want to make PKGBUILD for kernel in AUR with enabled options that I
> need.
> > I did
> > #ABSROOT=. abs core/linux
> > fixed config and I'm good, but to keep it updated with core/linux I need
> to
> > do it all the time (load latest core/linux, fix the config, rebuild). Is
> > there some tools/mechanisms to make a PKGBUILD that will do it
> > automatically?
> >
>
> ​PKGBUILDs are just bash scripts, so you can do pretty much anything in
> them.​ If you can automate updating the config, then you could stick that
> in the prepare() function of the PKGBUILD and you should be set I think.
>
> ​Regards,
> Maarten​
>

Re: [arch-general] PKGBUILD inheritance

[Roman Rader]

Well yes, basically what I want to do is to download the PKGBUILD in the
prepare function and do makepkg inside my build function. I can then unpack
content of resulting package and put it to destination directory to pack it
again, but it's hacky and doesn't look nice. I'm looking for the proper way
to do it.

Thanks,
Roman

Пт, 1 трав. 2015 13:06 Maarten de Vries  пише:

On 29 April 2015 at 22:48, Roman Rader  wrote:
>
> > Hi
> >
> > I want to make PKGBUILD for kernel in AUR with enabled options that I
> need.
> > I did
> > #ABSROOT=. abs core/linux
> > fixed config and I'm good, but to keep it updated with core/linux I need
> to
> > do it all the time (load latest core/linux, fix the config, rebuild). Is
> > there some tools/mechanisms to make a PKGBUILD that will do it
> > automatically?
> >
>
> ​PKGBUILDs are just bash scripts, so you can do pretty much anything in
> them.​ If you can automate updating the config, then you could stick that
> in the prepare() function of the PKGBUILD and you should be set I think.
>
> ​Regards,
> Maarten​
>

Re: [arch-general] at times when booting system-logind fails and system hangs

[nandan.reddev...@gmail.com]

I ran into the same problem this morning unfortunately :(

The relevant journalctl output (it is quite similar to Javier's)

= journalctl ===
May 01 12:50:13 archM systemd[1]: Listening on D-Bus System Message Bus
Socket.
May 01 12:50:13 archM systemd[1]: Starting D-Bus System Message Bus Socket.
...
May 01 12:50:13 archM systemd[1]: Starting Login Service...
...
May 01 12:50:42 archM dbus[286]: [system] Activating via systemd: service
name='org.freedesktop.PolicyKit1' unit='polkit.service'
May 01 12:50:42 archM systemd[1]: Unit systemd-logind.service entered
failed state.
May 01 12:50:42 archM systemd[1]: systemd-logind.service failed.
May 01 12:50:42 archM systemd[1]: systemd-logind.service has no holdoff
time, scheduling restart.
...
May 01 12:50:41 archM systemd-logind[279]: Failed to enable subscription:
Connection timed out
May 01 12:50:41 archM systemd-logind[279]: Failed to fully start up daemon:
Connection timed out
May 01 12:50:41 archM systemd[1]: Failed to register name: Connection timed
out
May 01 12:50:41 archM systemd[1]: Failed to set up API bus: Connection
timed out
May 01 12:50:41 archM systemd[1]: Starting D-Bus System Message Bus...
...
May 01 12:51:05 archM systemd[1]: Looping too fast. Throttling execution a
little.
May 01 12:51:07 archM systemd[1]: Looping too fast. Throttling execution a
little.
May 01 12:51:07 archM dbus[286]: [system] Failed to activate service
'org.freedesktop.PolicyKit1': timed out
May 01 12:51:07 archM ModemManager[275]:   failed to create PolicyKit
authority: 'Error initializing authority:
 Error calling StartServiceByName for org.freedesktop.PolicyKit1: Timeout
was reached'

May 01 12:51:07 archM NetworkManager[276]:  [1430464867.217039]
[nm-auth-manager.c:449] _dbus_new_proxy_cb():
auth: could not get polkit proxy: Error calling StartServiceByName for
org.freedesktop.PolicyKit1:
GDBus.Error:org.freedesktop.DBus.Error.TimedOut: Activation of
org.freedesktop.PolicyKit1 timed out

May 01 12:51:05 archM systemd[1]: Looping too fast. Throttling execution a
little.
May 01 12:51:07 archM systemd[1]: Looping too fast. Throttling execution a
little.
May 01 12:51:07 archM dbus[286]: [system] Failed to activate service
'org.freedesktop.PolicyKit1': timed out

May 01 12:51:08 archM systemd[1]: Looping too fast. Throttling execution a
little.
May 01 12:51:08 archM dbus[286]: [system] Failed to activate service
'fi.w1.wpa_supplicant1': timed out
...



Every line in /etc/systemd/logind.conf is commented out
---
$ grep '^[^#]' /etc/systemd/logind.conf
[Login]
---


For me, multiple (3-4 times) reboots did not fix it, the systemd-logind
service failed each time and I never saw any of the ttys or display manager
(sddm).
I was able to do some basic debugging and use pacman via the systemd debug
shell on tty9

As per the pacman logs systemd 219 was installed on 24th and I have booted
up the machine at least a couple of times since then but only hit the issue
this morning.

-
[2015-04-24 23:31] [ALPM] upgraded systemd (219-5 -> 219-6)
-


I am not sure whether it is related but the difference between booting the
machine yesterday and this morning, were these package upgrades from last
night.



[2015-05-01 01:12] [PACMAN] starting full system upgrade
[2015-05-01 01:12] [ALPM] transaction started
[2015-05-01 01:12] [ALPM] upgraded readline (6.3.006-1 -> 6.3.008-1)
[2015-05-01 01:12] [ALPM] upgraded kdelibs (4.14.7-1 -> 4.14.7-2)
[2015-05-01 01:12] [ALPM] upgraded kdebase-lib (15.04.0-2 -> 15.04.0-3)
[2015-05-01 01:13] [ALPM] upgraded kio (5.9.0-1 -> 5.9.0-2)
[2015-05-01 01:13] [ALPM] upgraded kdebase-dolphin (15.04.0-2 -> 15.04.0-3)
[2015-05-01 01:13] [ALPM] upgraded kdebase-kdialog (15.04.0-2 -> 15.04.0-3)
[2015-05-01 01:13] [ALPM] upgraded kdebase-keditbookmarks (15.04.0-2 ->
15.04.0-3)
[2015-05-01 01:13] [ALPM] upgraded kdebase-kfind (15.04.0-2 -> 15.04.0-3)
[2015-05-01 01:13] [ALPM] upgraded kdebase-konqueror (15.04.0-2 ->
15.04.0-3)
[2015-05-01 01:13] [ALPM] upgraded kdebase-konq-plugins (15.04.0-2 ->
15.04.0-3)
[2015-05-01 01:13] [ALPM] upgraded mlt (0.9.6-5 -> 0.9.6-6)
[2015-05-01 01:13] [ALPM] transaction completed




Using pacman via the debug shell I downgraded these packages and
reinstalled polkit, polkit-kde and polkit-kde-frameworks.

[2015-05-01 14:41] [ALPM] transaction started
[2015-05-01 14:41] [ALPM] downgraded readline (6.3.008-1 -> 6.3.006-1)
[2015-05-01 14:41] [ALPM] downgraded kdelibs (4.14.7-2 -> 4.14.7-1)
[2015-05-01 14:41] [ALPM] downgraded kdebase-lib (15.04.0-3 -> 15.04.0-2)
[2015-05-01 14:41] [ALPM] downgraded kio (5.9.0-2 -> 5.9.0-1)
[2015-05-01 14:41] [ALPM] downgraded kdebase-dolphin (15.04.0-3 ->
15.04.0-2)
[2015-05-01 14:41] [ALPM] downgraded kdebase-kdialog (15.04.0-3 ->
15.04.0-2)
[2015-05-01 14:41] [ALPM] downgraded kdebase-keditbookmarks (15.04.0-3 ->
15.04.0-2)
[2015-05-01 14:41] [ALPM] downgraded kdebase-k

Re: [arch-general] at times when booting system-logind fails and system hangs

[Mark Lee]

On Thursday, April 30, 2015 11:29:18 AM Javier Vasquez wrote:

>   echo  'Loading Linux linux ...'
>   linux /vmlinuz-linux root=UUID= rw rootwait
> rootfstype=ext4 resume=UUID=
This is unrelated, but does Arch Linux respect rootwait?
>   echo  'Loading initial ramdisk ...'
>   initrd   /intel-ucode.img /initramfs-linux.img
> 
> My hooks line in /etc/mkinitcpio.conf:
> 
> HOOKS="base udev autodetect modconf block resume filesystems keyboard fsck"

I don't see any issues in there that'd cause your systemd issues. Did you try 
reinstalling systemd?

Regards,
Mark

signature.asc
Description: This is a digitally signed message part.

Re: [arch-general] PKGBUILD inheritance

[Maarten de Vries]

On 29 April 2015 at 22:48, Roman Rader  wrote:

> Hi
>
> I want to make PKGBUILD for kernel in AUR with enabled options that I need.
> I did
> #ABSROOT=. abs core/linux
> fixed config and I'm good, but to keep it updated with core/linux I need to
> do it all the time (load latest core/linux, fix the config, rebuild). Is
> there some tools/mechanisms to make a PKGBUILD that will do it
> automatically?
>

​PKGBUILDs are just bash scripts, so you can do pretty much anything in
them.​ If you can automate updating the config, then you could stick that
in the prepare() function of the PKGBUILD and you should be set I think.

​Regards,
Maarten​

Re: [arch-general] Audacity Library Mismatch (Wxgtk2.8)

[Ralf Mardorf]

On Fri, 01 May 2015 05:18:04 -0400, Mark Lee wrote:
>On Friday, May 01, 2015 11:07:39 AM Ralf Mardorf wrote:
>> IIUC the patch let a fatal error become a warning. Everything behind
>> - is removed and + is added. Isn't this already a workaround for an
>> issue caused by upstream?
>
>Sort of. According to the Arch commit, it's an issue caused with
>compiling with GCC 5.1 (we're on 4.9.2). It's a bit of future proofing.

Thank you for the explanation Mark,

$ pacman -Q gcc-multilib
gcc-multilib 4.9.2-4

and in testing there already is 5.1.0.

Regards,
Ralf

Re: [arch-general] Audacity Library Mismatch (Wxgtk2.8)

[Mark Lee]

On Friday, May 01, 2015 11:07:39 AM Ralf Mardorf wrote:
> IIUC the patch let a fatal error become a warning. Everything behind -
> is removed and + is added. Isn't this already a workaround for an issue
> caused by upstream?

Sort of. According to the Arch commit, it's an issue caused with compiling 
with GCC 5.1 (we're on 4.9.2). It's a bit of future proofing.
The Fedora link indicates it's an ABI issue : 

From devel list:

That is a WxGTK bug.  __GXX_ABI_VERSION can change, but usually the result
is still ABI compatible, g++ emits just some aliases when mangling has
changed.


Regards,
Mark

signature.asc
Description: This is a digitally signed message part.

Re: [arch-general] Audacity Library Mismatch (Wxgtk2.8)

[Ralf Mardorf]

On Fri, 01 May 2015 04:40:28 -0400, Mark Lee wrote:
>On Friday, May 01, 2015 10:32:05 AM Ralf Mardorf wrote:
>> Actually the warning says that "The library used 2.8 [snip] and your
>> program used 2.8" too.
>> 
>> So you could file a bug against upstream, it isn't an issue caused by
>> the package.
>
>It's caused by the package:
>https://projects.archlinux.org/svntogit/packages.git/plain/trunk/make-abicheck-non-fatal.patch?h=packages/wxgtk2.8

$ cat /var/abs/extra/wxgtk2.8/make-abicheck-non-fatal.patch 
[snip]
-wxLogFatalError(msg.c_str());
[snip]
+wxLogWarning(msg.c_str());
[snip]
 
[rocketmouse@archlinux ~]$ grep pkgrel /var/abs/extra/wxgtk2.8/PKGBUILD 
pkgrel=3
[rocketmouse@archlinux ~]$ pacman -Q wxgtk2.8
wxgtk2.8 2.8.12.1-3

IIUC the patch let a fatal error become a warning. Everything behind -
is removed and + is added. Isn't this already a workaround for an issue
caused by upstream?

Re: [arch-general] Audacity Library Mismatch (Wxgtk2.8)

[Grady Martin]

On 2015年05月01日 08時36分, Geoff wrote:
> I don't make much use of audacity, and have not tested extensively.  Starting
> from an xterm on my i686, fully patched, system, I see the error, but the
> program runs and appears to function normally.  I see a lot of other errors in
> the xterm - but they don't appear to be wxgtk related, and they may well have
> been there and unnoticed before.

Thank you for checking.  You gave me just enough self-doubt to try a few more 
things, and I ended up finding an audacity instance that had failed to close.  
Audacity seems to limit itself to a single instance, which created the illusion 
of immediately exiting.

Errors notwithstanding, I consider the issue solved.

Re: [arch-general] Audacity Library Mismatch (Wxgtk2.8)

[Mark Lee]

On Friday, May 01, 2015 04:40:28 AM you wrote:
> It's caused by the package:
> https://projects.archlinux.org/svntogit/packages.git/plain/trunk/make-abiche
> ck-non-fatal.patch?h=packages/wxgtk2.8

I should clarify here. Arch packagers included that patch because there are 
issues building with gcc 5.1. See :
 

Regards,
Mark


signature.asc
Description: This is a digitally signed message part.

Re: [arch-general] Audacity Library Mismatch (Wxgtk2.8)

[Mark Lee]

On Friday, May 01, 2015 10:32:05 AM Ralf Mardorf wrote:
> Actually the warning says that "The library used 2.8 [snip] and your
> program used 2.8" too.
> 
> So you could file a bug against upstream, it isn't an issue caused by
> the package.

It's caused by the package:
https://projects.archlinux.org/svntogit/packages.git/plain/trunk/make-abicheck-non-fatal.patch?h=packages/wxgtk2.8

Regards,
Mark

signature.asc
Description: This is a digitally signed message part.

Re: [arch-general] Audacity Library Mismatch (Wxgtk2.8)

[Ralf Mardorf]

On Fri, 2015-05-01 at 09:58 +0200, Ralf Mardorf wrote:
> You could try to build from ABS, perhaps you'll get rid of a few
> messages.
> 
> $ cp -r /var/abs/extra/audacity/ .
> $ cd audacity/
> 
> Perhaps editing the package release:
> $ nano PKGBUILD 
> $ grep pkgrel PKGBUILD 
> pkgrel=2
> 
> $ makepkg -s
> 
> Just for fun I'm building it at the moment.

I didn't compare all messages, but I guess I didn't get rid of even
one message.

$ audacity 
Warning: Mismatch between the program and library build versions
detected.
The library used 2.8 (no debug,Unicode,compiler with C++ ABI 1008,wx
containers,compatible with 2.6),
and your program used 2.8 (no debug,Unicode,compiler with C++ ABI
1002,wx containers,compatible with 2.6).
[snip]
$ pacman -Qi audacity
Name   : audacity
Version: 2.1.0-2
[snip]
Build Date : Fri 01 May 2015 10:12:14 AM CEST
Install Date   : Fri 01 May 2015 10:16:56 AM CEST
[snip]

It was build on my machine ;).

Actually the warning says that "The library used 2.8 [snip] and your
program used 2.8" too.

So you could file a bug against upstream, it isn't an issue caused by
the package.

Regards,
Ralf

Re: [arch-general] Audacity Library Mismatch (Wxgtk2.8)

[Ralf Mardorf]

Jackd users usually considers Audacity as crappy software.
However, 2.1.0-1 starts on x86_64 too.

$ audacity 
Warning: Mismatch between the program and library build versions
detected. The library used 2.8 (no debug,Unicode,compiler with C++ ABI
1008,wx containers,compatible with 2.6), and your program used 2.8 (no
debug,Unicode,compiler with C++ ABI 1002,wx containers,compatible with
2.6).
[snip]

You could try to build from ABS, perhaps you'll get rid of a few
messages.

$ cp -r /var/abs/extra/audacity/ .
$ cd audacity/

Perhaps editing the package release:
$ nano PKGBUILD 
$ grep pkgrel PKGBUILD 
pkgrel=2

$ makepkg -s

Just for fun I'm building it at the moment.

Re: [arch-general] Audacity Library Mismatch (Wxgtk2.8)

[Geoff]

I don't make much use of audacity, and have not tested extensively.  Starting
from an xterm on my i686, fully patched, system, I see the error, but the
program runs and appears to function normally.  I see a lot of other errors in
the xterm - but they don't appear to be wxgtk related, and they may well have
been there and unnoticed before.