Switch to DuckDuckGo Search
   March 14, 2019  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >

Toggle Join/Part | bottom
[00:10:05] *** phyre__ <phyre__!~phyre___@31.211.191.164> has joined #illumos
[00:11:19] <LeftWing> Nevermind, that turned out to be a silly question
[00:14:48] <jlevon> LeftWing: is this with your KVM instance that won't boot?
[00:15:50] <LeftWing> Yes
[00:16:27] <LeftWing> I have thus far (I think) confirmed that the guest seems to have loaded the MBR, then the per-partition fake MBR, then the beginning of some kind of blob referenced there
[00:17:00] <jlevon> you had ipxe in play there from the screenshot?
[00:17:31] <LeftWing> Well, only in that we ship an iPXE-derived option ROM for the Virtio NIC
[00:17:41] <LeftWing> But we're not trying to boot from it
[00:18:28] <LeftWing> I read the stage2_sector (LBA) and stage2_size (#blocks) from the MBR, and then again from the per-partition fake MBR referenced there
[00:18:58] <LeftWing> I'm now staring into what would have been loaded at 0x7c00 as a result of that
[00:19:22] <jlevon> wait, you're chain loading another mbr instead of gptzfsboot?
[00:19:37] <LeftWing> No, that's what the loader does
[00:19:49] <LeftWing> According to the comments in installboot.c and a reading of the assembly
[00:20:22] <LeftWing> At least, that's what it seems like so far
[00:20:40] <LeftWing> The stub has an LBA and a sector count baked into it, which it uses BIOS routines to load
[00:21:29] <LeftWing> And then jumps to the start of what it loaded -- which happens to be, itself, another copy of the MBR
[00:21:45] <LeftWing> but with a different LBA/size
[00:21:56] <jlevon> that's not what stage2 normally is on a disk...
[00:22:36] *** c4rc4s <c4rc4s!~c4rc4s@unaffiliated/c4rc4s> has quit IRC (Remote host closed the connection)
[00:23:00] <jlevon> LeftWing: did you spot disk_label.so? it's useful for checking for MBR sigs and the like
[00:23:16] <LeftWing> I didn't!
[00:23:19] <LeftWing> I've been doing it by hand
[00:23:24] <jlevon> one mo
[00:23:31] <LeftWing> https://github.com/illumos/illumos-gate/blob/ebee07ff4f102cbd3179db7c5070283da35a79f3/usr/src/cmd/boot/installboot/i386/installboot.c#L58-L95
[00:23:46] *** andy_js <andy_js!~andy@94.6.62.238> has quit IRC (Quit: andy_js)
[00:24:01] <jlevon> /jlevon/public/bits/disk_label.so
[00:24:08] *** c4rc4s <c4rc4s!~c4rc4s@unaffiliated/c4rc4s> has joined #illumos
[00:24:34] <jlevon> LeftWing: ah, yeah, it might do that outside of pcfs
[00:24:53] <LeftWing> Yeah this is rpool on an OI box
[00:25:50] <jlevon> so the stage2_lba in the 2nd mbr should point to gptzfsboot inside the zfs boot area. I think?
[00:26:00] <jlevon> there's never enough steps for boot
[00:26:17] *** Guest81495 <Guest81495!~themsay@149.254.234.214> has quit IRC (Read error: Connection reset by peer)
[00:26:23] *** sinner <sinner!~themsay@149.254.234.214> has joined #illumos
[00:26:46] *** sinner is now known as Guest33329
[00:28:44] <jlevon> LeftWing: I sitll wonder if an un-modified boot will still go straight to stage2 though.
[00:29:00] <jlevon> LeftWing: I think the chained one may only be used *if* loader's MBR is over-written.
