IRC channel logs

2022-10-30.log

back to list of logs

<damo22>strange, i moved things around to refactor kernel_page_dir and it hangs really early
<damo22>youpi: how do you know kernel_page_dir needs to be unique per cpu? why cant they share the kernel_page_dir ?
<damo22>it breaks when i separate them
<damo22>is there supposed to be a kernel_pmap per cpu as well?
***jpoiret9 is now known as jpoiret
<youpi>damo22: I'm not saying there should be one kernel_page_dir per cpu
<youpi>I'm saying that you shouldn't tinker with it
<youpi>while setting up paging/segmentation
<youpi>since the bsp and other aps will be using it while you're tinkering with it
<damo22>ok
<youpi>so you can just use another page_dir while setting up the cpu, and once done you can use the kernel_page_di
<damo22>i created a new one per cpu so they can have their own, and then i am switching back
<damo22>yes
<damo22>i eventually figured that one
***dragestil_ is now known as dragestil
<iska>damo22 is SMP working now?
<Pellescours>not yet, but there was progress
<iska>hmm
<iska>would doing it on Xen be easier?