[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Slowness problem



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Camaleón <noelamac@gmail.com> writes:

> On Tue, 26 Jun 2012 17:28:26 +0200, Merciadri Luca wrote:
>
>> Camaleón <noelamac@gmail.com> writes:
>> 
>>> Top command output would be interesting :-)
>
>> Tasks: 203 total,   2 running, 201 sleeping,   0 stopped,   0 zombie
>> Cpu(s): 14.0%us,  1.9%sy,  0.0%ni, 84.1%id,  0.0%wa,  0.1%hi,  0.0%si, 
>> 0.0%st Mem:   4150284k total,  1712896k used,  2437388k free,   376712k
>> buffers Swap:  6080560k total,        0k used,  6080560k free,   523364k
>> cached
>> 
>>   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
>>  3749 root      20   0 67588  36m  11m S   57  0.9  13:02.61 Xorg
>> 27653 merciadr  20   0  299m  68m  26m S   21  1.7   0:08.76 firefox-bin
>
> (...)
>
>> with a brand new firefox (i.e. virgin profile) on youtube.com; if I
>> scroll really quickly, I get up to 106 % CPU usage by Xorg (with 0.9 %
>> MEM).
>> 
>> As I said earlier, this problem seems to happen with some other apps
>> too, e.g. the system monitor.
>
> Mmm... interesting. Note there's also Xorg there, taking too much of your
> CPU cycles. I've seen this before, Google will confirm:
>
> http://bit.ly/LNMAzg
>
> But if you say you experience the same problem with another applications, 
> this points to a component that is shared between them, and I can only
> think in Xorg and the VGA drivers. There's a additional "but" here and is
> that as you are using Lenny, I can guess that none of these two could have
> changed recently so the mistery still remains unless you installed "something"
> recently :-)
There is something I don't understand in this. Xorg takes way too much
CPU cycles, which might be a cause of the slowness I'm currently
encountering. However, what is the link with the VGA drivers? I know
Xorg is related to the graphical part of the system, but even if Xorg
was eating way too much CPU cycles, how damn could it result in such a
slowness? I'm having a 4-core config. with thus 4 CPUs at 2.66 Ghz!

I just
checked the Nvidia website (for my GeForce 8400 GS), and it looks like
they have absolutely no driver for me. However, many years ago, I
installed a proprietary driver from Nvidia for my graphic
card. Together with this driver, I installed nvidia-settings with
TwinView. This is the way I'm having a multi-screen
configuration. Here is my xorg.conf:

==
# nvidia-settings: X configuration file generated by nvidia-settings
# nvidia-settings:  version 1.0  (buildd@ninsei)  Tue Jun 24 18:11:48 UTC 2008

# nvidia-xconfig: X configuration file generated by nvidia-xconfig
# nvidia-xconfig:  version 1.0  (buildd@ninsei)  Fri Sep  5 22:23:08 UTC 2008

Section "ServerLayout"
    Identifier     "Layout0"
    Screen      0  "Screen0" 0 0
    InputDevice    "Keyboard0" "CoreKeyboard"
    InputDevice    "Mouse0" "CorePointer"
EndSection

Section "Files"
    RgbPath         "/usr/X11R6/lib/X11/rgb"
EndSection

Section "Module"
    Load           "dbe"
    Load           "extmod"
    Load           "type1"
    Load           "freetype"
    Load           "glx"
EndSection

Section "ServerFlags"
    Option         "Xinerama" "0"
EndSection

Section "InputDevice"

    # generated from default
    Identifier     "Mouse0"
    Driver         "mouse"
    Option         "Protocol" "auto"
    Option         "Device" "/dev/psaux"
    Option         "Emulate3Buttons" "no"
    Option         "ZAxisMapping" "4 5"
EndSection

Section "InputDevice"

    # generated from default
    Identifier     "Keyboard0"
    Driver         "kbd"
    Option	   "XkbRules"	"xorg"
    Option	   "XkbModel"	"pc105"
    Option	   "XkbLayout"	"be"
EndSection

Section "Monitor"
    Identifier     "Monitor0"
    VendorName     "Unknown"
    ModelName      "Samsung SyncMaster"
    HorizSync       30.0 - 81.0
    VertRefresh     56.0 - 75.0
    Option         "DPMS"
EndSection

Section "Device"
    Identifier     "Device0"
    Driver         "nvidia"
    VendorName     "NVIDIA Corporation"
EndSection

Section "Device"
    Identifier     "Videocard0"
    Driver         "nvidia"
    VendorName     "NVIDIA Corporation"
    BoardName      "GeForce 8400 GS"
EndSection

Section "Screen"
    Identifier     "Screen0"
    Device         "Videocard0"
    Monitor        "Monitor0"
    DefaultDepth    24
    Option         "TwinView" "1"
    Option         "TwinViewXineramaInfoOrder" "DFP-0"
    Option         "metamodes" "CRT: nvidia-auto-select +1680+0, DFP: nvidia-auto-select +0+0; CRT: 1280x1024 +0+0, DFP: NULL; CRT: 1280x1024 +1680+0, DFP: nvidia-auto-select +0+0"
    SubSection     "Display"
        Depth       24
    EndSubSection
EndSection
==

It looks like my xorg.conf file has not been changed since 2010,
according to the output of `ls -al' in /etc/X11.
>
>>> Disable the flash plugin at all (and so the same foe whatever
>>> additional version you can have installed, e.g., "gnash") and retry. It
>>> looks weird seeing Firefox to render slowly a usual page like Youtube,
>>> even more with flash player disabled :-?
>
>> I just tried, using the plug-in menu. This has no influence, and it
>> looks logical as the slowness problem is already encountered on the
>> welcome page, which does not really contain flash stuff (these are just
>> thumbnails)!
>
> Well, yes, the Firefox start page shouldn't be problematic _unless_ you are
> using the new feature that displays the latest visited sites in
> thumbnails ;-)
I was speaking about the Youtube home page, that is, some kind of list
of videos you might view.
>
>>> You can also run the same tests you are doing with Firefox but using a
>>> different browser to compare both results.
>
>> Yes. However, I think my problem is related to something other than
>> Firefox. What about the System monitor problem that I explained before?
>
> GNOME system monitor can become an intensive task, I mean, is not a good tool
> to use to compare with, but you can try to reproduce the CPU pikes launching 
> different applications (e.g., OOo writer, gedit, any Qt based app that you may
> have installed, a java tool...).
Ok.

- -- 
Merciadri Luca
See http://www.student.montefiore.ulg.ac.be/~merciadri/
- -- 

It is better to die on one's feet than live on one's knees.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.8 <http://mailcrypt.sourceforge.net/>

iEYEARECAAYFAk/qBTUACgkQM0LLzLt8MhzykACdH2CWTXM4JfxCbkKbAyCmh4qq
6hAAmweoaA6Fj8hr5d91B8MAegwJEEQ6
=XumG
-----END PGP SIGNATURE-----


Reply to: