Switch to DuckDuckGo Search
   February 7, 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 | >


NOTICE: This channel is no longer actively logged.

Toggle Join/Part | bottom
[00:05:22] *** andy_js <andy_js!~andy@90.215.171.48> has quit IRC (Quit: andy_js)
[00:06:18] *** Kruppt <Kruppt!~Kruppt@104.169.26.46> has joined #illumos
[00:13:37] *** arnold_oree <arnold_oree!~arnoldore@ranoldoree.plus.com> has quit IRC (Quit: Leaving)
[00:16:27] *** Kruppt <Kruppt!~Kruppt@104.169.26.46> has quit IRC (Quit: Leaving)
[00:22:32] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)
[00:23:05] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-141-249.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[00:26:58] *** jamtorus <jamtorus!~quassel@s91904426.blix.com> has joined #illumos
[00:27:16] *** jellydonut <jellydonut!~quassel@s91904426.blix.com> has quit IRC (Read error: Connection reset by peer)
[00:27:34] *** jbk <jbk!jbk@syzygy.io> has quit IRC (Read error: Connection reset by peer)
[00:27:42] *** jbk <jbk!jbk@syzygy.io> has joined #illumos
[00:30:58] *** Mazon <Mazon!~Mazon@37.205.127.168> has quit IRC (Quit: meh)
[00:31:38] *** Mazon <Mazon!~Mazon@37.205.127.168> has joined #illumos
[00:39:34] *** jamtorus is now known as jellydonut
[00:59:57] *** dippy <dippy!~dippy@67.133.97.99> has quit IRC (Quit: Leaving)
[01:11:14] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[01:11:59] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[01:35:25] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[02:23:10] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-141-249.rochester.res.rr.com> has joined #illumos
[02:52:09] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[03:03:47] *** danmcd_ <danmcd_!~danmcd@static-71-174-113-16.bstnma.fios.verizon.net> has joined #illumos
[03:03:47] *** danmcd <danmcd!~danmcd@static-71-174-113-16.bstnma.fios.verizon.net> has quit IRC (Read error: Connection reset by peer)
[03:03:52] *** danmcd_ <danmcd_!~danmcd@static-71-174-113-16.bstnma.fios.verizon.net> has left #illumos
[03:04:12] *** danmcd <danmcd!~danmcd@static-71-174-113-16.bstnma.fios.verizon.net> has joined #illumos
[03:31:39] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[04:00:42] <shuryanc[m]2> !th help
[04:01:07] <shuryanc[m]2> !tg help
[04:01:22] <shuryanc[m]2> * test
[04:01:22] <shuryanc[m]2> * test
[04:07:18] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[04:57:06] *** superjen96 <superjen96!jenelizabe@gateway/vpn/privateinternetaccess/jenelizabeth> has joined #illumos
[05:00:02] *** jenelizabeth <jenelizabeth!~jenelizab@cpc155793-brmb11-2-0-cust474.1-3.cable.virginm.net> has quit IRC (Ping timeout: 240 seconds)
[05:37:50] *** triffid <triffid!triffid@cthulhu.mcclung.systems> has quit IRC (Quit: WeeChat 2.2)
[05:56:59] *** btibble <btibble!~brantibbl@c-69-94-200-89.hs.gigamonster.net> has quit IRC (Remote host closed the connection)
[06:00:58] *** btibble <btibble!~brantibbl@c-69-94-200-89.hs.gigamonster.net> has joined #illumos
[06:02:39] *** btibble <btibble!~brantibbl@c-69-94-200-89.hs.gigamonster.net> has quit IRC (Read error: Connection reset by peer)
[06:03:12] *** btibble <btibble!~brantibbl@c-69-94-200-89.hs.gigamonster.net> has joined #illumos
[06:44:04] *** nde_ is now known as nde
[06:58:09] *** xzilla <xzilla!~robert@pool-71-166-61-131.bltmmd.fios.verizon.net> has joined #illumos
[07:11:14] *** kahiru <kahiru!~quassel@ip-89-102-207-18.net.upcbroadband.cz> has quit IRC (Ping timeout: 240 seconds)
[07:12:39] *** kahiru <kahiru!~quassel@ip-89-102-207-18.net.upcbroadband.cz> has joined #illumos
[07:13:20] *** BOKALDO <BOKALDO!~BOKALDO@87.110.146.164> has joined #illumos
[07:13:23] *** xzilla <xzilla!~robert@pool-71-166-61-131.bltmmd.fios.verizon.net> has quit IRC (Ping timeout: 272 seconds)
[07:14:27] *** jenelizabeth <jenelizabeth!~jenelizab@cpc155793-brmb11-2-0-cust474.1-3.cable.virginm.net> has joined #illumos
[07:17:49] *** superjen96 <superjen96!jenelizabe@gateway/vpn/privateinternetaccess/jenelizabeth> has quit IRC (Ping timeout: 272 seconds)
[07:20:58] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[08:01:54] *** tsoome_ <tsoome_!~tsoome@91.209.240.229> has quit IRC (Read error: Connection reset by peer)
[08:05:55] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: This computer has gone to sleep)
[08:21:47] <leoric> andyf: we've seen it here as well
[08:32:45] *** jimklimov <jimklimov!~jimklimov@ip-86-49-254-26.net.upcbroadband.cz> has joined #illumos
[08:33:39] *** alanc <alanc!~alanc@inet-hqmc02-o.oracle.com> has quit IRC (Ping timeout: 260 seconds)
[08:35:35] *** wiedi <wiedi!~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de> has quit IRC (Quit: ^C)
[08:37:53] *** alanc <alanc!~alanc@inet-hqmc02-o.oracle.com> has joined #illumos
[08:42:19] *** tsoome <tsoome!~tsoome@89.219.128.66> has joined #illumos
[08:43:53] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[08:44:30] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Client Quit)
[08:44:40] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[09:30:37] *** wiedi <wiedi!~wiedi@185.85.220.199> has joined #illumos
[09:33:13] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has quit IRC (Ping timeout: 268 seconds)
[09:36:43] <tsoome> whee: tsoome@tribblix:/code/illumos-gate/usr/src/uts$ echo $?
