01:01:13 A sysadmin cohort is sending them to iocaine. I would block them. Instead of keeping them busy on the system that I want them off. 01:06:31 maybe send them a redirection to a different-system iocaine or nepenthes? 01:09:38 I have thought of that! :-) I would rather block them at the firewall. 02:47:42 can someone help me attach another disk to a zpool. --> https://termbin.com/wmmj 03:08:58 skered: if you install llvmXX from ports, it'll give you a toolchain file like /usr/local/share/toolchains/llvmXX.mk which you can use with "make CROSS_TOOLCHAIN=llvmXX buildworld" 03:09:12 (XX = 19, 20, 21, etc) 03:12:15 if you'd rather use your own home-made llvm, you should be able to set XCC, XCXX, XCPP and XLD to point to your versions... i've never actually tried doing it that way though 03:15:33 bradd: You can partition the disk to make it match, but if it's only for storage or something and doesn't require partitions (which it seems like it might?), you can juse use "zpool add" to add a disk to a pool (not "attach") 03:25:11 ek, I think add is to add another mirror set with the newly specified drives. as per my url, I want to add nda1 to mirror-0. (not create a new mirrored device) 03:33:24 bradd: Ah. I think I misunderstood. 03:33:46 You're just trying to add a 3rd disk to the mirror, it sounds like? 03:34:31 In that case, "zpool attach " should work. 03:35:56 However, I'm not sure if the partitioning will matter. It might. So, you may want to copy the disk config (say, nda0) to nda1 first using gpart. 03:43:39 ok. I did 'zpool attach zroot mirror-0 nda1' and I get -> cannot attach to mirror-0 : can only attach to mirrors and top-level disks.. isn't zpool/mirror-0 a mirror? 03:45:09 Yeah. You need to attach to the pool (zroot) but not the vdev (mirror-0). You're just telling it to mirror a disk (nda0 or nda2). 03:46:28 So, probably do a "gpart backup | gpart restore" from nda0 to nda1, and then a "zpool attach zroot nda0p4 nda1p4" or whatever disk you want to mirror. 03:46:50 this seems like a bug, since the manual page clearly indicates this should work 03:47:38 ivy: Are you talking about bradd's issue? 03:47:43 yes 03:47:48 ek, yes that worked. thanks for the help 03:48:04 Which manpage are you looking at? ZFS or FBSD? 03:48:07 the manual explictly mentions attaching to "mirror-7" as a valid option, so it should work 03:48:11 ek: zpool-attach(8) 03:48:11 bradd: Excellent! 03:49:06 ivy: Huh. I've never tried to add a an additional disk to a mirror without specifying the disk. Haven't checked the manpage. But, yes, something is amiss. 03:56:00 ivy: I think it might just be failing with 21. 03:56:55 ivy: I was reading about that var about an hour ago but I did have the full list excluing CROSS_BINUTILS_PREFIX and X_COMPILER_TYPE set from the comman line the Xvars 03:58:02 I did have a couple other things set though.. Try with a clean set not trying to disable anything. 03:58:22 fwiw, i regularly build main with llvm 22 and don't have any issues. i haven't tried 21 though 03:58:52 Ok. I'm going to guess it's something I was doing then. 03:59:06 er no.. there it failed. 04:00:01 Might as well try with 22 since it's the latest. (21 always was). 04:00:18 fwiw, if you're on a stable branch, it might be some fixes didn't get backported yet 04:00:52 this is master. 04:01:40 you mean main, right? (i only ask because there's an old "master" branch on github that hasn't been updated for years) 04:02:10 er yeah main. 04:02:43 Yeah. I got bit by the master -> main change on github just earlier today. Took me a bit to figure out why things were... strange. 04:47:23 ivy: 22 just worked. So my head banging today I guess was just 21. 05:17:22 now if only i could get the linux gcc build working... 05:33:32 hi, anyone noticed that cpuset(1) cpu selection with ranges doesn't work as expected? 06:14:52 having this issue with firefox again where if I allow it to use the pulseaudio backend for audio, the output is so loud that it clips. looking in pavucontrol though, nothing is cranked up past 100% 06:16:18 setting cubeb to use the OSS backend is not a viable solution, unfortunately, since it makes FF drop about 50% of all frames when playing back 60fps video 06:17:08 has anyone else run into this?