Patch Set 3:
Patch Set 3:
(1 comment)
Does this have a time penalty?
I am pretty sure it is the one the sorting algorithm has, and it surely depends on the number of CPUs.
It's bound by CONFIG_MAX_CPUS which (AFAICS) is at most 48 in our tree, in serengeti_cheetah_fam10. That would mean ~2500 swaps at most in highly localized code and data.
I guess we can deal with that when it becomes an issue? I have the suspicion that using timsort (another stable sort) won't fare much better on our data set because of setup and copying costs (whereas bubble sort works in-place).
Patch set 4:Code-Review +1
1 comment:
Patch Set #4, Line 162: apic_ids
add a bounds check for the array. It _should_ never happen, but...
To view, visit change 31545. To unsubscribe, or for help writing mail filters, visit settings.