[09:36:44] <tsoome> 0
[09:44:55] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has joined #illumos
[09:50:57] <andyf> tsoome - gcc9?
[09:51:17] <tsoome> no, 4.4.4 build on sparc:D
[09:51:21] <andyf> leoric - I opened an issue for it.
[09:51:29] <andyf> tsoome - Nice!
[09:51:32] <tsoome> smatch + NULL
[09:52:00] <tsoome> still have some cleanups in userland...
[09:54:38] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has quit IRC (Remote host closed the connection)
[09:57:43] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has joined #illumos
[10:00:18] *** arnold_oree <arnold_oree!~arnoldore@ranoldoree.plus.com> has joined #illumos
[10:13:43] *** arnold_oree <arnold_oree!~arnoldore@ranoldoree.plus.com> has quit IRC (Remote host closed the connection)
[10:14:25] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #illumos
[10:16:46] *** man_u <man_u!~manu@manu2.gandi.net> has joined #illumos
[10:24:20] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Remote host closed the connection)
[10:25:21] *** arnold_oree <arnold_oree!~arnoldore@ranoldoree.plus.com> has joined #illumos
[10:27:31] *** andy_js <andy_js!~andy@90.215.171.48> has joined #illumos
[10:57:12] *** jollyd2 <jollyd2!~alarcher@77.158.181.1> has quit IRC (Quit: Leaving.)
[11:09:24] *** amrfrsh <amrfrsh!~Thunderbi@185.212.171.68> has joined #illumos
[11:15:11] *** amrfrsh <amrfrsh!~Thunderbi@185.212.171.68> has quit IRC (Quit: amrfrsh)
[11:18:47] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #illumos
[11:26:50] *** psydruid <psydruid!psydruidma@gateway/shell/matrix.org/x-qbtjbqblrqspmuwf> has quit IRC (Quit: killed)
[11:26:52] *** GrahamPerrin[m] <GrahamPerrin[m]!grahamperr@gateway/shell/matrix.org/x-qruwuevofafahxdp> has quit IRC (Quit: killed)
[11:27:06] *** Ericson2314 <Ericson2314!ericson231@gateway/shell/matrix.org/x-kcdipmpnpekpifsr> has quit IRC (Quit: killed)
[11:27:17] *** teutat3s <teutat3s!jhotmatrix@gateway/shell/matrix.org/x-cltercrfifglqwgf> has quit IRC (Quit: killed)
[11:27:23] *** shuryanc[m]2 <shuryanc[m]2!shuryancsi@gateway/shell/matrix.org/x-gxtufmsrjjydifdz> has quit IRC (Quit: killed)
[11:40:16] *** arnold_oree <arnold_oree!~arnoldore@ranoldoree.plus.com> has quit IRC (Quit: Leaving)
[11:45:58] <jlevon> andyf: pong
[11:46:30] <andyf> jlevon: Morning, it was just about this kpti_trampolines.s thing (https://www.illumos.org/issues/12285)
[11:46:44] <andyf> I wanted to confirm my understanding and see if you had any advice
[11:48:56] *** telser <telser!~telser@209.195.6.92> has quit IRC (Ping timeout: 246 seconds)
[11:49:42] *** telser <telser!~telser@209.195.6.92> has joined #illumos
[11:52:15] *** nde <nde!uid414739@gateway/web/irccloud.com/x-mlrmdtrolixnzlfq> has quit IRC (Quit: Connection closed for inactivity)
[12:03:20] <tsoome> great. smatch does not recognize uint32_t as unsigned type.
[12:04:24] <jlevon> tsoome: uh are you sure?
[12:04:40] <jlevon> andyf: this makes no sense. we don't pass -mintel64 to gas
[12:04:48] <jlevon> so what's it complaining about?
[12:05:27] <tsoome> usr/src/cmd/picl/plugins/sun4u/lib/fruaccess/crcutils.c 152 and 153, compute_checksum32() warn: potential negative subtraction from max '4294967295 - regval'
[12:05:38] <tsoome> fixed when regval is defined as unsigned
[12:06:32] <jlevon> 116 #define UINT32_T_MAX 0xFFFFFFFF
[12:06:42] <jlevon> that's signed is it not?
[12:06:50] <tsoome> I did add U to it.
[12:07:35] <tsoome> and did test with UINT_MAX
[12:08:27] <andyf> jlevon: it's complaining that the .s file uses sysenter/sysexit which are not valid on AMD in 64-bit mode
[12:08:42] <andyf> and since the default target is 'generic64', it fails to assemble
[12:08:42] <jlevon> but that's not what it should be doing
[12:08:46] <jlevon> unless we are passing -mintel64
[12:08:58] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has quit IRC (Remote host closed the connection)
[12:09:03] <andyf> with -mintel64 it builds fine, with -mgeneric64 (or no -m option), it aborts
[12:09:06] <jlevon> This option specifies that the assembler should accept only AMD64 or Intel64 ISA in 64-bit mode. The default is to accept both.
[12:09:17] <jlevon> looks like a gas bug to me
[12:09:41] <andyf> It's definitely a deliberate change from them but maybe ill advised
[12:10:19] <jlevon> did they forget to update the docs?
[12:11:28] <jlevon> found the specific change?
[12:11:33] <andyf> https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=d835a58baae720abe909795cb68763040d1750a8
[12:12:46] <jlevon> that's the disasm?
[12:13:37] <jlevon> or is that .tbl file count for asm too?
[12:13:50] <andyf> not sure - I have it checked out so I'll see what else has changed
[12:14:21] <jlevon> can you please report this upstream? It's definitely against the documentation
[12:14:23] <jlevon> so I think it's a bug
[12:14:40] <andyf> Yes, will do
[12:15:03] <andyf> Am I right that this code in kpti_trampolines.s does not get executed on AMD CPUs?
[12:16:27] <jlevon> yes
[12:16:33] <jlevon> please Cc: me in the mail
[12:16:49] <andyf> ok, thanks.
[12:17:31] <jlevon> does -march=generic64 work to re-enable both amd and intel compiling?
[12:18:10] <andyf> nope, and I thought that using arch-specific instructions when doing generic assembly was what they had decided to prevent
[12:18:17] <andyf> but it definitely doesn't match that documentation
[12:18:55] <jlevon> that's stupid if so
[12:20:12] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has joined #illumos
[12:22:02] <jlevon> so I have a slight preference to just accept -mintel64 directly in aw if it's the hack we need.
[12:22:11] <jlevon> and both places would need comments when we have an answer from upstream
[12:22:51] <andyf> and I thought about moving the file into an intel64/ subdirectory, just to make the makefile rule more generic
[12:23:05] <andyf> but we aren't in a rush to upgrade binutils (and OI can likely wait too)
[12:25:38] <jlevon> it's not an intel specific file though of course
[12:25:58] <jlevon> (ignoring our grossly mis-named directory structure of course)
[12:26:47] <jlevon> during the i86xpv days I spent quite a lot of time trying to pry apart intel versus i86pc versus amd64 better
[12:26:55] <jlevon> but it's way too much tedious work
[12:27:07] <jlevon> and i86xpv stopped being relevant of course
[12:28:18] <andyf> well it contains intel-specific instructions... I suppose splitting it up is harder than just temporarily hacking the makefile
[12:33:26] <jlevon> it would make sense if the same thing became two different byte sequences depending on the vendor
[12:33:33] <jlevon> maybe that's true now for some obscure corner
[12:37:11] *** TelegramBridge[m <TelegramBridge[m!telegramt2@gateway/shell/matrix.org/x-ynkudolxfwaerxer> has joined #illumos
[12:37:11] *** GrahamPerrin[m] <GrahamPerrin[m]!grahamperr@gateway/shell/matrix.org/x-dvpbevlegzwciyud> has joined #illumos
[12:37:12] *** teutat3s <teutat3s!jhotmatrix@gateway/shell/matrix.org/x-sfinfeautslfjqxo> has joined #illumos
[12:37:12] *** psydruid <psydruid!psydruidma@gateway/shell/matrix.org/x-oozvlwbdiwyzmqjs> has joined #illumos
[12:37:12] *** Ericson2314 <Ericson2314!ericson231@gateway/shell/matrix.org/x-sycczcxvcinmvlto> has joined #illumos
[12:37:18] *** shuryanc[m] <shuryanc[m]!shuryancma@gateway/shell/matrix.org/x-iqdxsyqdtscfmwas> has joined #illumos
[12:37:19] *** shuryanc1 <shuryanc1!shuryancsi@gateway/shell/matrix.org/x-qdnkilqunqiathtb> has joined #illumos
[12:38:33] <andyf> I opened the bug but you don't have an account on there so I can't add you as CC
[12:38:39] <andyf> https://sourceware.org/bugzilla/show_bug.cgi?id=25516
[12:40:03] <jlevon> I don't? huh.
[12:40:11] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has quit IRC (Ping timeout: 265 seconds)
[12:40:21] <andyf> well, not with your joyent address..
[12:40:32] <jlevon> ah levon at movementarian dot org there
[12:40:33] <jlevon> thanks
[13:03:52] <tsoome> jlevon wanna cry?:) see the crcmodel.h from http://src.illumos.org/source/search?q=SOLARIS_UNIX&defs=&refs=&path=&hist=&project=illumos-gate
[13:03:56] <tsoome> :>D
[13:03:59] <tsoome> :D
[13:05:18] <tsoome> that was reason why smatch did note recognize uint32_t.
[13:06:18] <jlevon> jesus
[13:06:32] <jlevon> I love that it's SOLARIS_UNIX and still not defined for us
[13:06:33] <jlevon> haha
[13:07:27] <tsoome> and in both instances.
[13:15:46] <ptribble> delving into some of the murkier recesses of the sparc code can be entertaining
[13:15:53] <ptribble> if you allow "entertaining" to include giving you nightmares, that is
[13:18:17] <tsoome> some of those are not limited to sparc really, just heavily gagged.
[13:30:48] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-141-249.rochester.res.rr.com> has quit IRC (Remote host closed the connection)
[13:31:25] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[13:46:49] *** zero-one <zero-one!zero-one@zer0-one.net> has joined #illumos
[13:52:02] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has joined #illumos
[13:54:09] *** BOKALDO <BOKALDO!~BOKALDO@87.110.146.164> has quit IRC (Quit: Leaving)
[13:59:35] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has quit IRC (Remote host closed the connection)
[14:03:46] *** jimklimov <jimklimov!~jimklimov@ip-86-49-254-26.net.upcbroadband.cz> has quit IRC (Quit: Leaving.)
[14:12:33] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has joined #illumos
[14:15:25] *** toasterson <toasterson!~toastmatr@85-195-235-249.fiber7.init7.net> has quit IRC (Read error: Connection reset by peer)
[14:15:43] *** toasterson <toasterson!~toastmatr@85-195-235-249.fiber7.init7.net> has joined #illumos
[14:21:32] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has joined #illumos
[14:48:36] <gitomat> [illumos-gate] 12266 loader: rewrite zfs reader zap code to use malloc -- Toomas Soome <tsoome at me dot com>
[14:51:36] *** robertdfrench <robertdfrench!~robertdfr@h69-11-139-135.cncrtn.dedicated.static.tds.net> has joined #illumos
[15:03:16] *** BOKALDO <BOKALDO!~BOKALDO@87.110.146.164> has joined #illumos
[15:08:13] *** Kruppt <Kruppt!~Kruppt@104.169.26.46> has joined #illumos
[15:08:27] *** Kruppt <Kruppt!~Kruppt@104.169.26.46> has quit IRC (Client Quit)
[15:09:19] *** Kruppt <Kruppt!~Kruppt@50.111.22.26> has joined #illumos
[15:09:52] *** shuryanc38 <shuryanc38!8a137c94@138019124148.ctinets.com> has joined #illumos
[15:17:33] *** shuryanc38 <shuryanc38!8a137c94@138019124148.ctinets.com> has quit IRC (Remote host closed the connection)
[15:18:20] *** shuryanc_solaris <shuryanc_solaris!8a137c94@138019124148.ctinets.com> has joined #illumos
[15:21:01] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 268 seconds)
[15:23:53] *** shuryanc_solari1 <shuryanc_solari1!shuryanc_s@gateway/shell/sameroom/x-cjsqfjosoqugjbtl> has joined #illumos
[15:41:39] *** shuryanc_solaris <shuryanc_solaris!8a137c94@138019124148.ctinets.com> has quit IRC (Remote host closed the connection)
[15:48:20] *** tsoome <tsoome!~tsoome@89.219.128.66> has quit IRC (Quit: tsoome)
[15:49:19] *** lblume <lblume!~lblume@greenviolet/laoyijiehe/lblume> has quit IRC (Quit: Leaving.)
[15:52:47] *** robertdfrench <robertdfrench!~robertdfr@h69-11-139-135.cncrtn.dedicated.static.tds.net> has quit IRC (Ping timeout: 240 seconds)
[15:54:58] <gitomat> [illumos-gate] 12235 Add libzutil for libzfs or libzpool consumers -- Don Brady <don.brady at delphix dot com>
[15:55:15] *** tsoome_ <tsoome_!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Ping timeout: 272 seconds)
[16:00:42] <jbk> i think it was libfru that i rewrote to be pure c, but then had no way to test...
[16:01:07] <jbk> but it seems to largely be useful only for oracle branded hw anyway, so...
[16:02:32] *** robertdfrench <robertdfrench!~robertdfr@h69-11-139-135.cncrtn.dedicated.static.tds.net> has joined #illumos
[16:12:02] *** man_u <man_u!~manu@manu2.gandi.net> has quit IRC (Ping timeout: 240 seconds)
[16:14:14] *** robertdfrench <robertdfrench!~robertdfr@h69-11-139-135.cncrtn.dedicated.static.tds.net> has quit IRC (Ping timeout: 268 seconds)
[16:14:48] *** lblume <lblume!~lblume@greenviolet/laoyijiehe/lblume> has joined #illumos
[16:15:48] *** man_u <man_u!~manu@78.194.27.39> has joined #illumos
[16:22:48] *** amrfrsh <amrfrsh!~Thunderbi@134.19.189.92> has joined #illumos
[16:28:49] *** man_u_ <man_u_!~manu@manu2.gandi.net> has joined #illumos
[16:30:16] *** man_u <man_u!~manu@78.194.27.39> has quit IRC (Ping timeout: 268 seconds)
[16:30:16] *** man_u_ is now known as man_u
[16:35:21] *** tsoome_ <tsoome_!~tsoome@f3c7-cf9f-d5b9-114e-2f80-4a40-07d0-2001.sta.estpak.ee> has joined #illumos
[16:35:47] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[16:43:48] *** nde <nde!uid414739@gateway/web/irccloud.com/x-adzerbnvowtkavmq> has joined #illumos
[16:45:40] <nde> danmcd: Hi. I have a machine running an Intel i350 and sending a lot of pause frames. Do you think this could be igb driver or firmware related?
[17:04:10] <danmcd> Could also be your switch? (Pause frames on GigE seems... wrong.)
[17:04:55] <danmcd> (Pardon latency now. Pardon upcoming latency, early-lunch here today.)
[17:10:18] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[17:16:44] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Remote host closed the connection)
[17:17:24] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[17:19:51] *** xzilla <xzilla!~robert@67.133.224.222> has joined #illumos
[17:44:47] *** mnrmnaugh <mnrmnaugh!~mnrmnaugh@unaffiliated/mnrmnaugh> has quit IRC (Ping timeout: 260 seconds)
[17:51:32] *** tsoome__ <tsoome__!~tsoome@91.209.240.229> has joined #illumos
[17:54:07] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Ping timeout: 240 seconds)
[17:54:07] *** tsoome__ is now known as tsoome
[17:58:05] *** mnrmnaugh <mnrmnaugh!~mnrmnaugh@unaffiliated/mnrmnaugh> has joined #illumos
[18:06:21] *** man_u <man_u!~manu@manu2.gandi.net> has quit IRC (Ping timeout: 272 seconds)
[18:11:06] *** morgib[m] <morgib[m]!morgibmatr@gateway/shell/matrix.org/x-nxjpxttfslwfhgqp> has joined #illumos
[18:11:21] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[18:17:55] <nde> danmcd: agree it's just running on GigE
[18:18:08] <nde> I totally get it.
[18:38:27] <danmcd> I350 only is GigE. Why there are pause frames is beyond me.
[18:39:27] *** wiedi <wiedi!~wiedi@185.85.220.199> has quit IRC (Ping timeout: 240 seconds)
[18:45:44] <nde> I know. That's confusing to me too. Do you think we should scrap those cards or is it most likely the switch at this point? I would think a decent amount of folks run the i350.
[18:49:02] *** Lirion <Lirion!~kesselink@wikimedia-commons/Lirion> has quit IRC (Remote host closed the connection)
[19:05:56] <danmcd> Do NOT scrap I350. It's well-tested both w.r.t. HW and driver.
[19:06:16] <danmcd> As a test, put two back-to-back and see if the pause-frame phenomenon still occurs. (It shouldn't.)
[19:06:16] <nde> Excellent. That's amazing feedback.
[19:17:04] *** arnold_oree <arnold_oree!~arnoldore@ranoldoree.plus.com> has joined #illumos
[19:17:33] *** wiedi <wiedi!~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de> has joined #illumos
[19:20:35] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Remote host closed the connection)
[19:21:04] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[19:23:43] *** xzilla <xzilla!~robert@67.133.224.222> has quit IRC (Ping timeout: 260 seconds)
[19:43:03] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has quit IRC (Remote host closed the connection)
[19:48:58] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Remote host closed the connection)
[19:49:26] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[19:49:54] *** basvdlei <basvdlei!~basvdlei@2001:980:a4c3:1:b88b:1590:d6bf:fbee> has joined #illumos
[20:07:48] <gitomat> [illumos-gate] 11083 support NFS server in zone -- Evan Layton <evan.layton at nexenta dot com>
[20:08:00] <jbk> \o/
[20:09:20] <danmcd> THUMP!
[20:10:36] <bahamat> Awesome work danmcd!
[20:11:12] <danmcd> Thanks. Also thanks to others who did the initial stuff, the others who additionally tested it (hey vgusev!), etc. etc.
[20:15:49] <tsoome> am..
[20:16:07] <tsoome> nice work:)
[20:16:39] <tsoome> I am a bit in trouble: /code/illumos-gate/usr/src/tools/proto/root_sparc-nd/opt/onbld/bin/sparc/ctfconvert -L VERSION kmod/gcore_isadep.o
[20:16:40] <tsoome> ctfconvert: file does not contain DWARF data
[20:18:42] <danmcd> Is LeftWing here? Because I marked 11083 in progress, the auto-close-after-push appears to not work.
[20:19:44] <tsoome> this mdb source is mdb/sparc/modules/genunix/gcore_isadep.c
[20:21:15] <ptribble> tsoome: there's a whole bunch of CTF issues like that on sparc
[20:21:47] <tsoome> this one happens because we do have -D_KMDB
[20:22:34] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has joined #illumos
[20:35:56] <LeftWing> danmcd: No it's because of the subtask
[20:36:03] <danmcd> Oh... oh...
[20:36:06] <danmcd> Thanks.
[20:36:36] <LeftWing> If you're trying to capture a dependent relationship, use a related issue instead
[20:57:13] <Woodstock> tsoome: i have a usb thumb drive with both a zpool and a ufs on it. how do i configure loader to boot either one at my request?
[20:58:07] <Woodstock> (and wtf is the fb console so slow? a vt100 at 9600 baud would be faster than that...)
[21:14:26] <tsoome> Woodstock: fb speed depends on gfx card actually. of course higher resolution and color depth will affect it too.
[21:14:47] <jbk> i remember when I was amazed at how fast 9600 baud was :)
[21:14:56] <jbk> (coming from 1200 then 2400 baud)
[21:17:36] <tsoome> Woodstock: if your ufs slice is set up properly, then you can use comand line chain command (like chain disk0s2: ) or add chain_disk=disk0s2: to have menu entry, or you can add entry to rpool/boot/menu.lst
[21:21:59] <richlowe> huh, -z guidance hates me again
[21:22:11] <jbk> heh
[21:22:17] <tsoome> :)
[21:22:48] <nde> danmcd: hey there. It turned out to be an upstream switch.
[21:22:52] <richlowe> The initial deflib work was smartos safe
[21:22:58] <richlowe> We had it checked
[21:23:10] <jbk> yeah, but the last change was not
[21:23:29] <richlowe> but this seems like it might just be busted?
[21:23:30] <richlowe> somewhere?
[21:23:51] <jbk> well it seems like there's perhaps the ever-present bad/wrong dependencies
[21:23:54] <jbk> that this somehow tickled
[21:24:06] <jbk> possibly/mostly with some of the add-on stuff
[21:24:17] <richlowe> Well, I'm getting errors for libraries being used (erroneously, admittedly) as interposes on libraries
[21:25:27] <richlowe> I think that's 90% of it, and then a couple of really unused dependencies
[21:25:27] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 240 seconds)
[21:25:27] <Woodstock> tsoome: whats the proper setup for the ufs slice? i just copied over the OI text install usb image, but all i get when i chainload it is a blank screen
[21:25:33] <richlowe> tsoome: did you miss a commit in your push?
[21:27:38] <richlowe> ugh, or is my ld busted :\
[21:28:05] <richlowe> does someone have a clean build since that integrated?
[21:28:37] <richlowe> 'cos yeah, these warnings are lies, but I can't see how I would have broken it, either.
[21:29:47] <tsoome> Woodstock: run installboot with ufs slice so you get the gptzfsboot installed there (with gpt you need boot slice there). also you need phys_path variable to be set. the chainload will read stage1 from partition, hece the installboot. Other option is to set currdev=disk0s2: and then just enter start (to read conf files)
[21:29:59] <richlowe> tsoome: I would have expected you to have had to change usr/src/cron to not link -lbsm into the parser tests added by andyf
[21:30:01] <danmcd> Thanks nde
[21:30:05] <richlowe> usr/src/cmd/cron
[21:30:15] <richlowe> tsoome: and you didn't(?), and I can't see how it could have worked for you without doing it
[21:30:25] <richlowe> tsoome: did you resolve the wrong side of a merge somewhere before pushing?
[21:30:27] <tsoome> richlowe: um I should not had — i had full builds done many times to be 100% sure
[21:31:40] <richlowe> tsoome: I'm about 95% sure I didn't pick the wrong side of a merge here, and in the cron parser test case at least ld is correct
[21:31:51] <tsoome> ah, I got ld: guidance: removal of unused dependency recommended: libbsm.so.1 too on sparc … interesting.
[21:32:15] <richlowe> tsoome: did you get others, largely about libraries trying to use interposers?
[21:32:29] <richlowe> but also a couple that look probably like real mistakes
[21:33:35] <tsoome> grr, and now x86 cron is telling the same.
[21:34:55] <richlowe> well, at least I'm not insane/haven't broken this ld in some weird quantum fashion
[21:35:22] <tsoome> this build system...
[21:35:49] <tsoome> well, just need to cut off those errors:)
[21:37:10] <Woodstock> tsoome: i tried installboot on the ufs slice (s7), but it installed gptzfsboot into the boot slice (s8). where and how do i set phys_path?
[21:37:34] <richlowe> tsoome: you might want to double check with other wise people about whether the right thing to do is stop the libraries (trying to) use the interposers.
[21:38:05] <tsoome> Woodstock: yes, because ufs has no space for gptzfsboot.
[21:38:13] <richlowe> but you could gag them and file followup bugs, to fix the build faster, too.
[21:40:08] <Woodstock> tsoome: once i ran installboot, the thing would boot from ufs by default, and no longer from zfs. so that didn't really solve the problem, just turned it around...
[21:40:26] <tsoome> Woodstock: the path can be set in either eeprom (bootenv.rc) or loader conf files
[21:41:08] <tsoome> yes, you need to run installboot again with default boot partition, so the PMBR will get restored.
[21:42:31] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC ()
[21:42:42] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:42:49] <tsoome> the device path is needed for rootfs mount as our kernel can not discover it otherwise (with zfs you get it from pool label - devid and phys_path)
[21:43:18] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:43:28] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:44:06] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:44:17] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:44:53] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:44:57] <Woodstock> tsoome: can you please summarize what i have to add to which file in order to be able to chainload the ufs slice, or how to get a menu item to switch currdev accordingly?
[21:45:04] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:45:41] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:45:51] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:46:28] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:46:39] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:46:54] <tsoome> echo fstype=ufs > /boot/conf.d/rootfs ; echo chain_disk=disk0s3: >> /boot/conf.d/rootfs
[21:47:16] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:47:25] <tsoome> and same file can be used for device path
[21:47:27] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:47:46] <Woodstock> and thats the /boot on the ufs, right?
[21:48:03] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:48:14] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:48:48] <tsoome> let me check over the variable name for device path
[21:48:51] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:48:54] <tsoome> yes
[21:49:43] <Woodstock> ok. s7 is actually disk0s1h:, i assume i should be using that?
[21:50:07] <tsoome> bootpath=/…
[21:50:11] <sjorge> congrats danmcd!
[21:50:20] <sjorge> There still needs to be a smartos change right?
[21:50:52] <tsoome> lsdev will show the exact name
[21:50:57] <Woodstock> tsoome: what should bootpath look like?
[21:52:55] <tsoome> yea, it was bootpath — traditionally it was stored in eeprom (bootenv.rc), but it does not really matter for simple setup. stmsboot may be surprised, however:D
[21:53:40] <tsoome> bootpath is physical path, that is, without /devices
[21:54:04] <Woodstock> ok. where would a plain OI text install usb thingy get it from?
[21:54:04] <tsoome> like /pci@0,0/pci15d9,805@1f,2/disk@0,0:a
[21:54:23] <tsoome> ls -l /dev/rdsk/c3t0d0s0
[21:54:23] <tsoome> lrwxrwxrwx 1 root root 53 Oct 15 2015 /dev/rdsk/c3t0d0s0 -> ../../devices/pci@0,0/pci15d9,805@1f,2/disk@0,0:a,raw
[21:55:05] <tsoome> format does show too, but without the slice
[21:55:20] <Woodstock> yes, i understand that part. but on a usb thumb drive the boot path is different from one system to another, and from one port to another on the same system
[21:55:43] <tsoome> Yes, thats why solaris does not support moving root disk around the systems
[21:55:50] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:56:10] <Woodstock> well, then tell me please the a usb image with ufs works fine normally?
[21:56:42] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:56:44] <tsoome> and LeftWing’s patch to teach the kernel to discover the disk devies before rootfs is mounted is still not there:)
[21:56:52] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:57:29] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:57:39] <tsoome> you mean the usb stick with install image? there the rootfs is in boot archive, loaded with kernel.
[21:57:40] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:58:06] <tsoome> and init will discover the install media
[21:58:08] <Woodstock> thats what i have in that usb image
[21:58:17] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:58:27] <Woodstock> and thats what i wrote into that slice
[21:58:28] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:58:44] <Woodstock> only difference is that originally it's in s0, but now it's in s7
[21:58:57] <tsoome> ah, then you do not need to set bootpath at all - that will leave boot archive to be used as rootfs
[21:59:04] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[21:59:15] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[21:59:52] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[22:00:02] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[22:00:18] <tsoome> kernel and boot_archive will be loaded and then the rest depends on how the BA was built - smartos is working like that
[22:00:39] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[22:00:51] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[22:01:27] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[22:01:38] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[22:01:47] <Woodstock> ok, that thing with /boot/conf.d/rootfs didn't do the trick. still chainloading disk0s1h: results in a blank screen and nothing else
[22:02:14] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[22:02:25] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[22:03:02] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[22:03:12] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has joined #illumos
[22:03:49] *** _alhazred <_alhazred!~Alex@mobile-access-bcee5b-130.dhcp.inet.fi> has quit IRC (Client Quit)
[22:04:11] <tsoome> am, s1h?
[22:04:21] <tsoome> you have mbr + vtoc?
[22:04:24] *** BOKALDO <BOKALDO!~BOKALDO@87.110.146.164> has quit IRC (Quit: Leaving)
[22:04:28] <Woodstock> yes
[22:06:04] <tsoome> then it is a bit more complicated; we install stage1 to MBR and partition start, but not in vtoc slice start.
[22:08:43] <tsoome> so, when you run installboot with ufs slice, you get pmbr stored in partition sector 0, and gptzfsboot stored after it and before s0. so, you need to read out the sector 0 from solaris partition, to some file, then you can use chain command with that file
[22:09:10] <tsoome> like chan /boot/pcode
[22:10:22] <tsoome> when you installboot with zfs slice, the solaris partition sector 0 will be updated for zfs slice.
[22:10:36] <Woodstock> so: 1. run installboot on ufs slice (s7), 2. dd s7 pmbr into file, 3. put file somewhere in /boot on zfs and use it as argument to chain_disk, and 4. installboot again on zfs slice
[22:11:19] <tsoome> chain command in bios system only reads 512B from pointed location and jumps on it
[22:11:44] <Woodstock> ok
[22:12:00] <Woodstock> so it's not s7 pmbr that i need to copy but the p1 pmbr?
[22:12:07] <Woodstock> because s7 doesn't have a pmbr, but p1 has?
[22:12:08] <tsoome> not s7, but 1 sector from s2 assuming s2 starts from 0
[22:12:49] <Woodstock> that should be equal to p1 in this case
[22:12:50] <tsoome> actually yes, you can use dd to store it to s7, indeed. ufs does have space for it
[22:12:57] <tsoome> yes
[22:13:03] <Woodstock> hm, thats another idea. hm.
[22:13:36] <Woodstock> btw., all that would be much easier if loader could just scan ufs filesystems for bootable systems and present them like bootenvs
[22:13:43] <tsoome> and then chainload from s7 should work as now you do have boot code in s7
[22:13:51] <Woodstock> or even if it would just allow me to add a custom menu entry for a ufs filesystem
[22:16:22] <tsoome_> you can add boots disk0s1h: into rpool/boot/menu.lst (with matching title line) - then it will appear in BE menu
[22:16:40] <tsoome_> bootfs disk0s1h:
[22:16:51] <tsoome_> stupid autocorrect
[22:18:54] <Woodstock> will that work without copying the pbr into s7?
[22:20:13] <tsoome_> loader menu.lst can have 3 types of entries, chain devicename: or bootfs device name: or bootfs pool/dataset
[22:20:43] *** jollyd <jollyd!~alarcher@aaubervilliers-682-1-58-169.w90-88.abo.wanadoo.fr> has joined #illumos
[22:20:58] <tsoome_> yes,bootfs means that we switch to that device and read the fs on it.
[22:22:51] <tsoome_> richlowe, indeed, somehow I have missed the list of ld: messages from build log. Apparently those have not been fatal as the nightly is happy, but I still feel stupid...
[22:24:09] <tsoome_> guess some followups are in order...
[22:24:29] <tsoome_> ok time to et sleep
[22:29:43] <richlowe> I think there was a reason we don't have fatal guidance
[22:29:53] <richlowe> it might be nitpicking about the definition of "guidance"
[22:29:58] <richlowe> rmustacc: do you remember?
[23:15:27] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[23:32:15] *** arnold_oree <arnold_oree!~arnoldore@ranoldoree.plus.com> has quit IRC (Quit: Leaving)
[23:42:49] *** arnoldoree <arnoldoree!~arnoldore@ranoldoree.plus.com> has quit IRC (Quit: Leaving)
[23:46:45] *** mnowak_ <mnowak_!~mnowak_@94.142.238.232> has quit IRC (Quit: Leaving)
[23:49:48] *** mnowak_ <mnowak_!~mnowak_@94.142.238.232> has joined #illumos
top

   February 7, 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 | >