X System clock haywire ...

[AnoobIS]

click the kMENU and in the "Run Command" section type

keyboard

to go to the keyboard menu and edit the type rate etc.

hope this helps

[HughR]

I too have an HP Pavilion a1250n with these problems:
- time-of-day clock advances at twice the rate it should
- in dmesg output I get some of these messages: APIC error on CPU0: 40(40)
- keyboard repeat rate is way too fast (theory: due to time-of-day clock speedup)
- /proc/interrupts shows way to many parport0 interrupts. Probably too many usb interrupts as well.

The clock problem is discussed in the bugzila entry:
http://bugme.osdl.org/show_bug.cgi?id=3927

I think that the problem is a mixture of bugs in some or all of
(1) ATI Radeon Xpress 200 chipset (note: not the graphics part, the APIC wiring)
(2) the BIOS, especially the ACPI tables
(3) the Linux kernel
(I list the most likely suspects first.)

I'm running Fedora Core 4 for x86-64 with updates up to a week ago.

I'd love to hear how others are doing with these problems.

[carl0ski]

I too have an HP Pavilion a1250n with these problems:
- time-of-day clock advances at twice the rate it should
- in dmesg output I get some of these messages: APIC error on CPU0: 40(40)
- keyboard repeat rate is way too fast (theory: due to time-of-day clock speedup)
- /proc/interrupts shows way to many parport0 interrupts. Probably too many usb interrupts as well.

The clock problem is discussed in the bugzila entry:
http://bugme.osdl.org/show_bug.cgi?id=3927

I think that the problem is a mixture of bugs in some or all of
(1) ATI Radeon Xpress 200 chipset (note: not the graphics part, the APIC wiring)
(2) the BIOS, especially the ACPI tables
(3) the Linux kernel
(I list the most likely suspects first.)

I'm running Fedora Core 4 for x86-64 with updates up to a week ago.

I'd love to hear how others are doing with these problems.

I personally dont believe APIC is responsible.

type
hwclock&&date

watch what happens,

hwclock = you RTC clock the deep core
date = what your software relies upon.

my date counts excessively HWCLOCK runs as it should


if someone can dig up how date works this can be worked around or fixed.
Does anyone know where date acquires the time and count?



noAPIC no localAPIC have no effect on the problem

notsc no_timer_check apm=off
sometimes work depending on the kernel ,
turning off Cool'N'Quiet sometimes works.

date is the key to whats happening
xine for example relies on date to play media and runs fast,

mplayer however must use hwclock since it runs normal.

[carl0ski]

Please no more of these silly NTP suggestions
HH.MM:SS
P.S do NOT!!! use ntp to workaround this issue.

ntp sets the clock to a time even if you set ntp to change the clock every second.

the timing rate (the beat ) of the clock will still run fast and will still cause chaos and system interupts (keyboard repeat, refresh rates etc) will still be irratic, and in my case the 15 seconds for network transfers will still cause timeouts.


but a new danger will arise using NTP to fix it.
1. High network usage
2. Rapid time shift into the Past will cause software to become unstable and dangerous (10.00:11 1hr later 12.00.11 ntp changes it to 11.00:11)
thats a full 30minutes in which files can invoke time mismatch errors.

--Files will have creation modification and access times in the Future
--Copying moving or Compiling software while this eventy occurs will cause data loss errors and Filesystem faults

3. The further apart the NTP update the more likely errors will occur

say your clock starts at

10.00:11 seconds 5 seconds later 10.00:21 seconds
to fix this you run ntp every 5 seconds


10.00:11 seconds 5 seconds later 10.00:21 seconds ntp makes it 10.00:16 seconds
what happens when a file is written at 10.00:21 and again just after ntp. again at 10.00:18

fatal error file exists in the future,
which depending on the software may result in data loss.

[p0ptart]

I am running Mandriva 2006 on a Compaq (HP) Presario SR1703WM, motherboard is an Asus A8AE-LE, or as HP calls it the AmberineM-GL6E. Chipset is ATI Radeon Xpress 200. CPU is Sempron (P) 3200+ 1.8 GHz. I am have the same problem as you are all having, with the clock going entirely too fast, messing up sound, making my monitor or pc not shut off or restart correctly, and making my video/sound go entirely too fast. I have yet to find a fix for it, and I am pretty tired of searching. If anyone could help me I would appreciate it greatly.

[dragonlor20]

I have the same system (HP a1250n) with Mandriva 2006 x64 and googling led to the following solultion which solved my fast-clock and hyper-responsive keyboard problems:
Add "notsc" to the append line in /etc/lilo.conf. Remember to run lilo as root to apply the changes.

I'm a true newbie to linux but i finally got fed up with windows and just deleted windows in a sink or swim effort... Anyways, which one of the append lines in lilo.conf should i add notsc to? There are 3.

default="linux"
boot=/dev/hda
map=/boot/map
keytable=/boot/us.klt
menu-scheme=wb:bw:wb:bw
prompt
nowarn
timeout=100
message=/boot/message
image=/boot/vmlinuz
label="linux"
root=/dev/hda1
initrd=/boot/initrd.img
append=" resume=/dev/hda5 splash=silent"
vga=788
image=/boot/vmlinuz
label="linux-nonfb"
root=/dev/hda1
initrd=/boot/initrd.img
append=" resume=/dev/hda5"
image=/boot/vmlinuz
label="failsafe"
root=/dev/hda1
initrd=/boot/initrd.img
append=" failsafe resume=/dev/hda5"
other=/dev/sda1
label="windows"
table=/dev/sda
map-drive=0x80
to=0x81
map-drive=0x81
to=0x80

[doczee]

I'm running Mandriva 2006 x86_64 on an athlon x2 machine. What worked was to add notsc to the append as recommended, but it appears this may not work for everyone.

I'd recommend making these kind of changes in one section only and if it works, propogate it to the rest. For your case, I'd change the linux entry (your default) and if things fail, select failsafe next time you boot to get back to a working system.

Your linux section should look like this then:
image=/boot/vmlinuz
label="linux"
root=/dev/hda1
initrd=/boot/initrd.img
append="notsc resume=/dev/hda5 splash=silent"
vga=788

Once you know it works, you can change the other append lines to match. Don' forget to run lilo after making the changes, or it will have no effect.

Looks like you didn't entirely give up windows either . I've been using Linux since version 1.x of the kernel and the reality is that as long as we have the MS world driving all computing, you will need windows in some form for something.

I'm a true newbie to linux but i finally got fed up with windows and just deleted windows in a sink or swim effort... Anyways, which one of the append lines in lilo.conf should i add notsc to? There are 3.

default="linux"
boot=/dev/hda
map=/boot/map
keytable=/boot/us.klt
menu-scheme=wb:bw:wb:bw
prompt
nowarn
timeout=100
message=/boot/message
image=/boot/vmlinuz
label="linux"
root=/dev/hda1
initrd=/boot/initrd.img
append=" resume=/dev/hda5 splash=silent"
vga=788
image=/boot/vmlinuz
label="linux-nonfb"
root=/dev/hda1
initrd=/boot/initrd.img
append=" resume=/dev/hda5"
image=/boot/vmlinuz
label="failsafe"
root=/dev/hda1
initrd=/boot/initrd.img
append=" failsafe resume=/dev/hda5"
other=/dev/sda1
label="windows"
table=/dev/sda
map-drive=0x80
to=0x81
map-drive=0x81
to=0x80

[dragonlor20]

Thanks for the reply, but no luck... I have heard that there is a fix for this problem under SuSe? The whole thing wouldn't be so bad, except that it speeds up animations (such as the little circle in the upper right corner of this browser) and advertisements, the clock runs fast, and programs like kaffiene that are somewhere along the line reliant on the system clock run super fast too... So far what I have found are work-arounds as opposed to solutions. Do you think that using Mandriava 32 as opposed to Mandriva 86- 64x or whatever could have something to do with it? I have an AMD Turion64 installed on this Compaq Presario laptop...

[dombas]

Dragonlor20 wrote:
Thanks for the reply, but no luck... I have heard that there is a fix for this problem under SuSe? The whole thing wouldn't be so bad, except that it speeds up animations (such as the little circle in the upper right corner of this browser) and advertisements, the clock runs fast, and programs like kaffiene that are somewhere along the line reliant on the system clock run super fast too... So far what I have found are work-arounds as opposed to solutions. Do you think that using Mandriva 32 as opposed to Mandriva 86- 64x or whatever could have something to do with it? I have an AMD Turion64 installed on this Compaq Presario laptop...

I had the same troubles, i have a hp Pavilion zv6000 running Mandriva 2006.
The clock system under linux was running at double speed, while under Windows worked correctly.
I followed the advice of Doczee, but nothing happened.
After i went to HP Customer Care site: http://h10025.www1.hp.com/ewfrf/wc/s...oduct=1152597&
and i found a flash utility for updating the BIOS : sp31533.exe (1.61 mb).
I ran this utility (obviously, under Windows), and the clock system began to work fine.
I hope that my experience may be useful for somebody.
Good luck