© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
0e1f616
(Get the code!)
Dear Tim,
i wonder about this TSC issue the patch was designed for happens on my (and the other) machines, because it's have a single CPU without HT.
The patch says "return some last value until the normal used one is higher than". From that: Will the TSC reset, if the CPU is suspended (powered down?). Then, after resume, the two calls will return the "last" values forever and *all* loops driven by that will hang.
If this is the core reason, maybe the TSC value should be compared with some low value indicating that the CPU have a low uptime yet. And the "last" value should be lowered (or zeroed) in this case before the committed compare. Or -- even better, but may be more difficult -- the "last" value should be zeroed on some early initialisation stage of the CPU(s)
greetings
Guido