Switch to DuckDuckGo Search
   April 18, 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 | >


NOTICE: This channel is no longer actively logged.

Toggle Join/Part | bottom
[00:04:56] <am11> LeftWing: switching to 2019 vagrant box did the trick.
[00:10:55] *** phyre <phyre!~phyre___@78.30.23.107> has quit IRC (Remote host closed the connection)
[00:27:26] <LeftWing> Ah yes, that will do it.
[01:02:15] *** bahamas10 <bahamas10!~dave@cpe-72-231-182-75.nycap.res.rr.com> has quit IRC (Remote host closed the connection)
[01:05:14] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[01:09:11] <jbk> am11: i just pushed an update to that PR that should avoid the need of a patched crt1.o
[01:09:49] <jbk> no idea if the upstream will take it or not
[01:10:03] <jbk> (they could argue they are following the spec and we're not, and they'd have a case)
[01:10:15] <jbk> so dunno
[01:10:27] <jbk> never dealt with them
[01:14:14] <am11> jbk: thank you. i pulled and tested it, the results have improved:
[01:14:15] <am11> # TOTAL: 33# PASS: 25# SKIP: 0# XFAIL: 0# FAIL: 8
[01:15:02] <jbk> hrm.. i was getting 28 passing
[01:15:44] <jbk> though there are a few tests that i'm not 100% sure are correct..
[01:15:54] <jbk> i guess once you're further along, we'll know if it's enough or not :)
[01:18:30] <am11> yup, here is the result on my smartos box http://sprunge.us/yLhZV3
[01:22:12] <am11> jbk: right on, and this is something which the maintainer also mentioned in another thread: https://github.com/libunwind/libunwind/issues/164#issuecomment-604746890
[01:23:02] <am11> these regression tests can be flaky, but depends on which one are failing :)
[01:25:20] <am11> jbk: if you could update the `README`'s solaris section, and from `Initial support` to `✓` in the markdown table as part of this PR, that would be cool. 8-)
[01:27:49] *** amrmesh <amrmesh!~Thunderbi@134.19.189.92> has joined #illumos
[01:28:22] *** amrfrsh <amrfrsh!~Thunderbi@95.174.67.172> has quit IRC (Ping timeout: 256 seconds)
[01:28:23] *** amrmesh is now known as amrfrsh
[01:39:09] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[01:40:12] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[01:42:58] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[01:44:02] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[01:46:17] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[01:47:50] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[01:50:29] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[01:50:51] <jbk> done
[01:51:41] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[01:53:14] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[01:53:17] *** neirac <neirac!~cneira@pc-4-149-45-190.cm.vtr.net> has quit IRC (Read error: Connection reset by peer)
[01:54:07] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[01:54:31] *** wacki <wacki!~wacki@i577A5B2E.versanet.de> has joined #illumos
[01:54:37] *** neirac <neirac!~cneira@pc-4-149-45-190.cm.vtr.net> has joined #illumos
[01:57:08] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Remote host closed the connection)
[01:57:23] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[01:59:15] *** wacki <wacki!~wacki@i577A5B2E.versanet.de> has quit IRC (Ping timeout: 250 seconds)
[02:04:03] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[02:08:14] <am11> jbk: thanks, there is also this section about solaris test results status: https://github.com/libunwind/libunwind#expected-results-on-solaris-x86-64.
[02:08:39] <am11> i think we can update it based on 28 passing.
[02:18:08] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Remote host closed the connection)
[02:19:33] <ypankov> or rather create an illumos entry there?
[02:20:54] <jbk> it's using autotools to detect the platform, so they're one in the same with that library
[02:21:56] <ypankov> too bad we are still "sunos"
[02:21:59] <jbk> and now that you're here :)
[02:22:17] <jbk> can you relook at the w/whodo cr?
[02:23:33] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Ping timeout: 256 seconds)
[02:25:00] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[02:26:27] <ypankov> yep, nitpicking as requested, other than that, just ship it :)
[02:27:35] <jbk> that is what the original did :)
[02:28:14] <jbk> granted, I mean techincally i suppose malloc could fail with EAGAIN, but I don't know if anyone's actually seen that in practice
[02:31:38] <jbk> though since you gave the +1, I'm guessing you'd be ok with deferring that to a cleanup
[02:31:54] <jbk> (I'd like to combine the two and some other non-security bits)
[02:32:57] <ypankov> yes, no need to stall this further
[02:34:53] <am11> jbk: could you please pass me the link to OS ticket for USRSTACK? i will add a comment in the vicinity of this hack `lpSystemInfo->lpMaximumApplicationAddress = (PVOID) 0xfffffd7fffe00000ul;`. :)
[02:43:18] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[02:43:30] *** richlowe <richlowe!~richlowe@2605:a000:160c:8b5b::bb7> has quit IRC (Ping timeout: 256 seconds)
[02:49:26] *** richlowe <richlowe!~richlowe@cpe-65-27-234-53.cinci.res.rr.com> has joined #illumos
[02:52:55] *** pcd <pcd!~pcd@openzfs/developer/pcd> has quit IRC (Ping timeout: 260 seconds)
[02:53:18] *** pcd <pcd!~pcd@openzfs/developer/pcd> has joined #illumos
[02:54:52] <jbk> i don't know if one's been filed on that
[02:55:18] <am11> jbk: test result from smartos 2019 (previous one was from 2018): http://sprunge.us/OSlVsm
[02:55:24] <am11> 27 passing :)
[02:55:55] <am11> guess with 2020 image i will also get 28/33.
[03:04:06] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[03:08:35] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Remote host closed the connection)
[03:15:04] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[03:18:36] *** MnrMcGregor <MnrMcGregor!~mnrmnaugh@unaffiliated/mnrmnaugh> has quit IRC (Quit: Leaving)
[03:19:48] *** amrfrsh <amrfrsh!~Thunderbi@134.19.189.92> has quit IRC (Quit: amrfrsh)
[03:20:12] *** amrfrsh <amrfrsh!~Thunderbi@134.19.189.92> has joined #illumos
[03:28:25] *** pcd <pcd!~pcd@openzfs/developer/pcd> has quit IRC (Ping timeout: 264 seconds)
[03:28:56] *** pcd <pcd!~pcd@openzfs/developer/pcd> has joined #illumos
[03:55:28] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has joined #illumos
[03:58:44] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[04:00:18] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has quit IRC (Ping timeout: 256 seconds)
[05:13:10] *** aaro1 is now known as aaro
[05:17:27] *** Alasdair <Alasdair!~alasdair@al.cloud.ec> has quit IRC (Quit: ZNC 1.7.5 - https://znc.in)
[05:56:28] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has joined #illumos
[06:01:48] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has quit IRC (Ping timeout: 265 seconds)
[06:52:41] *** BOKALDO <BOKALDO!~BOKALDO@46.109.206.117> has joined #illumos
[06:56:09] <rmustacc> am11, jbk: FWIW, we don't want to expose something like that as a macro. So if we want to add something, it feels more appropriate via a runtime thing ala sysconf, similar to how you can get a page size.
[06:59:47] *** leoric <leoric!~leoric@46.147.108.58> has joined #illumos
[07:00:00] <leoric> rmustacc: https://github.com/jurikm/illumos-fusefs
[07:03:46] <jbk> i have copies of the work gordon (and probably others) did that used a door based interface but never messed with it
[07:03:51] <jbk> https://github.com/jasonbking/illumos-fuse2
[07:04:19] <jbk> https://github.com/jasonbking/illumos-fuse1 -- that might be based on the original fuse impl as well
[07:04:35] <jbk> i figured maybe one of these days I could mess with it
[07:04:39] <jbk> just hasn't happened yet :)
[07:33:17] *** kev009 <kev009!~kev009@ip72-222-200-117.ph.ph.cox.net> has quit IRC (Remote host closed the connection)
[07:57:30] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has joined #illumos
[08:01:54] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has quit IRC (Ping timeout: 240 seconds)
[08:16:20] *** tsoome_ <tsoome_!~tsoome@91.209.240.229> has quit IRC (Quit: tsoome_)
[08:16:32] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[08:17:35] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Client Quit)
[08:22:40] *** leoric <leoric!~leoric@46.147.108.58> has quit IRC (Ping timeout: 256 seconds)
[08:31:25] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Ping timeout: 250 seconds)
[08:42:45] *** leoric <leoric!~leoric@46.147.108.23> has joined #illumos
[08:51:30] *** alanc <alanc!~alanc@inet-hqmc01-o.oracle.com> has quit IRC (Remote host closed the connection)
[08:51:57] *** alanc <alanc!~alanc@inet-hqmc01-o.oracle.com> has joined #illumos
[08:56:43] *** nde <nde!uid414739@gateway/web/irccloud.com/x-xhgfzpxccdzleqya> has quit IRC (Quit: Connection closed for inactivity)
[09:31:42] *** psarria <psarria!~psarria@37.red-79-146-98.dynamicip.rima-tde.net> has quit IRC (Quit: Leaving)
[09:33:42] *** psarria <psarria!~psarria@37.red-79-146-98.dynamicip.rima-tde.net> has joined #illumos
[09:37:43] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has joined #illumos
[09:38:05] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has quit IRC (Client Quit)
[09:50:28] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has joined #illumos
[09:54:37] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 256 seconds)
[09:58:05] *** neirac_ <neirac_!~cneira@pc-4-149-45-190.cm.vtr.net> has joined #illumos
[09:58:25] *** neirac <neirac!~cneira@pc-4-149-45-190.cm.vtr.net> has quit IRC (Read error: Connection reset by peer)
[09:59:21] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[10:38:33] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #illumos
[10:49:50] <am11> rmustacc: for linux, user code is setting `lpMaximumApplicationAddress` to `(PVOID) (1ull << 47)`. should i use the same on Illumos, instead of setting it to `0xfffffd7fffe00000ul` on illumos (value which i copied from https://github.com/illumos/illumos-gate/blob/4e0c5ef/usr/src/uts/i86pc/sys/machparam.h#L235, header which is not available in base
[10:49:51] <am11> install)?
