8:06 a.m.
Hi,
I've noticed that there is a some kind of "delay" in the screen refresh.
For instance, under VIM when I move the cursor (that is a black
rectangle) under a character, it takes between 1 and 4 secs before
the character appears inside the cursor.
I've experienced the same issue under OpenOffice and FireFox.
This is really annoying. :(
Under F12 no problem
Does any other experienced this behaviour?
My system is:
* OS: F13 x86_64 - up-to-dated
* WM: OpenBox
* CPU: Intel Core 2 Duo
* GPU: Intel GM965/GL960
* RAM: 2GB
Thank you so much!
-- Marco
4:24 a.m.
On Mon, May 31, 2010 at 10:51 AM, Mike Fleetwood
<mike.fleetwood(a)googlemail.com> wrote:
[... cut ...]
Hi Marco,
I'm no expert but it sound to me like you have a latency issue. For
some reason your machine is sometimes having to wait 1 to 4 seconds
before it can give a program (vim, bash, etc.) its CPU time slice,
hence the apparent delay.
I have seen this kind of thing before when the kernel has to re-try
disk I/O multiple times on a failing hard drive.
Suggestions:
1) Play music! If it pauses when your app pauses it is an OS wide
issue effecting everything.
I will try tonight at home
2) Run top in another window. When top refreshes after the pause
does
any process jump to the top?
No process seems to have a great impact on CPU usage.
These below are the more time consuming processes:
1563 root 20 0 172m 42m 11m S 3.6 2.1 4:39.59 Xorg
2751 marco 20 0 942m 173m 30m S 3.6 8.7 4:32.69 firefox
2790 marco 20 0 419m 19m 10m S 2.0 1.0 2:11.23 npviewer.bin
After refresh, Xorg goes from 1.0 to 2.5 % of CPU. This percentage
increase to ~16% when I switch between wm virtual desktops.
3) Check /var/log/messages for any errors.
Done, but no meaningful information (or maybe I don't know what to look for).
Vague ideas:
4) Google for investigating latency issues.
5) With modern kernels I guess it should be possible to dynamically
trace it to find the lateny issue, but I have no idea how.
Sorry I can't be more help,
Mike
Thank you very much for your help.
Hope someone other can help me :)
In addition to system information provided in previous email I add
that my notebook is a DELL Latitude D830
Cheers,
-- Marco
1:26 a.m.
On Mon, May 31, 2010 at 10:07 PM, Marco Guazzone
<marco.guazzone(a)gmail.com> wrote:
On Mon, May 31, 2010 at 5:29 PM, Marco Guazzone
[cut]
The problem appears also under GNOME.
I've tried with xterm, ViM, and OpenOffice.
For instance, with oowriter the "latency" problem happens during
tooltip drawing (e.g., go to File -> Recent Documents and then move
the mouse up and down along the list of open documents slow enough
(but not too slow) in order to let the tooltipo appears; in this case,
I can note a delay which make appear an empty cyan rectangle and after
a few the tooltip text)
:(((
FYK
Finally, it seems that some of the recent updates solved this annoying problem.
Cheers,
-- Marco
Cheers,
-- Marco
3:07 p.m.
On Mon, May 31, 2010 at 5:29 PM, Marco Guazzone
<marco.guazzone(a)gmail.com> wrote:
On Mon, May 31, 2010 at 4:43 PM, Genes MailLists
<lists(a)sapience.com> wrote:
> On 05/31/2010 09:58 AM, Marco Guazzone wrote:
>>
>> Here below is the information you have asked
>>
>> But before you dig into it I want to say that it seem a problem of the
>> OpenBox window manager.
>> I say so since I've tried to use GNOME and apparently the problem
>> seems to be disappeared (for instance under ViM there is no delay in
>> showing characters under the cursor).
>
> Ahah ... yes indeed sounds like openbox is busted for some reason ...
> good catch.
>
> I found no obvious issues with mem pressure etc ... file a bug def ...
> I know nothing of openbox (honestly had not even heard of it till this
> email!)
>
:)
It's a light-weight window manager and you can try It with LXDE Fedora
spin: http://spins.fedoraproject.org/lxde/
> Good luck.
>
> Glad you found the source of the problem ... and can use gnome as a
> work around.
I speak too quickly! :(
The problem appears also under GNOME.
I've tried with xterm, ViM, and OpenOffice.
For instance, with oowriter the "latency" problem happens during
tooltip drawing (e.g., go to File -> Recent Documents and then move
the mouse up and down along the list of open documents slow enough
(but not too slow) in order to let the tooltipo appears; in this case,
I can note a delay which make appear an empty cyan rectangle and after
a few the tooltip text)
:(((
Cheers,
-- Marco
10:29 a.m.
On Mon, May 31, 2010 at 4:43 PM, Genes MailLists <lists(a)sapience.com> wrote:
On 05/31/2010 09:58 AM, Marco Guazzone wrote:
>
> Here below is the information you have asked
>
> But before you dig into it I want to say that it seem a problem of the
> OpenBox window manager.
> I say so since I've tried to use GNOME and apparently the problem
> seems to be disappeared (for instance under ViM there is no delay in
> showing characters under the cursor).
Ahah ... yes indeed sounds like openbox is busted for some reason ...
good catch.
I found no obvious issues with mem pressure etc ... file a bug def ...
I know nothing of openbox (honestly had not even heard of it till this
email!)
:)
It's a light-weight window manager and you can try It with LXDE Fedora
spin: http://spins.fedoraproject.org/lxde/
Good luck.
Glad you found the source of the problem ... and can use gnome as a
work around.
Thank you very much for the help!
Cheer,
-- Marco
9:43 a.m.
On 05/31/2010 09:58 AM, Marco Guazzone wrote:
Here below is the information you have asked
But before you dig into it I want to say that it seem a problem of the
OpenBox window manager.
I say so since I've tried to use GNOME and apparently the problem
seems to be disappeared (for instance under ViM there is no delay in
showing characters under the cursor).
Ahah ... yes indeed sounds like openbox is busted for some reason ...
good catch.
I found no obvious issues with mem pressure etc ... file a bug def ...
I know nothing of openbox (honestly had not even heard of it till this
email!)
Good luck.
Glad you found the source of the problem ... and can use gnome as a
work around.
8:58 a.m.
On Mon, May 31, 2010 at 2:52 PM, Genes MailLists <lists(a)sapience.com> wrote:
On 05/31/2010 05:24 AM, Marco Guazzone wrote:
Weird - I've seen similar when something is either occupying large CPU
or large memory.
If something has sucked up the memory and the application is paging
off disk for example. Look for any processes which are taking up a lot
of memory (run away firefox, chrome etc)
Hi thank you for replying.
Here below is the information you have asked
But before you dig into it I want to say that it seem a problem of the
OpenBox window manager.
I say so since I've tried to use GNOME and apparently the problem
seems to be disappeared (for instance under ViM there is no delay in
showing characters under the cursor).
While the problem reappears if I turn to OpenBox (without GNOME).
Very strange, under F12 I had no problem even with the same version of OpenBox.
I think I should post a bug report, do you?
Can you run
vmstat 1
in a terminal window while this happening too please.
Sure here below the output. This is taken while moving the cursor
inside ViM character-by-character (I confirm the delay of some second
in showing the character under the cursor)
--- [vmstat] ---
procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu-----
r b swpd free buff cache si so bi bo in cs us sy id wa st
2 0 4216 124544 108840 912952 0 0 33 21 448 520 7 5 87 1 0
0 0 4216 124412 108840 913244 0 0 0 0 516 1107 1 2 97 0 0
0 0 4216 124412 108840 913240 0 0 0 0 316 465 1 0 99 0 0
0 0 4216 124412 108840 913044 0 0 0 0 515 905 1 2 97 0 0
0 0 4216 124412 108840 913044 0 0 0 0 293 455 1 0 99 0 0
1 0 4216 124412 108840 913044 0 0 0 0 409 569 1 0 98 0 0
0 0 4216 124412 108840 913292 0 0 0 0 419 966 1 1 98 0 0
0 0 4216 124412 108840 913348 0 0 0 0 523 1237 2 2 97 0 0
0 0 4216 124412 108840 913348 0 0 0 0 447 806 1 2 97 0 0
1 0 4216 124412 108840 913028 0 0 0 0 428 624 2 0 98 0 0
0 0 4216 124412 108840 913028 0 0 0 0 324 469 1 0 99 0 0
0 0 4216 124412 108840 913028 0 0 0 0 384 564 1 0 99 0 0
0 0 4216 124412 108840 913028 0 0 0 0 317 500 2 1 97 0 0
0 0 4216 124412 108840 913032 0 0 0 0 657 958 5 1 93 0 0
1 0 4216 124412 108840 913032 0 0 0 0 307 455 1 0 99 0 0
0 0 4216 124412 108840 913032 0 0 0 0 386 559 1 0 99 0 0
0 0 4216 124412 108840 913280 0 0 0 0 690 1228 2 3 94 0 0
--- [/vmstat] ---
I've also seen this with driver problems causing the kernel to have
some kind of problem (page faults, oops etc).
For /var/log/messages I'd
egrep 'warn|err' /var/log/message
--- [messages] ---
May 31 09:08:51 scramble kernel: ACPI: Using IOAPIC for interrupt routing
May 31 09:08:51 scramble kernel: ACPI: PCI Interrupt Link [LNKA] (IRQs 9 10 *11)
May 31 09:08:51 scramble kernel: ACPI: PCI Interrupt Link [LNKB] (IRQs 5 7) *3
May 31 09:08:51 scramble kernel: ACPI: PCI Interrupt Link [LNKC] (IRQs 9 *10 11)
May 31 09:08:51 scramble kernel: ACPI: PCI Interrupt Link [LNKD] (IRQs
5 7 9 10 *11)
May 31 09:08:51 scramble kernel: ACPI: PCI Interrupt Link [LNKE] (IRQs
3 4 5 6 7 9 *10 11 12 14 15)
May 31 09:08:51 scramble kernel: ACPI: PCI Interrupt Link [LNKF] (IRQs
3 4 5 6 7 *9 10 11 12 14 15)
May 31 09:08:51 scramble kernel: ACPI: PCI Interrupt Link [LNKG] (IRQs
3 4 *5 6 7 9 10 11 12 14 15)
May 31 09:08:51 scramble kernel: ACPI: PCI Interrupt Link [LNKH] (IRQs
3 4 5 6 7 9 10 11 12 14 15) *0, disabled.
May 31 09:08:58 scramble libvirtd: 09:08:58.498: warning :
qemudStartup:1150 : Unable to create cgroup for driver: No such device
or address
May 31 09:09:00 scramble libvirtd: 09:09:00.374: warning :
lxcStartup:1748 : Unable to create cgroup for driver: No such device
or address
May 31 09:09:09 scramble dbus: [system] Rejected send message, 1
matched rules; type="method_call", sender=":1.20" (uid=42 pid=1696
comm="gnome-power-manager)
interface="org.freedesktop.Hal.Device.LaptopPanel"
member="GetBrightness" error name="(unset)" requested_reply=0
destination=":1.1" (uid=0 pid=1303 comm="hald))
May 31 09:09:09 scramble dbus: [system] Rejected send message, 1
matched rules; type="method_call", sender=":1.20" (uid=42 pid=1696
comm="gnome-power-manager)
interface="org.freedesktop.Hal.Device.LaptopPanel"
member="SetBrightness" error name="(unset)" requested_reply=0
destination=":1.1" (uid=0 pid=1303 comm="hald))
May 31 09:33:08 scramble python: hp-makeuri[3157]: error: Device not found
--- [/messages] ---
Also try looking in the X log file(s) (/var/log/Xorg.0.log and
~/.xsession-errors)
--- [xsession-errors] ---
--> I: 0 [ 13 ]
--> I: 1 [ 13 ]
--> I: 2 [ 13 ]
--> I: 3 [ 13 ]
--> I: 4 [ 12 ]
--> I: 5 [ 11 ]
--> I: 6 [ 10 ]
--> I: 7 [ 9 ]
(polkit-gnome-authentication-agent-1:4833): GLib-GObject-WARNING **:
cannot register existing type `_PolkitError'
(polkit-gnome-authentication-agent-1:4833): GLib-CRITICAL **:
g_once_init_leave: assertion `initialization_value != 0' failed
xnots error: Unable to read file /home/marco/.xnotsrc
Conky: /home/marco/.conkyrc: 25: no such configuration: 'on_bottom'
Conky: /home/marco/.conkyrc: 91: no such configuration: 'border_margin'
Idesk starting in :0.0
[idesk] Background's file not found.
[idesk] Background's source not found.
(deja-dup-monitor:4831): GLib-GObject-CRITICAL **: g_value_get_uint:
assertion `G_VALUE_HOLDS_UINT (value)' failed
Conky: forked to background, pid is 5109
X Error of failed request: BadWindow (invalid Window parameter)
Major opcode of failed request: 20 (X_GetProperty)
Resource id in failed request: 0x1400001
Serial number of failed request: 118
Current serial number in output stream: 118
(gnome-settings-daemon:4824): GdkPixbuf-CRITICAL **:
gdk_pixbuf_format_get_name: assertion `format != NULL' failed
(mail-notification:5112): Gtk-WARNING **: Error loading theme icon
'stock_mail' for stock: Icon 'stock_mail' not present in theme
Received event ClientMessage for unknown window (0x00000000)
Unhandled event ReparentNotify on note 0
Unhandled event ReparentNotify on note 1
Unhandled event ReparentNotify on note 2
Unhandled event ReparentNotify on note 3
Unhandled event ReparentNotify on note 4
Unhandled event ReparentNotify on note 5
Unhandled event ReparentNotify on note 6
Unhandled event ReparentNotify on note 7
Unhandled event ReparentNotify on note 8
Unhandled event ReparentNotify on note 9
tint2 : nb monitor 1, nb monitor used 1, nb desktop 4
(gnome-settings-daemon:4824): GdkPixbuf-CRITICAL **:
gdk_pixbuf_format_get_name: assertion `format != NULL' failed
(mail-notification:5112): Gtk-WARNING **: Error loading theme icon
'stock_mail' for stock: Icon 'stock_mail' not present in theme
Received event MappingNotify for unknown window (0x00000000)
(gnome-settings-daemon:4824): Gtk-WARNING **: gtk_window_set_wmclass:
shouldn't set wmclass after window is realized!
(gnome-settings-daemon:4824): GLib-GObject-WARNING **:
IA__g_object_notify: object class `GkbdStatus' has no property named
`name'
(gnome-settings-daemon:4824): GdkPixbuf-CRITICAL **:
gdk_pixbuf_format_get_name: assertion `format != NULL' failed
Received event ClientMessage for unknown window (0x02a00362)
Received event ClientMessage for unknown window (0x02a00362)
Received event ClientMessage for unknown window (0x02a00362)
Received event ClientMessage for unknown window (0x02a00362)
Received event ClientMessage for unknown window (0x02a00362)
Received event ClientMessage for unknown window (0x02a00362)
Received event ClientMessage for unknown window (0x02a00362)
Warning: Cannot convert string "nil2" to type FontStruct
Received event ClientMessage for unknown window (0x02a00390)
Received event ClientMessage for unknown window (0x02a00390)
Received event ClientMessage for unknown window (0x02a00390)
Received event ClientMessage for unknown window (0x02a003ea)
Received event ClientMessage for unknown window (0x02a003ea)
Received event ClientMessage for unknown window (0x02a003ea)
Received event ClientMessage for unknown window (0x02a003ea)
Received event ClientMessage for unknown window (0x02a003ea)
Received event ClientMessage for unknown window (0x02a003ea)
Received event ClientMessage for unknown window (0x02a003ea)
App startup
FoxyProxy settingsDir:
/home/marco/.mozilla/firefox/bcgz7zpb.default/foxyproxy.xml
Received event ClientMessage for unknown window (0x02a00418)
Received event ClientMessage for unknown window (0x02a00418)
Received event ClientMessage for unknown window (0x02a00418)
--- [/xsession-errors] ---
It is also possible your disk is dying and the errors cannot be
written
out.
You could try checking if there are hardware problems (perhaps run
your dell diagnostics disk).
I've tried under Win Vista (the OS installed on another disk
partition) and I had no problem.
Thank you very much
Cheers,
-- Marco
7:52 a.m.
On 05/31/2010 05:24 AM, Marco Guazzone wrote:
Weird - I've seen similar when something is either occupying large CPU
or large memory.
If something has sucked up the memory and the application is paging
off disk for example. Look for any processes which are taking up a lot
of memory (run away firefox, chrome etc)
Can you run
vmstat 1
in a terminal window while this happening too please.
I've also seen this with driver problems causing the kernel to have
some kind of problem (page faults, oops etc).
For /var/log/messages I'd
egrep 'warn|err' /var/log/message
Also try looking in the X log file(s) (/var/log/Xorg.0.log and
~/.xsession-errors)
It is also possible your disk is dying and the errors cannot be written
out.
You could try checking if there are hardware problems (perhaps run
your dell diagnostics disk).
gene
3:02 a.m.
On Sat, May 29, 2010 at 3:06 PM, Marco Guazzone
<marco.guazzone(a)gmail.com> wrote:
Hi,
I've noticed that there is a some kind of "delay" in the screen refresh.
For instance, under VIM when I move the cursor (that is a black
rectangle) under a character, it takes between 1 and 4 secs before
the character appears inside the cursor.
I've experienced the same issue under OpenOffice and FireFox.
This is really annoying. :(
Under F12 no problem
Does any other experienced this behaviour?
My system is:
* OS: F13 x86_64 - up-to-dated
* WM: OpenBox
* CPU: Intel Core 2 Duo
* GPU: Intel GM965/GL960
* RAM: 2GB
Thank you so much!
Hi!
It seems that this problem is only affecting me.
In the meanwhile I've noticed that this delay seems to be really X
dependent, since almost every applications exhibits it.
For instance, I've tried with:
* bash inside xterm (e.g., execute "ls" and then move up with Shift+PgUp)
* gvim
...
Please, can anyone help me?
Thank you very much!!
Cheers,
-- Marco
3:51 a.m.
On 31 May 2010 09:02, Marco Guazzone <marco.guazzone(a)gmail.com> wrote:
On Sat, May 29, 2010 at 3:06 PM, Marco Guazzone
<marco.guazzone(a)gmail.com> wrote:
> Hi,
>
> I've noticed that there is a some kind of "delay" in the screen
refresh.
> For instance, under VIM when I move the cursor (that is a black
> rectangle) under a character, it takes between 1 and 4 secs before
> the character appears inside the cursor.
> I've experienced the same issue under OpenOffice and FireFox.
> This is really annoying. :(
>
> Under F12 no problem
>
> Does any other experienced this behaviour?
>
> My system is:
> * OS: F13 x86_64 - up-to-dated
> * WM: OpenBox
> * CPU: Intel Core 2 Duo
> * GPU: Intel GM965/GL960
> * RAM: 2GB
>
> Thank you so much!
Hi!
It seems that this problem is only affecting me.
In the meanwhile I've noticed that this delay seems to be really X
dependent, since almost every applications exhibits it.
For instance, I've tried with:
* bash inside xterm (e.g., execute "ls" and then move up with Shift+PgUp)
* gvim
...
Please, can anyone help me?
Thank you very much!!
Cheers,
-- Marco
Hi Marco,
I'm no expert but it sound to me like you have a latency issue. For
some reason your machine is sometimes having to wait 1 to 4 seconds
before it can give a program (vim, bash, etc.) its CPU time slice,
hence the apparent delay.
I have seen this kind of thing before when the kernel has to re-try
disk I/O multiple times on a failing hard drive.
Suggestions:
1) Play music! If it pauses when your app pauses it is an OS wide
issue effecting everything.
2) Run top in another window. When top refreshes after the pause does
any process jump to the top?
3) Check /var/log/messages for any errors.
Vague ideas:
4) Google for investigating latency issues.
5) With modern kernels I guess it should be possible to dynamically
trace it to find the lateny issue, but I have no idea how.
Sorry I can't be more help,
Mike
5043
days inactive
5081
days old
9 comments
3 participants
participants (3)
-
Genes MailLists -
Marco Guazzone -
Mike Fleetwood