IRC channel logs


back to list of logs

***Server sets mode: +nt
<AlmuHS>i'm noticing that the implementation of cpu_number() will be a problem. I don't find any way to write an arch-independant implementation for this
<AlmuHS>store the kernel ID in a per-cpu segment can be ease the implementation for x86, in fact. But It's not generic
<AlmuHS>I've just finished to fix all issues in my patch. Tomorrow I'll do a couple of test to check that I've not broken anything
<jrtc27>for when AlmuHS gets back:
<jrtc27>yes, it is generic, it's a very common way to do it
<jrtc27>you'll always have a per-cpu segment, so you can always shove an int in there