[11:32:48] *** BOKALDO <BOKALDO!~BOKALDO@46.109.206.117> has quit IRC (Quit: Leaving)
[11:35:03] *** cneira__ <cneira__!~cneira@pc-4-149-45-190.cm.vtr.net> has joined #illumos
[11:35:03] *** neirac_ <neirac_!~cneira@pc-4-149-45-190.cm.vtr.net> has quit IRC (Read error: Connection reset by peer)
[11:44:38] <am11> on the topic of linkage-time issues; with VERBOSE=1 and j=1, the first error turned out to be about relocs http://sprunge.us/w10tqT. there are some adjuster thunks or stubs written in assembly, which manually adjust the calls to underlying abi.
[11:45:07] <am11> porting guide has some related information https://github.com/dotnet/runtime/blob/master/docs/design/coreclr/botr/guide-for-porting.md#assembly-stubs
[11:46:49] <am11> i am trying to determine if it is due to the difference in native linker vs. gnu linker, which happily compiles it on linux/bsds, or some other platform/abi difference.
[11:51:46] *** mnowak_ <mnowak_!~mnowak_@94.142.238.232> has quit IRC (Ping timeout: 256 seconds)
[11:53:43] *** mnowak_ <mnowak_!~mnowak_@94.142.238.232> has joined #illumos
[12:02:44] *** mnrmnaugh <mnrmnaugh!~mnrmnaugh@unaffiliated/mnrmnaugh> has joined #illumos
[12:21:24] *** neuroserve <neuroserve!~toens@ip-95-222-233-106.hsi15.unitymediagroup.de> has quit IRC (Ping timeout: 258 seconds)
[12:59:43] *** BOKALDO <BOKALDO!~BOKALDO@46.109.206.117> has joined #illumos
[13:04:04] <am11> ok, added `-mimpure-text -Wl,-ztextoff` to get by that reloc error. i will upstream the issue anyway as they might want to take some action.
[13:15:41] <am11> https://github.com/dotnet/runtime/issues/35154
[13:15:59] <sjorge> jbk can you take another peak at the pam_list change?
[13:16:08] <sjorge> I think I fixed your 2nd wave of feedback
[13:27:20] *** tsoome_ <tsoome_!~tsoome@91.209.240.229> has joined #illumos
[13:52:28] *** roryrjb <roryrjb!~rory@bl9-21-218.dsl.telepac.pt> has joined #illumos
[14:07:03] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has joined #illumos
[14:10:45] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[14:25:27] *** roryrjb <roryrjb!~rory@bl9-21-218.dsl.telepac.pt> has quit IRC (Quit: leaving)
[14:36:49] *** cneira__ is now known as neirac
[15:23:08] *** HotSwap <HotSwap!~hotswap@unaffiliated/hotswap/x-1719459> has quit IRC (Ping timeout: 256 seconds)
[15:23:19] *** HotSwap <HotSwap!~hotswap@unaffiliated/hotswap/x-1719459> has joined #illumos
[15:23:42] *** jthan <jthan!~jonathan@maxilaria.jrdemasi.com> has quit IRC (Ping timeout: 256 seconds)
[15:23:58] *** jthan <jthan!~jonathan@maxilaria.jrdemasi.com> has joined #illumos
[15:46:01] *** neirac <neirac!~cneira@pc-4-149-45-190.cm.vtr.net> has quit IRC (Read error: Connection reset by peer)
[15:46:03] *** neirac_ <neirac_!~cneira@pc-4-149-45-190.cm.vtr.net> has joined #illumos
[15:53:04] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[16:11:00] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[16:33:31] *** cneira__ <cneira__!~cneira@pc-4-149-45-190.cm.vtr.net> has joined #illumos
[16:33:37] *** neirac_ <neirac_!~cneira@pc-4-149-45-190.cm.vtr.net> has quit IRC (Remote host closed the connection)
[16:34:36] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has joined #illumos
[16:39:56] *** jamtorus <jamtorus!~quassel@s91904423.blix.com> has joined #illumos
[16:40:19] *** jellydonut <jellydonut!~quassel@s91904423.blix.com> has quit IRC (Read error: Connection reset by peer)
[16:42:28] *** leoric <leoric!~leoric@46.147.108.23> has quit IRC (Quit: Konversation terminated!)
[16:43:17] *** cneira__ <cneira__!~cneira@pc-4-149-45-190.cm.vtr.net> has quit IRC (Remote host closed the connection)
[16:43:32] *** cneira__ <cneira__!~cneira@pc-4-149-45-190.cm.vtr.net> has joined #illumos
[16:46:22] *** ChrisBF <ChrisBF!56890402@host86-137-4-2.range86-137.btcentralplus.com> has joined #illumos
[17:10:59] *** cneira__ is now known as neirac
[17:14:07] *** neirac <neirac!~cneira@pc-4-149-45-190.cm.vtr.net> has quit IRC (Quit: Leaving)
[17:14:30] *** jamtorus is now known as jellydonut
[17:24:51] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has quit IRC (Quit: leaving)
[17:25:32] *** neirac <neirac!~neirac@pc-4-149-45-190.cm.vtr.net> has joined #illumos
[17:27:27] *** jimklimov1 <jimklimov1!~jimklimov@ip-86-49-254-26.net.upcbroadband.cz> has quit IRC (Quit: Leaving.)
[17:32:20] *** tsoome_ <tsoome_!~tsoome@91.209.240.229> has quit IRC (Read error: Connection reset by peer)
[17:34:17] *** tsoome_ <tsoome_!~tsoome@91.209.240.229> has joined #illumos
[17:44:45] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[17:56:33] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[17:58:04] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[18:06:31] *** vozhyk <vozhyk!~vozhyk@89-74-155-109.dynamic.chello.pl> has joined #illumos
[18:15:01] *** pvs <pvs!uid19318@lopsa/member/pvs> has joined #illumos
[18:15:09] <vozhyk> Is there a way to run /boot/loader through GRUB2 without chainloading the pool partition?
[18:17:20] <vozhyk> I have a few Linux and OpenIndiana BEs on one pool and boot Linux normally with GRUB2 (it has a generated menu entry for every BE), but booting illumos is a bit inconvenient: I have to set bootfs to an illumos BE and chainload the pool's partition.
[18:18:27] <vozhyk> it would be nice if I could make GRUB2 pass the BE name to /boot/loader directly (bypassing gptzfsboot)
[18:20:46] <vozhyk> then I'd be able to choose any BE (Linux or illumos) in GRUB2 without having to change the active BE beforehand
[18:23:14] <vozhyk> I'd also like to add some FreeBSD BEs, and I suppose I would need to install the FreeBSD bootloader on the pool's partition, which would clash with the illumos one
[18:23:52] <vozhyk> but, if I can start the illumos/FreeBSD bootloader as files, that wouldn't be a problem
[18:27:25] <vozhyk> another way I tried was booting the illumos kernel with GRUB2, but it doesn't seem to work currently
[18:28:03] <vozhyk> and booting the illumos loader would be better anyway, as it has pretty menus for setting kernel options
[18:45:24] <tsoome_> vozhyk: have you attempted to load linux kernel from loader?
[18:45:48] <vozhyk> no
[18:46:11] <vozhyk> how can I do that?
[18:47:05] <tsoome_> it is definitely possible to pass BE name to loader, however, I’m not too sure if grub can actually deal with starting locader out of the box.
[18:47:33] <neirac> is the mdb support for go still working ? https://github.com/joyent/mdb_go
[18:48:54] <tsoome_> from ok load path/to/linux
[18:55:27] <vozhyk> ok, linux loaded
[18:55:34] <vozhyk> initramfs left
[18:57:55] <vozhyk> nice, the same command is used to load the initramfs
[19:02:08] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[19:05:32] <vozhyk> it boots nicely
[19:07:01] <vozhyk> would adding menu entries for Linux BEs to loader be practical?
[19:11:26] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[19:11:51] <vozhyk> what I'm actually trying to do is to make my beadm on Linux (a port of illumos C code) be able to manage other OSs' BEs to an extent
[19:12:54] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)
[19:12:57] <vozhyk> that port makes GRUB2 regenerate the boot menu on every change to BEs, but does nothing for GRUB1 or loader, as I've #define'd that code out
[19:17:51] <vozhyk> the most important part I want to achieve is to be able to reboot into another OS with just "beadm activate ...; reboot"
[19:18:10] <vozhyk> but to also be able to choose a BE at boot
[19:18:23] <vozhyk> I have the former so far
[19:19:48] <vozhyk> I'd be willing to make GRUB2 able to start loader just for fun, although I'm completely unfamiliar with code that might do that
[19:21:27] <LeftWing> neirac: I do not expect that code works these days, no
[19:21:55] <LeftWing> From what I recall it depended on a lot of very version specific information
[19:22:20] <am11> sjorge: coreclr build has succeeded. :)
[19:22:22] <LeftWing> I imagine you would need to use the DWARF information in Go binaries these days
[19:22:31] <LeftWing> am11: Neat!
[19:22:53] <neirac> LeftWing thanks a lot
[19:23:06] <am11> i will polish the patch and issue a PR.
[19:23:29] <vozhyk> another option would be to change my Linux libbe to use loader - I suspect I'd only need to make the menu.lst-editing code work
[19:24:01] <vozhyk> but I have no idea how an entry for Linux would look
[19:24:40] <vozhyk> the entries the illumos libbe generates contain only the name and the BE name, but no boot instructions
[19:25:32] <tsoome_> vozhyk: you can create custom entry for your needs, seek for chain in /boot/forth/menu*
[19:26:54] <vozhyk> tsoome_: nice, I'll look into that
[19:28:14] <neirac> LeftWing, will be too much work to make it work again?, are there mdb modules for rust ?
[19:30:18] <neirac> LeftWing I was trying to gather the mdb modules that exists for other languages to include them in omnios
[19:30:28] <LeftWing> neirac: There is not currently an MDB module for Rust of which I am aware, though the model is a bit different -- there isn't as much of a runtime as there is with Go
[19:30:38] <LeftWing> Binaries look at lot more like C/C++ binaries
[19:31:27] <LeftWing> I suspect you would need on some level to start again for Go. Not completely, probably, but you would need to understand what we were doing all those years ago and try and apply it to current Go toolchains
[19:31:58] <LeftWing> We were hard coding a bunch of stuff that you would really want to get from the DWARF, for instance
[19:32:45] <tsoome_> vozhyk: if you can load and start linux fine, then you can look on the problem as how to specify kernel and initramfs as module for your menu entry.
[19:34:50] <tsoome_> vozhyk: as for grub, you need 2 things - how to make grub to load and start loader binary and to pass arguments (gptzfsboot code does show how)
[19:35:43] *** ikonia_ is now known as ikonia
[19:36:51] <neirac> LeftWing oh, ok seems a lot of work,not just a couple of days. I'll revisit when I have more time. one question about go-kstats do we need to ask for the repo owner to hand over it ?
[19:38:21] <LeftWing> There are a number of options. You could ask cks to transfer ownership to the illumos org, and I suspect the GitHub redirects would make everything basically keep working
[19:38:29] <LeftWing> Or we can just fork it
[19:38:41] <LeftWing> But if we just fork it, you will need to update any consumers to use the new URL
[19:38:49] <LeftWing> Probably polite to reach out and ask first, I suspect
[19:39:16] <vozhyk> tsoome_: Is what gptzfsboot does related to multiboot or chainload? Trying to guess if there's related code in GRUB already.
[19:39:59] <tsoome_> no, gptzfsboot is just small stage we can store on relatively small space
[19:40:32] <vozhyk> I see
[19:45:10] *** Riastradh <Riastradh!~riastradh@netbsd/developer/riastradh> has quit IRC (Remote host closed the connection)
[19:46:03] <toastersonerson1> neirac (IRC): If you want debugging adapting delve will definetly be easier. the only thing missing is c code to locate the process entry point of the process to debug. a lot of it is already present in our psutils and should be copy paste.
[19:47:02] <tsoome_> vozhyk: bios loader has 3 stages, 512B pmbr, then gptzgdboot and finally loader. grub has 2 stages, but their stage2 is modular, so you have to go through the pain of configuration management.
[19:53:10] <vozhyk> do you mean using gptzfsboot as a different kind of GRUB stage2?
[19:53:55] <vozhyk> I was thinking of it being something like a "loader $file $args" command
[19:54:02] <vozhyk> I was thinking of it being something like a "loader $loader_file $args" command
[19:54:40] <vozhyk> oops, got too used to editing messages on non-IRC
[19:58:06] <tomww> 
[20:01:23] *** Riastradh <Riastradh!~riastradh@netbsd/developer/riastradh> has joined #illumos
[20:10:56] *** ChrisBF <ChrisBF!56890402@host86-137-4-2.range86-137.btcentralplus.com> has quit IRC (Ping timeout: 240 seconds)
[20:15:17] <vozhyk> tsoome_: how do the boot stages work on EFI?
[20:15:46] <tsoome_> UEFI only has loader.efi:)
[20:16:36] <vozhyk> does it search for the root pool in a similar way to gptzfsboot?
[20:16:51] <tsoome_> yes
[20:16:52] <vozhyk> is it possible to just tell it which pool/BE to use?
[20:17:46] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has quit IRC (Read error: Connection reset by peer)
[20:18:54] <tsoome_> I’m not sure our current version does.
[20:19:13] <tsoome_> it would be easy to add, if it does not.
[20:19:51] <vozhyk> I see, so there is a reasonable way to pass arguments to EFI programs
[20:20:57] <tsoome_> yes, chainload usually does allow that.
[20:21:22] <vozhyk> nice
[20:21:40] <vozhyk> I don't see that used in the random examples I found
[20:23:46] <tsoome_> to make it usable, would need some code:)
[20:26:29] * ypankov eyes tsoome_
[20:26:47] <tsoome_> woot
[20:26:48] <ypankov> speaking about efi, when are you going to fix the freebsd's one on vmware? :)
[20:27:35] <tsoome_> if im right, it is rather annoying amount of work.
[20:29:15] <vozhyk> are you referring to passing arguments to loader or the question about freebsd above?
[20:29:29] <tsoome_> I *think*, the issue is about moving that huge chunk of memory blindly into the place and hitting something bad while doing that.
[20:29:42] <tsoome_> freebsd one.
[20:29:58] <tsoome_> args is relatively easy one:)
[20:31:23] <vozhyk> nice
[20:31:55] <tsoome_> re: fbsd, first I thought the allocation of memory map just before switching off the BS is the cause, but it was not that easy.
[20:32:36] <tsoome_> but even that allocation is also just trap.
[20:35:16] <vozhyk> I originally started the project without EFI in mind. I have another system with Linux and Windows, and it's pretty easy to choose which bootloader to start on boot, but managing the EFI boot menu (when I have to) is a pain. So having a single (GRUB) boot menu and a single way of choosing which to boot by default (beadm activate) would be nice even on EFI.
[20:39:40] *** phyre <phyre!~phyre___@78.30.23.107> has joined #illumos
[21:19:05] *** BOKALDO <BOKALDO!~BOKALDO@46.109.206.117> has quit IRC (Quit: Leaving)
[21:25:50] <neirac> toastersonestor1 that seems an easier route, I'll check delve after I finish with the termios comments that you leaved on nomad
[21:45:41] <gitomat> [illumos-gate] 12523 Buffer overflow in w and whodo -- Jason King <jason.king at joyent dot com>
[21:49:38] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[21:51:30] <jbk> sjorge: two small suggestions: put 'grbuflen' as a local variable inside pam_sm_acct_mgmt() and make sure sysconf() doesn't return -1 (and error if it does)..
[21:51:41] <am11> if someone could shed light on reloc issue https://github.com/dotnet/runtime/issues/35154, that would be huge help. i have a workaround, but wanted to address it properly. :)
[21:52:43] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[21:53:33] <jbk> if you're building a shared library, AFAIK, all the constitutent objects must be compiled with -fPIC
[21:53:43] <jbk> you can't just specify that at the end when you're linking
[21:54:28] <am11> jbk: all code is compiled with fPIC, afaict.
[21:57:04] <jbk> hmm.. those are both assembly files
[21:58:15] <jbk> do you have a link to the source for those?
[22:01:43] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Read error: Connection reset by peer)
[22:01:55] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[22:01:57] <am11> yup, https://github.com/dotnet/runtime/blob/3fc245e/src/coreclr/src/vm/amd64/jithelpers_fast.S#L68 and https://github.com/dotnet/runtime/blob/3fc245e/src/coreclr/src/vm/amd64/unixasmhelpers.S#L29
[22:03:42] <am11> these thunks or stubs are basically used to align the misaligned function calls to the abi calling conventions.
[22:04:04] <am11> some info about them in porting guide - https://github.com/dotnet/runtime/blob/master/docs/design/coreclr/botr/guide-for-porting.md#assembly-stubs
[22:05:19] <jbk> the definition of that C_FUNC macro would be interesting
[22:06:39] <am11> it is just a pass-thru https://github.com/dotnet/runtime/blob/3fc245e/src/coreclr/src/pal/inc/unixasmmacros.inc#L12 :)
[22:08:13] <jbk> are you sure it's actually generating 'jmp ThePreStub@PLT' and not just 'jmp ThePreStub' ?
[22:08:33] <am11> on Ubuntu it is PLT, on SmartOS without PLT
[22:09:08] <jbk> but that macro should be expanding it if 'HOST_AMD64' or HOST_X86 is set
[22:10:05] <am11> C_FUNC doesn't have PLT, EXTERNAL_C_FUNC has it
[22:10:29] <am11> ThePreStub is wrapped with C_FUNC
[22:17:49] *** vozhyk <vozhyk!~vozhyk@89-74-155-109.dynamic.chello.pl> has quit IRC (Quit: WeeChat 2.7.1)
[22:18:39] *** neuroserve <neuroserve!~toens@ip-95-222-233-106.hsi15.unitymediagroup.de> has joined #illumos
[22:19:42] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[22:22:12] *** neuroserve <neuroserve!~toens@ip-95-222-233-106.hsi15.unitymediagroup.de> has quit IRC (Remote host closed the connection)
[22:22:24] *** neuroserve <neuroserve!~toens@ip-95-222-233-106.hsi15.unitymediagroup.de> has joined #illumos
[22:22:40] <am11> jbk: theprestub is defined here: https://github.com/dotnet/runtime/blob/3fc245e/src/coreclr/src/vm/amd64/theprestubamd64.S
[22:26:00] <jbk> i'm not sure offhand.. maybe richlowe might be able to say better
[22:26:09] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[22:29:18] <am11> jbk: thanks. other than this workaround (-mimpure-text for these two functions), the coreclr src is compiling fine.
[22:55:11] *** phyre <phyre!~phyre___@78.30.23.107> has quit IRC (Remote host closed the connection)
[23:01:22] <ypankov> jbk: I no longer can use `w` as user? w: failed to enable privilege bracketing: Error 0
[23:03:42] <ypankov> (BTW, looks like here errx() would be better)
[23:08:28] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Read error: Connection reset by peer)
[23:08:30] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has joined #illumos
[23:08:58] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[23:09:05] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Read error: Connection reset by peer)
[23:09:39] *** wacki <wacki!~wacki@i577B0D15.versanet.de> has quit IRC (Quit: Lingo: www.lingoirc.com)
[23:09:58] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[23:10:29] <LeftWing> ypankov: I suspect it is actually that __priv_bracket() should set errno when it is failing
[23:10:53] <LeftWing> It is otherwise a wrapper around a system call which presumably *will* set errno, but I suspect it is failing in the initial check there
[23:11:23] <LeftWing> Do we need to back 12523 out?
[23:12:02] <ypankov> I'm not sure, it could be that I'm doing something stupid, running w from proto after `dmake install` :) it would be nice if someone else would confirm
[23:14:48] <LeftWing> hmm
[23:15:04] <LeftWing> That might not work
[23:15:20] <ypankov> Right, still waiting for a full build
[23:16:10] <LeftWing> Though honestly not really sure why it would not work, as I only see changes in w and whodo
[23:16:48] <LeftWing> Oh wait
[23:16:52] <LeftWing> It is supposed to be setuid root
[23:17:45] <LeftWing> It should probably detect when it is not and print some kind of error message, but I imagine that is what is going on for you there
[23:18:14] <ypankov> Grrr, it's also isaexec'ed?
[23:18:47] <ypankov> I was looking at one in /usr/bin/ and *that* doesn't have any bits set :)
[23:19:11] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[23:19:49] <LeftWing> It does in fact seem to be!
[23:20:04] <ypankov> Wonder if we should un-isaexec it, as i386 one doesn't work anyway
[23:20:48] <jbk> i ran that as a normal unprivileged user
[23:20:51] <jbk> and as aroot
[23:21:02] <jbk> so let me see if I can figure out what I missed
[23:21:06] <ypankov> jbk: right, I run it from proto, sorry for the noise :)
[23:21:09] <LeftWing> You did not miss anything
[23:21:11] <jbk> oh ok
[23:21:11] <LeftWing> Well
[23:21:22] <LeftWing> We could make it error out more gracefully if the setuid bit is not on
[23:21:28] <LeftWing> But it is not otherwise broken
[23:21:33] <jbk> oh phew
[23:22:04] <LeftWing> I do suspect __priv_bracket() should set EINVAL or something for the error case it has in it
[23:22:19] <jbk> ahh ok
[23:22:36] <LeftWing> And w should probably complain if not running as root
[23:22:45] <LeftWing> But I suspect those are preexisting conditions haha
[23:23:11] <ypankov> EPERM?
[23:23:37] <LeftWing> Well I think there are multiple holes in the error reporting here haha
[23:23:51] <LeftWing> The thing I was concerned about was just from looking at the function
[23:24:01] <LeftWing> It seems to return -1 without setting errno at least once
[23:24:19] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has joined #illumos
[23:24:19] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has quit IRC (Client Quit)
[23:24:22] <jbk> previously __priv_backet errors were ignored.. except then you'd get errors if someone other than yourself is logged in (if not suid root)
[23:24:40] <jbk> since it was unlikely to be able to open the necessary files in /proc
[23:24:44] <LeftWing> Mmm
[23:25:42] *** roryrjb <roryrjb!~rory@209.250.250.222> has joined #illumos
[23:26:26] <jbk> and yeah, the isaexec bit was a bit of a surprise for me too
[23:27:09] <jbk> part of the suggested cleanup (besides combining whodo + w, and utilizing libproc more) was to just ship a 64-bit binary
[23:29:26] <jbk> a somewhat tangential thing i've thought might be interesting (though i'm sure there's some unforseen consequences that might make it a bad idea) would be if you could assign startup privs to binaries (maybe in extended attributes) -- you'd need to have some checks on changing and when you execute (e.g. can't add privs not in your permitted set, can run unless your limit set contains the necessary privs..
[23:29:32] <jbk> things like that)
[23:31:07] <ypankov> and someone (TM) needs to fix it to correctly report the 'what' in pseudo-terminals
[23:31:51] <ypankov> https://pastebin.com/pKpjkkn7 <- that doesn't look good
[23:32:13] <jbk> existing issue?
[23:32:42] <ypankov> Try it in screen, having several windows open
[23:34:15] <LeftWing> Only you can prevent forest fires!
[23:35:11] *** heroux <heroux!sandroco@gateway/shell/insomnia247/x-zjoagjvsfoxwfeot> has quit IRC (Ping timeout: 260 seconds)
[23:35:53] <andyf> mine looks ok..
[23:36:08] <ypankov> andyf: can you pastebin it please?
[23:36:14] <andyf> https://paste.ec/paste/iqXKnxpG#KJ1BdNJIDFi5QgRjTg9zff3DYMbOeHxjEi7vxyQfN1n
[23:36:19] <andyf> yeah, was getting there
[23:36:48] <jbk> i did open a ticket for improving things further for the two (including combinging the two)... might be a good place to note any pre-existing things
[23:39:26] <ypankov> andyf: for me it works inconsistently with screen, some lines are ok, some show the jobs from other windows
[23:40:05] <ypankov> in any case, I'm looking into that utempter thing, and tmux panes aren't showing the 'what' correctly
[23:41:22] *** roryrjb <roryrjb!~rory@209.250.250.222> has quit IRC (Quit: leaving)
[23:41:36] *** heroux <heroux!sandroco@gateway/shell/insomnia247/x-lfureqhxxlxpmzkf> has joined #illumos
[23:41:37] *** roryrjb <roryrjb!~rory@209.250.250.222> has joined #illumos
[23:42:52] <LeftWing> Something is definitely off with `w` in my SmartOS zone
[23:46:35] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[23:49:49] <jbk> w/ the change?
[23:50:35] <LeftWing> No just the same behaviour ypankov sees -- this is pre change
[23:50:41] <jbk> oh ok
[23:50:49] <LeftWing> It says "w" on rather a lot of lines where it should probably say something else
top

   April 18, 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 | >