[00:29:16] <LeftWing> Well, I'm using dd to load the same sector it asks the BIOS to load
[00:29:19] <LeftWing> and it's definitely the chained MBR
[00:29:44] <jlevon> ok
[00:29:48] <LeftWing> in the root: STAGE1_STAGE2_LBA: 524544 STAGE1_STAGE2_SIZE: 1
[00:30:03] <LeftWing> at that location: STAGE1_STAGE2_LBA: 525568 STAGE1_STAGE2_SIZE: 332
[00:30:04] <jlevon> yeah the size is a clue :)
[00:30:06] <LeftWing> yup
[00:38:09] *** amrfrsh <amrfrsh!~Thunderbi@185.180.15.226> has quit IRC (Ping timeout: 246 seconds)
[00:41:19] <LeftWing> Yeah I think the gptzfsboot is just garbage
[00:41:55] <jlevon> something stomped on the boot area?
[00:43:13] <LeftWing> I guess it must have!
[00:43:15] <LeftWing> https://gist.github.com/jclulow/48e97e79e890ae376d544454adaafe0b
[00:43:41] <LeftWing> Jerry is testing some ZFS bits I think
[00:43:45] <LeftWing> so... maybe there are problems there.
[00:45:10] <LeftWing> That's the result of dumping a file I got from dd if=ZVOL_RDSK of=/tmp/file bs=512 skip=525568 count=332
[00:45:32] <LeftWing> Indeed that is also what appears at 7C00 in the guest
[00:45:35] <LeftWing> So I'm pretty sure we loaded it!
[00:45:52] <jlevon> there should be the fake mb header magic visible at least
[00:46:12] <jlevon> maybe look for that too?
[00:47:42] <LeftWing> Do you know what that value is
[00:48:25] <jlevon> one mo
[00:48:48] <jlevon> 0x1badb002
[00:48:52] <jlevon> it's like they knew
[00:48:55] <LeftWing> ha
[00:49:29] <LeftWing> 7a0: 02 b0 ad 1b 00 00 01 00 fe 4f 51 e4 00 00 00 00 .........OQ.....
[00:49:31] <LeftWing> there?
[00:50:36] <LeftWing> That string of bytes looks the same as it does on a host that isn't ruinous
[00:50:51] <jlevon> looks like it
[00:54:24] <jlevon> 0: 31 c9 xor %ecx,%ecx
[00:54:25] <jlevon> 2: 8e c1 mov %ecx,%es
[00:54:25] <jlevon> 4: 8e d9 mov %ecx,%ds
[00:54:30] <jlevon> I tihnk that's what it should start like
[00:54:34] <jlevon> from gptldr.S
[00:54:44] <LeftWing> Indeed it does on a working host
[00:54:55] <LeftWing> > 0::dis
[00:54:57] <LeftWing> 0: xorw %cx,%cx
[00:54:59] <LeftWing> 2: movw %cx,%es
[00:57:21] <LeftWing> The corruption seems limited to the first two blocks here
[00:57:36] <LeftWing> At least, by eye, 0x400 onward looks right
[00:58:09] <jlevon> yeah
[00:58:55] <LeftWing> So either installboot did something unfortunate, or (I'll check later) the gptzfsboot file is itself busted
[00:58:57] <LeftWing> or
[00:59:08] <LeftWing> it's the in progress port of the new MMP bits from ZOL
[00:59:35] <jlevon> does that have boot area changes?
[00:59:42] <LeftWing> Apparently it's a big patch
[01:00:05] <LeftWing> I could imagine a world in which it writes things to funny places in the partition
[01:00:09] <LeftWing> as a kind of reservation
[01:00:48] <LeftWing> I have to head out. I've synced up with Jerry on what I've found so far.
[01:01:38] <jlevon> wouldn't be amazed if that bumps on stuff we thought we owned
[01:01:39] <jlevon> there
[01:17:50] *** joltman <joltman!~znc@unaffiliated/joltman> has quit IRC (Ping timeout: 250 seconds)
[01:37:06] *** joltman <joltman!znc@gateway/vpn/privateinternetaccess/joltman> has joined #illumos
[01:42:33] *** phyre__ <phyre__!~phyre___@31.211.191.164> has quit IRC (Ping timeout: 245 seconds)
[02:20:00] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[02:20:07] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[02:21:25] *** danmcd <danmcd!~danmcd@static-71-174-113-16.bstnma.fios.verizon.net> has joined #illumos
[03:01:22] <LeftWing> Oh it's definitely the MMP thing
[03:01:43] <LeftWing> I found the MMP magic haha
[03:02:08] <LeftWing> Also it by design is writing verrrry close to the boot block
[03:03:08] <LeftWing> I guess we'll have to check the porting.
[03:17:36] *** Bdragon <Bdragon!~bdragon@drupal.org/user/53081/view> has quit IRC (Ping timeout: 268 seconds)
[03:24:21] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[03:28:19] *** merzo <merzo!~merzo@8-50-133-95.pool.ukrtel.net> has quit IRC (Ping timeout: 252 seconds)
[04:31:35] *** phyre__ <phyre__!~phyre___@31.211.191.164> has joined #illumos
[05:02:47] *** Diagon <Diagon!~DiagonalA@unaffiliated/diagonalarg> has joined #illumos
[05:04:13] *** Diagon <Diagon!~DiagonalA@unaffiliated/diagonalarg> has quit IRC (Remote host closed the connection)
[05:26:31] *** phyre__ <phyre__!~phyre___@31.211.191.164> has quit IRC (Ping timeout: 244 seconds)
[06:34:27] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has joined #illumos
[06:54:33] <tsoome> ou, so the boot block area was corrupted?
[07:30:58] *** argulp <argulp!~argulp@95.129.115.145> has joined #illumos
[07:34:56] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: tsoome)
[07:40:30] *** digrouz <digrouz!~digrouz@81.80.47.150> has joined #illumos
[08:12:25] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has quit IRC (Remote host closed the connection)
[08:14:50] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has joined #illumos
[08:51:41] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has quit IRC (Remote host closed the connection)
[08:56:26] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[08:57:53] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has joined #illumos
[09:29:51] *** _joes_ <_joes_!~Thunderbi@82-132-246-37.dab.02.net> has joined #illumos
[09:41:07] *** _joes_ <_joes_!~Thunderbi@82-132-246-37.dab.02.net> has quit IRC (Ping timeout: 255 seconds)
[09:42:24] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 268 seconds)
[09:42:33] *** andy_js <andy_js!~andy@94.6.62.238> has joined #illumos
[09:44:18] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[09:48:29] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[09:53:22] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Read error: Connection reset by peer)
[09:53:25] *** jimklimov1 <jimklimov1!~jimklimov@31.7.243.238> has joined #illumos
[09:54:01] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #illumos
[10:07:25] *** jimklimov1 <jimklimov1!~jimklimov@31.7.243.238> has quit IRC (Read error: Connection reset by peer)
[10:07:43] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[10:18:42] *** edef <edef!edef@NixOS/user/edef> has quit IRC (Read error: error:1408F10B:SSL routines:ssl3_get_record:wrong version number)
[10:19:30] *** man_u <man_u!~manu@manu2.gandi.net> has joined #illumos
[10:20:01] *** edef <edef!edef@NixOS/user/edef> has joined #illumos
[10:46:14] *** merzo <merzo!~merzo@8-50-133-95.pool.ukrtel.net> has joined #illumos
[10:47:28] <jlevon> LeftWing: yuck
[11:01:50] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 244 seconds)
[11:03:50] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[11:18:33] *** psarria <psarria!~psarria@98.red-79-150-188.dynamicip.rima-tde.net> has quit IRC (Read error: Connection reset by peer)
[11:20:00] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[11:20:07] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[11:21:55] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Read error: Connection reset by peer)
[11:22:25] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[11:26:51] *** neuroserve <neuroserve!~toens@195.71.113.124> has joined #illumos
[11:26:55] *** psarria <psarria!~psarria@98.red-79-150-188.dynamicip.rima-tde.net> has joined #illumos
[11:33:28] *** _joes_ <_joes_!~Thunderbi@cpc1-watf9-0-0-cust236.watf.cable.ntl.com> has joined #illumos
[12:02:24] *** _joes_ <_joes_!~Thunderbi@cpc1-watf9-0-0-cust236.watf.cable.ntl.com> has quit IRC (Ping timeout: 246 seconds)
[12:08:12] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)
[12:53:17] *** _Tenchi_ <_Tenchi_!~phil@207-255-80-203-dhcp.aoo.pa.atlanticbb.net> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[13:00:18] *** amrfrsh <amrfrsh!~Thunderbi@185.180.15.226> has joined #illumos
[13:05:14] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has joined #illumos
[13:10:42] *** _joes_ <_joes_!~Thunderbi@cpc1-watf9-0-0-cust236.watf.cable.ntl.com> has joined #illumos
[13:16:03] *** snuff-work <snuff-work!~snuff-wor@202-161-112-134.tpgi.com.au> has quit IRC (Read error: Connection reset by peer)
[13:23:40] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Ping timeout: 250 seconds)
[13:26:21] *** Guest33329 <Guest33329!~themsay@149.254.234.214> has quit IRC (Read error: Connection reset by peer)
[13:26:39] *** sinner <sinner!~themsay@149.254.234.214> has joined #illumos
[13:27:02] *** sinner is now known as Guest93128
[13:28:12] *** MarcelT <MarcelT!~marcel@tyhog-gua01a-216.cronon.net> has quit IRC (Remote host closed the connection)
[13:31:43] *** _joes_ <_joes_!~Thunderbi@cpc1-watf9-0-0-cust236.watf.cable.ntl.com> has quit IRC (Quit: _joes_)
[13:31:50] *** _joes_1 <_joes_1!~Thunderbi@cpc1-watf9-0-0-cust236.watf.cable.ntl.com> has joined #illumos
[13:55:19] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Read error: Connection reset by peer)
[14:01:58] *** cantstanya <cantstanya!~chatting@gateway/tor-sasl/cantstanya> has quit IRC (Remote host closed the connection)
[14:03:43] *** cantstanya <cantstanya!~chatting@gateway/tor-sasl/cantstanya> has joined #illumos
[14:08:17] *** MarcelT <MarcelT!~marcel@tyhog-gua01a-216.cronon.net> has joined #illumos
[14:13:17] *** Kurlon <Kurlon!~Kurlon@98.13.72.207> has quit IRC (Ping timeout: 252 seconds)
[14:35:55] *** rsully <rsully!~rsully@unaffiliated/rsully> has joined #illumos
[14:40:52] *** merzo <merzo!~merzo@8-50-133-95.pool.ukrtel.net> has quit IRC (Ping timeout: 268 seconds)
[14:40:55] *** phyre__ <phyre__!~phyre___@31.211.191.164> has joined #illumos
[14:58:58] *** gh34 <gh34!~textual@rrcs-162-155-144-114.central.biz.rr.com> has joined #illumos
[15:01:53] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[15:31:28] *** rsully <rsully!~rsully@unaffiliated/rsully> has quit IRC (Read error: Connection reset by peer)
[15:32:28] *** rsully <rsully!~rsully@unaffiliated/rsully> has joined #illumos
[15:34:35] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[15:42:46] *** phlm <phlm!~phlm@46.240.183.153> has joined #illumos
[16:01:00] *** tomoyat1 <tomoyat1!~tomoyat1@tomoyat1.com> has quit IRC (Quit: ZNC 1.6.5 - http://znc.in)
[16:01:21] *** _Tenchi_ <_Tenchi_!~phil@207-255-80-203-dhcp.aoo.pa.atlanticbb.net> has joined #illumos
[16:01:31] *** tomoyat1 <tomoyat1!~tomoyat1@133.130.119.65> has joined #illumos
[16:05:08] *** polishdub <polishdub!~polishdub@207.86.38.254> has joined #illumos
[16:09:21] <gitomat> [illumos-gate] 9874 Add fksmbcl development tool -- Gordon Ross <gwr at nexenta dot com>
[16:09:23] <gitomat> [illumos-gate] 9875 SMB client connection setup rework -- Gordon Ross <gwr at nexenta dot com>
[16:09:23] <gitomat> [illumos-gate] 9735 Need to provide SMB 2.1 Client -- Gordon Ross <gwr at nexenta dot com>
[16:12:56] <andyf> Woohoo!!!
[16:23:02] <jimklimov> yuippee
[16:23:26] <jimklimov> now... what's next... zfs TRIM support? :)
[16:23:39] <jimklimov> ARC surviving reboots?
[16:23:41] <jimklimov> L2ARC
[16:24:19] <sensille> persistent l2arc isn't merged yet?
[16:25:14] <jimklimov> I am not sure really... but as of FOSDEM diuscussions seems it was not yet
[16:25:34] <andyf> No it isn't, and it's not on anyone's priority list (it was discussed on a recent ZFS call)
[16:25:49] <jimklimov> and on our OmniOS server updated in late 2018, L2ARC did not survive reboots and had to be re-heated
[16:26:48] <jimklimov> we actually have a script to initiate a time-shifted series of `find /hotdir -type f -exec cat '{}' \; >/dev/null` upon boot
[16:27:45] <jimklimov> takes an hour or two for a few gigabytes scattered across many small files (NFS-shared ccache index for build farm workers), and likely needlessly wears out the SSD, but afterwards a rare I/O goes to spinning disks
[16:27:50] <jbk> there's nothing stopping anyone from doing the work
[16:28:03] *** phlm <phlm!~phlm@46.240.183.153> has quit IRC (Quit: WeeChat 2.4)
[16:28:19] <jimklimov> and that's part of why the OS updates are only applied rarely ;)
[16:28:23] <jimklimov> vicious circle :)
[16:34:01] *** digrouz <digrouz!~digrouz@81.80.47.150> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[16:34:53] <andyf> We just need a volunteer to dust off the old persistent L2ARC patches and rebase them on recent changes... and probably on encryption too
[16:35:29] <igork> andyf: encryption still in progress by lundman :)
[16:36:12] <igork> https://github.com/openzfs/openzfs/pull/489
[16:36:21] <igork> he is still active
[16:39:49] <tsoome> w have those 2 zfs panic cases too:)
[16:40:45] <igork> we have a little more panic :)
[16:43:22] *** lystra <lystra!~lystra@d53-64-11-169.nap.wideopenwest.com> has quit IRC (Ping timeout: 272 seconds)
[16:50:48] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Ping timeout: 250 seconds)
[16:58:03] *** _joes_1 <_joes_1!~Thunderbi@cpc1-watf9-0-0-cust236.watf.cable.ntl.com> has quit IRC (Ping timeout: 252 seconds)
[16:59:30] *** lystra <lystra!~lystra@d53-64-11-169.nap.wideopenwest.com> has joined #illumos
[17:04:07] *** amrfrsh <amrfrsh!~Thunderbi@185.180.15.226> has quit IRC (Ping timeout: 240 seconds)
[17:05:33] *** amrfrsh <amrfrsh!~Thunderbi@185.180.15.226> has joined #illumos
[17:09:42] *** neuroserve <neuroserve!~toens@195.71.113.124> has quit IRC (Ping timeout: 246 seconds)
[17:18:17] *** neirac <neirac!~cneira@190.162.109.53> has quit IRC (Quit: Lost terminal)
[17:19:53] *** richlowe <richlowe!~richlowe@2605:a000:160c:8b5b::bb7> has quit IRC (Ping timeout: 268 seconds)
[17:21:55] *** richlowe <richlowe!~richlowe@2605:a000:160c:8b5b::bb7> has joined #illumos
[17:22:08] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 245 seconds)
[17:31:42] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[17:41:33] *** eki <eki!~eki@dsl-hkibng41-567327-143.dhcp.inet.fi> has quit IRC (Ping timeout: 246 seconds)
[17:47:09] *** amrfrsh <amrfrsh!~Thunderbi@185.180.15.226> has quit IRC (Ping timeout: 246 seconds)
[17:49:39] *** eki <eki!~eki@dsl-hkibng41-567327-143.dhcp.inet.fi> has joined #illumos
[17:56:14] *** rsully <rsully!~rsully@unaffiliated/rsully> has quit IRC (Ping timeout: 250 seconds)
[17:56:36] *** rsully_ <rsully_!~rsully@unaffiliated/rsully> has joined #illumos
[17:56:52] *** rsully_ is now known as rsully
[18:00:15] *** morgib <morgib!uid342182@gateway/web/irccloud.com/x-czemwjdwnxxyhtyd> has joined #illumos
[18:23:30] *** merzo <merzo!~merzo@185.39.197.205> has joined #illumos
[18:39:18] *** merzo <merzo!~merzo@185.39.197.205> has quit IRC (Read error: Connection reset by peer)
[18:39:34] *** merzo <merzo!~merzo@95.67.66.254> has joined #illumos
[18:39:55] *** man_u <man_u!~manu@manu2.gandi.net> has quit IRC (Quit: man_u)
[18:58:31] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has quit IRC (Quit: Leaving)
[19:03:48] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has joined #illumos
[19:11:44] <alanc> protip for the next person to touch https://github.com/illumos/illumos-gate/blob/master/usr/src/lib/libc/port/gen/calloc.c - you might want to leave a note that things like the bourne shell & libbsdmalloc depend on that version of calloc calling *their* version of malloc, and bad stuff will happen in subtle & confusing ways if you ever change it to call into the libc malloc internals directly
[19:18:12] <igork> alanc: do you have updates for file? :) or just update for comment?
[19:18:56] <alanc> I'm just suggesting a commment to warn anyone who wants to change the code in the future - your copy of that code is fine as is
[19:19:06] * alanc isn't allowed to send you code changes anyway
[19:22:36] *** kkantor <kkantor!~kkantor@156.99.40.14> has joined #illumos
[19:29:47] *** merzo <merzo!~merzo@95.67.66.254> has quit IRC (Ping timeout: 240 seconds)
[19:35:22] <jeffpc> not that it really matters to me, but:
[19:35:22] <jeffpc> * A version number is included in the x86 SVR4 stat and mknod interfaces
[19:35:22] <jeffpc> * so that SVR4 binaries can be supported. An LP64 kernel that supports
[19:35:22] <jeffpc> * the i386 ABI need to be aware of this too.
[19:35:29] <jeffpc> SVR4 x86?
[19:35:47] <jeffpc> did that ever actually exist?
[19:35:52] *** Bdragon <Bdragon!~bdragon@drupal.org/user/53081/view> has joined #illumos
[19:36:39] <jeffpc> I suppose that's a question for alanc and others knowledgeable with history
[19:40:24] <richlowe> They all expect their own malloc to be called when interposed
[19:40:31] <richlowe> that's part of the (implicit) contract.
[19:45:23] *** amrfrsh <amrfrsh!~Thunderbi@185.180.15.226> has joined #illumos
[19:45:46] *** amrfrsh <amrfrsh!~Thunderbi@185.180.15.226> has quit IRC (Read error: Connection reset by peer)
[19:48:12] *** amrfrsh <amrfrsh!~Thunderbi@185.180.15.226> has joined #illumos
[19:49:31] *** atrius <atrius!~atrius@2600:1700:22d1:518:955:ac4a:d1eb:4bc7> has quit IRC (Ping timeout: 252 seconds)
[19:50:32] <alanc> richlowe: unfortunately, the person who added ADI support to ours didn't know about that implicit contract, so it would have helped us for it to be explicitly noted
[19:50:50] <richlowe> yuck
[19:50:59] *** atrius <atrius!~atrius@99-160-139-232.lightspeed.nsvltn.sbcglobal.net> has joined #illumos
[19:51:34] <alanc> jeffpc: yes: https://en.wikipedia.org/wiki/UNIX_System_V#SVR4
[19:52:28] <richlowe> and yes, several x86 svr4s existed
[19:53:08] <jeffpc> uhuh
[19:53:28] <richlowe> for a while, it was like a cottage industry
[19:54:25] <richlowe> alanc: do I want to know what goes wrong? :)
[19:54:59] <richlowe> alanc: also, you're 100% going to take this chance to strip the bourne shell of its custom allocator, right? :)
[19:55:12] <alanc> god no, I'm not insane enough to touch that
[19:55:24] <alanc> I just fixed bsdmalloc and left bourne shell to casper
[19:55:52] <jeffpc> alanc, richlowe: but there was a sparc-only time for solaris, no? IOW, this comment (and code) survived an obsolescence, removal, and subsequent re-introduction of x86 support, right?
[19:57:04] <richlowe> that's more complicated
[19:57:07] <richlowe> and I'm not sure I know the real answer
[19:58:32] <richlowe> I know Sun bought one those boutique x86 svr4s
[19:58:38] <alanc> Solaris 2.0 was the only real sparc only time for Solaris
[19:58:56] <alanc> after the x86 code came in at 2.1, it never left
[19:59:10] <alanc> richlowe: Interactive Unix, aka "SunSoft South"
[19:59:46] <jeffpc> alanc: ah, so the rewrite around sol 9(?) timeframe was just that - a rewrite?
[19:59:59] <richlowe> I think that was more a lie in the press.
[20:00:03] <richlowe> by the press.
[20:00:05] <alanc> Sun didn't ship Solaris 9 for x86 at first, but the engineers never got around to removing the code and kept skunkworks builds alive during the dark times until management decided to bring it back
[20:00:20] <alanc> yeah, there was no "rewrite" there
[20:00:29] <jeffpc> ah
[20:01:02] <richlowe> it's never been really clear which of the "there used to be separate source trees" statements are true, lies, or reflections of interactive either.
[20:01:06] <richlowe> to outsiders, that is.
[20:01:08] <alanc> there was some repair work to bring it back up to shape after being officially neglected and not fully QA'ed for a year or two, but nothing major
[20:01:52] <jeffpc> interesting
[20:02:00] <alanc> it depends on your definition of "source tree" - there used to be platform-specific consolidations for hardware bits, especially drivers, that had their own source gates
[20:02:03] <jeffpc> thanks for the history lesson :)
[20:02:21] <jeffpc> alanc: when did it become a "monorepo"?
[20:02:26] <alanc> never
[20:02:51] <alanc> we've reduced the number of consolidations, but still have about a dozen, each with their own repo
[20:03:03] <jeffpc> I meant os/net specifically
[20:03:09] <jeffpc> with the hw bits
[20:03:13] <alanc> our ON gate even has 4 separate hg repos at the moment
[20:03:25] <richlowe> I think jmcp killed the last x86 separate
[20:03:32] <richlowe> with the ihv gates
[20:03:43] <richlowe> so that'd be... 2009?
[20:04:02] <jeffpc> what a mess ;)
[20:04:08] <alanc> the x86 platform consolidation merged parts in around Solaris 8, the rest in Solaris 10 - the sparc graphics & networking (GAN) consolidation merged in parts in Solaris 10, the rest EOL'ed or merged in Solaris 11
[20:04:30] <alanc> oh yeah, forgot the IHV gate for NDA code - 2009 sounds about right for that
[20:04:40] <richlowe> there's also the time prior to 2.6 mentioned in the QDS document, when things were named "gate" because they were things (plural) one went to through into the tree
[20:04:46] <jeffpc> but it does explain why there seems to be so much duplication
[20:05:02] <alanc> and answering a question from a screen or so back: "/usr/sunos/bin/sh dies in setlocale() because it implements malloc() but not calloc()"
[20:05:49] <richlowe> jeffpc: I think "people suck" explains a lot more of the surprising ones
[20:06:01] <richlowe> though I guess nobody is lefting caring about the ATA sparcs
[20:06:26] <richlowe> s/lefting/left/
[20:06:44] <jeffpc> sure, people suck explains a lot
[20:07:27] <jeffpc> but having separate repos for different platform/hw bits means that you may have to copy more, or somehow factor things out which is harder because of the multi-repo layout
[20:07:32] *** phyre__ <phyre__!~phyre___@31.211.191.164> has quit IRC (Ping timeout: 250 seconds)
[20:08:51] <jeffpc> the short version is, that is ancient code trying to maintain compat with bins from 1988 (aka. pre-WWW)
[20:08:54] <jeffpc> ;)
[20:13:07] <richlowe> you're only allowed to call code ancient if it's older than you are
[20:13:50] <jeffpc> damn it
[20:13:52] <jeffpc> :)
[20:14:40] <jeffpc> can there be a 3 year fudge factor?
[20:14:42] <jeffpc> :)
[20:15:55] <alanc> dammit, that limits me to K&R C and Version 4 Unix
[20:16:30] <jeffpc> alanc: the good news is, nobody would disagree with you calling them ancient!
[20:29:26] *** morgib <morgib!uid342182@gateway/web/irccloud.com/x-czemwjdwnxxyhtyd> has quit IRC (Quit: Connection closed for inactivity)
[20:51:06] *** amrmesh <amrmesh!~Thunderbi@185.180.15.226> has joined #illumos
[20:52:18] *** amrfrsh <amrfrsh!~Thunderbi@185.180.15.226> has quit IRC (Ping timeout: 246 seconds)
[20:52:18] *** amrmesh is now known as amrfrsh
[21:03:01] <gitomat> [illumos-gate] 10524 wsdiff much slower after move from deprecated commands module -- Andy Fiddaman <omnios at citrus-it dot co.uk>
[21:20:01] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[21:20:08] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[21:26:09] *** freakazoid0223 <freakazoid0223!~matt@pool-108-52-159-210.phlapa.fios.verizon.net> has joined #illumos
[21:26:49] *** phyre__ <phyre__!~phyre___@31.211.191.164> has joined #illumos
[21:38:22] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has quit IRC (Ping timeout: 245 seconds)
[21:40:19] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has joined #illumos
[21:55:04] *** kkantor <kkantor!~kkantor@156.99.40.14> has quit IRC (Ping timeout: 246 seconds)
[21:58:51] *** Kurlon_ <Kurlon_!~Kurlon@bidd-pub-04.gwi.net> has joined #illumos
[22:02:25] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 246 seconds)
[22:03:10] *** Kurlon_ <Kurlon_!~Kurlon@bidd-pub-04.gwi.net> has quit IRC (Ping timeout: 255 seconds)
[22:05:05] *** rsully <rsully!~rsully@unaffiliated/rsully> has quit IRC (Quit: rsully)
[22:05:48] *** phyre__ <phyre__!~phyre___@31.211.191.164> has quit IRC (Ping timeout: 246 seconds)
[22:11:01] *** gh34 <gh34!~textual@rrcs-162-155-144-114.central.biz.rr.com> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[22:15:06] *** kkantor <kkantor!~kkantor@c-24-118-59-107.hsd1.mn.comcast.net> has joined #illumos
[22:18:12] <jlevon> richlowe: hey, just realised I forgot to commit the fix for usr/src/uts/intel/pcbe/core_pcbe.c: 1000: line > 80 characters which was my fault
[22:19:04] <richlowe> followup and (fix cstyle)?
[22:19:11] <jlevon> I didn't push yet
[22:19:32] <jlevon> I'll kick off another build and pbchk, but does your approval still stand?
[22:19:35] <jlevon> new webrev coming shortly
[22:20:30] <jlevon> http://cr.illumos.org/~webrev/jlevon/10146/usr/src/uts/intel/pcbe/core_pcbe.c.wdiff.html
[22:20:48] <richlowe> Yes.
[22:21:16] <jlevon> thanks
[22:21:48] <gitomat> [illumos-gate] 7780 mdb could extract NT_PRPSINFO information from core files -- Carlos Neira <cneirabustos at gmail dot com>
[22:29:23] *** alanc <alanc!~alanc@129.157.69.40> has quit IRC (Remote host closed the connection)
[22:29:50] *** alanc <alanc!~alanc@129.157.69.40> has joined #illumos
[22:38:55] *** Kurlon <Kurlon!~Kurlon@98.13.72.207> has joined #illumos
[22:52:36] *** snuff-work <snuff-work!~snuff-wor@202-161-112-134.tpgi.com.au> has joined #illumos
[23:27:27] *** merzo <merzo!~merzo@8-50-133-95.pool.ukrtel.net> has joined #illumos
[23:42:50] *** polishdub <polishdub!~polishdub@207.86.38.254> has quit IRC (Quit: leaving)
[23:43:06] *** merzo <merzo!~merzo@8-50-133-95.pool.ukrtel.net> has quit IRC (Ping timeout: 246 seconds)
[23:51:35] *** rsully <rsully!~rsully@unaffiliated/rsully> has joined #illumos
[23:59:32] *** merzo <merzo!~merzo@111-23-132-95.pool.ukrtel.net> has joined #illumos
top

   March 14, 2019  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >