Switch to DuckDuckGo Search
   May 5, 2020  
< | 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 | 31 | >

Toggle Join/Part | bottom
[00:01:30] <neirac> toastersonerson route + dns I think that's handled by the brand
[00:01:59] <toastersonerson> neirac (IRC): yes but I needed code to handle it for the pod brand
[00:02:27] <toastersonerson> so I need to exclude it if it's not that brand or for brands, that do it themselves
[00:02:43] <neirac> toastersonerson oh ok, on the pod brand or in the go code?
[00:02:55] *** andy_js <andy_js!~andy@90.218.209.121> has quit IRC (Quit: andy_js)
[00:02:58] <toastersonerson> it's in the go code
[00:03:20] <toastersonerson> but I need to check how I can best extract it.
[00:03:37] <KungFuJesus> tsoome__: can you tell me if the trays between the T4 and T5120 are identical? E.g., will the light pipes line up?
[00:06:27] <azarus> Okay, I think I got a bit of a grip on ppriv now. However, how do I grant a privilege when the only sets that can be modified are L (limit) and I (inheritable)?
[00:11:06] <azarus> I tried "L-proc_clock_highres" "L+proc_clock_highres" and the same with 'I'
[00:11:49] <KungFuJesus> I'm dangerous close to buying this, I already picked out a hostname...
[00:12:02] <KungFuJesus> dangerously*
[00:12:41] <rmustacc> azarus: So with ppriv you'd have to start rom a privileged user and drop things down to that point.
[00:14:08] <azarus> rmustacc: That makes me wonder what user would be suitable, because the only user in an LX zone with those kinds of privs is root
[00:14:27] <azarus> (at least as far as I know)
[00:14:48] <rmustacc> The mapping in lx isn't great, but I think at that intersection that's what you'd need to do.
[00:15:00] <rmustacc> Is start with root and run it in the context of limited privs.
[00:15:45] <azarus> Hm, okay. I'll have a go with that.
[00:16:02] <rmustacc> Especially since pfexecd isn't running there.
[00:16:12] <rmustacc> But I'm trying to go see if we can get that change in to reduce the need the privs.
[00:16:17] <rmustacc> So that might help you out there.
[00:16:34] <azarus> Yeah, that'd be great for the future. thank you ^^
[01:06:13] *** phyre <phyre!~phyre___@78.30.23.107> has quit IRC (Ping timeout: 264 seconds)
[01:06:16] <ypankov> Looking at rmustacc's review, I noticed the "LIST_CLASS_ENTRY.3head"-like symlinks; looks completely wrong to me.
[01:07:39] <ypankov> (unrelated to review itself, just looking at file contents in there)
[01:09:42] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[01:12:15] <rmustacc> Why is it wrong?
[01:13:22] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[01:13:23] <rmustacc> It is something the manual page documents, though it could be clearer.
[01:14:51] <ypankov> Those are not headers, so it's just confusing
[01:15:36] <rmustacc> They are a header.
[01:15:50] <ypankov> Probably queue.h.3head should just says that it contains the following macros, and macros themselves should be in section 3 or elsewhere
[01:15:56] <rmustacc> Or rather, queue.h is a header and you want to have synonyms.
[01:16:04] <LeftWing> Right, they're _in_ a header
[01:16:40] <LeftWing> And it's nice to be able to "man STAILQ_REMOVE_AFTER"
[01:17:36] <rmustacc> They are in a subsection of section 3. It sounds like the issue is that you don't feel 3HEAD is good for that. Admittidly this is one of the few interfaces that is entirely and only implemented in a header.
[01:17:48] <ypankov> Yes, man STAILQ_REMOVE_AFTER.3, not .3head, as there's no such file as STAILQ_REMOVE_AFTER.h
[01:17:57] <rmustacc> And therefore 3HEAD feels weird. But it's not clear what other subsction of 3 exists.
[01:18:58] <rmustacc> I don't know. If we had other things that were documented only in a header, it's not clear they should have a separate pointer / section. Though I can see how it is confusing in some ways
[01:19:10] <rmustacc> However, you'll note that all the things in 3head that are actually a header are foo.h.3head.
[01:19:24] <rmustacc> So the 3head section can have names that aren't headers.
[01:19:43] <rmustacc> If you're in a world where it's not all in just section 3, it's not clear that it's wrong for it be there.
[01:20:07] <LeftWing> I guess Solaris put them in 3EXT
[01:20:33] <LeftWing> We could move them there
[01:20:34] <rmustacc> But there are definitely different ways to approach and document a system.
[01:21:46] <ypankov> OK, it just somehow feels wrong to me, nothing needs to change.
[01:22:22] <rmustacc> I don't think either of us are saying it can't change. Just that header only things are also pretty uncommon so there's less of a standard pattern here.
[01:23:10] <ypankov> And I was a reviewer for a change that added it, so I'm a bit late I guess :)
[01:24:18] *** ypankov <ypankov!~ypankov@85.175.17.187> has quit IRC (Quit: brb)
[01:24:41] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[01:24:49] <rmustacc> Eh, it's fine. We can all change opinions.
[01:25:02] <rmustacc> No one is required to notice everything in review.
[01:25:09] <LeftWing> He ducked out
[01:28:50] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[01:29:05] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[01:35:48] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[01:36:18] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[01:40:42] *** ypankov <ypankov!~ypankov@85.175.17.187> has joined #illumos
[01:42:35] <ypankov> Having a review in gerrit, should I include the patch in RTI, or would the gerrit link be enough?
[01:43:03] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[01:43:17] <LeftWing> I think the Gerritt link is fine. Also, https://code.illumos.org/~diff/<sha> is wired up to take you to a diff
[01:43:23] <rmustacc> For the time being, I'd suggest including a patch, like you would if rb or otherwise, but including a link to the gerrit review is useful.
[01:43:49] <ypankov> Oh, updating the review with reviewers made all the +1 disappear
[01:43:53] * ypankov feels stupid
[01:43:59] <LeftWing> huh
[01:44:08] <LeftWing> Did you happen to rebase at the same time
[01:44:11] <ypankov> I mean I updated the commit message and pushed
[01:44:17] <ypankov> LeftWing: likely yes.
[01:44:47] <LeftWing> There's a small problem with that. If you rebase with no message changes, it can tell. If you change the message with no rebasing or code changes, it can tell.
[01:44:57] <LeftWing> It's only when you do them both together that it gives up and removes the votes
[01:45:05] <ypankov> Cool.
[01:45:14] <LeftWing> Because it is less confident about it I guess
[01:45:22] <ypankov> Something I need to remember then.
[01:45:25] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[01:45:26] <rmustacc> We should probably take a look at what is involved in fixing that.
[01:45:31] <rmustacc> Since I suspect a lot of folks'll hit that.
[01:45:50] <LeftWing> Yeah, we could do that. I suspect it will require patches to Gerrit.
[01:46:11] <LeftWing> I don't think I was able to configure it to do the right thing, at any rate, when I tried
[01:46:32] <rmustacc> Maybe worth filing a bug upstream?
[01:47:07] <LeftWing> I can't imagine there isn't one filed already
[01:47:32] <LeftWing> (Which is to say, whoever files should look first)
[01:51:32] <ypankov> and pbchk isn't happy about Change-Id line
[01:51:50] <ypankov> and I can't get rid of it as it's in hooks :)
[01:52:24] <LeftWing> The advocate will massage the patch on the way in anyway, I wouldn't worry about that
[01:52:33] <ypankov> OK, good then
[01:52:39] <rmustacc> (I've been deleting the hook so far)
[01:52:45] <LeftWing> haha
[01:52:53] <ypankov> Right!
[01:53:10] <LeftWing> I modified mine to allow me to override it with an environment variable
[01:53:24] *** vgusev <vgusev!~vgusev@188.187.60.230> has joined #illumos
[01:58:06] *** vgusev <vgusev!~vgusev@188.187.60.230> has quit IRC (Ping timeout: 260 seconds)
[02:00:51] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[02:07:59] *** baojg <baojg!~baojg@2604:a880:0:1010::16:1> has joined #illumos
[02:09:45] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[02:14:35] *** amrfrsh <amrfrsh!~Thunderbi@134.19.189.92> has quit IRC (Ping timeout: 260 seconds)
[02:20:00] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[02:21:31] *** amrfrsh <amrfrsh!~Thunderbi@134.19.189.92> has joined #illumos
[02:26:37] *** ypankov <ypankov!~ypankov@85.175.17.187> has quit IRC (Quit: leaving)
[02:37:24] *** baojg <baojg!~baojg@2604:a880:0:1010::16:1> has quit IRC (Remote host closed the connection)
[02:37:39] *** baojg <baojg!~baojg@2604:a880:0:1010::16:1> has joined #illumos
[02:38:11] *** baojg <baojg!~baojg@2604:a880:0:1010::16:1> has quit IRC (Remote host closed the connection)
[02:39:30] *** ypankov <ypankov!~ypankov@85.175.17.187> has joined #illumos
[02:50:45] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[02:58:01] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[03:03:38] <gitomat> [illumos-gate] 12454 remove lint library machinery -- Richard Lowe <richlowe at richlowe dot net>
[03:03:39] <gitomat> [illumos-gate] 12447 should install library compat symlinks from library makefiles, not Targetdirs -- Richard Lowe <richlowe at richlowe dot net>
[03:16:10] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[03:17:18] *** neirac <neirac!~neirac@pc-4-149-45-190.cm.vtr.net> has quit IRC (Quit: bye)
[03:19:19] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[03:21:58] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[03:25:07] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[03:35:48] <ypankov> That "Mail for Windows 10" program Garrett is using seems to break threading :\
[03:41:14] *** neirac_ <neirac_!~neirac@pc-4-149-45-190.cm.vtr.net> has joined #illumos
[03:41:59] *** neirac_ is now known as neirac
[03:46:55] <LeftWing> It sure does
[03:53:42] <rmustacc> liv3010m: If you're not watching it, I updated https://www.illumos.org/issues/12686 with some additional information.
[04:04:23] <gitomat> [illumos-gate] 11807 "private use area" characters should be marked as "printable" -- Yuri Pankov <ypankov at fastmail dot com>
[04:08:00] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[05:04:09] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[05:04:23] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Client Quit)
[05:25:55] *** jeffpc <jeffpc!~jeffpc@josefsipek.net> has quit IRC (Ping timeout: 258 seconds)
[05:26:07] *** jeffpc <jeffpc!~jeffpc@josefsipek.net> has joined #illumos
[05:37:10] *** wl_ <wl_!~wl_@2605:6000:1b0c:46ad::87c> has joined #illumos
[05:39:09] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[07:02:28] *** leoric <leoric!~leoric@46.147.99.173> has joined #illumos
[07:03:41] <leoric> Hi. Do I understand correctly that 'hma_vmx_init: CPU does not support VMX' notice means that on this platform hypervisors can't coexist?
[07:04:10] <leoric> I'm just interested if it should be on console or in logs...
[07:05:32] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Ping timeout: 256 seconds)
[07:13:35] <LeftWing> leoric: It means that hypervisors that need VMX can't run at all
[07:13:48] <LeftWing> But I agree it should go in the log, if it's printed at all
[07:14:22] <LeftWing> It seems more appropriate for the hypervisor software to tell you there's no support when you try to run it
[07:14:44] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[07:16:19] *** BOKALDO <BOKALDO!~BOKALDO@87.110.102.49> has joined #illumos
[07:26:48] *** leoric <leoric!~leoric@46.147.99.173> has quit IRC (Remote host closed the connection)
[07:28:43] *** leoric <leoric!~leoric@46.147.99.173> has joined #illumos
[07:50:21] *** nbhauke <nbhauke!~hauke@55d42c22.access.ecotel.net> has joined #illumos
[08:04:45] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has joined #illumos
[08:12:58] *** neuroserve <neuroserve!~toens@ip-88-152-243-25.hsi03.unitymediagroup.de> has quit IRC (Ping timeout: 272 seconds)
[09:02:15] *** tsoome__ <tsoome__!~tsoome@91.209.240.229> has quit IRC (Read error: Connection reset by peer)
[09:03:34] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[09:03:39] *** neuroserve <neuroserve!~toens@ip-95-222-232-201.hsi15.unitymediagroup.de> has joined #illumos
[09:03:58] *** BrownBear <BrownBear!~BrownBear@107.175.247.247> has quit IRC (Ping timeout: 256 seconds)
[09:03:58] *** freezing_ <freezing_!~weechat@ec2-13-57-230-193.us-west-1.compute.amazonaws.com> has quit IRC (Ping timeout: 256 seconds)
[09:06:19] *** freezing_ <freezing_!~weechat@ec2-13-57-230-193.us-west-1.compute.amazonaws.com> has joined #illumos
[09:06:46] *** edef_ <edef_!edef@NixOS/user/edef> has joined #illumos
[09:06:46] *** edef <edef!edef@NixOS/user/edef> has quit IRC (Killed (cherryh.freenode.net (Nickname regained by services)))
[09:06:46] *** edef_ is now known as edef
[09:09:41] *** andy_js <andy_js!~andy@90.218.209.121> has joined #illumos
[09:09:46] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (*.net *.split)
[09:09:46] *** patdk-lap <patdk-lap!~patrickdk@208.94.189.75> has quit IRC (*.net *.split)
[09:09:46] *** KungFuJesus <KungFuJesus!~adam@72.49.160.91> has quit IRC (*.net *.split)
[09:09:47] *** tsundoku <tsundoku!~tsundoku@irc.tsundoku.ne.jp> has quit IRC (*.net *.split)
[09:09:47] *** xzilla <xzilla!~robert@pool-71-166-61-141.bltmmd.fios.verizon.net> has quit IRC (*.net *.split)
[09:09:47] *** Mokou <Mokou!~hokuto@li1608-68.members.linode.com> has quit IRC (*.net *.split)
[09:09:47] *** rmustacc <rmustacc!~daemar@zelgadis.fingolfin.org> has quit IRC (*.net *.split)
[09:09:47] *** anthis <anthis!~anthis@essential.coach> has quit IRC (*.net *.split)
[09:09:47] *** nahamu <nahamu!~nahamu@67.158.54.193> has quit IRC (*.net *.split)
[09:12:46] *** Teknix <Teknix!~pds@69.41.134.110> has quit IRC (Ping timeout: 246 seconds)
[09:14:59] *** Teknix <Teknix!~pds@69.41.134.110> has joined #illumos
[09:15:57] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[09:15:57] *** patdk-lap <patdk-lap!~patrickdk@208.94.189.75> has joined #illumos
[09:15:57] *** KungFuJesus <KungFuJesus!~adam@72.49.160.91> has joined #illumos
[09:15:57] *** tsundoku <tsundoku!~tsundoku@irc.tsundoku.ne.jp> has joined #illumos
[09:15:57] *** xzilla <xzilla!~robert@pool-71-166-61-141.bltmmd.fios.verizon.net> has joined #illumos
[09:15:57] *** Mokou <Mokou!~hokuto@li1608-68.members.linode.com> has joined #illumos
[09:15:57] *** rmustacc <rmustacc!~daemar@zelgadis.fingolfin.org> has joined #illumos
[09:15:57] *** anthis <anthis!~anthis@essential.coach> has joined #illumos
[09:15:57] *** nahamu <nahamu!~nahamu@67.158.54.193> has joined #illumos
[09:15:57] *** barjavel.freenode.net sets mode: +o rmustacc
[09:17:12] *** BrownBear <BrownBear!~BrownBear@107.175.247.247> has joined #illumos
[09:19:47] *** pwinder <pwinder!~pwinder@86.11.191.8> has joined #illumos
[09:27:25] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 256 seconds)
[09:31:37] *** kohju <kohju!~kohju@gw.justplayer.com> has quit IRC (Excess Flood)
[09:32:10] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[09:32:10] *** Exagone313 <Exagone313!exa@3131313.xyz> has quit IRC (Quit: see ya!)
[09:32:22] *** Exagone313 <Exagone313!exa@3131313.xyz> has joined #illumos
[09:32:41] *** jimklimov <jimklimov!~jimklimov@ip-86-49-254-26.net.upcbroadband.cz> has joined #illumos
[09:33:52] *** kohju <kohju!~kohju@gw.justplayer.com> has joined #illumos
[09:37:11] *** phyre <phyre!~phyre___@78.30.23.107> has joined #illumos
[09:48:25] *** leoric <leoric!~leoric@46.147.99.173> has quit IRC (Quit: Konversation terminated!)
[09:49:42] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #illumos
[09:54:56] *** nbhauke <nbhauke!~hauke@55d42c22.access.ecotel.net> has quit IRC (Quit: nbhauke)
[10:02:31] *** leoric <leoric!~leoric@46.147.99.173> has joined #illumos
[10:07:25] *** tsoome__ <tsoome__!~tsoome@91.209.240.229> has joined #illumos
[10:08:15] *** jimklimov <jimklimov!~jimklimov@ip-86-49-254-26.net.upcbroadband.cz> has quit IRC (Quit: Leaving.)
[10:10:07] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Ping timeout: 260 seconds)
[10:31:22] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[10:37:03] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[10:37:32] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[10:44:11] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[10:46:25] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[10:54:36] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[11:03:23] *** cantstanya <cantstanya!~chatting@gateway/tor-sasl/cantstanya> has quit IRC (Ping timeout: 240 seconds)
[11:10:43] *** cantstanya <cantstanya!~chatting@gateway/tor-sasl/cantstanya> has joined #illumos
[11:26:51] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[11:32:00] <andyf> leoric: yes, hypervisors can co-exist if they use the framework properly
[11:32:25] <andyf> On OmniOS we send that message to the log and not the console (just added a ? to the start of the message)
[11:32:42] <andyf> So once bhyve lands, you can run KVM and bhyve alongside each other
[11:35:07] <andyf> The patch I did for virtualbox doesn't
[11:35:24] <andyf> know how to co-operate, so you have to shut down all KVM and bhyve VMs before starting a Virtualbox one
[11:35:37] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[11:36:17] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[11:38:26] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[11:38:27] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[11:41:37] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[11:51:04] *** jimklimov <jimklimov!~jimklimov@78.80.224.132> has joined #illumos
[12:03:52] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has quit IRC (Ping timeout: 260 seconds)
[12:14:52] <EisNerd> I have a really bad day
[12:16:29] <EisNerd> again yes, but was an accident :(
[12:16:37] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[12:18:26] <toastersonerson> EisNerd (IRC): That one server still bugging you?
[12:19:07] <EisNerd> someone an idea howto find a botlleneck in smb server, as I tried transfer from/to >200MB/s zfs over i40e qsfp+ link to an linux with connectx4 edr vpi (100G), according to iperf this link is god for 5Gbit/s in one and 20 to 24 GBit/s in the other direction (no idea why)
[12:19:24] <EisNerd> but over smb I only get 60 to 80MB/s
[12:19:51] <toastersonerson> tcpdump?
[12:19:56] <EisNerd> toastersonerson: yes, I just opened a case on HPE support as it looks like a problem with ilo integration in esxi
[12:20:35] <EisNerd> toastersonerson: I get same speeds when using a standard GBit link on the client end
[12:20:58] <EisNerd> and same client to out old 2012 file server gives me 100 to 110MB/s
[12:21:03] <toastersonerson> ough HPE. I lost all confidence in them when they denied a bug was real even though their power management sucked...
[12:21:10] <sensille> EisNerd: what is the RTT between client and server?
[12:21:24] <EisNerd> so there is sth really bad with SMB server on illumos / oi hipster
[12:21:25] <toastersonerson> I was thinking of a routing problem on the illumo server.
[12:21:48] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[12:22:13] <EisNerd> I disabled the tge if now and the server is completely running over the fge link with both IPs
[12:22:36] <EisNerd> you mean ping time?
[12:22:41] <EisNerd> sensille: ?
[12:23:14] <sensille> yeah, round trip time
[12:23:29] <EisNerd> let me check
[12:23:39] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[12:24:10] <sensille> are you testing with one big file?
[12:24:58] <EisNerd> from my ws (the other GBit client), from the linux in vbox: rtt min/avg/max/mdev = 0.373/0.481/0.681/0.118 ms
[12:25:13] <EisNerd> sensille: yes
[12:27:07] <EisNerd> according to iperf the link is fast enough, so where is the problem?
[12:27:14] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[12:28:01] <toastersonerson> EisNerd (IRC): is the illumos system multihomed?
[12:28:19] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[12:28:35] *** Tempt_ <Tempt_!~avenger@mexico.purplecow.org> has joined #illumos
[12:28:46] *** kovert <kovert!~kovert@204.141.173.249> has quit IRC (Ping timeout: 246 seconds)
[12:28:56] <sensille> there are 3 main bottlenecks with smb: a) tcp window b) smb request size c) number of smb requests in flight
[12:28:56] <sensille> i have zero experience with the illumos smb server, though
[12:29:03] *** kovert <kovert!~kovert@204.141.173.249> has joined #illumos
[12:29:12] *** Mazon <Mazon!~Mazon@37.205.127.168> has quit IRC (Read error: Connection reset by peer)
[12:29:12] *** Cthulhux <Cthulhux!~cthulhu@piratenpartei/ni/tux> has quit IRC (Read error: Connection reset by peer)
[12:29:23] *** Cthulhux <Cthulhux!cthulhu@rosaelefanten.org> has joined #illumos
[12:29:25] *** Tempt <Tempt!~avenger@unaffiliated/tempt> has quit IRC (Remote host closed the connection)
[12:29:33] *** wiedi <wiedi!~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de> has quit IRC (Read error: Connection reset by peer)
[12:29:39] *** wiedi_ <wiedi_!~wiedi@91.64.150.166> has joined #illumos
[12:29:41] *** danmcd <danmcd!~danmcd@static-71-174-113-16.bstnma.fios.verizon.net> has quit IRC (Read error: Connection reset by peer)
[12:29:45] <jimklimov> technically also the frame size (standard or one of Jumbo ethernet MTUs), so the portion of time and processing spent on overheads vs data
[12:29:49] *** kahiru <kahiru!~quassel@ip-89-102-207-18.net.upcbroadband.cz> has quit IRC (Remote host closed the connection)
[12:29:53] *** Mazon <Mazon!~Mazon@37.205.127.168> has joined #illumos
[12:30:00] <jimklimov> I assume in this test both clients had same settings?
[12:30:11] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Ping timeout: 246 seconds)
[12:30:11] *** tsoome__ is now known as tsoome
[12:30:24] <sensille> EisNerd: if the window size is too small, the link can't be utilized
[12:30:24] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has joined #illumos
[12:30:49] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Ping timeout: 264 seconds)
[12:30:55] <sensille> tcpdump can show you where the bottleneck is, but it's not that easy to read
[12:31:04] *** kahiru <kahiru!~quassel@ip-89-102-207-18.net.upcbroadband.cz> has joined #illumos
[12:31:22] <EisNerd> hm iperf also reports non perfect link speeds
[12:31:55] <EisNerd> so tcp window, how cna I check this on oi?
[12:32:26] <sensille> i think netstat -an can show windows of running connections
[12:34:08] *** jimklimov <jimklimov!~jimklimov@78.80.224.132> has quit IRC (Read error: Connection reset by peer)
[12:39:13] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[12:39:23] *** jimklimov <jimklimov!~jimklimov@78.80.224.132> has joined #illumos
[12:39:39] <sensille> if you're trying to understand network performance the concept of windowing is one of the most important ones
[12:40:14] <sensille> and with smb you basically have nested windowing. the inner window from tcp, and the outer from smb
[12:40:17] <sensille> same with ssh
[12:46:12] <ypankov> I think these are the properties now: ipadm show-prop -p send_buf,recv_buf tcp
[12:51:15] *** jollyd <jollyd!~alarcher@2a01:cb1d:611:bb00:8820:7cbd:8965:af56> has joined #illumos
[13:20:54] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[13:23:10] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[13:25:32] *** jollyd <jollyd!~alarcher@2a01:cb1d:611:bb00:8820:7cbd:8965:af56> has quit IRC (Quit: Leaving.)
[13:27:10] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[13:28:25] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[13:48:33] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[14:01:34] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[14:02:50] *** amrfrsh <amrfrsh!~Thunderbi@134.19.189.92> has quit IRC (Ping timeout: 260 seconds)
[14:15:14] *** Lirion <Lirion!~kesselink@wikimedia-commons/Lirion> has quit IRC (Remote host closed the connection)
[14:26:03] *** phyre <phyre!~phyre___@78.30.23.107> has quit IRC (Remote host closed the connection)
[14:35:04] <jlevon> can people build the gate right now?
[14:35:08] <jlevon> boot/solaris/bin/symdef: non-conforming mcs(1) comment <no $(POST_PROCESS)?>
[14:35:11] <jlevon> for everything
[14:37:57] <andyf> jlevon: yes, but only with the "nightly" from gate
[14:38:06] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[14:38:21] <jlevon> I did :/
[14:44:31] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[14:44:42] <EisNerd> hm would this be identifyable using wireshark?
[14:45:00] *** leoric <leoric!~leoric@46.147.99.173> has quit IRC (Quit: Konversation terminated!)
[14:49:09] <andyf> jlevon: hmm, let me try a build
[14:52:59] *** phyre <phyre!~phyre___@78.30.23.107> has joined #illumos
[14:57:34] <andyf> jlevon: have you checked `mcs -p proto/..../bin/ls` or something?
[14:57:44] <jlevon> no I'll dig
[14:58:14] <andyf> check_rtime in gate now looks for #(@) illumos ... or whatever the exact syntax is
[15:02:06] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has joined #illumos
[15:02:27] <liv3010m> @rmustacc: Hi! I registered yesterday there before you make the post. I'll run you new dtrace snippet and upload it
[15:10:34] *** amrfrsh <amrfrsh!~Thunderbi@134.19.189.92> has joined #illumos
[15:17:53] <toastersonerson> EisNerd (IRC): should be.
[15:20:45] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[15:28:24] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[15:30:32] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[15:31:07] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[15:32:53] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[15:33:48] <EisNerd> hm looks like the windows size thing is a good
[15:34:03] <EisNerd> copied same file (2G) to both servers
[15:34:22] <EisNerd> the old is more than 20MB/s faster
[15:34:52] <EisNerd> I picked some random frame out of the conversation
[15:35:57] <EisNerd> to the old server window size is given with 33017 to the illumos 8243
[15:43:07] <EisNerd> https://pasteboard.co/J6YcMud.png
[15:49:12] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[15:50:41] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: Leaving)
[15:54:23] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[15:55:16] <andyf> @jlevon: my gate build worked, clean mail_msg
[15:55:26] <jlevon> thanks
[15:57:07] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[16:01:04] <jlevon> jlevon:symdef $ mcs -p symdef
[16:01:04] <jlevon> symdef:
[16:01:04] <jlevon> @(#)illumos socket-race-0-g23307990f1 May 2020
[16:01:04] <jlevon> @(#)SunOS Internal Development: gk 2020-May-05 [2]
[16:01:10] <jlevon> I guess it's the latter there
[16:03:28] <andyf> symdef:
[16:03:29] <andyf> @(#)illumos gate-upstream_gate-3052595ab8 May 2020
[16:03:29] <andyf> @(#)illumos development build: af 2020-May-05 [illumos]
[16:04:36] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[16:04:41] <jlevon> didn't notice bldenv had changed
[16:04:50] <andyf> I presume you have DEV_CM set somewhere in your env file or whatever
[16:04:54] <jlevon> so had old DEV_CM set yeah
[16:04:59] <jlevon> came from bldenv actually
[16:06:14] <liv3010m> @rmustacc: I just posted new bge.out files on issue 12686 ;)
[16:06:22] <jlevon> richlowe: don't suppose you know history of _lib_version in some lib mapfiles?
[16:09:22] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[16:11:43] <toastersonerson> EisNerd (IRC): Are those two comparisons one with the windows server and one with the illumos one?
[16:12:56] <toastersonerson> Let me guess the second one is windows?
[16:13:13] *** BOKALDO <BOKALDO!~BOKALDO@87.110.102.49> has quit IRC (Quit: Leaving)
[16:16:34] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[16:18:54] *** jollyd <jollyd!~alarcher@lfbn-nic-1-142-60.w2-15.abo.wanadoo.fr> has joined #illumos
[16:21:17] <EisNerd> no, ...57 is windows and clusfile is illumos
[16:21:51] <EisNerd> I increased the windows sizes on OI and get now a lot better values I would say
[16:23:20] <LeftWing> jlevon: Sorry, I tried to call that out in the heads up mail
[16:26:59] <jlevon> LeftWing: yeah, my fault, I forgot I never updated my alias to point directly to new bldenv
[16:33:43] <EisNerd> hard to say this, but now the tcp statistics of the oi host in wireshark look now very similiar to the windows ones
[16:34:14] <EisNerd> never dreamed of tuning a *nix to behave more like a windows
[16:35:09] <EisNerd> so under the line, we have to admit, that microsoft did a good job in this matter, at least regarding the defaults
[16:35:31] <EisNerd> as I'm quite sure no one ever did any tuning to this box
[16:45:56] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[16:48:51] <EisNerd> I found a quite good summary about the smb params here https://www.samba.org/samba/docs/current/man-html/smb.conf.5.html
[16:49:25] <EisNerd> but which of this values are supported by the illumos smbd and how to check for their current values and how to tune them?
[16:49:53] *** jellydonut <jellydonut!~quassel@s91904423.blix.com> has quit IRC (Quit: jellydonut)
[16:51:07] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[16:51:25] <ypankov> EisNerd: if you are using samba, then everything should apply; if using native smb service, then nothing would apply.
[16:52:01] *** jellydonut <jellydonut!~quassel@s91904423.blix.com> has joined #illumos
[16:52:07] <EisNerd> credits and workers for example have their params
[16:52:25] <EisNerd> and yes it is the native smbd
[16:53:35] <EisNerd> and I'm referring to params like smb2 max write smb2 max read smb2 max trans
[16:54:19] <EisNerd> which should have their representation as they are given from the protocol
[16:54:35] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[16:55:02] <EisNerd> btw if I change any of the svc values, do I need to restart the smbd?
[16:56:53] <LeftWing> EisNerd: I think "sharectl get smb" shows you the configurable bits?
[16:57:47] <LeftWing> I'm not deeply familiar with smbsrv though
[16:58:14] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[16:58:19] *** BOKALDO <BOKALDO!~BOKALDO@87.110.102.49> has joined #illumos
[16:58:22] <toastersonerson> EisNerd (IRC): I don't think there is a guide on how to tune smbd. COuld you share your tuning work once you are happy. that would a lor of people a lot.
[16:59:15] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[17:00:51] <EisNerd> I would say it would help a lot if the defaults in illumos would be equal to the ones given for w2012 in the samba doc I referred to above
[17:03:56] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[17:13:18] <EisNerd> the read performance is for some reason awful, I have to check this too
[17:15:01] *** danmcd <danmcd!~danmcd@static-71-174-113-16.bstnma.fios.verizon.net> has joined #illumos
[17:16:49] *** wacki <wacki!~wacki@i577B80A9.versanet.de> has joined #illumos
[17:19:23] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[17:19:56] <EisNerd> the ipadm tuning works instantly?
[17:20:10] *** neuroserve <neuroserve!~toens@ip-95-222-232-201.hsi15.unitymediagroup.de> has quit IRC (Ping timeout: 272 seconds)
[17:20:34] <toastersonerson> should
[17:23:06] <EisNerd> as congestion_control and ecn may also be relevant
[17:25:34] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[17:27:38] *** alanc <alanc!~alanc@inet-hqmc01-o.oracle.com> has quit IRC (Remote host closed the connection)
[17:28:05] *** alanc <alanc!~alanc@inet-hqmc01-o.oracle.com> has joined #illumos
[17:31:18] *** man_u <man_u!~manu@89-92-19-81.hfc.dyn.abo.bbox.fr> has joined #illumos
[17:32:09] *** neuroserve <neuroserve!~toens@ip-88-152-243-25.hsi03.unitymediagroup.de> has joined #illumos
[17:40:00] <EisNerd> active ecn and newreno improves it again
[17:40:41] <EisNerd> so inreasing buffers to 8M and increase credits and max workers gives a lot better performance
[17:41:19] <EisNerd> thx
[17:46:31] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[17:50:11] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[17:52:11] *** pwinder <pwinder!~pwinder@86.11.191.8> has quit IRC (Quit: This computer has gone to sleep)
[17:54:29] <jlevon> pmooney: oops, sorry, meant to say here instead. https://gist.github.com/jlevon/a1bdd05ba4b6a7b3f7c24aa945a28ea9
[17:54:45] <jlevon> pmooney: I'm churning through illumos-joyent so not sure where they are on your train
[17:55:14] <jlevon> don't know offhand if that pragma works with earlier gcc's that don't understand the option
[17:55:26] <pmooney> OK
[17:55:35] <pmooney> perhaps just continue the churn in illumos-joyent?
[17:55:46] <pmooney> my target sync point is from november-ish right now
[17:55:54] <jlevon> well I think that's the last of the bhyve ones
[17:56:04] <pmooney> with the assumption that we'll make another round to pick up changes since then once it's all in the gate
[17:56:10] <pmooney> (including zeller's sync-up wad)
[17:56:12] <jlevon> ok, was just a heads up realy
[17:56:15] <jlevon> really
[17:56:19] <pmooney> sure thing, appreciate it
[17:56:23] *** tsoome_ <tsoome_!~tsoome@06c2-e8aa-94cd-4c50-2f80-4a40-07d0-2001.sta.estpak.ee> has joined #illumos
[17:56:42] <pmooney> I'll look into the emulate-cpuid thing
[17:56:48] <pmooney> that seems like something we could jsut fix
[17:56:52] <pmooney> rather than need the pragma stuff
[17:57:12] <pmooney> and the casting changes for mdb-bhyve aren't a concern
[17:57:17] <pmooney> (IMO)
[17:58:48] <jlevon> I don't think the svm code is wrong in this case, just gcc doesn't know that we happen to be aligned for other reasons
[17:58:50] <jlevon> ?
[17:58:55] <jlevon> but I didn't look much
[18:08:44] <pmooney> I don't think the svm code is wrong, but it could be more explicit about it's handling register changes for that emulation
[18:09:09] *** jollyd <jollyd!~alarcher@lfbn-nic-1-142-60.w2-15.abo.wanadoo.fr> has quit IRC (Quit: Leaving.)
[18:09:51] *** wacki <wacki!~wacki@i577B80A9.versanet.de> has quit IRC (Quit: Lingo: www.lingoirc.com)
[18:11:11] *** Riastradh <Riastradh!~riastradh@netbsd/developer/riastradh> has quit IRC (Remote host closed the connection)
[18:11:39] *** Riastradh <Riastradh!~riastradh@netbsd/developer/riastradh> has joined #illumos
[18:12:14] <pmooney> actually, I think what's there is wrong
[18:12:21] <pmooney> but just happens to work?
[18:12:46] <pmooney> cpuid should zero-extend into the high 32 bits of those registers, right?
[18:12:59] <pmooney> which that goofy cast isn't going to accomplish
[18:13:13] <jlevon> uh let me look again
[18:14:03] <pmooney> I can't recall if 'EAX <- value' implies the zero extension in the instruction specification
[18:14:56] <jlevon> hmm
[18:19:39] <pmooney> testing seems to suggest that the zeroing does occur
[18:21:13] <jlevon> 1. On Intel 64 processors, CPUID clears the high 32 bits of the RAX/RBX/RCX/RDX registers in all modes.
[18:21:15] <jlevon> found it
[18:21:31] <jlevon> so yeah, thanks gcc9, you were wrong in a useful way hah
[18:22:23] <pmooney> lol
[18:22:29] <pmooney> I'll get a patch drawn up to fix that
[18:22:32] <pmooney> thanks for pointing it out
[18:22:56] <jlevon> cool
[18:23:03] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[18:28:13] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[18:36:32] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[18:37:07] <neirac> toastersonerson I just saw you did a pr for upstreaming the docker term changes in nomad thanks a lot!.
[18:37:20] <toastersonerson> sure thing
[18:38:35] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has joined #illumos
[18:40:26] <Smithx10> Nice, I think it’s just gopsutil now
[18:55:07] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[19:01:50] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[19:05:29] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[19:15:04] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[19:30:58] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[19:36:39] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[19:38:52] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[19:42:39] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[19:43:35] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[19:47:32] *** mscheker <mscheker!uid437521@gateway/web/irccloud.com/x-cqwtuqbdxuvprobg> has joined #illumos
[19:48:16] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[19:54:53] *** jimklimov <jimklimov!~jimklimov@78.80.224.132> has quit IRC (Quit: Leaving.)
[19:55:18] *** cypa <cypa!~cypam]_@5.79.187.109> has joined #illumos
[19:56:13] *** cypa <cypa!~cypam]_@5.79.187.109> has quit IRC (Client Quit)
[19:56:33] *** cypa <cypa!~cypam]_@5.79.187.109> has joined #illumos
[19:57:55] *** pwinder <pwinder!~pwinder@86.11.191.8> has joined #illumos
[20:00:20] *** pwinder <pwinder!~pwinder@86.11.191.8> has quit IRC (Client Quit)
[20:02:46] <toastersonerson> and a lot of syscalls in sys/unix it looks like
[20:03:55] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[20:07:35] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[20:12:47] *** man_u <man_u!~manu@89-92-19-81.hfc.dyn.abo.bbox.fr> has quit IRC (Quit: man_u)
[20:21:45] <toastersonerson> neirac (IRC): Did your unix syscall work ever land in sys/unix?
[20:21:55] <toastersonerson> for illumos
[20:24:11] <neirac> toastersonerson no, I started working in nomad before doing that, it was flock support I recall
[20:24:38] <neirac> toastersonerson I completed that work, but needed to add tests
[20:25:15] <toastersonerson> ah did you add MNT_DETACH and Statfs?
[20:31:47] *** nbhauke <nbhauke!~hauke@55d42c22.access.ecotel.net> has joined #illumos
[20:40:15] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[20:44:25] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[20:51:53] <EisNerd> btw no luck with my rpool mirror even with latest OI bits
[20:52:21] <azarus> oh yea, btw, the issue I had yesterday was fixed by just creating a lipkg brand zone instead of a lx brand zone. works well now ^_^
[20:56:21] <neirac> toastersonerson no, I just worked on flock
[20:57:04] <toastersonerson> neirac (IRC): did you work on getting the mkall.sh file to generate more of the syscall files?
[20:57:48] <neirac> toastersonerson oh, yes I ended up creating illumos syscalls separated from the solaris ones
[20:58:04] <neirac> toastersonerson maybe that was not a good aproach
[20:58:20] <neirac> toastersonerson I'll push the changes if you want to take a look
[20:58:28] *** nbhauke <nbhauke!~hauke@55d42c22.access.ecotel.net> has quit IRC (Quit: nbhauke)
[20:58:42] <toastersonerson> no that works. there is illumos work in sys/unix after all
[20:58:50] <toastersonerson> yes please do a push
[20:59:24] <toastersonerson> I can use it to add the other syscalls
[21:00:40] *** phyre <phyre!~phyre___@78.30.23.107> has quit IRC (Remote host closed the connection)
[21:09:12] *** BOKALDO <BOKALDO!~BOKALDO@87.110.102.49> has quit IRC (Quit: Leaving)
[21:09:32] <Smithx10> toastersonerson: that will be great
[21:19:14] *** nbhauke <nbhauke!~hauke@55d42c22.access.ecotel.net> has joined #illumos
[21:26:28] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[21:26:34] <pmooney> jlevon: can confirm that the in-guest bhyve behavior is wrong
[21:26:54] <pmooney> (on INTC)
[21:28:42] <toastersonerson> Do we have a sysnumfile in the same format as freebsd and openbsd?
[21:29:08] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[21:30:35] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[21:35:37] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[21:38:49] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[21:39:00] <KungFuJesus> tsoome_: so, I ordered that t4 last night...
[21:52:24] <toastersonerson> KungFuJesus (IRC): oh dear
[21:53:09] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[21:54:47] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[22:00:33] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[22:00:34] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[22:05:14] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[22:06:05] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[22:08:49] *** nbhauke <nbhauke!~hauke@55d42c22.access.ecotel.net> has quit IRC (Quit: nbhauke)
[22:14:35] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[22:14:52] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[22:19:39] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[22:19:44] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[22:20:25] <jlevon> pmooney: good to know
[22:20:40] <jlevon> pmooney: presumably no guests really care
[22:20:52] <pmooney> clearly not
[22:21:05] <pmooney> like, if you want to go looking for magic feature bits in the upper 32, go right ahead
[22:21:43] <pmooney> also interesting: CPUs (at least the AMD ones I tested on) don't care about garbage in the upper 32 as input
[22:25:26] <jlevon> quick, stuff a speculation mitigation in there
[22:25:30] <tsoome_> KungFuJesus, gz:) I hope you enjoy it:)
[22:26:56] <pmooney> jlevon: if you open a ticket on that, I'll submit a PR against illumos-joyent
[22:30:13] <jlevon> OS-8168
[22:30:16] <jlevon> thanks
[22:31:08] *** vgusev <vgusev!~vgusev@188.187.60.230> has joined #illumos
[22:32:59] <KungFuJesus> tsoome_: what were the issues you were having trying to build the latest ISO?
[22:34:27] *** phyre <phyre!~phyre___@78.30.23.107> has joined #illumos
[22:34:30] <pmooney> jlevon: cool, thanks
[22:35:17] <andyf> I should probably get back to reviewing the outstanding gcc9 fixes that tsoome has in the queue
[22:36:23] <pmooney> I'm not going to do this with a bunch of ifdefs, but operate under the assumption that upstream bhyve will want it
[22:40:07] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[22:42:20] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Remote host closed the connection)
[22:43:23] <KungFuJesus> hah, well that looks like one of them you just posted to developer
[22:44:29] *** vgusev <vgusev!~vgusev@188.187.60.230> has quit IRC (Remote host closed the connection)
[22:44:29] <tsoome> new ones:)
[22:45:00] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[23:02:09] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[23:05:25] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 246 seconds)
[23:16:59] <LeftWing> toastersonerson: If you mean a file with system call numbers, then no, we do not have that
[23:17:16] <LeftWing> Our system call interface is not stable; you need to use libc
[23:18:08] <toastersonerson> ah thats why it's not used in the make script for go
[23:18:30] <LeftWing> I imagine so
[23:18:40] *** vgusev <vgusev!~vgusev@188.187.60.230> has joined #illumos
[23:18:48] <toastersonerson> I think I can get by without it.
[23:19:01] *** vgusev <vgusev!~vgusev@188.187.60.230> has quit IRC (Remote host closed the connection)
[23:19:04] <LeftWing> Is there a particular routine you're trying to call?
[23:19:49] <toastersonerson> adding more of our API to sys/unix as I am missing some in a few places
[23:20:17] <LeftWing> Ah. There, you should be able to do something like this: https://github.com/jclulow/wireguard-go-illumos-wip/blob/jclulow/tun/tun_illumos.go#L24-L103
[23:20:34] <LeftWing> It uses this assembly stub: https://github.com/jclulow/wireguard-go-illumos-wip/blob/jclulow/tun/asm_solaris_amd64.s
[23:20:45] <LeftWing> I lifted that from either the core of Go or from the sys/unix bits -- I can't remember which
[23:22:13] <toastersonerson> ah nice. yeah I am doing something similar. But using go tool cgo -godefs to import the consts automatically
[23:22:56] <toastersonerson> I am also making gy types with that from c types
[23:28:13] *** kahiru <kahiru!~quassel@ip-89-102-207-18.net.upcbroadband.cz> has quit IRC (Read error: Connection reset by peer)
[23:42:40] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[23:49:29] <gitomat> [illumos-gate] 12684 Inject zinject a percentage amount of dev errs -- Don Brady <don.brady at intel dot com>
[23:55:16] *** vgusev <vgusev!~vgusev@188.187.60.230> has joined #illumos
[23:55:18] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)
[23:59:56] *** vgusev <vgusev!~vgusev@188.187.60.230> has quit IRC (Ping timeout: 256 seconds)
[23:59:57] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
top

   May 5, 2020  
< | 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 | 31 | >