Switch to DuckDuckGo Search
   November 12, 2019  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >


NOTICE: This channel is no longer actively logged.

Toggle Join/Part | bottom
[00:03:00] <zsj> the way ZoL does it if you give it a full disk is a GPT partition for the pool and a small one at the end of the disk for the uefi esp, formatted fat. it's enough for multiple copies of grub at least, and that suffices to boot
[00:03:41] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-141-249.rochester.res.rr.com> has joined #illumos
[00:03:53] <zsj> I was thinking of giving freebsd and illumos separate small ssds until I figured out a way to successfully integrate them
[00:04:57] <zsj> and by illumos I mean OpenIndiana
[00:07:46] *** CME_ <CME_!~CME@ip5f5b13b8.dynamic.kabel-deutschland.de> has joined #illumos
[00:08:41] *** CME <CME!~CME@2a02:810a:8c80:3301:ba27:ebff:fe8d:4857> has quit IRC (Ping timeout: 250 seconds)
[00:09:30] *** CME_ is now known as CME
[00:29:00] <Reinhilde> ah
[00:42:37] *** Qatz <Qatz!~db@2601:187:8400:5::42d> has quit IRC (Quit: Gone looking for beer)
[00:48:27] *** sjorge <sjorge!~sjorge@unaffiliated/sjorge> has quit IRC (Quit: 410 Gone)
[00:49:02] *** sjorge <sjorge!~sjorge@unaffiliated/sjorge> has joined #illumos
[00:54:53] <LeftWing> jlevon: I've merged it
[00:54:55] <LeftWing> Sorry about that
[00:55:17] <LeftWing> rmustacc: ^^
[00:56:16] <LeftWing> Reinhilde: To make sure I'm being clear, I suspect you'll bump into issues trying to do it. I've not done it myself, so I don't know for sure what they would be.
[00:56:50] <Reinhilde> LeftWing: So I should install Tribblix and hope it doesn't crap out
[00:57:34] <LeftWing> It's your computer -- you should do whatever will make you happy! Just want to make clear that it's probably not an explicitly supported configuration, and there might be issues you'll need to sort out.
[00:58:31] <Reinhilde> Explicitly nothing is supported when your Unix vendor is a one man band.
[00:58:36] <Reinhilde> I'm used to it.
[01:00:00] <Reinhilde> Basically I just want to have an easier place to debug the epilepsy bug, and to see if the atheros driver is just missing one line like the fbsd wacom driver was
[01:00:16] <Reinhilde> or if it's a more serious deficiency
[01:00:19] <LeftWing> I don't think that's true. I would say that it's generally expected that installing onto a blank disk using the regular installer is supported.
[01:00:43] <LeftWing> It'd be best to not use terms like "epilepsy" to describe issues like this.
[01:00:55] <Reinhilde> words lawyering...
[01:01:04] <LeftWing> No, just asking you to be sensitive to people who have epilepsy.
[01:01:24] <LeftWing> Just say "the SIGBUS issue"? It has the added bonus that people will know what issue you're referring to.
[01:02:15] <Reinhilde> I also wonder if it's KVM specific
[01:02:42] <Reinhilde> Which would be a surprise, were that true
[01:03:04] <LeftWing> Unclear
[01:03:11] <Reinhilde> LeftWing: Have you ever had the SIGBUS ENOMEM/"illumos brainfog" issue?
[01:03:21] <LeftWing> I have not
[01:03:30] <LeftWing> I promise I would have mentioned it, if I had!
[01:04:43] <Reinhilde> Never had a compiler die on you SIGBUS ENOMEM while you were running b or gmake -j4 on a single core proc (I do this up to -j6 on single core procs and only have slowness, not SIGBUS ENOMEM scattershot brainfog)?
[01:04:45] <LeftWing> The only time I can remember hitting SIGBUS like that is in a zone with a tight swap cap, where a bunch of allocations were made using the MAP_NORESERVE flag to mmap(@)
[01:05:02] <Reinhilde> (on other OSes)
[01:06:49] <Reinhilde> LeftWing: huh, woah. Interesting parallel, I think in the FreeBSD mdconfig(8) man page, it refers to malloc memdisks with (or without? would need to consult the man page) noreserve being an easy way to panick the system.
[01:07:11] <jlevon> LeftWing: thanks
[01:07:22] <Reinhilde> So maybe this "bug" stretches back to the Berkeley/NJ split.
[01:07:51] <Reinhilde> Or even before?
[01:08:22] <LeftWing> There's really no point in guessing -- tracing the source of the ENOMEM will point us to the part of the kernel that's causing it, and we can walk backwards from there to see why
[01:08:22] *** Qatz <Qatz!~db@2601:187:8400:5::42d> has joined #illumos
[01:09:30] <LeftWing> It could be any number of things, and possibly exacerbated by the small RAM size in your guest. Not that we shouldn't be able to make 1GB of RAM work, just that it's probably less common these days -- certainly for the large scale deployments of illumos.
[01:10:31] <Reinhilde> Well Tribble made a deployment in 768 work with ZFS, 512 work with UFS, 256 struggles a bit, 192 struggles even harder, and can't install under that, but with an optimised image he could have bootup ram down to the 90s and still have A System.
[01:10:51] <Reinhilde> I'm not sure he really tested it that much?
