[SOLVED] Some strange things happening on Slackware 14.0

raphaeldavidf · 09-29-2012, 01:32 AM

Hi guys,

I've just installed 14.0 and I started noticing some strange stuff(comparing to my 13.37). First of all my processor temperature is ,in average, 4 ºC hotter in idle (now is 50 ºC against 46 ºC in the 13.37), my processor is an Intel i5-2410M @ 2.3 GHz. The second strange thing is probably related to the first but I'll include it anyway, i'm felling that the system is responding slower compared to the 13.37. And the last strange thing is that some font files(like ConkySymbols.ttf) that i used before aren't working(when i try to open with font viewer all is blank!). I don't know if I screwed up something, but i tried to configure just like i did in the 13.37. Excluding those things I liked very much what Pat and the slack crew did in this new release!

Thanks!

Follow my ps -e command:

raphael@DeathStar:~$ ps -e
PID TTY TIME CMD
1 ? 00:00:00 init
2 ? 00:00:00 kthreadd
3 ? 00:00:00 ksoftirqd/0
6 ? 00:00:00 migration/0
7 ? 00:00:00 migration/1
9 ? 00:00:00 ksoftirqd/1
11 ? 00:00:00 migration/2
13 ? 00:00:00 ksoftirqd/2
14 ? 00:00:00 migration/3
16 ? 00:00:00 ksoftirqd/3
17 ? 00:00:00 cpuset
18 ? 00:00:00 khelper
19 ? 00:00:00 kdevtmpfs
20 ? 00:00:00 netns
227 ? 00:00:00 sync_supers
229 ? 00:00:00 bdi-default
231 ? 00:00:00 kblockd
368 ? 00:00:00 ata_sff
375 ? 00:00:00 khubd
381 ? 00:00:00 md
483 ? 00:00:00 rpciod
484 ? 00:00:03 kworker/2:1
533 ? 00:00:00 khungtaskd
539 ? 00:00:00 kswapd0
608 ? 00:00:00 khugepaged
610 ? 00:00:00 fsnotify_mark
614 ? 00:00:00 nfsiod
620 ? 00:00:00 crypto
664 ? 00:00:00 kthrotld
757 ? 00:00:00 kworker/1:1
758 ? 00:00:01 kworker/3:1
823 ? 00:00:00 scsi_eh_0
826 ? 00:00:00 scsi_eh_1
829 ? 00:00:00 scsi_eh_2
832 ? 00:00:00 scsi_eh_3
835 ? 00:00:00 scsi_eh_4
838 ? 00:00:00 scsi_eh_5
844 ? 00:00:00 kworker/u:5
1045 ? 00:00:00 devfreq_wq
1173 ? 00:00:00 jbd2/sda1-8
1174 ? 00:00:00 ext4-dio-unwrit
1228 ? 00:00:00 udevd
1323 ? 00:00:00 irq/46-mei
1325 ? 00:00:00 kpsmoused
1327 ? 00:00:00 cfg80211
1334 ? 00:00:00 hd-audio0
1364 ? 00:00:00 kworker/3:2
1416 ? 00:00:05 mount.ntfs-3g
1419 ? 00:00:00 mount.ntfs-3g
1443 ? 00:00:00 flush-8:0
1450 ? 00:00:00 syslogd
1454 ? 00:00:00 klogd
1521 ? 00:00:00 kworker/1:2
1667 ? 00:00:00 inetd
1674 ? 00:00:00 acpid
1687 ? 00:00:02 dbus-daemon
1690 ? 00:00:00 console-kit-dae
1758 ? 00:00:00 polkitd
1783 ? 00:00:00 crond
1786 ? 00:00:07 wicd
1787 ? 00:00:00 atd
1797 ? 00:00:00 udevd
1805 ? 00:00:00 udevd
1810 ? 00:00:00 gpm
1813 tty2 00:00:00 agetty
1814 tty3 00:00:00 agetty
1815 tty4 00:00:00 agetty
1816 tty5 00:00:00 agetty
1817 tty6 00:00:00 agetty
1820 ? 00:00:02 wicd-monitor
1911 ? 00:00:00 dhcpcd
2015 ? 00:00:00 udisksd
2092 ? 00:00:00 upowerd
7107 ? 00:00:00 kworker/2:0
7671 ? 00:00:00 kworker/0:0
10464 ? 00:00:00 udisks-daemon
10465 ? 00:00:00 udisks-daemon
10674 tty1 00:00:00 bash
10796 tty1 00:00:00 startx
10812 tty1 00:00:00 xinit
10813 tty7 00:01:51 X
10818 tty1 00:00:00 ck-launch-sessi
10826 tty1 00:00:00 sh
10829 tty1 00:00:00 dbus-launch
10830 ? 00:00:01 dbus-daemon
10836 tty1 00:00:00 xfce4-session
10838 ? 00:00:00 xfconfd
10841 ? 00:00:00 gpg-agent
10842 tty1 00:00:04 xfwm4
10844 tty1 00:00:01 Thunar
10846 ? 00:00:00 gvfsd
10848 ? 00:00:00 gvfs-fuse-daemo
10852 tty1 00:00:04 xfce4-panel
10856 ? 00:00:00 xfsettingsd
10857 tty1 00:00:00 xfdesktop
10859 ? 00:00:00 xfce4-power-man
10866 tty1 00:00:01 wicd-client
10868 tty1 00:00:00 panel-6-systray
10870 tty1 00:00:21 conky
10872 tty1 00:00:00 polkit-gnome-au
10874 tty1 00:00:00 panel-2-actions
10880 ? 00:00:00 xfce4-volumed
10889 ? 00:00:00 xfce4-power-man
10893 ? 00:00:00 gvfs-udisks2-vo
10897 tty1 00:00:21 conky
10906 ? 00:00:00 gvfs-gphoto2-vo
10928 ? 00:00:00 gvfsd-trash
11157 ? 00:00:00 gconfd-2
11200 ? 00:00:00 kdeinit4
11201 ? 00:00:00 klauncher
11203 ? 00:00:00 kded4
11205 ? 00:00:00 gam_server
11209 ? 00:00:00 knotify4
12336 tty1 00:02:34 firefox
13765 tty1 00:00:34 plugin-containe
17230 ? 00:00:00 kworker/0:2
18921 ? 00:00:00 kworker/u:1
20228 ? 00:00:00 kworker/u:0
21489 tty1 00:00:00 konsole
21491 pts/0 00:00:00 bash
21634 pts/0 00:00:00 ps
24970 ? 00:00:00 xfce4-appfinder

suttiwit · 09-29-2012, 03:57 AM

You might consider:
1. Trying another distribution that is based off slackware 14.
or if this problem still exists even on many other distributions based on slackware 14 that you have tried:
2. Reporting this bug to Slackware and to some distributions you have tried. (See: http://www.linuxquestions.org/questi...e-bugs-410562/)

raphaeldavidf · 09-29-2012, 04:03 AM

Quote:

Originally Posted by suttiwit

You might consider:
1. Trying another distribution that is based off slackware 14.
or if this problem still exists even on many other distributions based on slackware 14 that you have tried:
2. Reporting this bug to Slackware and to some distributions you have tried. (See: http://www.linuxquestions.org/questi...e-bugs-410562/)

I've tried another distribution, the comparison that I made was between the slackware 14.0(that i've just installed) and slackware 13.37(that i had installed before)

I don't know if maybe some kernel module that should be loaded isn't...

suttiwit · 09-29-2012, 04:20 AM

better go for slackware 13.37

vulcan59 · 09-29-2012, 04:36 AM

Have you tried running top and powertop to see if they show anything obvious? Maybe the machine isn't as idle as you think. Post the results here if you want some more opinions.

tallship · 09-29-2012, 05:05 AM

powertop is good, but htop is now part of Slackware mainline too so I would recommend using that over plain old top.

Also, are you running the 'huge' kernel or the 'generic' kernel?

Although huge will run just fine, it's recommended that you use one of the generic kernels in production machines (this will require the creation of an mkinitrd, a quick two or three line edit in lilo.conf, and then running lilo before rebooting).

Since there's a little bit of /.'ing going on right now because of the new release, you might want to look for the equivilant location on your CD/DVD: http://taper.alienbase.nl/mirrors/sl.../README.initrd

Make *sure* you choose the SMP kernel, and note that you don't have to actually change those lines, but you can add a section with each label in order to still boot under the old config and kernel if need be.

If you just add another section then you'll see both boot options when the splash page appears upon system boot.

At the very least, this will give you something to compare against between various installs of Slackware 14 itself.

I hope that helps!

Kindest regards,

.

raphaeldavidf · 09-29-2012, 05:35 AM

Quote:

Originally Posted by vulcan59

Have you tried running top and powertop to see if they show anything obvious? Maybe the machine isn't as idle as you think. Post the results here if you want some more opinions.

Quote:

Originally Posted by tallship

powertop is good, but htop is now part of Slackware mainline too so I would recommend using that over plain old top.

Also, are you running the 'huge' kernel or the 'generic' kernel?

Although huge will run just fine, it's recommended that you use one of the generic kernels in production machines (this will require the creation of an mkinitrd, a quick two or three line edit in lilo.conf, and then running lilo before rebooting).

Since there's a little bit of /.'ing going on right now because of the new release, you might want to look for the equivilant location on your CD/DVD: http://taper.alienbase.nl/mirrors/sl.../README.initrd

Make *sure* you choose the SMP kernel, and note that you don't have to actually change those lines, but you can add a section with each label in order to still boot under the old config and kernel if need be.

If you just add another section then you'll see both boot options when the splash page appears upon system boot.

At the very least, this will give you something to compare against between various installs of Slackware 14 itself.

I hope that helps!

Kindest regards,

.

So, my htop output is in the attachment.
I'm using the only generic kernel that was in my /boot folder after installation (changing kernel was the first thing i did, following all the steps from the README.initrd). I really don't know if my kernel is the SMP kernel (it was the only one in my /boot folder), how can I make sure? I'll add a label to the huge kernel and try to boot with this label to see if the temperature drops, as soon as i reboot i'll post the result =D

raphaeldavidf · 09-29-2012, 05:48 AM

Booting with the huge kernel didn't make any perceptible difference....

unSpawn · 09-29-2012, 06:12 AM

Quote:

Originally Posted by suttiwit

better go for slackware 13.37

Like I said before: please do not post if you have nothing constructive to contribute.

TobiSGD · 09-29-2012, 06:26 AM

If you don't have a SMP kernel in your /boot then you are running the 64 bit version (it is a good idea to always put information like that in the first posts

). The 64 bit kernels are all SMP, so don't worry about that.
Was your 13.37 install 64 bit also?
Is your power-saving working? Please post the output of

Code:

lsmod | grep cpufreq

Please use code-tags (the #-symbol above the editor) for any output of commands, that will make your posts more readable.

raphaeldavidf · 09-29-2012, 06:36 AM

Quote:

Originally Posted by TobiSGD

If you don't have a SMP kernel in your /boot then you are running the 64 bit version (it is a good idea to always put information like that in the first posts

). The 64 bit kernels are all SMP, so don't worry about that.
Was your 13.37 install 64 bit also?
Is your power-saving working? Please post the output of

Code:

lsmod | grep cpufreq

Please use code-tags (the #-symbol above the editor) for any output of commands, that will make your posts more readable.

Sorry about the lack of information, yes i'm using the 64 bit version and the 13.37 was 64 bit too. Sorry about the code too, i'm really new at the forum xD.

Code:

raphael@DeathStar:~ $ lsmod | grep cpufreq
cpufreq_ondemand        6252  4 
acpi_cpufreq            5773  1 
mperf                   1171  1 acpi_cpufreq
freq_table              2475  2 acpi_cpufreq,cpufreq_ondemand
processor              25592  1 acpi_cpufreq

TobiSGD · 09-29-2012, 06:54 AM

No need to apologize, we can't know everything.

So the output lets us know that your powersaving modules are loaded and that you use the ondemand-governor, which I would recommend.
So You should really have a look at htop to see if your machine is really idling.
You should also be aware that such differences in temperature may be caused by the newer kernel version or, if you use open source drivers for your video card, by a different video driver version.

raphaeldavidf · 09-29-2012, 08:12 AM

Quote:

Originally Posted by TobiSGD

No need to apologize, we can't know everything.

So the output lets us know that your powersaving modules are loaded and that you use the ondemand-governor, which I would recommend.
So You should really have a look at htop to see if your machine is really idling.
You should also be aware that such differences in temperature may be caused by the newer kernel version or, if you use open source drivers for your video card, by a different video driver version.

I'm monitoring the system in idle and I realized that there are some duplicated process(with the same command), they are using almost none CPU but just to make sure here they go:

Code:

root      1540     1  0 07:38 ?        00:00:00 /sbin/udevd --daemon
root      2135  1540  0 07:38 ?        00:00:00 /sbin/udevd --daemon
root      2136  1540  0 07:38 ?        00:00:00 /sbin/udevd --daemon

Code:

raphael   2328     1  0 07:39 ?     00:00:00 xfce4-volumed
raphael   2347     1  0 07:39 ?        00:00:00 xfce4-volumed

I'm using Xfce and I already had disabled the "save session" checkbox.

All you have said to me leads to only one way to go: check the modules, so I'm going to install the slackware64 13.37 in my other partition an see if there's any diference between the loaded modules. If that gives me any new information I'll post soon. Thanks to all for the great help and the patience!

raphaeldavidf · 09-30-2012, 11:45 PM

Hey guys, just reporting some findings here.

I've finally fixed the temperature problem! Turned out that the problem is the kernel's(actually any version higher than 2.6.37.6) power management systems. There was a major power regression in the vanilla kernels[1], that affects Sandy Bridge systems(my case, but didn't find any information claiming that all Sandy Bridge systems are afected), and then they no longer can manage the power usage from the cpu, making the temperature go considerably up. I've found some ways to fix it, but I'll just put the way that , after some tests, made my system run cooler(all the ways that I found are in the references). So let's go to the fix ;D. First off all run :

Code:

yes "" | sensors-detect

then:

Code:

vim /etc/lilo.conf

and put this in the append comand:

Code:

"acpi_osi=Linux apm=1 pcie_aspm=force i915.i915_enable_rc6=1 i915.i915_enable_fbc=1 i915.lvds_downclock=1 vt.default_utf8=0"

save and run

Code:

lilo

to apply the changes. Finally run:

Code:

vim /etc/rc.d/rc.local

and add this lines:

Code:

# Start cpu power management stuff
/sbin/modprobe coretemp
/sbin/modprobe cpufreq_conservative
/sbin/modprobe cpufreq_powersave
/usr/bin/sensors -s

now you just have to reboot and its done =D!

Thanks everybody for huge help!

References:
[1] http://www.phoronix.com/scan.php?pag...le_uffda&num=1
[2] http://www.phoronix.com/scan.php?pag...solution&num=1
[3] http://www.phoronix.com/scan.php?pag...tem&px=MTAxNDM
[4] http://crunchbanglinux.org/forums/topic/10817/why-drive-when-you-can-fly-liquorixzen-kernel/page/6/ (the post from user ej64)
[5] https://bbs.archlinux.org/viewtopic.php?id=136933 (the post from user BRAXS69)

willysr · 10-01-2012, 02:24 AM

You might want to write an article on Slackware Documentation Project about this issue

i'm sure some people will experience the same problem