Switch to DuckDuckGo Search
   May 28, 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 | >


NOTICE: This channel is no longer actively logged.

Toggle Join/Part | bottom
[00:08:27] <rmustacc> jollyd: I have mixed feelings. One one hand, it feels like you want a full tuple then and to start working programs to know about that tuple. On the other hand, often I feel like software would be better if it had feature tests to look for the feature rather than assume the implementation. But there are tradeoffs all around and working through what those are, what transition plans are, etc. would be
[00:08:33] <rmustacc> helpufl.
[00:13:52] *** andy_js <andy_js!~andy@94.11.185.123> has quit IRC (Quit: andy_js)
[00:20:36] <yuripv> More so as we don't have versions for illumos itself.
[00:21:33] <rmustacc> Sure, but there can still be use as seen with the __linux style defs, and the majority of the BSD defs in source code aren't with versions.
[00:21:44] <rmustacc> Certainly, if it existed, I'd have no doubt if folks used it.
[00:21:53] <rmustacc> no doubt that folks would use it
[00:22:34] *** Mazon <Mazon!~Mazon@37.205.127.168> has quit IRC (Ping timeout: 272 seconds)
[00:22:51] <yuripv> I mean when checking for features, you can't decide that feature is present only by having the __illumos defined.
[00:23:00] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[00:23:14] <rmustacc> That's true. But also doesn't stop anyone from how they use it for other platforms today.
[00:23:17] *** Mazon <Mazon!~Mazon@37.205.127.168> has joined #illumos
[00:23:45] <rmustacc> I'm just trying to point out the general difference between it'd be great if everyone did proper feature detection and the reality of platform specific #ifdefs that aren't versioned at all.
[00:24:16] <rmustacc> And even if you and I believe that it should always use proper feature detection people are going to do what they've done for years, which is the opposite. ;)
[00:24:34] *** phyre <phyre!~phyre___@78.30.22.107> has joined #illumos
[00:25:01] <rmustacc> Especially as build-time vs. run-time feature detection is another whole can of worms.
[00:26:32] *** Kurlon <Kurlon!~Kurlon@167-48.gwi.net> has quit IRC (Ping timeout: 258 seconds)
[00:41:13] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[00:41:50] *** Kurlon <Kurlon!~Kurlon@167-48.gwi.net> has joined #illumos
[00:43:47] *** phyre <phyre!~phyre___@78.30.22.107> has quit IRC (Remote host closed the connection)
[01:38:34] *** Kurlon <Kurlon!~Kurlon@167-48.gwi.net> has quit IRC (Read error: Connection reset by peer)
[01:38:49] *** Kurlon <Kurlon!~Kurlon@167-48.gwi.net> has joined #illumos
[01:43:25] <jbk> ok.. i really need to get these pkcs11 bugs fixed.. would it be worth an email to illumos-developer to outline the general approach? there's enough issues that there's going to be several bugs (maybe quite a few..as i fix the bugs that cause the test to crash, i see more, so until all of those are fixed, it's hard to know how many total).. it probably also needs to be done against pkcs11_softtoken,
[01:43:31] <jbk> pkcs11_kernel, libpkcs11, and the metaslot code
[01:44:16] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Ping timeout: 256 seconds)
[01:45:16] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[01:45:17] <rmustacc> Depends on what all is involved. It sounds like you're doing major architectural work to fix a large number of issues?
[01:45:25] <rmustacc> And what are you looking for from others?
[01:45:39] <jbk> well mostly just to set expectations during review
[01:46:01] <rmustacc> I'm not sure what expectations you're trying to set.
[01:46:21] <rmustacc> Is there a patch stream of things that you have and you're trying to avoid certain classes of changes from review?
[01:46:36] <rmustacc> Are you trying to just set expectations that there are going to be additional changes coming?
[01:46:47] <jbk> a bit of both
[01:46:47] <rmustacc> And that not everything will magically get better in one change?
[01:47:16] <rmustacc> I mean, no one has reviewed these before right?
[01:47:34] <rmustacc> So it feels a little different from say, the bhyve set. So probably flexibility from everyone will be appreciated.
[01:49:06] <rmustacc> I'm not sure how much you're trying to just give folks additional background versus trying to say take it or leave it.
[01:50:25] <rmustacc> So if you can help describe a bit more of what you're trying to get across, I can probably give better suggestions.
[01:52:03] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[01:54:40] <jbk> mostly, i'm planning to tackle the provider libaries individually, starting with pkcs11_softtoken (since that's used the most), then pkcs11_kernel, then libpkcs11.. so there may be similar bugs in all of them, but would be handled as it's own issue as i tackle that library (vs. trying to look and test all of the libraries at once for the same bugs since it ends up being tricker), and just to expect there's
[01:54:46] <jbk> going to be some number of changes (i.e. don't expect one giant change that fixes everything)
[01:55:54] <rmustacc> OK. I would probably just add additional color as you put up the first reviews, rather than pre-announcing things unless you would change your path based on someone's feedback.
[01:56:54] <rmustacc> I would say just be open to the fact that we may realize a different way of doing something in a subsequent provider library.
[01:57:14] <jbk> and things can of course adjust as i go along if it turns out some tweaking would make the process go better..
[01:57:26] <rmustacc> Which is part of the advantage of breaking it up as you're proposing.
[01:57:39] <jbk> but i'll go with that -- right now i'm not really looking to change things too much as much as just fix issues
[01:57:53] <rmustacc> Then I wouldn't worry about it too much.
[02:00:17] <LeftWing> Yeah if you've just got lots of bugs, I'd just file them
[02:00:32] <LeftWing> If there's a broader architectural change you're seeking to make, over many commits, I'd write an IPD
[02:01:00] <LeftWing> But like, if it's just that nobody has been in here for a while, and there are thus a lot of latent issues, that doesn't seem particularly different to say, Toomas' cleanup work
[02:01:23] <LeftWing> Break it up into whatever chunks you're able to get reviewed and then tested
[02:47:57] *** yuripv <yuripv!~yuripv@91.240.124.138> has quit IRC (Remote host closed the connection)
[03:17:16] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[03:20:39] *** Kurlon <Kurlon!~Kurlon@167-48.gwi.net> has quit IRC (Ping timeout: 265 seconds)
[03:22:41] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Quit: Leaving...)
[03:34:56] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[04:28:46] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[04:30:19] *** arnoldoree <arnoldoree!~arnoldore@2001:d08:1a81:7d61:6923:f108:5b14:38a7> has quit IRC (Ping timeout: 252 seconds)
[04:43:52] *** nde <nde!uid414739@gateway/web/irccloud.com/x-eeqhyugrjvkygzpg> has joined #illumos
[04:44:55] *** arnoldoree <arnoldoree!~arnoldore@113.210.193.7> has joined #illumos
[05:53:06] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[06:14:00] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has quit IRC (Ping timeout: 256 seconds)
[06:29:02] *** MillerBoss <MillerBoss!~shit@107.191.50.190> has quit IRC (Ping timeout: 260 seconds)
[06:38:33] <jbk> heh.. how did this ever work..
[06:53:42] *** Blkt <Blkt!~Blkt@2a01:4f8:200:2425::adda> has quit IRC (Remote host closed the connection)
[06:58:38] *** Blkt <Blkt!~Blkt@2a01:4f8:200:2425::adda> has joined #illumos
[07:00:05] *** Blkt <Blkt!~Blkt@2a01:4f8:200:2425::adda> has quit IRC (Remote host closed the connection)
[07:10:51] *** Blkt <Blkt!~Blkt@2a01:4f8:200:2425::adda> has joined #illumos
[07:18:10] *** BOKALDO <BOKALDO!~BOKALDO@91.105.119.143> has joined #illumos
[07:21:16] <tsoome> jbk also considering disabling weak algorithms?
[07:24:08] <jbk> that would work around this :)
[07:24:14] <jbk> well removing them
[07:24:45] <jbk> but at least for userland (i'd need to look for the kernel).. what's supposed to happen is that the cryptoadm command is supposed to be used to configure/manage all of that
[07:25:00] *** arnoldoree <arnoldoree!~arnoldore@113.210.193.7> has quit IRC (Ping timeout: 272 seconds)
[07:25:38] <jbk> though it'd probably be nice to have a way to specify an alternate config file for that (mostly for testing)
[07:37:50] *** arnoldoree <arnoldoree!~arnoldore@113.210.191.186> has joined #illumos
[07:54:55] *** wacki <wacki!~wacki@i577A5D84.versanet.de> has joined #illumos
[07:58:33] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: Leaving)
[08:01:41] *** nbhauke <nbhauke!~hauke@55d4989d.access.ecotel.net> has joined #illumos
[08:29:54] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[08:39:45] *** phyre <phyre!~phyre___@78.30.22.107> has joined #illumos
[08:40:52] *** alanc <alanc!~alanc@inet-hqmc02-o.oracle.com> has quit IRC (Remote host closed the connection)
[08:41:19] *** alanc <alanc!~alanc@inet-hqmc02-o.oracle.com> has joined #illumos
[08:41:36] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has joined #illumos
[08:44:29] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: Leaving)
[08:45:55] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[08:51:03] *** KeiraT <KeiraT!~k4ra@gateway/tor-sasl/k4ra> has quit IRC (Ping timeout: 240 seconds)
[08:54:23] *** KeiraT <KeiraT!~k4ra@gateway/tor-sasl/k4ra> has joined #illumos
[09:39:38] *** jimklimov <jimklimov!~jimklimov@78-80-21-240.nat.epc.tmcz.cz> has joined #illumos
[09:43:19] *** neuroserve <neuroserve!~toens@ip-84-118-131-4.unity-media.net> has quit IRC (Ping timeout: 260 seconds)
[09:51:48] *** neuroserve <neuroserve!~toens@ip-94-114-253-33.unity-media.net> has joined #illumos
[09:59:18] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #illumos
[10:00:12] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has joined #illumos
[10:02:44] *** jimklimov1 <jimklimov1!~jimklimov@ip-86-49-254-26.net.upcbroadband.cz> has joined #illumos
[10:07:38] *** jimklimov <jimklimov!~jimklimov@78-80-21-240.nat.epc.tmcz.cz> has quit IRC (Ping timeout: 256 seconds)
[10:11:03] *** KeiraT <KeiraT!~k4ra@gateway/tor-sasl/k4ra> has quit IRC (Ping timeout: 240 seconds)
[10:13:47] *** pwinder <pwinder!~pwinder@86.11.191.8> has joined #illumos
[10:16:00] *** KeiraT <KeiraT!~k4ra@gateway/tor-sasl/k4ra> has joined #illumos
[10:16:38] *** kerberizer <kerberizer!~luchesar@wikipedia/Iliev> has quit IRC (Ping timeout: 272 seconds)
[10:21:01] *** andy_js <andy_js!~andy@94.11.185.123> has joined #illumos
[10:23:20] <gitomat> [illumos-gate] 12785 grep -h ignored with -r -- Peter Tribble <peter.tribble at gmail dot com>
[11:08:02] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 265 seconds)
[11:19:45] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[12:05:22] *** kerberizer <kerberizer!~luchesar@wikipedia/Iliev> has joined #illumos
[12:23:38] *** nde <nde!uid414739@gateway/web/irccloud.com/x-eeqhyugrjvkygzpg> has quit IRC (Quit: Connection closed for inactivity)
[12:34:14] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 240 seconds)
[12:47:29] <igork> andyf: did you test your microcode updates on SkyLake CPU ?
[12:48:37] <igork> previous one has issue with reboot - stuck on bios - and i have used older one
[12:52:29] *** eki <eki!~eki@dsl-hkibng41-54f858-46.dhcp.inet.fi> has quit IRC (Quit: leaving)
[12:56:36] *** eki <eki!~eki@dsl-hkibng41-54f858-46.dhcp.inet.fi> has joined #illumos
[13:17:59] <EisNerd> pwinder: sounds very good, maybe you are the right one to look after the i40e driver or maybe help me with my specific trouble with it. Btw did anyone ever tested if the i40e is able to reach the full link speed, as Paul did for the mlx?
[13:19:39] <igork> EisNerd: we are using i40e on supermicro box - 4 links, dual in aggr
[13:20:46] <EisNerd> igork: which device?
[13:20:53] <igork> 722
[13:21:52] <igork> supermicro SSG-6048R-E1CR90L
[13:24:53] <EisNerd> how to check in oi
[13:25:05] <igork> EisNerd: sorry, i use dilos
[13:29:55] <EisNerd> we have X722 and XL710
[13:30:37] <EisNerd> and both do not work as perfect as expected
[13:35:36] <EisNerd> we see a lot of retries
[13:50:54] <andyf> igork - no, as I said in the review and the RTI, I don't have those parts
[13:51:05] <andyf> igork - which is the "previous one" and which is the "older one"?
[13:51:15] <igork> EisNerd: i have a lot of updates from smartos on dilos - probably some updates missed on illumos
[14:01:18] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[14:01:35] <igork> andyf: i'm using now 20190618
[14:01:35] <sjorge> LeftWing poorly
[14:02:00] <sjorge> I was messing around with the attach detatch stuff but I never managed to get them to trigger when plugging in the usb device
[14:02:01] <igork> andyf: you have updated another = previous
[14:02:56] <sjorge> LeftWing: I assume I am missing a bit somewhere where i tell it vendorId + deviceId ==> this driver here
[14:04:31] *** phyre <phyre!~phyre___@78.30.22.107> has quit IRC (Remote host closed the connection)
[14:09:27] *** phyre <phyre!~phyre___@78.30.22.107> has joined #illumos
[14:32:56] *** andy_js <andy_js!~andy@94.11.185.123> has quit IRC (Quit: andy_js)
[14:56:18] *** rzezeski <rzezeski!uid151901@gateway/web/irccloud.com/x-fonrknrgumwkdqys> has quit IRC (Quit: Connection closed for inactivity)
[15:01:34] *** nbhauke <nbhauke!~hauke@55d4989d.access.ecotel.net> has quit IRC (Ping timeout: 240 seconds)
[15:07:43] *** andy_js <andy_js!~andy@94.11.185.123> has joined #illumos
[15:08:00] *** BOKALDO <BOKALDO!~BOKALDO@91.105.119.143> has quit IRC (Quit: Leaving)
[15:13:10] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[15:21:42] *** nbhauke <nbhauke!~hauke@55d48c7f.access.ecotel.net> has joined #illumos
[15:27:46] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has joined #illumos
[15:28:12] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Ping timeout: 256 seconds)
[15:29:46] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[15:31:56] *** EisNerd <EisNerd!~manschwet@mail.cs-software-gmbh.de> has quit IRC (Read error: Connection reset by peer)
[15:32:54] *** EisNerd <EisNerd!~manschwet@mail.cs-software-gmbh.de> has joined #illumos
[15:52:24] *** nbhauke <nbhauke!~hauke@55d48c7f.access.ecotel.net> has quit IRC (Read error: Connection reset by peer)
[15:52:44] *** nbhauke <nbhauke!~hauke@55d4869b.access.ecotel.net> has joined #illumos
[16:09:31] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[16:10:09] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[16:11:56] <LeftWing> sjorge: Ah that's usually specified through "add_drv"
[16:13:45] <LeftWing> What command have you been using to install the module thus far?
[16:20:09] *** BOKALDO <BOKALDO!~BOKALDO@91.105.119.143> has joined #illumos
[16:21:54] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[16:25:43] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Client Quit)
[16:25:52] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 272 seconds)
[16:28:52] *** phyre <phyre!~phyre___@78.30.22.107> has quit IRC (Remote host closed the connection)
[16:36:51] *** jellydonut <jellydonut!~quassel@s91904422.blix.com> has quit IRC (Quit: jellydonut)
[16:40:38] *** jellydonut <jellydonut!~quassel@s91904421.blix.com> has joined #illumos
[16:45:36] <sjorge> onu over the existing omnios install
[16:45:44] <sjorge> I can force load it with modload
[16:45:50] <sjorge> But ofcourse it doesn't attach either :p
[16:46:03] <sjorge> So far I just have the attrach hardcoded to usb_failure
[16:46:09] <sjorge> So I am expecting to see that in dmsg
[16:52:38] *** man_u <man_u!~manu@89-92-19-81.hfc.dyn.abo.bbox.fr> has joined #illumos
[16:55:17] *** blackwood821 <blackwood821!~blackwood@c-174-49-16-176.hsd1.tn.comcast.net> has joined #illumos
[17:10:31] *** cantstanya <cantstanya!~chatting@gateway/tor-sasl/cantstanya> has quit IRC (Remote host closed the connection)
[17:12:46] *** cantstanya <cantstanya!~chatting@gateway/tor-sasl/cantstanya> has joined #illumos
[17:16:52] *** man_u <man_u!~manu@89-92-19-81.hfc.dyn.abo.bbox.fr> has quit IRC (Quit: man_u)
[17:18:25] <LeftWing> sjorge: Did you add an IPS package manifest for it, with the aliases for USB binding for that device?
[17:20:11] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 260 seconds)
[17:38:10] <vgusev> tsoome: Could you please look at commit a3874b8b1f "10405 Implement ZFS sorted scans". It looks it had to remove those: zfs_resilver_delay zfs_scrub_delay zfs_scan_idle
[17:39:22] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[17:46:02] <sjorge> LeftWing: most certainly not xD example somewhere I can copy/base it on?
[17:46:51] *** patdk-lap <patdk-lap!~patrickdk@208.94.189.75> has quit IRC (Remote host closed the connection)
[17:49:18] *** patdk-lap <patdk-lap!~patrickdk@208.94.189.75> has joined #illumos
[17:50:17] <LeftWing> sjorge: It's usually easiest to look at an existing driver's packaging; e.g., for "usbsacm": usr/src/pkg/manifests/driver-serial-usbsacm.mf
[17:50:43] <rmustacc> For doing this kind of development if you're only touching your driver, you don't need onu or ips at all, fwiw.
[17:51:07] <LeftWing> Right, that's also true. If you've got your clean workspace built and you've onu-ed to that, you can just load the driver now
[17:51:18] <rmustacc> sjorge: Do you have a prtconf -v output for that device?
[17:51:31] <rmustacc> You should see a compatible field in the listing for that device.
[17:51:50] <rmustacc> The compatible array is a list of names that the system has enumerated for the device that it will try to match and bind a driver to.
[17:52:23] <rmustacc> So if you can provide that we can give you the add_drv command to use.
[17:52:50] <rmustacc> Then you should be able to use something like a dtrace -Zn 'fbt::<your attach func>:entry' and load the driver and see your function called.
[17:52:55] <rmustacc> Or put a cmn_err in your attach.
[17:54:04] <LeftWing> If you've got the /devices path for your USB device, you can, e.g., "prtconf -v /devices/pci@0,0/pci8086,2064@14:17 | grep -A1 compat" to get that information I believe
[17:58:01] *** rzezeski <rzezeski!uid151901@gateway/web/irccloud.com/x-qolxfazvovupaudl> has joined #illumos
[17:59:22] *** jimklimov1 <jimklimov1!~jimklimov@ip-86-49-254-26.net.upcbroadband.cz> has quit IRC (Quit: Leaving.)
[18:01:53] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[18:05:09] <sjorge> Here is a diff with it plugged in and without
[18:05:09] <sjorge> https://gist.github.com/sjorge/cdc4d8f81af448f29a95f06756366444
[18:05:27] <sjorge> I'll see if I can get the manifest made in a bit... I assume more makefile editing will be involved xD
[18:05:37] <rmustacc> No, just don't bother with a manifest sjorge.
[18:05:43] <rmustacc> Keep your life simple.
[18:05:53] <rmustacc> What's the name of your driver?
[18:06:00] <sjorge> uchom
[18:06:05] <sjorge> *uchcom
[18:06:16] <sjorge> I kept it the same as FreeBSD/OpenBSD
[18:06:30] <rmustacc> Fair enough. So, there are two steps to iterating for development.
[18:06:37] <rmustacc> 1) Copying the driver into the right place
[18:06:40] <rmustacc> 2) Adding the device id.
[18:07:01] *** arnoldoree <arnoldoree!~arnoldore@113.210.191.186> has quit IRC (Quit: Leaving)
[18:07:38] *** neuroserve <neuroserve!~toens@ip-94-114-253-33.unity-media.net> has quit IRC (Ping timeout: 246 seconds)
[18:08:30] <rmustacc> For 2, you need to run the following: add_drv -i '"usb1a86,7523"' uchcom
[18:08:39] <rmustacc> Note the quoting with the single and double quotes is quite important.
[18:09:01] <rmustacc> So what I do is a create a symlink from /kernel/drv/amd64/uchcom to /tmp/uchcom
[18:09:35] <rmustacc> I then will use dmake install to iterate on the driver. When satisfied, I copy it to /tmp/uchcom. Then I run the add_drv to test.
[18:09:52] <rmustacc> If attach is at the point it works and detach also works, then I run rem_drv uchcom, iterate, and then repeat.
[18:10:14] <rmustacc> No need to iteratively onu or do other things.
[18:10:31] <rmustacc> Unless you're changing something outside of that device, which I don't think you will be.
[18:23:39] *** Kurlon <Kurlon!~Kurlon@167-48.gwi.net> has joined #illumos
[18:38:01] *** am11 <am11!~am11@dsl-hkibng22-54f985-221.dhcp.inet.fi> has quit IRC (Ping timeout: 264 seconds)
[18:52:22] *** am11 <am11!~am11@dsl-hkibng22-54f985-221.dhcp.inet.fi> has joined #illumos
[18:58:32] <sjorge> ack
[18:58:34] <sjorge> lets try
[19:00:41] <sjorge> that didn't work, it can't find the driver it says... but maybe that is because it's a weird dsd thingy for usbser?
[19:03:21] <sjorge> nvm, I had it in the wrong spot
[19:03:34] *** roryrjb <roryrjb!~rory@209.250.250.222> has quit IRC (Ping timeout: 272 seconds)
[19:04:04] <rmustacc> You will need to properly declare a dependency. But that usually fails in a different way.
[19:04:36] <sjorge> I now get a failure to attach, which I think is to be exacted as I just return USB_FAILURE in the attach function
[19:04:49] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has joined #illumos
[19:05:44] <rmustacc> Is your function actually being called?
[19:05:57] <rmustacc> I would use dtrace or add a cmn_err statement to the attach function to verify.
[19:10:32] <sjorge> It's being called "May 28 18:44:51 omniosce uchcom: [ID 891739 kern.warning] WARNING: uchcom_attach called"
[19:11:08] <sjorge> now the hart part begins of trying to figure out what the BSD's do in there attach
[19:19:06] *** dopplerg- <dopplerg-!~dop@85.10.56.227> has joined #illumos
[19:19:33] *** dopplergange <dopplergange!~dop@192.111.134.213> has quit IRC (Ping timeout: 256 seconds)
[19:21:22] <rmustacc> I would focus on a few things. One is understanding how the chip works.
[19:21:39] <rmustacc> e.g. what pipes do you actually need. Control, interrupt, bulk, etc.
[19:21:46] <rmustacc> And understand how they are used.
[19:22:00] <rmustacc> Once you have a sense of how the device works, the actual API bits are the easy bits.
[19:24:40] <sjorge> It's just the one page pdf I could find
[19:25:23] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Ping timeout: 265 seconds)
[19:26:55] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[19:26:59] <sjorge> The BSD code I was looking at is full "Note: sc_lsr is always zero", "work as chip default - CS8, no parity, !CSTOPB, other modes are not supported."
[19:27:11] <sjorge> So looks like it's pretty picky
[19:55:55] <gitomat> [illumos-gate] 12709 Support custom URI schemes for the keylocation property -- Jason King <jason.king at joyent dot com>
[20:22:49] *** amrmesh <amrmesh!~Thunderbi@134.19.189.92> has joined #illumos
[20:23:10] *** amrfrsh <amrfrsh!~Thunderbi@134.19.189.92> has quit IRC (Ping timeout: 246 seconds)
[20:23:11] *** amrmesh is now known as amrfrsh
[20:42:09] <sjorge> well enough for tonight
[20:42:28] <sjorge> But I managed to get the vm to dump core... so it's doing SOMETHING now
[20:42:32] <sjorge> Not the thing I want obviously
[20:42:59] <sjorge> oh... now it's stuck in a loop
[20:43:12] <sjorge> ah I can just unplug, thank got for USB
[20:46:16] <sjorge> Whats the easiest way to recover dmesg from a dump?
[20:47:12] <tsoome> vgusev, I'm sure there are followup changes, but I do not have time to attend those right now. Any spare hands are welcome:)
[20:47:25] <sjorge> anyway, thats for tomorrow
[20:47:27] <sjorge> Gnite folks!
[20:47:31] <sjorge> And thanks for the extra help
[20:54:14] <sjorge> Oh ::msgbuf it seems
[20:54:26] <sjorge> it's a NULL deref, I guess that is a classic first oopsie to have
[20:54:32] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has quit IRC (Ping timeout: 256 seconds)
[20:59:28] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 272 seconds)
[21:01:35] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[21:03:37] <sjorge> Ooh nice! it did detect the correct of the two chips! "Chip Type: CH341" got printed in msgbuf
[21:03:50] <sjorge> as wel as all my debug prints... ahem...
[21:04:23] <sjorge> Aah the NULL deref makes sense too as I stupped out the dev_attach for now!
[21:04:35] <sjorge> Neat, I know where to poke more tomorrow or in the weekend
[21:06:30] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 256 seconds)
[21:07:09] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[21:21:47] <LeftWing> sjorge: Sounds like a great start!
[21:22:00] <bahamat> rmustacc: Would illumos#12496 cause an interface to come up with a different mac than expected for that device?
[21:23:06] *** lystra <lystra!~lystra@d53-64-11-169.nap.wideopenwest.com> has quit IRC (Ping timeout: 258 seconds)
[21:23:11] <bahamat> Someone came to me with a case where ipxe boots and the mac ends with 50, but once the kernel boots dladm reports the mac ends with 4e.
[21:23:29] <jlevon> love a good "leaks like a sieve" bug. been filing those since '03
[21:24:43] *** lystra <lystra!~lystra@d53-64-11-169.nap.wideopenwest.com> has joined #illumos
[21:31:55] <LeftWing> ha
[21:33:44] <vgusev> tsoome: yes, sure. At least they can be deleted by following updates from openzfs.
[21:33:53] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[21:34:24] <tsoome> aye:)
[21:37:25] *** Kurlon <Kurlon!~Kurlon@167-48.gwi.net> has quit IRC (Ping timeout: 264 seconds)
[21:40:30] *** strugee <strugee!~strugee@strugee.net> has quit IRC (Ping timeout: 256 seconds)
[21:45:25] *** BOKALDO <BOKALDO!~BOKALDO@91.105.119.143> has quit IRC (Quit: Leaving)
[21:48:22] *** strugee <strugee!~strugee@strugee.net> has joined #illumos
[21:55:50] *** lystra <lystra!~lystra@d53-64-11-169.nap.wideopenwest.com> has quit IRC (Ping timeout: 272 seconds)
[21:56:38] *** lystra <lystra!~lystra@d53-64-11-169.nap.wideopenwest.com> has joined #illumos
[21:57:02] *** DocMnaugh <DocMnaugh!~mnrmnaugh@unaffiliated/mnrmnaugh> has quit IRC (Ping timeout: 260 seconds)
[21:57:58] *** DocMnaugh <DocMnaugh!~mnrmnaugh@unaffiliated/mnrmnaugh> has joined #illumos
[21:59:13] <rmustacc> bahamat: No, I don't believe so.
[21:59:59] <rmustacc> bahamat: You'd probably want to see if it's a multi-function device.
[22:02:29] <gitomat> [illumos-gate] 12795 Remove remaining bhyveconsole wait code -- Andy Fiddaman <omnios at citrus-it dot co.uk>
[22:05:31] <blackwood821> rmustacc: it's a BRCM 2P 1G BT + 2P 10G BT rNDC
[22:07:48] <rmustacc> I think the question I'd have are what are all four macs on the device.
[22:08:24] <blackwood821> Only the two 1G ports are showing up in dladm
[22:08:27] <blackwood821> [root@b0-26-28-c0-d7-4e ~]# dladm show-phys -m
[22:08:27] <blackwood821> LINK SLOT ADDRESS INUSE CLIENT
[22:08:27] <blackwood821> bge0 primary b0:26:28:c0:d7:4e no --
[22:08:29] <blackwood821> bge1 primary b0:26:28:c0:d7:4f no --
[22:09:29] <blackwood821> The network/early-admin service log says "Error: ERROR: admin mac address b0:26:28:c0:d7:50 not found on system"
[22:09:52] <blackwood821> When it iPXE boots it says: "Configuring (net0 b0:26:28:c0:d7:50)"
[22:11:45] <bahamat> rmustacc: Yeah, I figured it was a stretch because the image he's booted to has the commit for that fix.
[22:12:07] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Read error: Connection reset by peer)
[22:12:46] <blackwood821> I just checked the iDRAC and the mac address under "Integrated NIC 1 Port 1 Partition 1" is "B0:26:28:C0:D7:50" which is what iPXE is using
[22:14:40] <bahamat> Well that's something at least.
[22:15:38] <bahamat> But shouldn't dladm show-phys show it even if it's not plumbed?
[22:17:05] <blackwood821> Yea usually I see all ports listed in dladm show-phys even if there is no ethernet plugged in but not in this case
[22:17:35] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[22:20:47] <blackwood821> So in the iDRAC I see:
[22:20:47] <blackwood821> Port 1
[22:20:47] <blackwood821> MAC: B0:26:28:C0:D7:50
[22:20:49] <blackwood821> Link Status: Up
[22:20:51] <blackwood821> Partition: Disabled
[22:20:53] <blackwood821> Switch Port Connection ID: gi4
[22:20:55] <blackwood821> Port 2
[22:20:57] <blackwood821> MAC: B0:26:28:C0:D7:51
[22:20:59] <blackwood821> Link Status: Up
[22:21:01] <blackwood821> Partition: Disabled
[22:21:03] <blackwood821> Switch Port Connection ID: gi1
[22:21:05] <blackwood821> Port 3
[22:21:07] <blackwood821> MAC: B0:26:28:C0:D7:4E
[22:21:09] <blackwood821> Link Status: Up
[22:21:11] <blackwood821> Partition: Not Capable
[22:21:13] <blackwood821> Switch Port Connection ID: Not Supported
[22:21:15] <blackwood821> Port 4
[22:21:17] <blackwood821> MAC: B0:26:28:C0:D7:4F
[22:21:19] <blackwood821> Link Status: Up
[22:21:21] <blackwood821> Partition: Not Capable
[22:21:23] <blackwood821> Switch Port Connection ID: Not Supported
[22:21:27] <blackwood821> So dladm is listing only ports 3 and 4 but iPXE is using the MAC address of port 1
[22:21:48] <toastersonerson> My question is what is this Partition thing iDRAC speaks of
[22:22:06] <toastersonerson> Because that is a common denominator of those nics
[22:22:37] <blackwood821> Looking into that now
[22:26:46] <toastersonerson> Also from the MAC's that looks like two PCI cards or one onboard and two extension cards. Are they the same NIC type?
[22:33:10] <blackwood821> I believe it's one onboard card with 2 10G ports and 2 1G ports
[22:33:38] <toastersonerson> Yeah then the Onboard card is likely not supported.
[22:34:49] <toastersonerson> But to verify what Device is the onboard card? the 2P 10G BT rNDC
[22:36:22] <toastersonerson> If they worked before or you know they should be supported can you see anything in /var/adm/messages about a driver problem?
[22:36:38] <rmustacc> So in prtconf what PCI devices actually show up related to that NIC?
[22:37:14] <rmustacc> But if you have multiple cables plugged in, it's not uncommon to boot from something that's not waht you expect.
[22:37:50] <rmustacc> Traditionally those style parts are basically 2 bge and 2 bnxe or something like that.
[22:38:30] <rmustacc> The partitions are a feature of the NetExtreme II line, IIRC which was sold off.
[22:39:01] <rmustacc> But they're usually not something that shows up.
[22:39:14] <rmustacc> I guess the early-admin service assumes that the NIC you're going to pxe on is the one that has to be used.
[22:39:40] <blackwood821> This is the first time trying a Broadcom card in Triton
[22:39:42] <rmustacc> so the thing is to look at prtconf to see if we even see PCI Devices for the 10 GbE ports or not.
[22:40:16] <rmustacc> If you can gist/pastebin/whatever your prtconf -dD I can answer that quickly.
[22:40:19] <rmustacc> But I'm going to have to drop shortly.
[22:41:34] <blackwood821> rmustacc: https://gist.github.com/blackwood821/d96a91845a194e9a416b6651bffca577
[22:42:53] <rmustacc> Ah, it's the NetExtreme-E with bnxt.
[22:43:00] <rmustacc> We don't support that, unfortunately.
[22:43:15] <rmustacc> Basically what's happened is that you've booted with some ipxe supports, but illumos doesn't.
[22:43:22] <rmustacc> And Triton seems to not be handling that correctly, unfortuntaely.
[22:43:27] <rmustacc> *unfortunately
[22:43:47] <rmustacc> I started a rough bit of that driver, but no longer have hw for it, so it hasn't progressed for some time.
[22:43:56] <rmustacc> So, there is no MAC re-ordering going on.
[22:44:07] <rmustacc> bahamat: Does that above make sense?
[22:45:34] <rmustacc> I've got to bounce. I can answer questions in a couple hours.
[22:45:42] <blackwood821> Thanks for the help
[22:45:55] <blackwood821> We can put another card in
[22:53:33] *** neuroserve <neuroserve!~toens@ip-94-114-237-8.unity-media.net> has joined #illumos
[23:03:27] *** pwinder <pwinder!~pwinder@86.11.191.8> has quit IRC (Quit: This computer has gone to sleep)
[23:08:10] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Ping timeout: 260 seconds)
[23:08:56] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[23:15:11] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has joined #illumos
[23:22:11] *** wacki <wacki!~wacki@i577A5D84.versanet.de> has left #illumos
[23:22:11] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)
[23:34:30] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[23:44:29] <bahamat> blackwood821: Before you do that, if there's a network.json for that CN in your tftp root, I'd like to take a look at it.
[23:44:32] <am11> LeftWing: i've opened a PR in sysroot repo. not sure if that's good idea to have a helper script there. please take a look when time permits. :)
[23:44:54] <LeftWing> am11: Hey, I did see it -- I'm not sure if we want to put that script in there to be honest
[23:45:38] <LeftWing> My biggest issue is that we're not going to be testing it ourselves, and that different projects will end up with different requirements
[23:45:55] <LeftWing> e.g., you've bumped the binutils version but I don't think we're going to do that for Rust right now
[23:46:29] <LeftWing> The extra files in the sysroot archive, though, definitely something I should be able to get to
[23:47:13] <blackwood821> bahamat: on the dhcpd zone on the HN?
[23:47:44] *** andy_js <andy_js!~andy@94.11.185.123> has left #illumos
[23:49:11] <am11> LeftWing: thanks, with next release of sysroot, if we could get those headers in, we can subsequently clean up in downstream.
[23:50:03] <blackwood821> bahamat: https://gist.github.com/blackwood821/8556eb9def524ce76fa276578d705acd
[23:51:02] <am11> LeftWing: i can copy over the script. is there a license for the toolchain script?
[23:51:13] <bahamat> blackwood821: thanks.
[23:51:25] <LeftWing> am11: Err, it's under whatever licence Rust is under I guess!
[23:51:45] <LeftWing> Or you can just have it and say you wrote it tbh
[23:51:50] <LeftWing> I don't really care
[23:52:36] <am11> i can try to inline the bits and adapt to their local style. thanks.
[23:52:43] <LeftWing> Yeah that sounds like the right idea
top

   May 28, 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 | >