[01:11:13] <Reinhilde> But I'm well above the Tribble limit (which is the limit below which you can't even install the operating system)
[01:11:41] <LeftWing> Sure, but you're not running Tribblix for starters. His experiments were also done some time ago with whatever illumos was current then
[01:11:49] <Reinhilde> aye
[01:11:55] <Reinhilde> I'm running omnios, which is tuned for, bigger things
[01:12:02] <Reinhilde> and doesn't have a vanilla gate
[01:12:23] <LeftWing> But all this just means we need to find out exactly what's going on -- which is totally possible.
[01:12:27] <Reinhilde> yeah
[01:12:32] <Reinhilde> it's possible. I just don't know how
[01:12:38] <Reinhilde> because of my caffeine-induced brain fog
[01:19:47] <Reinhilde> let's pull up the dtrace guide
[01:21:12] <Reinhilde> ... does this seem like something someone would want to probe? or will it return ENOTENOUGHINFO
[01:21:14] <Reinhilde> 16513 fbt unix trap entry
[01:21:16] <Reinhilde> 16514 fbt unix trap return
[01:28:10] <LeftWing> As I recall, if you look at the code, the value that gets put in the signal errno is extracted from part of the value that comes back from pagefault()?
[01:28:36] <LeftWing> But the challenge is that the value eventually returned by pagefault() is created in like ten other places
[01:28:43] <LeftWing> And we need to know which one
[01:29:08] <Reinhilde> I'm so sorry, I'm not a literal God like you are
[01:29:24] <Reinhilde> Like I don't understand this dialect of English at all
[01:30:32] <Reinhilde> ... that's curious.
[01:30:34] <Reinhilde> 475 faultcode_t pagefault(), res, errcode;
[01:31:06] <Reinhilde> I have NEVER written this in a C program before
[01:31:38] <Reinhilde> Is there something happening here that I don't understand? Is pagefault() setting something or is that not somewhere it's being called
[01:35:00] <LeftWing> Err that's a somewhat unfortunate declaration of a function from another compilation unit, I think
[01:35:31] <LeftWing> It doesn't look like pagefault() is declared in a header
[01:37:15] <Reinhilde> right, I found the actual location of faultcode_t pagefault() (at least on i86pc) and it's ../vm/vm_machdep.c from where I was in (directory avalanche)/i86pc/os (if I'm not mistaken)
[01:37:25] <Reinhilde> Not sure that really matters
[01:37:38] <Reinhilde> but somehow I feel like I have to understand the entire $%&@^ing codepath
[01:39:37] <LeftWing> So you already know that FC_CODE(res) must == FC_OBJERR
[01:40:37] <LeftWing> And if you look under usr/src/uts for uses of FC_OBJERR I think the only time we'd actually use FC_OBJERR to create a value is in the FC_MAKE_ERR() macro
[01:41:20] <LeftWing> Looking in uts/common/vm/faultcode.h the comment for FC_OBJERR says: /* underlying object returned errno value */
[01:41:28] <Reinhilde> ok
[01:41:40] <Reinhilde> i'm gonna just ../.. and look for FC_OBJERR then
[01:41:40] <LeftWing> "object" in this case is probably the vnode that underpins the mapping
[01:41:47] <Reinhilde> kinda fell into a rabbit hole there
[01:42:02] <LeftWing> Well, you probably want to look for FC_MAKE_ERR() calls
[01:42:16] <LeftWing> Because I think that's the only place we actually construct a value that includes FC_OBJERR
[01:43:15] <LeftWing> There are a bunch of places where we do something like FC_MAKE_ERR(EIO) -- which is not ENOMEM, the code you're looking for, so you can probably ignore that one
[01:43:49] <Reinhilde> aye
[01:43:58] <LeftWing> You want to find all the places where we use either an explicit value like FC_MAKE_ERR(ENOMEM), or any place where we do something like FC_MAKE_ERR(x)
[01:44:05] <LeftWing> Because x might be ENOMEM
[01:44:26] <Reinhilde> ok
[01:44:36] <Reinhilde> i was just about to say, this doesn't look related
[01:44:38] <Reinhilde> 1191 static faultcode_t
[01:44:40] <Reinhilde> 1192 acquire_kpmem_lock(struct ddi_umem_cookie *cookie, size_t npages)
[01:44:42] <Reinhilde> 1193 {
[01:45:28] <LeftWing> Note the definition of FC_CODE: #define FC_CODE(fc) ((fc) & 0xff)
[01:45:53] <LeftWing> You can do that mask in D and check that the masked value is ENOMEM so that you only get relevant firings
[01:46:11] <Reinhilde> My eyes are glazing over
[01:47:38] <LeftWing> By way of example, segkp_fault() returns several different ways, either 0 for success or FC_MAKE_ERR() of some error code
[01:48:07] <LeftWing> So something like... dtrace -n 'fbt::segkp_fault:return /(arg1 & 0xff) == ENOMEM/ { stack(); }' might help
[01:50:13] <Reinhilde> so should I mock up a quick d program based on the one you just wrote in this channel and include every function that I believe could ever return with enomem?
[01:50:47] <LeftWing> I think that's a good idea!
[01:51:07] <Reinhilde> i found a few places where err = ENOMEM, where it'd be believable that FC_MAKE_ERR(err) would be invoked
[01:51:14] *** igork <igork!~igork@178.162.50.254> has quit IRC (Ping timeout: 240 seconds)
[01:51:52] *** igork <igork!~igork@91.204.56.74> has joined #illumos
[01:52:00] <Reinhilde> seg_dev, seg_spt, seg_vn. I don't have details on which functions though
[01:53:17] <LeftWing> Using cscope can help find all the places; e.g., https://gist.github.com/jclulow/6d858e13b8a89684da5d499f3622a107
[01:55:41] <Reinhilde> if we ever meet, i owe you a lamb chop
[01:57:12] <Reinhilde> like seriously
[02:00:41] <Reinhilde> https://umbrellix.net/~ellenor/thismaynotmakesense.txt does this look like a good starting place for dtraceage?
[02:03:48] <Reinhilde> right so now I'm gonna see if I can trigger the bug
[02:04:47] <Reinhilde> Getting the load average up above 4 might do it.
[02:05:32] <Reinhilde> it kaboom',d
[02:12:00] <Reinhilde> but i didn't actually catch what was up
[02:20:00] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[02:20:07] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[02:25:45] <jbk> LeftWing: I can add additional emails for me to code.illumos.org (e.g. joyent email), and that'll allow me to push w/ that email, correct?
[02:25:58] <LeftWing> jbk: I think so!
[02:26:25] <LeftWing> Are you pushing a review, or just trying to push to master
[02:26:44] <LeftWing> I thought I enabled Forge Author for pushes to the regular master branch
[02:27:03] <LeftWing> jlevon wants me to do it for reviews, too, I think -- which I will look into
[02:27:07] <jbk> push to master (#11946)
[02:27:11] <LeftWing> huh
[02:27:20] <LeftWing> Can you pm me the command and the output
[02:27:33] <jbk> i got a lack 'force committer' because the author is correct (Jordan), but the git repo has my joyent email setup as teh default
[02:27:47] <LeftWing> oh committer
[02:28:00] <LeftWing> Yeah add whatever other email
[02:28:04] <jbk> i was able to add my joyent email to gerrit (though i've not tried the push again)
[02:28:23] <jbk> though i'll try it now
[02:28:36] <jbk> and that seems to have worked
[02:28:37] <gitomat> [illumos-gate] 11946 clean up improper use of mdb_getopts -- Jordan Paige Hendricks <jordan.hendricks at joyent dot com>
[02:29:01] <jbk> i double checked everything, and asked dan for a sanity check, so hopefully that's all correct :)
[02:29:52] <jbk> as much as i use git, it still has lots of footguns, so i'm always a bit nervous/cautious on a new workflow :)
[02:29:57] <LeftWing> Looks good! Thank you
[02:45:01] <jbk> now if i can keep my mailing lists straight :)
[02:54:31] *** despair86 <despair86!~despair@2605:6000:1515:45b:b62e:99ff:fea1:3bad> has joined #illumos
[02:55:38] <despair86> what's so significant about ZOL that everyone including solaris-kernel systems like illumos are moving to it. pretty disappointing
[02:56:33] <Reinhilde> agreed.
[02:57:04] <Reinhilde> ZOL seems like a step backwards
[02:58:25] <despair86> in off-topic findings, i accidentally punched a hole in a tape recording by absentmindedly setting my earphones down on top of the tape ;-;
[02:58:49] <LeftWing> despair86: Lots of folks are doing good work on ZOL
[02:59:19] <Reinhilde> LeftWing: which is not a reason to rebase OZFS on it.
[02:59:23] <LeftWing> In the limit, it won't be "ZOL" per se -- it'll be a cross-platform ZFS repository
[02:59:39] <LeftWing> With support for Linux, FreeBSD, illumos -- eventually I imagine Windows and Mac OS X
[02:59:46] <LeftWing> That's the plan, anyway
[02:59:55] <LeftWing> So we can all share in the new features and bug fixes
[03:00:02] <despair86> heh, saw the BSOD on an early windows nt zfs build the other day
[03:01:08] <LeftWing> As discussed at the summit, the FreeBSD folks are working on getting their platform-specific bits into the ZOL repo and then it will be renamed to openzfs and there'll be a 2.0 release, I believe
[03:01:35] <LeftWing> I expect that once the FreeBSD bits settle down a bit we'll look at how to get illumos in there
[03:01:35] <despair86> woooo
[03:05:57] <Reinhilde> in the realm of NIH syndrome... I'd pick smf over sysd any day, if I had a choice. Systemd as used on Debian doesn't seem to have a "maintenance mode" it can put service units in if they keep restarting. Guess how large my /var/log/daemon.log was. Additionally - their equivalent to svcadm (systemctl) is confusingly named on a system that also has sysctl (which has been deleted from Solaris, a change that
[03:05:59] <Reinhilde> persists in illumos).
[03:06:24] <Reinhilde> I thought yall would enjoy that fun anecdote.
[03:06:52] <despair86> > /bin/ksh: sysctl: not found
[03:06:58] <despair86> hmm
[03:06:59] <LeftWing> I'm not sure that SVR4 ever had sysctl
[03:10:02] <Reinhilde> LeftWing: you might be right
[03:10:13] <Reinhilde> despair86: was that on someone's loonix system
[03:10:43] <despair86> no that was my crippled ryzen box
[03:10:46] <LeftWing> I think Linux systems generally have a sysctl(1)
[03:10:51] <LeftWing> err (8)
[03:11:13] <despair86> uname -a
[03:11:13] <despair86> SunOS sun-pc 5.11 illumos-fec6629365 i86pc i386 i86pc
[03:12:04] <Reinhilde> LeftWing: it'd be sysctl(1m) on a linux distribution were I shipping one of those
[03:12:34] <Reinhilde> just to confuse everyone a little more
[03:13:02] <despair86> > heirloom toolchest contains oracle packaging tools from nokia toolchest
[03:13:16] <despair86> i wonder why there isn't a loonix distro with SVR4 packaging
[03:13:52] <Reinhilde> i wonder why there isn't a bsd distro with svr4 packaging
[03:14:00] <LeftWing> Probably because the tools were historically not open source, and it didn't really offer anything that competing solutions like dpkg and rpm didn't also offer
[03:14:39] <despair86> well i mean _now_, ofc
[03:14:46] <despair86> but yes
[03:16:17] <Reinhilde> should I port smf (and the required kernel machinery) to a bsd if I ever get a chance?
[03:16:36] <despair86> why? bsd expects /etc/rc like it's 1979
[03:16:48] <Reinhilde> most BSD users expect /etc/rc like it's 1979
[03:16:50] <Reinhilde> yes
[03:17:03] <Reinhilde> there's been people trying to port systemd, I think, which disgusts me
[03:17:34] <despair86> i thought that was just userland shims for systemd cancer to interact with saner traditional UNIX interfaces
[03:18:01] <Reinhilde> but the BSD kernel cares about as much as the illumos kernel what it's running, i.e. only that it's a valid executable under that kernel
[03:18:19] <despair86> because i'd imagine GNOME doesn't want to support BSD or Solaris any longer at this point (i.e. they're going all-in on systemd+linux)
[03:18:35] <Reinhilde> for all I care the first process could quite literally be your grandma's toothbrush
[03:18:41] <despair86> heh
[03:18:52] <Reinhilde> (that's an image i don't want to ever evoke again)
[03:22:09] <Reinhilde> despair86: it'd be nuts if GNOME reversed course for vers. 4 and decided to only support illumos
[03:22:24] <Reinhilde> (and would make my idea of porting SMF to non-illumos unices COMPLETELY SENSICAL)
[03:22:38] <Reinhilde> i don't feel like they will tho
[03:22:48] <despair86> well in the past they were at least neutral on supporting UNIX machines
[03:22:56] <despair86> not so much anymore
[03:23:01] <Reinhilde> now they're anti UNIX
[03:23:06] <Reinhilde> and only liking unicoid robots
[03:23:32] <Reinhilde> (linux and MAAAAYYBE, IF YOU GIVE US A LOT OF CONCESSIONS, the berkeley students)
[03:23:46] <rmustacc> In general, I suggest focusing not on how others are making things worse, but on how you can make things better. When you look at things from others perspective, it's not the most unreasonable.
[03:24:57] <despair86> yeah freebsd basically has to support alsa userland shim for OSS just to get a bunch of desktop stuff to work. we at least have pulseaudio v14 (which i've heard is a lot more improved now that poettering isn't working on it anymore)
[03:25:16] <Reinhilde> pöttering left pa?
[03:25:32] <despair86> yeah that's what i've heard. think he moved on to pipewire
[03:25:39] <despair86> or w/e
[03:25:53] <Reinhilde> is it bad that I think poettering should be on OSS projects' shun lists (if any OSS projects have shun lists)
[03:26:04] <despair86> > blacklists
[03:26:25] <despair86> heh, they put all the *wrong* people on them :^)
[03:27:05] <Reinhilde> really the only people who should be on OSS projects' shun lists are richard stallman and lennard poettering
[03:27:14] <Reinhilde> and maybe esr
[03:33:43] <rmustacc> Let's focus on illumos, please.
[03:35:13] *** _despair86 <_despair86!~despair@2600:1700:8f61:163f:36e8:94ff:fede:7ce2> has joined #illumos
[03:37:33] *** despair86 <despair86!~despair@2605:6000:1515:45b:b62e:99ff:fea1:3bad> has quit IRC (Ping timeout: 250 seconds)
[03:39:58] <Reinhilde> how about just nothing gets said
[03:40:01] <Reinhilde> because i have brain fog
[03:41:31] *** despair86 <despair86!~despair@75-60-137-0.lightspeed.rcsntx.sbcglobal.net> has joined #illumos
[03:41:33] *** _despair86 <_despair86!~despair@2600:1700:8f61:163f:36e8:94ff:fede:7ce2> has quit IRC (Ping timeout: 245 seconds)
[03:49:12] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[03:52:56] *** despair86 <despair86!~despair@75-60-137-0.lightspeed.rcsntx.sbcglobal.net> has quit IRC (Ping timeout: 240 seconds)
[05:11:01] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[05:15:17] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[05:25:02] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[05:35:18] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 245 seconds)
[05:49:01] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[05:56:56] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[06:02:37] *** _Tenchi_ <_Tenchi_!~phil@d-207-255-80-203.paw.cpe.atlanticbb.net> has quit IRC (Read error: Connection reset by peer)
[06:18:04] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[06:18:37] *** MarcelT <MarcelT!~marcel@tortuga.telka.sk> has joined #illumos
[06:22:20] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 246 seconds)
[06:47:19] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[06:51:36] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[07:06:40] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[07:15:18] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 245 seconds)
[07:28:16] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[07:35:36] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[07:45:59] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[07:50:50] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 276 seconds)
[08:02:29] <tsoome> zsj: you can not share illumos boot pool at this time
[08:03:14] <tsoome> zsj: or, to be exact, you can but with great pain.
[08:03:16] <Reinhilde> is there any technical reason
[08:04:06] <tsoome> yes, we depend on devid and phys_path values.
[08:04:32] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[08:05:04] <Reinhilde> like i can't just unzip a tribblix install minus kernel on a new dataset, set that to mountpoint=legacy, and have the tribblix kernel loaded off of a FAT partition mount root off that? (that's mechanistically identical to what i do with freebsd)
[08:06:43] <tsoome> sure you can, but kernel will not find the pool.
[08:07:17] <Reinhilde> or would i have to have the minimal viable illumos in a partition with init as a shim that mounts overlay the real root and exec's init from there?
[08:07:51] <Reinhilde> that could totally work, riiiiight?
[08:07:58] <tsoome> there is work going on to fix it, but for now, we depend on the values in pool label and if anything will change those (other os or you moving disk to another controller/system), kernel will not be able to use the pool.
[08:08:20] <Reinhilde> like at all, or just for boot?
[08:08:35] <tsoome> just for boot.
[08:08:36] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[08:09:39] <tsoome> well, it actually will go deeper than that but the boot is what we see first:)
[08:09:50] <Reinhilde> tsoome: so my hack with the minimal viable boot system on, say, a ufs could possibly work?
[08:10:05] <tsoome> ufs has the same issue.
[08:10:32] <Reinhilde> except that it'll always be at the same path
[08:10:50] <Reinhilde> unlike the zfs pool which will be schizophrenic
[08:10:57] <tsoome> ufs depends on bootpath variable containing proper device path provided to kernel.
[08:11:10] <tsoome> no it is not always the same path
[08:11:20] <Reinhilde> okaythen
[08:11:30] <tsoome> the same thing - move disk to different controller and you have different device path
[08:11:47] <zsj> tsoome: thanks for the info
[08:12:21] <Reinhilde> zsj: tl;dr, it's possible ... but it's a very bad idea
[08:12:29] <Reinhilde> and very doesn't work
[08:12:42] <tsoome> the work is going on to discover disk devices in early kernel, that will allow us to search the disks (for pool or ufs)
[08:13:42] <Reinhilde> is it so complicated that a lot of work really needs to be put into it
[08:14:08] <tsoome> ufs root would also need the way to identify the actual root file system, but since we do default to zfs root, ufs issues are not priority atm:)
[08:16:30] <Reinhilde> isn't there some way you can pass a /devices/pci-whatever/sd@0 path to the kernel and have it use that
[08:17:18] <tsoome> with uefi we actually can to some extent.
[08:18:01] <tsoome> as there we can access devicepaths
[08:18:32] <tsoome> translation is pain anyhow and im not sure it is worth the effort:)
[08:18:35] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[08:18:40] <tsoome> but we shall see.
[08:19:04] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: tsoome)
[08:23:20] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 276 seconds)
[08:26:27] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[08:30:43] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 245 seconds)
[08:35:50] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #illumos
[08:45:45] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[08:50:02] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 246 seconds)
[09:02:35] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[09:03:26] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[09:06:51] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 252 seconds)
[09:09:45] <Reinhilde> the berkeley students aren't right on everything, but a lot can be learned from them
[09:12:01] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[09:16:17] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[09:31:29] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[09:31:35] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[09:35:43] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 245 seconds)
[09:41:59] *** pwinder <pwinder!~pwinder@86.2.210.254> has quit IRC (Quit: Leaving)
[09:42:46] *** pwinder <pwinder!~pwinder@86.2.210.254> has joined #illumos
[09:46:31] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[09:48:24] *** andy_js <andy_js!~andy@97e29e78.skybroadband.com> has joined #illumos
[09:50:51] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 252 seconds)
[09:52:30] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[09:56:37] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[10:03:10] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[10:08:03] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 265 seconds)
[10:12:22] *** man_u <man_u!~manu@manu2.gandi.net> has joined #illumos
[10:18:40] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[10:22:56] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[10:33:24] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[10:37:47] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 252 seconds)
[10:39:59] *** wilbury <wilbury!otis@m68k.sinclairql.eu> has quit IRC (Ping timeout: 252 seconds)
[10:41:14] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Ping timeout: 240 seconds)
[10:41:23] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[10:43:20] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[10:48:17] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 276 seconds)
[11:01:25] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[11:05:36] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[11:06:44] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Read error: Connection reset by peer)
[11:07:21] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[11:12:44] *** wilbury <wilbury!otis@m68k.sinclairql.eu> has joined #illumos
[11:20:00] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[11:20:07] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[11:21:27] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[11:25:49] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 252 seconds)
[11:41:08] *** kahiru <kahiru!~quassel@ip-89-102-207-18.net.upcbroadband.cz> has quit IRC (Remote host closed the connection)
[11:41:55] *** kahiru <kahiru!~quassel@ip-89-102-207-18.net.upcbroadband.cz> has joined #illumos
[11:44:57] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[11:51:17] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[12:08:38] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[12:13:14] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 265 seconds)
[12:23:09] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[12:23:54] *** neirac <neirac!~cneira@pc-184-104-160-190.cm.vtr.net> has joined #illumos
[12:27:16] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[12:40:24] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[12:45:17] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 276 seconds)
[12:46:06] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[12:50:31] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 252 seconds)
[13:16:57] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[13:21:16] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[13:25:07] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Read error: Connection reset by peer)
[13:25:52] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[13:26:21] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Read error: Connection reset by peer)
[13:26:39] *** neirac <neirac!~cneira@pc-184-104-160-190.cm.vtr.net> has left #illumos ("Leaving")
[13:26:52] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[13:30:27] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[13:34:53] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 245 seconds)
[13:39:09] *** Ericson2314 <Ericson2314!ericson231@gateway/shell/matrix.org/x-zikrtcvpfdqnvskv> has quit IRC (Write error: Connection reset by peer)
[13:39:09] *** GrahamPerrin[m] <GrahamPerrin[m]!grahamperr@gateway/shell/matrix.org/x-uvqdmglzomcfgyht> has quit IRC (Write error: Connection reset by peer)
[13:39:21] *** jhot[m] <jhot[m]!jhotmatrix@gateway/shell/matrix.org/x-yciyrijngwbpqaby> has quit IRC (Remote host closed the connection)
[13:39:26] *** psydroid <psydroid!psydroidma@gateway/shell/matrix.org/x-odnfmblxnmbcfzzu> has quit IRC (Remote host closed the connection)
[13:41:14] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Ping timeout: 240 seconds)
[13:46:56] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[13:47:56] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has joined #illumos
[13:51:23] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 252 seconds)
[14:10:48] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[14:11:42] *** steph <steph!steph@minos.ber.rdev.info> has joined #illumos
[14:15:38] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 276 seconds)
[14:21:39] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[14:25:51] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 252 seconds)
[14:28:09] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: tsoome)
[14:37:28] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[14:41:27] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has joined #illumos
[14:41:58] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 245 seconds)
[14:45:36] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Ping timeout: 240 seconds)
[14:47:54] *** psydroid <psydroid!psydroidma@gateway/shell/matrix.org/x-kixxqbpzpkbnsvsd> has joined #illumos
[14:47:54] *** jhot[m] <jhot[m]!jhotmatrix@gateway/shell/matrix.org/x-zcstpuyhefhuaevr> has joined #illumos
[14:47:55] *** GrahamPerrin[m] <GrahamPerrin[m]!grahamperr@gateway/shell/matrix.org/x-uvccgllqinjighsb> has joined #illumos
[14:47:55] *** Ericson2314 <Ericson2314!ericson231@gateway/shell/matrix.org/x-swaljhoccubovsyw> has joined #illumos
[14:56:07] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-141-249.rochester.res.rr.com> has quit IRC (Ping timeout: 265 seconds)
[14:59:17] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[15:03:37] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 252 seconds)
[15:23:19] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[15:27:48] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 245 seconds)
[15:40:21] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[15:44:37] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[15:45:11] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[15:54:24] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[15:58:57] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 265 seconds)
[16:18:21] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[16:25:37] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[16:25:56] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[16:26:00] *** mnrmnaugh is now known as notmnrmnaugh
[16:26:08] *** notmnrmnaugh is now known as mnrmnaugh
[16:45:54] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[16:49:15] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has joined #illumos
[16:49:38] *** dopplergange <dopplergange!~dop@45.12.223.70> has quit IRC (Ping timeout: 240 seconds)
[16:50:58] *** dopplergange <dopplergange!~dop@45.12.223.70> has joined #illumos
[16:51:06] *** KungFuJesus <KungFuJesus!~adamstyli@207.250.97.74> has quit IRC (Quit: leaving)
[16:55:32] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 276 seconds)
[16:59:51] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: tsoome)
[17:04:08] *** KungFuJesus <KungFuJesus!~adamstyli@207.250.97.74> has joined #illumos
[17:21:15] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[17:25:37] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[17:33:01] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 268 seconds)
[17:39:44] *** myrkraverk_ <myrkraverk_!~chatzilla@180.190.57.141> has quit IRC (Ping timeout: 276 seconds)
[17:57:54] <danmcd> An old IETF contact asked me if there's any illumos for GCE. I know we have Azure and AWS covered-ish, but AFAIK not GCE.
[17:58:08] <danmcd> If I'm missing something, I'd love to know it.
[17:58:51] <jlevon> LeftWing was getting GCE working. dunno of status
[18:00:31] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[18:00:38] *** _Tenchi_ <_Tenchi_!~phil@207.255.80.203> has joined #illumos
[18:05:40] <danmcd> Okay. I told someone GCE was a no-go, use AWS or Azure instead.
[18:06:02] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 240 seconds)
[18:11:21] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[18:19:41] <rmustacc> danmcd: Right now it's still a WIP. We'll need to finish up virtio scsi and a few other things.
[18:29:51] <gitomat> [illumos-gate] 11825 PKCS#11 CKM_AES_CBC_PAD decryption can fail -- Jason King <jason.king at joyent dot com>
[18:33:43] *** man_u <man_u!~manu@manu2.gandi.net> has quit IRC (Quit: man_u)
[18:35:14] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[18:37:27] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Remote host closed the connection)
[18:52:57] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[18:57:23] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 245 seconds)
[19:09:18] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)
[19:12:35] <danmcd> Cool, so my view of reality IS reality.
[19:12:39] <danmcd> Thx rmustacc
[19:17:36] <alanc> Happy End of Embargo!
[19:18:16] <alanc> https://blogs.intel.com/technology/2019/11/ipas-november-2019-intel-platform-update-ipu/
[19:20:20] <rmustacc> There are some new side channels in this one. Most notably TAA. However, for most platforms, the prior work should have you covered, excepting certain CLX steppings, IIRC.
[19:21:40] <rmustacc> jlevon: Do you happen to have the TAA changes?
[19:21:51] <jlevon> sending now
[19:22:12] <rmustacc> How much changed from what I had originally?
[19:22:31] <jlevon> basically just to always disable tsx if we can
[19:23:24] <rmustacc> Seems a reasonable choice.
[19:24:57] <jlevon> I don't have the ucode update ready to go though anyway.
[19:40:01] *** danmcd <danmcd!~danmcd@static-71-174-113-16.bstnma.fios.verizon.net> has quit IRC (Read error: Connection reset by peer)
[19:50:05] *** danmcd <danmcd!~danmcd@static-71-174-113-16.bstnma.fios.verizon.net> has joined #illumos
[19:50:41] <danmcd> (Of course I get disconnected right after alanc's celebration of end-of-embargo day. Time to go check the archives...)
[19:51:29] <danmcd> Oh, I didn't miss anything. phew.
[19:52:56] <alanc> I see intel has also published the gas patches now so that you don't lose performance on Skylake & later by just installing the new microcode, due to one of the non-security errata fixes
[19:53:43] <alanc> That's https://www.intel.com/content/www/us/en/support/articles/000055650/processors/intel-xeon-processors.html & https://sourceware.org/ml/binutils/2019-11/msg00173.html
[19:54:44] <LeftWing> danmcd, jlevon: There are not many things left to go. I just uncovered another apparent issue with their PIT emulation, but one we can easily work around. Then I have vioscsi roughly working, need to polish. There's an issue with MSI-X interrupts seemingly not arriving for vioif
[19:54:54] <jlevon> cool
[19:54:56] <LeftWing> And the DHCP support for MTU and (off-subnet) Classless Routes
[19:55:12] <danmcd> Very cool. Thx LeftWing
[19:55:47] <danmcd> Oh and if anyone is bored, I'm looking for reviewers (and in some cases reproducing testing) for illumos#11928, illumos#11927, illumos#2988, and illumos#11945
[19:56:08] <danmcd> I know MarcelT has been helping, e.g.
[20:09:02] *** despair86 <despair86!~despair@75-60-137-0.lightspeed.rcsntx.sbcglobal.net> has joined #illumos
[20:09:26] <despair86> hi
[20:09:56] <despair86> does illumos link editor have the same bug
[20:09:57] <despair86> # 64-bit Solaris/x86 ld breaks calls to __tls_get_addr with non-TLS
[20:09:58] <despair86> # relocs. Work around by disabling TLS transitions. Not necessary
[20:09:58] <despair86> # on 32-bit x86, but cannot be distinguished reliably in specs.
[20:13:49] *** pwinder <pwinder!~pwinder@86.2.210.254> has quit IRC (Quit: This computer has gone to sleep)
[20:14:15] <jbk> there's been some TLS related bugs fixed with ld
[20:14:34] <jbk> though it might help to have a bit more context
[20:14:37] <despair86> yeah, was building GDC because openindiana doesn't ship it
[20:14:44] <despair86> the d compiler
[20:14:59] <jbk> oh
[20:15:06] <despair86> and apparently GDC doesn't play well with either oracle as or ld
[20:15:09] <jbk> most of the stdlib still needs work
[20:15:36] <jbk> i kinda gave up on it when they kept saying they're a systems language, but refuse to even entertain anythin approaching a stable ABI
[20:16:02] <jbk> and they've kinda gone template crazy
[20:16:19] <despair86> eyah
[20:16:23] <despair86> *yeah
[20:16:41] <jbk> where even the most the most trivial bit of code requires using at least 12 different templates
[20:17:12] <jbk> the big thing (this has been several years) is that a lot of the tests might fail
[20:17:30] <jbk> because they need to enable c99 behavior in libc
[20:18:08] <jbk> but if they have a specific test case, that would help
[20:18:21] <jbk> because there were several TLS related bugs that rich fixed
[20:21:52] <despair86> oh
[20:21:58] <jbk> 10471 ld(1) amd64 LD->LE TLS transition causes memory corruption
[20:22:03] <jbk> is one that comes to mind
[20:23:13] <jbk> and 10267
[20:23:32] <despair86> even worse is that LDC claimed to support smartos back in 2016, even has a CI test runner, but LDC 1.x crashes LLVM after the first-stage bootstrap
[20:23:55] <despair86> and the LDC unit tests haven't passed in ages
[20:24:25] <despair86> DMD ofc dates back to the Symantec era and doesn't even support UNIX at all, just linux i think
[20:29:13] <LeftWing> danmcd, jlevon: I filed https://www.illumos.org/issues/11968
[20:57:44] <despair86> i was told that the "betterC" mode of operation was intended to be stable
[20:59:46] *** mxzcrit <mxzcrit!uid382946@gateway/web/irccloud.com/x-auogsikxrynycxlo> has joined #illumos
[21:02:57] <jbk> if i get some free time (ha!) i might get back to it at some point
[21:03:06] <jbk> though there's a lot of things ahead of it on the list now :)
[21:20:06] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[21:23:19] *** ESol5 <ESol5!~a@unaffiliated/asol5> has joined #illumos
[21:32:59] *** despair86 <despair86!~despair@75-60-137-0.lightspeed.rcsntx.sbcglobal.net> has quit IRC (Remote host closed the connection)
[21:36:40] <gitomat> [illumos-gate] 11952 large USB hard disks experience I/O failures -- Joshua M. Clulow <josh at sysmgr dot org>
[21:48:19] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 268 seconds)
[21:49:45] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[21:50:06] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[22:15:42] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has joined #illumos
[22:16:38] *** Yogurt <Yogurt!~Yogurt@c-73-189-45-147.hsd1.ca.comcast.net> has quit IRC (Remote host closed the connection)
[22:20:18] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 265 seconds)
[22:24:17] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[22:27:16] *** MarcelT <MarcelT!~marcel@tortuga.telka.sk> has quit IRC (Ping timeout: 240 seconds)
[22:30:55] *** steph <steph!steph@minos.ber.rdev.info> has quit IRC (Quit: quit)
[22:31:55] *** steph <steph!~steph@minos.ber.rdev.info> has joined #illumos
[22:36:08] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 276 seconds)
[22:45:32] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[22:57:36] *** Kurlon_ <Kurlon_!~Kurlon@bidd-pub-04.gwi.net> has joined #illumos
[23:01:05] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 268 seconds)
[23:02:21] *** Kurlon_ <Kurlon_!~Kurlon@bidd-pub-04.gwi.net> has quit IRC (Ping timeout: 265 seconds)
[23:07:34] <Reinhilde> is there a code path that runs the device scanning routines we would need to mount a zfs pool without relying on dev_id and physpath
[23:11:35] <rzezeski> Can anyone tell me what the hell the CALLB_CPR_SAFE_BEGIN/END stuff is for?
[23:13:04] <jlevon> the proximal answer is cpr_stop_kernel_threads() but ...
[23:14:19] <rzezeski> For anyone wondering why I'm asking, mac has places where it will guard a cv_wait with those macros, but other places where it won't. And in the past I've just ignored it. Today I'm curious to know its history and if it matters anymore.
[23:14:46] <jlevon> I'm wondering if it used to be the case that some old sparc CPR stuff really had to stop all threads from doing something before doing cpr
[23:15:10] <jlevon> the modern* way is just to quiesce all the CPUs by forcing them to spin somewhere 'safe'
[23:15:33] <jlevon> but cpr_suspend() just describes what it's doing rather than why
[23:15:52] <jbk> int x = 1; /* assign 1 to x */
[23:15:53] <jbk> :)
[23:16:05] <rzezeski> lolwut
[23:16:08] <rzezeski> is that a real comment
[23:16:28] <jbk> i hope not :)
[23:17:56] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has quit IRC (Quit: Leaving)
[23:18:57] *** sjorge <sjorge!~sjorge@unaffiliated/sjorge> has quit IRC (Quit: 410 Gone)
[23:19:09] <jlevon> rzezeski: so for sure we don't go through cpr_suspend() for VM-style suspend/resume/live migration
[23:19:18] <jlevon> I'm also pretty sure we don't for S3 on x86 either
[23:20:24] <rzezeski> So I guess the point of the macro is to mark where it's safe to suspend/resume?
[23:21:06] <rzezeski> so something would call into this facility and wait for each thread to call the "safe" entry?
[23:21:09] <jlevon> yeah, if the thread is already sleeping, it'll stay sleeping
[23:21:27] <jlevon> I'm 90% sure it's junk and can all be ripped out but
[23:21:33] <rzezeski> okay, now this makes a _bit_ more sense
[23:22:05] <rzezeski> if we could rip it out that would be cool, because it's a bit ugly
[23:22:05] <jlevon> like you say, who knows when this last worked
[23:22:11] <jlevon> kill it with fire
[23:22:56] <rzezeski> well, based on what you just said I think having the one being/end call is fine, but yea, I'm not sure how well this has all been kept up given that no one ever said a thing to me about it and it's not mentioned in WDD at all or any of the man pages.
[23:23:41] <jlevon> quick email to the list suggesting to kill it?
[23:23:57] <rzezeski> Yea I'll send out an email and see what the deal is.
[23:24:08] *** sjorge <sjorge!~sjorge@unaffiliated/sjorge> has joined #illumos
[23:28:38] *** andy_js <andy_js!~andy@97e29e78.skybroadband.com> has quit IRC (Quit: andy_js)
[23:30:05] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[23:43:30] <LeftWing> ;oz
[23:43:32] <LeftWing> .. sigh
[23:45:36] *** alanc <alanc!~alanc@inet-hqmc02-o.oracle.com> has quit IRC (Remote host closed the connection)
[23:46:04] *** alanc <alanc!~alanc@inet-hqmc02-o.oracle.com> has joined #illumos
top

   November 12, 2019  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >