On a Thinkpad T41p running Fedora Core 3 kept up-to-date since November. Recently, in the past two or three weeks, the clock has been incorrect every time I resume from ACPI suspend. Before that it was always correct when it woke up. Now the clock is always fast when it wakes up. It appears to be proportional to how long it has been sleeping, as if the clock were running consistently extremely fast while asleep, but I haven't run any tests to see if it is reproducible. I've been fixing it by restarting ntpd. I'm currently running kernel 2.6.10-1.741, but I've had all of the kernel updates since FC3 came out. I can't say for sure that the problem started with a kernel update, but it may have coincided with the first 2.6.10 kernel.
Is this a known problem? Is there a fix?
On Fri, 21 Jan 2005, Eric Benson wrote:
On a Thinkpad T41p running Fedora Core 3 kept up-to-date since November. Recently, in the past two or three weeks, the clock has been incorrect every time I resume from ACPI suspend. Before that it was always correct when it woke up. Now the clock is always fast when it wakes up. It appears to be proportional to how long it has been sleeping, as if the clock were running consistently extremely fast while asleep, but I haven't run any tests to see if it is reproducible. I've been fixing it by restarting ntpd. I'm currently running kernel 2.6.10-1.741, but I've had all of the kernel updates since FC3 came out. I can't say for sure that the problem started with a kernel update, but it may have coincided with the first 2.6.10 kernel.
Is this a known problem? Is there a fix?
Noticed this as well. (with APM/600E). This is more pronounced with Vmware.
[haven't looked at bugzilla yet to see if this is reported]
One workarround is to rebuild the kernel with HZ=100 instead of the default HZ=1000 (CONFIG_X86_HZ flag is available for -ac kernels)
Satish
I have exactly the same problem on a DELL 600m. It was driving me wacky until I figured out why the clock was always wrong.
I also have the impression the amount of time shift is proportional to the amount of time the computer is asleep.
Philip
On Fri, 2005-01-21 at 12:22 -0500, Eric Benson wrote:
On a Thinkpad T41p running Fedora Core 3 kept up-to-date since November. Recently, in the past two or three weeks, the clock has been incorrect every time I resume from ACPI suspend. Before that it was always correct when it woke up. Now the clock is always fast when it wakes up. It appears to be proportional to how long it has been sleeping, as if the clock were running consistently extremely fast while asleep, but I haven't run any tests to see if it is reproducible. I've been fixing it by restarting ntpd. I'm currently running kernel 2.6.10-1.741, but I've had all of the kernel updates since FC3 came out. I can't say for sure that the problem started with a kernel update, but it may have coincided with the first 2.6.10 kernel.
Is this a known problem? Is there a fix?
-
Eric Benson -
Philip Balister -
Satish Balay