Switch to DuckDuckGo Search
   March 28, 2020  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >

Toggle Join/Part | bottom
[00:42:21] *** Kruppt <Kruppt!~Kruppt@50-111-39-191.drhm.nc.frontiernet.net> has quit IRC (Quit: Leaving)
[01:06:13] *** xzilla <xzilla!~robert@pool-71-166-61-141.bltmmd.fios.verizon.net> has quit IRC (Read error: Connection reset by peer)
[01:12:01] *** xzilla <xzilla!~robert@pool-71-166-61-141.bltmmd.fios.verizon.net> has joined #illumos
[01:42:43] *** alanc <alanc!~alanc@inet-hqmc01-o.oracle.com> has quit IRC (Remote host closed the connection)
[01:43:09] *** alanc <alanc!~alanc@inet-hqmc01-o.oracle.com> has joined #illumos
[02:01:32] <alanc> and of course, now that richlowe has made me aware of getdate(3C), I've found it documented in the interface to useradd/usermod -e for setting expiration dates
[02:04:51] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[02:10:47] <rmustacc> I've got another AMD Zen 2 update for cpu perf counters, that if anyone wants to help me test, would be grealy appreciated.
[03:16:16] *** hawk <hawk!~hawk@d.qw.se> has quit IRC (Ping timeout: 256 seconds)
[03:38:02] *** varna_ <varna_!~varna@183.157.11.110> has quit IRC (Ping timeout: 256 seconds)
[03:38:21] *** varna <varna!~varna@183.157.11.110> has joined #illumos
[04:12:30] *** Psy0rz <Psy0rz!psy@lounge.datux.nl> has joined #illumos
[04:17:40] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[04:50:10] *** pvs <pvs!uid19318@lopsa/member/pvs> has quit IRC (Quit: Connection closed for inactivity)
[05:14:18] *** Asgaroth <Asgaroth!~Asgaroth@51.37.122.90> has joined #illumos
[05:15:54] *** Asgaroth_ <Asgaroth_!~Asgaroth@64.43.141.170> has quit IRC (Ping timeout: 240 seconds)
[05:53:19] *** varna <varna!~varna@183.157.11.110> has quit IRC (Ping timeout: 250 seconds)
[06:10:37] *** nikolam_ is now known as nikolam
[07:13:30] *** BOKALDO <BOKALDO!~BOKALDO@81.198.21.4> has joined #illumos
[07:15:08] *** neuroserve <neuroserve!~toens@ip-178-202-217-1.hsi09.unitymediagroup.de> has quit IRC (Ping timeout: 256 seconds)
[07:27:52] *** mnrmnaugh is now known as find
[07:44:52] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[08:06:02] *** neuroserve <neuroserve!~toens@ip-94-114-237-187.unity-media.net> has joined #illumos
[08:11:32] *** jellydonut <jellydonut!~quassel@s91904423.blix.com> has quit IRC (Quit: jellydonut)
[08:13:36] *** jellydonut <jellydonut!~quassel@s91904423.blix.com> has joined #illumos
[09:06:26] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[09:41:37] *** amrmesh <amrmesh!~Thunderbi@134.19.189.92> has joined #illumos
[09:45:20] *** amrmesh <amrmesh!~Thunderbi@134.19.189.92> has quit IRC (Client Quit)
[10:15:48] *** man_u <man_u!~manu@89-92-19-81.hfc.dyn.abo.bbox.fr> has joined #illumos
[10:35:09] *** nbhauke <nbhauke!~hauke@55d41b60.access.ecotel.net> has joined #illumos
[10:41:24] *** nbhauke <nbhauke!~hauke@55d41b60.access.ecotel.net> has quit IRC (Quit: nbhauke)
[10:55:19] *** nbhauke <nbhauke!~hauke@55d41b60.access.ecotel.net> has joined #illumos
[10:56:17] *** nbhauke <nbhauke!~hauke@55d41b60.access.ecotel.net> has quit IRC (Client Quit)
[10:56:42] *** ovi <ovi!~sh42@fsf/member/zeroSignal> has joined #illumos
[11:07:28] *** Teknix <Teknix!~pds@69.41.134.110> has quit IRC (Ping timeout: 256 seconds)
[11:08:45] *** Teknix <Teknix!~pds@69.41.134.110> has joined #illumos
[11:10:47] *** varna <varna!~varna@36.24.137.145> has joined #illumos
[11:57:16] *** andy_js <andy_js!~andy@51.146.99.40> has joined #illumos
[12:13:38] *** pwinder <pwinder!~pwinder@86.11.191.8> has joined #illumos
[12:16:54] *** pwinder <pwinder!~pwinder@86.11.191.8> has quit IRC (Client Quit)
[12:19:57] *** neirac <neirac!~cneira@pc-215-101-46-190.cm.vtr.net> has quit IRC (Read error: Connection reset by peer)
[13:08:20] *** neuroserve <neuroserve!~toens@ip-94-114-237-187.unity-media.net> has quit IRC (Ping timeout: 256 seconds)
[13:17:20] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[13:19:59] *** pwinder <pwinder!~pwinder@86.11.191.8> has joined #illumos
[13:25:50] *** pwinder <pwinder!~pwinder@86.11.191.8> has quit IRC (Quit: This computer has gone to sleep)
[14:01:23] *** wacki <wacki!~wacki@i577B81D6.versanet.de> has joined #illumos
[14:17:44] *** igork <igork!~igork@91.204.56.74> has joined #illumos
[14:27:22] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[15:14:13] *** andy_js <andy_js!~andy@51.146.99.40> has quit IRC (Read error: Connection reset by peer)
[15:14:42] *** andy_js <andy_js!~andy@51.146.99.40> has joined #illumos
[15:23:08] *** AndrewG <AndrewG!~andrew@a20.cucumber.me.uk> has quit IRC (Ping timeout: 265 seconds)
[15:23:20] *** amrmesh <amrmesh!~Thunderbi@134.19.189.92> has joined #illumos
[15:23:38] *** amrmesh <amrmesh!~Thunderbi@134.19.189.92> has quit IRC (Remote host closed the connection)
[15:26:13] *** AndrewG <AndrewG!~andrew@a20.cucumber.me.uk> has joined #illumos
[15:28:08] *** hawk <hawk!~hawk@d.qw.se> has joined #illumos
[15:28:34] *** jenelizabeth <jenelizabeth!~jenelizab@cpc155793-brmb11-2-0-cust474.1-3.cable.virginm.net> has quit IRC (Ping timeout: 240 seconds)
[15:40:19] *** jenelizabeth <jenelizabeth!~jenelizab@cpc155793-brmb11-2-0-cust474.1-3.cable.virginm.net> has joined #illumos
[15:44:34] *** jenelizabeth <jenelizabeth!~jenelizab@cpc155793-brmb11-2-0-cust474.1-3.cable.virginm.net> has quit IRC (Ping timeout: 240 seconds)
[15:46:27] *** jenelizabeth <jenelizabeth!~jenelizab@185.92.25.103> has joined #illumos
[15:47:02] *** find is now known as mnrmnaugh
[15:54:34] *** jenelizabeth <jenelizabeth!~jenelizab@185.92.25.103> has quit IRC (Ping timeout: 240 seconds)
[15:57:54] *** jenelizabeth <jenelizabeth!~jenelizab@185.92.25.103> has joined #illumos
[16:48:28] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[16:55:03] *** elegast <elegast!~elegast@83-161-172-44.mobile.xs4all.nl> has joined #illumos
[16:55:27] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has joined #illumos
[17:02:22] <rmustacc> liv3010m: I've been looking at the surrounding code and there's a bunch of suspect stuff in the area thta you're in that leaves me uncertain how the rings/groups work ever worked when it was added.
[17:03:03] <rmustacc> The good and bad news is that it's not specific to that chipset. So I think I'll try to at least get support for your bits wrapped up and then we'll try to untangle this issue, if that sounds ok to you?
[17:05:22] <liv3010m> Hi @rmustacc. Many thanks. Yes, it sounds right for me.
[17:06:29] <rmustacc> The issue you're seeing looks pretty much a given with some of the code I'm seeing. So it doesn't quite all make sense to me yet.
[17:07:53] <rmustacc> Basically, what's happened is that you're panicking due to an assertion (since I had built a debug version of the driver).
[17:08:01] <rmustacc> There are four slots for mac addreses on the card that we can program.
[17:08:34] <rmustacc> In the code path we think we have them free and therefore can assign them al.
[17:08:36] <rmustacc> *all
[17:09:17] <liv3010m> Is there anything I can do to help?
[17:09:18] <rmustacc> However, another piece of code when we set up a driver, seems to have claimed all four of those registers and filled them in.
[17:09:37] <liv3010m> I see
[17:10:16] <rmustacc> And because of the way the code is structured, the worrying thing is that on non-debug bits we're corrupting memory as a result without that assertion.
[17:11:23] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has quit IRC (Quit: Leaving...)
[17:12:12] <liv3010m> so this is going the be more complicated than planned because the driver has to be fixed as a whole?
[17:12:28] <rmustacc> Well, first step is we can get support for the part working
[17:12:37] <rmustacc> So then that'll work, but not with an aggr/vnic.
[17:12:46] <rmustacc> What I'm confused with is how it works for others today.
[17:13:20] <rmustacc> But if you don't use a link aggregation or vnics or other fancier features, then it'd work just fine I guess.
[17:13:54] <rmustacc> It's all fixable, but will require me diving in a bit more and I'd like to prove my theory first.
[17:14:33] <rmustacc> Which at least should be pretty easy in kmdb.
[17:15:01] <liv3010m> Actually I need them, in smartos i use a "lot" of zones so VNICs are a must :(
[17:15:19] <rmustacc> No worries. Makes sense.
[17:15:39] <rmustacc> That's what I'd expect.
[17:16:12] <liv3010m> we can try the kmdb stuff the moment you have time if you want
[17:16:23] <rmustacc> Sure, that'd be useful to prove what's going on.
[17:16:25] <liv3010m> its not urgent of course
[17:16:40] <liv3010m> I'm dual booting the system
[17:16:56] <liv3010m> so when we have to try i reboot and thats it
[17:17:23] <rmustacc> OK. Well, let's give it a shot then. Because if I can prove it, then I can write up a bug and know where to focus on, rather than guessing at the proble.
[17:17:48] <liv3010m> sounds good to me
[17:17:48] <rmustacc> If you boot, there should be some loader menu options for kmdb and debug.
[17:17:59] <rmustacc> The goal is to get us to the prompt at the beginning. I don't remember what they're called off hand.
[17:18:01] <liv3010m> yup I made that yesterday
[17:19:04] <rmustacc> OK. When you get to that first prompt, we want to run '::bp bge`bge_addmac' to set a break point on bge_addmac and then we can run ':c' to continue to execution.
[17:19:06] <liv3010m> the first option (kmdb) gave mi this (searching for photo..)
[17:19:56] <liv3010m> https://i.ibb.co/5FMHH80/IMG-1966.jpg
[17:20:15] <liv3010m> OK let me boot into OMNIOS
[17:20:30] <rmustacc> OK, if you end up there again, that'll work too.
[17:20:40] <rmustacc> Though it might need a bit more back and forth.
[17:32:59] <liv3010m> OK I did the ::bp bge`bge_addmac but didn't notice anything
[17:33:13] <rmustacc> It ended up at the same final prompt?
[17:33:19] <liv3010m> yes
[17:33:35] <liv3010m> I can boot "to kmdb" directly option if you want
[17:33:38] <rmustacc> Hmm, weird. That's fine. We cna do stuff from this point anyways.
[17:33:48] <liv3010m> I think its the same with les stuff loader
[17:33:53] <liv3010m> *loeaded
[17:34:02] <liv3010m> okey
[17:34:16] <liv3010m> *less
[17:34:28] <rmustacc> So you saw the assertion message right and the panic?
[17:36:45] <rmustacc> So in this case, let's run the following command '::prtconf -d bge | ::print struct dev_info devi_driver_data | ::print bge_t cur_addr[0] cur_addr[1] cur_addr[2] cur_addr[3]'
[17:36:59] <rmustacc> That's a bit of a mouthful. I'm hoping I transcribed it correctly.
[17:39:57] <liv3010m> "So you saw the assertion message right and the panic?" correct
[17:41:00] <liv3010m> If I boot into the last option (I don't remember if its called drop to kmdb) then I only see the kmdb prompt and no panic
[17:41:16] <liv3010m> I'll try the command you provided me
[17:41:26] <rmustacc> Ah, that's the place where we would set the break point. But let's just do that command above from the place you saw the panic.
[17:41:36] <liv3010m> okey
[17:43:34] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-136-97.rochester.res.rr.com> has joined #illumos
[17:48:23] <rmustacc> Did that command work by chance?
[17:48:35] <liv3010m> partially
[17:49:06] <liv3010m> it works till this '::prtconf -d bge | ::print struct dev_info devi_driver_data | ::print bge_t'
[17:49:23] <liv3010m> once I add cur_addr stuff it fails with the following message
[17:49:37] <rmustacc> Let me double check.
[17:49:42] <liv3010m> failed to find member cur_addr
[17:49:48] <rmustacc> Ugh, the struct member has two 'r's.
[17:49:52] <rmustacc> So it's curr_addr
[17:50:01] <liv3010m> of gbe_t: no such member of structure or nion
[17:50:25] <liv3010m> Ok let me test, one moment please
[17:50:32] <liv3010m> ;
[17:50:36] <liv3010m> ;)
[17:53:05] <liv3010m> it worked
[17:53:12] <liv3010m> we have lots of output
[17:53:48] <rmustacc> OK, great.
[17:53:50] <liv3010m> should I upload a pic?
[17:53:57] <rmustacc> Sure, if you don't mind.
[17:54:24] <liv3010m> well two because I had to break it with crtl+c...
[17:56:04] <liv3010m> part1: https://i.ibb.co/Yc6g2fR/IMG-1971.jpg
[17:56:04] <liv3010m> part2: https://i.ibb.co/cDS4Bfb/IMG-1972.jpg
[17:58:12] <rmustacc> Thanks! This confirms my hypothesis.
[17:58:45] <rmustacc> So I'm at least correct about why it's happening.
[17:58:49] <rmustacc> This is really helpful.
[17:58:49] <liv3010m> glad I could help
[17:59:04] <rmustacc> I'll try to put together a fix for this over soon.
[17:59:28] <rmustacc> Basically the driver code pre-programmed and used all the slots for the actual device mac address.
[17:59:42] <rmustacc> But it didn't adjust the other things that track how the slots are used.
[17:59:53] <rmustacc> So we think we have free mac addresses, but the state says otherwise.
[17:59:58] <rmustacc> So we don't find a free slot and blow the assert.
[18:00:14] <rmustacc> Which is better than the otherwise memory corruption we'd have from dereferencing beyond the array!
[18:00:16] <liv3010m> I understand
[18:00:51] <rmustacc> Thank you for your patience as we work through this.
[18:01:15] <rmustacc> I'll let you know when I have a fix ready.
[18:01:25] <liv3010m> you are welcome, many thanks to you for willing to fix this
[18:55:40] *** mgerdts <mgerdts!~textual@2600:6c44:c7f:ec89:84b:ccaa:4c3a:36c3> has joined #illumos
[19:00:25] *** mgerdts <mgerdts!~textual@2600:6c44:c7f:ec89:84b:ccaa:4c3a:36c3> has quit IRC (Ping timeout: 240 seconds)
[19:01:28] *** mgerdts <mgerdts!~textual@2600:6c44:c7f:ec89:c54c:48c2:f7f4:f79a> has joined #illumos
[19:10:22] <gitomat> [illumos-gate] 12310 Add demangle(1) command -- Jason King <jason.king at joyent dot com>
[19:37:26] *** andy_js <andy_js!~andy@51.146.99.40> has quit IRC (Read error: Connection reset by peer)
[19:38:22] *** andy_js <andy_js!~andy@51.146.99.40> has joined #illumos
[19:47:30] *** Mazon <Mazon!~Mazon@37.205.127.168> has joined #illumos
[19:47:33] *** Mazon_ <Mazon_!~Mazon@37.205.127.168> has quit IRC (Ping timeout: 272 seconds)
[19:49:03] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has quit IRC (Ping timeout: 260 seconds)
[19:50:29] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has joined #illumos
[19:55:32] <gitomat> [illumos-gate] 12404 audioemu10k: variable may be used uninitialized -- Toomas Soome <tsoome at me dot com>
[20:03:07] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has quit IRC (Ping timeout: 256 seconds)
[20:04:54] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has joined #illumos
[20:48:05] *** man_u <man_u!~manu@89-92-19-81.hfc.dyn.abo.bbox.fr> has quit IRC (Quit: man_u)
[20:51:03] *** igork <igork!~igork@91.204.56.74> has quit IRC (Quit: Leaving.)
[21:24:25] *** roryrjb <roryrjb!rory@gateway/vpn/privateinternetaccess/roryrjb> has quit IRC (Quit: Lost terminal)
[21:33:15] *** BOKALDO <BOKALDO!~BOKALDO@81.198.21.4> has quit IRC (Quit: Leaving)
[21:42:27] *** neuroserve <neuroserve!~toens@ip-178-202-217-1.hsi09.unitymediagroup.de> has joined #illumos
[21:56:45] *** jollyd <jollyd!~alarcher@2a01:cb1d:611:bb00:d531:fbb0:5595:f512> has quit IRC (Ping timeout: 240 seconds)
[21:57:37] *** jollyd <jollyd!~alarcher@2a01:cb1d:611:bb00:2949:cca9:6c2c:fc> has joined #illumos
[22:21:54] <Woodstock> libschily. hmpf.
[22:22:12] <LeftWing> Mmm?
[22:23:01] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 264 seconds)
[22:23:20] *** elegast <elegast!~elegast@83-161-172-44.mobile.xs4all.nl> has quit IRC (Remote host closed the connection)
[22:23:56] <Woodstock> LeftWing: i'm trying to figure out why cdrdao from pkgsrc segfaults instead of doing anything usefil
[22:24:05] <LeftWing> Ahh
[22:24:12] <LeftWing> On which platform?
[22:24:14] <Woodstock> LeftWing: looks like libschily's implementation of getfp() is the culprit
[22:24:16] <Woodstock> amd64
[22:24:20] <Woodstock> OI
[22:27:36] <jollyd> I did not even remember that it is packaged :S
[22:29:04] <Woodstock> in pkgsrc
[22:29:54] *** kovert <kovert!~kovert@204.141.173.249> has quit IRC (Ping timeout: 240 seconds)
[22:33:18] *** kovert <kovert!~kovert@204.141.173.249> has joined #illumos
[22:40:05] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Ping timeout: 256 seconds)
[22:40:13] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[22:41:40] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[23:16:55] *** neirac <neirac!~neirac@pc-215-101-46-190.cm.vtr.net> has joined #illumos
[23:19:57] *** andy_js <andy_js!~andy@51.146.99.40> has quit IRC (Quit: andy_js)
[23:26:09] <Woodstock> so ok, cdrdao includes a rather old version of libschily, and that has a check for stack scanning that doesn't detect that it fails
[23:27:27] <Woodstock> ironically the newer libschily in cdrtools properly fails this check, but for a completely bogus reasons: libschily's own special getfp() implementation is just broken.
[23:27:47] <Woodstock> i just wonder, why am i doing this on a saturday night?
[23:28:48] <LeftWing> Sounds like my kind of Saturday :P
[23:28:54] <Woodstock> well :)
[23:30:44] <LeftWing> Though a bogus getfp() implementation does sound pretty tiresome haha
[23:31:03] <LeftWing> I have to wonder why it even needs that
[23:31:10] <Woodstock> yes, that too
[23:31:20] <LeftWing> Feels like an attempt to be too clever
[23:31:54] <Woodstock> especially since it's not even used in cdrtools, and probably wouldn't be in cdrdao if the failure was detected during configuration
[23:37:29] *** wacki <wacki!~wacki@i577B81D6.versanet.de> has quit IRC (Quit: Lingo: www.lingoirc.com)
top

   March 28, 2020  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >