14:36:57 <gitomat> [illumos-gate] 18017 fcsm: pointer used after free in fcsm_alloc_cmd() -- Toomas Soome <tsoome⊙mc>
15:44:36 <jbk> so that last change did prompt a bigger question -- the comments in pci_boot.c suggest the PCI unit address cache was needed because the pre-dboot code would enumerate child buses differently (thus the cache was needed to preserve the numbering)
15:45:14 <jbk> which if true, means there's no systems that can run illumos that'd need it
15:46:09 <jbk> or at least where you'd end up with different numbering within the OS than what the BIOS or UEFI configures
15:46:29 <jbk> since IIUC, dboot is from (at least) solaris 10
15:46:32 <tsoome> dboot got introduced with grub, that is, with solaris 10 ~2005.
18:13:56 <gitomat> [illumos-gate] 18013 sockfs: possible memory leak of 'laddr' -- Toomas Soome <tsoome⊙mc>
19:20:50 <richlowe> dboot is from post-xen onnv
19:21:08 <richlowe> prior to that there was a separate 'multiboot' executable that did similar things
19:21:26 <richlowe> "that" would be early onnv, and s10u1 "newboot"
19:21:41 <richlowe> prior to _that_ there was a native mode horror show built with masm on nt4 or something
19:24:00 <danmcd> The dboot stuff is actually better than its predecessor.
19:24:23 <danmcd> ("I'll take Damning with Faint Praise for $200, Ken."
19:24:25 <danmcd> )
19:25:01 <jbk> so faint it's practically translucent :P
20:00:17 <Reinhilde> :(
23:42:15 <snek> does dhcpagent automatically edit /etc/resolv.conf?
23:42:25 <snek> i don't see it mentioned in the man pages at all
23:44:44 <richlowe> lsls
23:45:00 <richlowe> ls
23:45:02 <richlowe> oops
23:45:18 <richlowe> snek: if it's configured to take the DNS info etc, yes.
23:45:31 <snek> kk