00:01:58 <jbk> oooh 'overconstrained allocation' from vmem_xalloc... that's a new one...
02:37:18 <jbk> drat.. redmine is still having problems...
17:14:09 <alanc>  anyone feeling old today? https://github.com/illumos/illumos-gate/commit/fa9e4066f08beec538e775443c5be79dd423fcab
17:14:10 <fenix> → GitHub commit fa9e406: PSARC 2002/240 ZFS 6338653 Integrate ZFS PSARC 2004/652 - (committed)
18:00:35 <toasterson> Just 3 years before my Apprenticeship started
18:40:36 <cross> I feel old every day. It's mostly the knees, to be honest.
19:17:17 <jbk> i regularly joke about yelling at clouds
19:23:10 <jbk> cross: so interestingly, after getting past that panic (tl;dr iommu was wanting a separate ipi vector for every iommu, system runs out, we ignore that, other ipis get messed up, sadness ensues)
19:23:36 <jbk> basically by just using one ipi vector for all of the iommus, sending an nmi now resets the box :)
19:24:26 <jbk> (since we only seem to use the iommu interrupts for events, and not like on-demand paging of dma memory, i'm going to try to skip it to see if that makes any difference)
19:24:39 <jbk> at least for testing...
19:25:26 <jbk> was also going to write some mdb modules to make it a bit easier to view some of this data...
22:30:56 <richlowe> for my own education (and happy halloween!), how big _is_ the nvidia driver now?
22:31:11 <richlowe> I just caght up to Hans doubling HEAPTEXT_SIZE to make space for it
22:31:16 <richlowe> but it's had 2 years since then...
22:33:31 <alanc> clearly when your company is worth $5T, the price of memory means nothing