Recent changes to bugs

g15macro cpu usage and mbutton lights

Anonymous — Fri, 30 Apr 2010 05:49:38 -0000

I used to have g15macro installed a while back when I noticed the CPU usage creeping up. I read on the G15Tools forums that inserting a usleep(5000) in the main thread solved the issue for one user so I did the same on mine. So far I haven't seen any usage increases, but I haven't been running it for too long. I'll check again in the morning.

Apart from that, I noticed that the M buttons didn't light up when switching and would only light up whenever I finished a recording. They would function just fine, only the lights would switch only when ending recordings. This also effected the MR button so it would never light up (by the time you end recording the light is off anyway).

I ran in debug mode and found that the record_cleanup() method always ran:

g15_send(g15screen_fd,(char *)canvas->buffer,G15_BUFFER_LEN);

before:

g15_send_cmd (g15screen_fd,G15DAEMON_MKEYLEDS,mled_state);

I added the same line to the handle_mkey_switch method and to occur on the MR keypress and it has solved the issue. Now my buttons light up as they are expected to.

Does someone want to implement this change or should I be able to commit my file? I'm attaching my g15macro.c file here for anyone to compile.

Kernel 2.6.33 + udev

CHaoSlayeR — Thu, 04 Mar 2010 06:53:21 -0000

Hi,

I'm currently using the mainline kernel 2.6.33 with Kubuntu 10.04 Alpha3 for testing (and because only with 2.6.33 my Radeon HD4770 [RV740] works correctly). As I installed g15daemon from the repositories (version 1.9.5.3-8) I noticed these in the /var/log/daemon.log:

Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:7
Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:10
Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:13
Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:16
Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:19
Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:7
Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:10
Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:13
Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:16
Mar 3 23:44:24 ladydeath udevd[2942]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /lib/udev/rules.d/40-g15daemon.rules:19

After the next reboot those messages also appear on the console just before X starts up. But instead of getting the clock display as the default the LCD panel is now empty, not even the G15 logo as it shows when g15daemon is not installed.

I don't really think that those are related because the messages just look like the usual deprecation warnings, but not like errors. Nevertheless I'll attach the full dmesg startup here.

Thanx

audacious plugin does not build

Götz Waschk — Fri, 08 Jan 2010 10:09:33 -0000

g15daemon-audacious 2.5.7 does not build against audacious 2.2:

g15daemon_audacious_spectrum.c:56:32: error: audacious/playlist.h: No such file or directory

The header file was removed in audacious 2.2. If you remove that include line, the plugin builds fine.

dmesg flood uhci_hcd release reserve

nitso — Fri, 30 Oct 2009 09:40:43 -0000

When I start g15daemon I get streaming messages in dmesg like:

Oct 30 12:26:55 localhost [ 904.756839] uhci_hcd 0000:00:1d.1: reserve dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 904.778737] uhci_hcd 0000:00:1d.1: release dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 904.818928] uhci_hcd 0000:00:1d.1: reserve dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 904.840736] uhci_hcd 0000:00:1d.1: release dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 904.880857] uhci_hcd 0000:00:1d.1: reserve dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 904.902744] uhci_hcd 0000:00:1d.1: release dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 904.942868] uhci_hcd 0000:00:1d.1: reserve dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 904.964746] uhci_hcd 0000:00:1d.1: release dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 905.005242] uhci_hcd 0000:00:1d.1: reserve dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 905.027756] uhci_hcd 0000:00:1d.1: release dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 905.067877] uhci_hcd 0000:00:1d.1: reserve dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 905.089760] uhci_hcd 0000:00:1d.1: release dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 905.130007] uhci_hcd 0000:00:1d.1: reserve dev 5 ep81-INT, period 1, phase 0, 36 us
Oct 30 12:26:55 localhost [ 905.152766] uhci_hcd 0000:00:1d.1: release dev 5 ep81-INT, period 1, phase 0, 36 us

LCD works properly, just problem with dmesg.

# g15daemon -v
G15Daemon version 1.9.5.3 - Not Running
compiled with libg15 version 1.201

# uname -a
Linux localhost 2.6.30-gentoo-r8 #1 SMP PREEMPT Fri Oct 30 01:46:15 MSK 2009 i686 Intel(R) Pentium(R) 4 CPU 3.00GHz GenuineIntel GNU/Linux

Wrong PID path

Ejis — Sun, 19 Jul 2009 09:54:00 -0000

Hello,

I am trying to package g15daemon for my distribution, but g15daemon won't start because of this error:

open(/usr/var/run/g15daemon.pid): No such file or directory
Unable to create PID File! Exiting

I'd wish to create the PID file in /var/run instead of /usr/var/run, and the "./configure" part won't seem to help.

How can I modify the PID path? Has it something to do with g15daemon/main.c?

No picture in screen on kernel 2.6.30

Anonymous — Wed, 17 Jun 2009 17:55:29 -0000

Can't get any picture to the screen using Arch linux with 2.6.30-ARCH kernel. Downgraded back to 2.6.29 and got picture back. Can't get any error messages.

"g15stats -i eth0" consumes 100% cpu if no network

Guy Dierx — Tue, 07 Apr 2009 19:13:17 -0000

"g15stats -i eth0" consumes 100% cpu if there is no network activity.

Reason (function void network_watch(void *iface) in g15stats.c):
If there is no network activity the if statement "if(previous_in+previous_out==0)" yields true and executes "goto last". At "last:" it restarts the while(1) without executing "sleep(1)", thus executing a busy wait.
while(1) {
goto last;
...
sleep (1);
last:
previous_in = netload.bytes_in;
previous_out = netload.bytes_out;
}

Fix:
change the sequence of these 4 lines into:
last:
previous_in = netload.bytes_in;
previous_out = netload.bytes_out;
sleep (1);

g15macro cpu usage increases over time

SteelSide — Fri, 30 Jan 2009 18:47:02 -0000

At first i got a cpu usage of g15macro that's at 1% or lower, but after time it starts to increase.
After 6 hours it's usually up at 10% (this with cpu throttled down to 1ghz all the time) for no apparent reason (have not recorded macros, have not hit any M/G keys at all).

last letter in macros is repeated endlessly

Jarl Stefansson — Fri, 29 Jun 2007 17:14:07 -0000

On Ubuntu Feisty Fawn with most recent stable (Fri Jun 29) the last keystroke in a macro is not released and as such the key press is repeated until you break it with a manual keystroke.

g15macro exits if run as system (from UDEV)

Jarl Stefansson — Fri, 29 Jun 2007 16:45:39 -0000

It's impossible to automatically run g15macro for hotplug events because it will for some reason exit right away unless launched by in a terminal by a user.

To install g15daemon to launch via udev automatically see:/usr/local/bin/g15daemon-hotplug
http://ubuntuforums.org/showpost.php?p=2933822&postcount=339

It would be very nice to have g15macro able run run from there with a simple DISPLAY:0.0 but alternatively create a shell script like /usr/local/bin/g15macro-hotplug with the contents:

#!/bin/bash
export DISPLAY=:0.0
/usr/bin/g15macro

and edit /usr/local/bin/g15daemon-hotplug and add the following after starting the daemon:

sleep 1
sudo g15macro-hotplug &