December 20, 2006 
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
December 20, 2006 [00:07:21] <gisburn> !seen alhopper [00:07:25] <Drone> AlHopper (AlHopper!n=chatzill at logical dot logical-approach.com) was last seen in #opensolaris on Wed 14 Jun 2006 20:17 GMT, saying 'stefanp: it's becoming an increasing issue ... some Northbridges are already dissipating 20Watts ... and rising!'. [00:09:04] *** estibi has quit IRC [00:15:25] *** qdk has quit IRC [00:23:20] <delewis> is there any reason why Sendmail isn't making use of privileges? [00:23:47] <delewis> seems like all that's stopping both processes from running as smmsp is granting net_privaddr [00:24:56] <delewis> same goes for Bind [00:25:08] *** anthony79 has joined #opensolaris [00:26:12] *** _william_ has quit IRC [00:32:54] <dwc-> what about writing to mail spools [00:33:12] <dwc-> I can't think of a reason bind needs it other than for the port [00:35:53] <delewis> hmm, yeah, the mail spools make it tricky. [00:36:48] <jamesd> dwc-, also it needs access to the encryption code that is read only.. not sure its owned by root or bind [00:37:03] <wamty> "A devout Christian who said an accident at work boosted his libido and wrecked his marriage as he turned to prostitutes and pornography was awarded more than 3 million pounds ($5.89 million) in damages Tuesday." [00:37:10] <wamty> What on earth kind of an accident, can cause that? [00:37:27] <wamty> lol [00:37:36] <delewis> you could always grant file_dac_write, too [00:37:38] <jamesd> if you have to ask your wife doesn't want you to know ;-p [00:37:41] <delewis> while not ideal [00:37:46] <delewis> it's still better than granting all :-) [00:38:07] *** qdk has joined #opensolaris [00:38:38] <LeftWing> delewis: You could play with ACLs. [00:38:41] <delewis> what would be really nice is if you had the granularity to say which files it could write to or which ports it could bind to [00:39:02] <wamty> jamesd: lol yea [00:39:02] <wamty> but what? [00:39:06] <delewis> LeftWing, and grant smmsp access to write to the mail spools? that'd work. [00:39:27] <LeftWing> delewis: Yeah. With NFSv4 ACLs you could probably even grant very specific kinds of access like append, or some such. [00:42:54] <delewis> UFS doesn't support NFSv4 ACLs, does it? (only ZFS does, if I'm not mistaken) [00:43:15] <jamesd> yeah ufs is limited to the old and moldy acl's [00:43:35] <dwc-> jamesd: root:bind 640 would be sufficient to eliminate that need ... or you can just turn off remote [00:44:14] <delewis> I think someone needs to go through all of the SMF services and see which ones can be least privileged-ized. [00:44:25] *** wamty has quit IRC [00:44:26] <hali> cool, b54 comes with java6 :) [00:44:27] <dwc-> delewis: you'd need user homedirs through .forward to be readable too [00:44:27] <delewis> Bind is an obvious canidate, IMO. [00:44:32] <delewis> Sendmail would be tricky. [00:44:33] <dwc-> unless you move your .forward files into another dir [00:44:52] <delewis> dwc-, I still think granting file_dac_write would be the best solution, though, not exactly ideal. [00:44:53] <dwc-> and you wouldn't be able to support |procmail, |vacation, etc. [00:45:08] <delewis> well, file_dac_read, in that case :-) [00:45:17] <dwc-> because those need to run as the unprivileged user [00:45:34] <delewis> ugh. [00:46:00] <dwc-> now, I think you could do it more easily with postfix or qmail, or one of those 29837592384 parts smtpds [00:46:03] <delewis> and then you'd get into the mess about granting fork() which makes the entire least privilege model pointless. [00:46:14] <delewis> almost, anyway. [00:46:24] <dwc-> but I'm not really fond of either of those anyways [00:46:54] <onbot> commit by Rick McNeal: 6495328 daemon should use msync on mmap'd files; 6496416 iscsitadm modify admin -d <basedir> will report error if there are targets created by zvol directly; 6496424 targets can not be seen if created by zvol with shareiscsi property; 6499244 libiscsitgt leaks like a sieve [00:46:58] <dwc-> mail zone might be the easiest way to compartmentalize all the prileves for sendmail [00:47:15] <delewis> yes [00:47:27] <delewis> doing it with privileges is messy. [00:47:34] <delewis> if not impossible. [00:51:20] *** yarihm has quit IRC [00:53:23] *** Cass has quit IRC [00:53:35] *** Cass has joined #opensolaris [00:53:44] *** Error_404 has quit IRC [00:57:59] *** uebayasi has joined #opensolaris [00:58:02] *** pseudoXh4 has quit IRC [00:58:24] *** axisys has joined #opensolaris [01:08:59] *** vmhobbes has quit IRC [01:20:24] <onbot> commit by Raghuram Kothakota: 6502268 ntpdate(ntp) fails to connect to XSCF via DSCP [01:23:06] *** RaD|Tz has quit IRC [01:29:40] *** alfism has quit IRC [01:31:52] *** Peanut has quit IRC [01:34:16] *** lloy0076 has joined #opensolaris [01:38:11] *** alfism has joined #opensolaris [01:38:14] *** kb7sqi has joined #opensolaris [01:40:11] *** edwardocallaghan has joined #opensolaris [01:40:21] <edwardocallaghan> Evening [01:43:38] *** slowhog has joined #opensolaris [01:44:51] *** piwi has joined #opensolaris [01:47:55] *** lloy0076 has quit IRC [01:55:52] *** simford has joined #OpenSolaris [01:58:19] *** edwardocallaghan has quit IRC [02:00:01] *** Cass has quit IRC [02:00:02] *** peteh has quit IRC [02:00:02] *** mnowak_ has quit IRC [02:00:21] *** Cass has joined #opensolaris [02:00:34] *** peteh has joined #opensolaris [02:05:03] *** mnowak_ has joined #opensolaris [02:12:44] *** sahafeez has quit IRC [02:17:29] *** aliquis has quit IRC [02:17:38] *** nbkk6fo has joined #opensolaris [02:17:55] *** nbkk6fo is now known as phus [02:31:02] *** ily___sch has joined #opensolaris [02:37:27] *** sommerfeld has quit IRC [02:37:54] *** jsubl2 has joined #opensolaris [02:39:52] *** jimgris has joined #opensolaris [02:41:05] <Tpenta> hi jim [02:41:24] <jimgris> hey, alan [02:41:37] <jimgris> how's things down there? [02:42:12] <jmcp> hi jim [02:42:17] <jimgris> I'm trying to look into LWExpo Tokyo for the spring [02:42:58] <jimgris> hey, james [02:44:02] <jmcp> when's LWExpo Tokyo scheduled for? [02:44:39] <jimgris> May 30 - June 1 [02:45:22] <jmcp> that's only just Spring [02:45:33] <jimgris> I don't think Sun Japan is sponsoring (not sure yet), but I'd like to go as OpenSolaris in the .org area like we did in SF last time around [02:45:56] * gisburn grumbles [02:46:11] <gisburn> make sources released but still no dmake... ;-( [02:47:01] <elektronkind> ah man. what an exercise this has been. converting the openafs client kernel module to not poke cred_t structs directly and instead use the awesome new cool (and safe) ddi_cred routines. [02:47:12] <Tpenta> just about to go and bbq lunch [02:47:21] <gisburn> jimgris: Hi! [02:47:26] <jmcp> ah .. .I knew I was missing something up in the office [02:47:37] <gisburn> jimgris: Do you come to berlin in 2007 ? [02:47:38] *** ly___schi has quit IRC [02:47:49] <Tpenta> just bout a 250gm new york cut [02:48:05] <jimgris> gisburn: I would love to come but I can't make the trip [02:48:14] *** adp` has quit IRC [02:48:19] <gisburn> jimgris: Why ? [02:48:36] * Tpenta is getting realy annoyed by http://bugs.opensolaris.org/view_bug.do?bug_id=6404018 on his notebook [02:49:12] <gisburn> Tpenta: who had the clever idea to deploy ZFS on jurassic ? [02:49:35] <Tpenta> it's a sensible idea, it's what jurassic is for [02:49:38] <jimgris> gisburn: next time for sure, though. the timing is wrong for me. I'll be in India tne previous week and I need to be back in Japan the following week [02:49:41] <jmcp> gisburn: pretty much everybody did [02:49:54] <onbot> commit by apersson: PSARC 2006/314 Updated MIBs; 6501024 Add support for new IP/TCP/UDP MIBs [02:49:54] <onbot> commit by Govinda Tatti: 6504229 prtconf -v core dumps with snv_54 [02:50:03] <gisburn> grrrrr [02:50:12] <Tpenta> ooo that's a good oen to fix ;) [02:50:17] <gisburn> onbot: why on earth forget people SCTP all the time ? [02:50:17] <onbot> gisburn: Error: "why" is not a valid command. [02:50:19] <Tpenta> 6504229 [02:50:24] * gisburn kicks onbot [02:50:43] <jimgris> gisburn: sorry. :) I can't be in two places at once. [02:51:00] <jmcp> ye canna change the laws of physics! [02:51:20] <gisburn> jmcp: no, but chainsaws can be used for cheating. [02:51:25] <jimgris> i wonder how long the flight is to Berlin from Tokyo [02:51:42] <gisburn> via ICBM 18mins [02:51:47] <jimgris> and als oFrom Hyderabad [02:51:48] <Tpenta> my train trip to sydney airport from home took longer than the fligt to canberra yesterday [02:52:26] <gisburn> (okok, bad joke) [02:52:27] <jimgris> gisburn: yah, well, we don't have ICBM's here but maybe some day [02:52:36] <jimgris> :) [02:52:52] <jimgris> gisburn: how is the conf shaping up? [02:53:11] * gisburn wonders whether he could lure mike or april to berlin [02:53:45] <jmcp> that's the conf in feb/mar 2007, right? [02:53:48] <gisburn> jimgris: no idea. I've been asked three times whether I could do a presentation. [02:53:52] <gisburn> jmcp: yes [02:54:01] <jmcp> I thought about trying to go to it, but the airfare cost was insane - like close to AUD9000 [02:54:13] <gisburn> jimgris: either the project is very popular OR they still lack speakers [02:54:40] <gisburn> jmcp: 1AUD=10 yen ? [02:54:41] <gisburn> =:-) [02:54:47] <jmcp> gisburn: not sure [02:54:55] <jmcp> at the moment, AUD1 =~ USD0.80 [02:55:12] <jbalint> whats the routines for shared library initialization? init/fini, something like that? [02:55:13] <jmcp> so about USD7200 or so just for the flights [02:55:18] <Tpenta> that much? [02:55:21] <jimgris> gisburn: how many people are expected at the conf? [02:55:25] <jmcp> Tpenta: yeah ... like I said, insane [02:55:30] <jbalint> i looked at man _init but it seems to be specific to DDK [02:55:55] <gisburn> jimgris: *shug* ... I am not part of organising it. To be honestly I didn't even visit their pages yet. [02:55:58] <gisburn> er [02:55:59] <gisburn> shrug [02:56:11] * gisburn hits his rrrrrrrrrrr key [02:56:20] * gisburn hits harder [02:56:22] <jimgris> gisburn: ok [02:57:12] <gisburn> jimgris: sorry, but I am slightly uninformed about this... right now I having a little bit trouble elsewhere since your institute is going to close in januar or feb and I need a new job somehow... ;-( [02:57:41] <asyd> \_o< [02:57:50] <gisburn> asyd: ?! [02:58:05] <asyd> yes ? [02:58:37] <gisburn> "\_o<" == ? [02:58:46] <jamesd> gisburn, it == duck [02:58:51] <jimgris> gisburn: sorry about that [02:59:08] * gisburn wishes that the person who broke dbx should spend 15mins in a pit filled with komodo dragons [02:59:23] * asyd is a bit sad this week [03:00:34] <jimgris> Any of you guys in Asia? [03:00:53] <jimgris> North, I mean :) Sorry [03:01:32] <moazamraja> re [03:03:00] <jbalint> is there a command to print all predefined CPP amcros? [03:03:37] <gisburn> alanc: is kupfer around ? [03:03:56] <alanc> dunno - not many people here now [03:04:10] * gisburn looks at gman [03:04:15] <alanc> it's 18:00 and lots of people have already started holiday breaks [03:04:32] <alanc> Gman's on your continent right now [03:04:49] *** phus has quit IRC [03:05:02] *** dmick has joined #opensolaris [03:05:06] *** phus has joined #opensolaris [03:06:58] <alanc> jbalint: I usually look at the lists in the cc and cpp man pages [03:07:17] <gisburn> alanc: thats not enougth... ;-( [03:07:22] <jbalint> thanks, i'll look [03:07:28] <dmick> pre-#defines? [03:07:51] <gisburn> alanc: I had the same problem with AST's msgcpp and did some trial&error tests to get the list. [03:07:54] <gisburn> very nasty [03:07:59] <jbalint> dmick: yes [03:08:08] <dmick> docs.sun.com has a full list [03:08:11] *** anthony79 has quit IRC [03:08:20] <jbalint> mmm, any idea which "book" or the search keyword? [03:08:31] <dmick> Sun Studio from the top [03:08:36] <jbalint> i could have sworn there was some command... [03:08:39] <dmick> find your way to the right version of C User's Guide [03:08:47] <dmick> gcc has a command; SS never has, to my konwledge [03:08:49] <dmick> it should [03:08:58] <jbalint> ok, maybe i was thinking of the gcc one [03:09:05] <alanc> using -### to cc will show the ones it passes to cpp, but not cpp's pre-defines [03:09:38] <dmick> for example: [03:09:40] <dmick> http://docs.sun.com/source/819-3688/sun.specific.html [03:09:45] <dmick> look for 2.9 Predefined Names [03:10:13] <jbalint> thanks dmick [03:10:43] <jbalint> what's the deal with #pragma init? should i only use that if i dont use _init() and _fini() as the symbol names? [03:10:48] <dmick> although I don't see anything there for amd64. feh. [03:12:49] <dmick> it was my half-impression that naming it _init wasn't enough to put it in the .init section [03:13:00] <dmick> you had also to specify #pragma init (_init) [03:13:01] <dmick> but I could be high [03:13:46] <jbalint> well, i tried it and it worked with an so [03:13:55] <jbalint> wihtout the pragma that is [03:14:11] <jbalint> but i dont know if its called by the linker or what... [03:15:18] <dmick> would be interesting to see what nm shows for _init [03:15:19] <dmick> in that .so [03:15:32] <dmick> I just tried cc -S on a test with a routine named _init [03:15:39] <jbalint> yeah, i just am now looking at nm [03:15:42] <dmick> and it ended up in .text, like I'd expect [03:16:40] *** yongsun has joined #opensolaris [03:17:10] *** notibooki has quit IRC [03:17:41] <jbalint> here's what i get. left side is no pragma. http://www.improvedideas.com/files/init_fini.PNG [03:18:07] <jbalint> so there are extra entries, but the name is blank. not sure a better way to check [03:19:12] <dmick> now I gotta remember how sect symbols look [03:19:35] <jbalint> ah, i found it in the .o with dump [03:20:05] <jbalint> i guess the linker would call it if there is no .init section maybe [03:20:24] <dmick> you'd think that would be documented [03:20:25] <dmick> but [03:20:30] <dmick> are you sure it's not in the .init section? [03:21:14] <dmick> nm shows a diff section from _START_, at least [03:21:45] <jbalint> there is no .init section without the pragma [03:22:22] <dmick> ok, but [03:22:26] <dmick> ther'es a section 5 [03:22:32] <dmick> that only _init and _fini are in [03:22:45] <dmick> don't know why nm doesn't show section names [03:24:23] *** sparc-kly has quit IRC [03:24:36] <jbalint> is there a restriction on the name to be exactly "_init" [03:25:01] <dmick> well, certainly not with the #pragma [03:25:12] <dmick> but I've yet to find any indication of special behavior for that symbole [03:25:19] <dmick> other than what you're saying [03:25:58] <jbalint> here's nm -s [03:25:59] <jbalint> http://rafb.net/paste/results/Aq4ZwQ86.html [03:27:02] <Stric> I wrote code on some solaris 2.6 machines many years ago that used the fact that _init got called before main().. trying to run that on sol8 (could have been 9), did not give expected results.. _init wasn't run.. [03:27:03] <dmick> even more bizarre. not in .init even with pragma. [03:27:23] <Stric> although what I've read from man pages etc, _init is only available for kernel drivers.. [03:27:28] <jbalint> but this output from dump shows them http://rafb.net/paste/results/bnBo3w19.html [03:27:35] <Stric> so it was for sure using an unsupported api.. [03:27:40] <dmick> _init and _fini are definitely magical for kernel modules [03:27:43] <dmick> that's documented [03:27:58] <Stric> but for userspace programs, it's undocumented (afaik?) [03:28:01] <jbalint> but i tried calling them, even "init" and "fini" (no underscore) and it didnt work, even with the pragma [03:28:15] <dmick> well, it shows that there *is* an .init section [03:28:18] <jbalint> is there any more portable way for library initialization [03:28:26] <jamesd> fire up gdb or dbx put a break point in main, and get a stack trace to get an idea what called main [03:28:33] <jbalint> yeah, and what are the .rela.* sections? [03:28:57] <dmick> jbalint: forty questions at a time please. :) [03:28:58] <Stric> I used it for such purpose... an LD_PRELOAD library that was called before any app that was started in that process tree.. [03:29:48] <dmick> from the linker code: [03:30:02] <jbalint> i tried with dbx and main is the only stack frame. the init() functio was called already [03:30:23] *** slowhog has left #opensolaris [03:30:28] <dmick> "Two special symbols are _init and _fini. If these are supplied by crti.o then they are used to represent the total concatenation of the .init and .fini sections." [03:30:29] <dmick> hm [03:31:17] <dmick> it may do to look at the docs.sun.com linker manual too [03:31:19] <jbalint> but that's not for shared libs, is it? [03:31:24] <onbot> commit by cg149915: 6490277 AD1988 hd audio codec support; 6501554 audiohd should support Analog Devices AD1986 hd audio codecs; Contributed by Juergen Keil <jk at tools dot de> [03:31:39] <jbalint> here's the stack from _init http://rafb.net/paste/results/u0ONmx46.html [03:32:42] <dmick> docs.sun.com [03:32:50] <dmick> title search "linker" for the linker & libraries guide [03:32:59] <dmick> look for a section titled "Initialization and Termination Routines" [03:33:56] <jbalint> i have that one downloaded, just can remember which number combination it is... [03:34:07] <dmick> 817-19984 [03:34:11] <dmick> er, 1984 [03:35:14] <dmick> it looks like _init and _fini are reserved for the compiler [03:35:33] <jbalint> because they arent mentioned? [03:35:40] <dmick> and all .init routines get called from a routine crti.o provides, named _init [03:35:45] <dmick> for dynlibs [03:35:47] <jbalint> yeah, i saw that... [03:35:52] <dmick> similarly, .init sections rule [03:36:04] <dmick> so I would think #pragma init has to work [03:36:14] <dmick> there are some ldd tricks in there to try [03:36:41] <jbalint> ok, i'll look through it [03:36:41] <dmick> in the Nov 2006 version, p36 has info, as does p82 [03:38:11] <jbalint> great. Thanks alto [03:38:56] <dmick> gl [03:44:56] *** aFlag has joined #opensolaris [03:46:12] *** jimgris has left #opensolaris [03:57:16] *** piwi has quit IRC [04:00:07] <richlowe> While I guess it's in part due to being on opensolaris.org, the fact the MIB update folks had update_sma on opensolaris.org in advance is real cool. [04:00:14] <aFlag> hello, could someone point me to a guide teaching how to upgrade opensolaris from build 53 to build 54? [04:02:25] <onbot> commit by apersson: PSARC 2006/314 Updated MIBs; 6501024 Add support for new IP/TCP/UDP MIBs [04:07:33] <dmick> aFlag: don't know that there's anything special about "opensolaris" (maybe you mean the Community Release) [04:07:42] <dmick> but with "normal Solaris Express", there's no guide per se [04:07:47] <dmick> you just select Upgrade when given the optinos [04:07:50] <dmick> *option [04:08:20] *** mnowak_ has quit IRC [04:08:20] *** peteh has quit IRC [04:08:23] <aFlag> when what gives me that option? [04:08:30] *** mnowak_ has joined #opensolaris [04:08:33] *** peteh has joined #opensolaris [04:08:49] <alanc> the installer that comes up when you boot from the install media [04:09:15] <aFlag> so I have to download all the cds to upgrade? [04:09:24] <alanc> unfortunately, yes [04:09:33] <aFlag> there's no way to do it through the net? [04:09:44] <dmick> nope [04:09:48] <alanc> the upgrade process was designed for people doing it no more than a couple of times a year, not every other week [04:10:12] <alanc> it hasn't caught up to the Solaris Express model yet [04:11:41] <aFlag> hum, it seems like it doesn't support my video driver yet, so I was hoping an upgrade would solve my problem [04:12:14] <alanc> I don't think we changed any video drivers from nv_53 to nv_54 [04:12:34] <aFlag> hum [04:12:42] <alanc> which one do you need? [04:12:53] <aFlag> i815 [04:13:02] <alanc> that's been supported for years [04:13:03] *** chollien has joined #opensolaris [04:13:09] <aFlag> hum [04:13:20] *** rachel has joined #opensolaris [04:13:21] <alanc> nv_54 did add i945 to the existing driver [04:13:22] * dmick says "hmmmmmmmmmmm" [04:13:33] <alanc> almost forgot about that change [04:13:51] <aFlag> X is not working here and I read on the net that it wasn't supported [04:14:25] <alanc> sometimes the net is wrong, or is just cached from years ago [04:14:28] <dmick> certainly should be. I nkow many happy users. [04:14:52] <aFlag> yeah, it's probably wrong [04:15:18] <alanc> i815 is even one of the better supported video families on solaris x86 - it was the first to get DRI support, second to get 3D acceleration [04:15:56] <alanc> but given the x86 graphics market is pretty much ati, intel & nvidia these days, that's about what you'ld expect [04:17:05] *** swa_mobil has joined #opensolaris [04:17:42] *** chollien has quit IRC [04:17:45] <alanc> suppose I should get used to saying x86 graphics is AMD, intel and nvidia now [04:18:06] <dmick> ATMDI [04:18:18] <aFlag> it says (EE) GARTInit: Unable to open /dev/agpgart (Resource temporarily unavailable) (EE) I810(0): AGP GART support is not available. Make sure your kernel has agpgart support or that agpgart kernel module is loaded. [04:20:44] <alanc> that should be working on i810 through i915 in build 53 [04:21:08] <alanc> not sure how to debug it when it doesn't [04:21:13] <Tpenta> alanc: heard anything on the javascript probes in firefox issue now tyhat you've done the xorg/xsun probes? [04:21:21] <aFlag> it does load a i810 driver [04:21:30] <dmick> EAGAIN is a weird error from agpgart [04:22:03] <alanc> Tpenta: john rice is meeting with them next month and said he'd ask about it [04:22:11] <Tpenta> :-D [04:22:36] <alanc> they've been busy getting TBird 2.0 ready to go into Solaris [04:22:37] <dmick> aFlag: do you perhaps have very little system memory? [04:23:19] <aFlag> hum, I have 256mb [04:23:38] <dmick> total in the system??? [04:23:49] <aFlag> of ram yes [04:24:02] <dmick> I'll bet that's the problem then [04:24:07] <dmick> I doubt we even officially support 256 [04:24:15] <richlowe> You do, I believe. [04:24:19] <dmick> but with an i815, a pile of that has to go to the graphics [04:24:24] <richlowe> it's the *very* minimum, for x86 post newboot, but I believe you support it. [04:24:26] <aFlag> I have some swap also [04:24:27] <richlowe> (as a company, not you personally :)) [04:24:38] <dmick> richlowe: I bet not with an i815 :) [04:24:44] <dmick> aFlag: try more memory [04:25:29] <alanc> i815 would be trying to steal something like 8-16Mb of that for video RAM, depending on your screen size [04:25:35] <dmick> but: what does the 'Chipset:" line in Xorg.0.log say? [04:26:10] <dmick> also, this thread seems to imply that error is not fatal: [04:26:14] <dmick> http://www.opensolaris.org/jive/thread.jspa?messageID=27895 [04:26:32] <alanc> had a system in our lab with the i945 before they fixed agpgart to work with it that wouldn't come up with a 20" (1600x1200) monitor, but would with a 19" (1280x1024), since that was the limit the BIOS allocated to VRAM [04:26:35] <aFlag> it says Chipset: "i815" [04:26:47] <jsubl2> lower resolution and/or less colors might help also [04:29:43] <alanc> signs your computer is lying to you: "[monitor.power.critical:EMERGENCY]: All power supplies failed" [04:30:12] <dmick> hehehe [04:30:17] <alanc> amazing how it can write the log message to the disk with no power [04:30:28] <dmick> "The AE-135 unit is malfunctioning, Dave" [04:31:27] *** fedorared has joined #opensolaris [04:32:00] <dmick> gah. *AE-35. I suck. :) [04:39:10] <dmick> lol hey jbalint: [04:39:21] <dmick> check out Rod's latest blog entry: [04:39:26] <dmick> http://blogs.sun.com/rie/entry/init_fini_not_found_use [04:41:03] <jbk> alanc: sounds similar to my server with no network connections that is pingable :) [04:43:02] <richlowe> Hm. [04:43:16] <richlowe> stuff line-wrapping Subject: at 80 columns is getting really annoying. [04:45:17] <richlowe> (or at the very least thunderbirds insisting on representing them with the n/l character thingy is) [04:54:36] <dmick> Tbird 1.5.0.5 doesn't do that [04:54:38] <dmick> I get ... [04:54:57] <dmick> maybe the subject actually *has* nl in it? [05:02:40] *** hile_ has quit IRC [05:06:21] *** laca has quit IRC [05:07:58] <richlowe> It does, that's the irritant. [05:08:38] <richlowe> given it's displayed on one line regardless, tbird should s/\n/ /, rather than use a one char representation of 'n/l' (letters, not literal). [05:09:01] <dmick> so you're objecting to the extra char representation? [05:09:08] <dmick> I guess I can see that [05:09:20] <dmick> and yes, there it is [05:09:24] <dmick> ew, in fact [05:09:35] <dmick> I have one that has two N/Ls within five chars of each other [05:09:41] <dmick> time to trundle off and look at the raw data [05:12:54] <dmick> so the subject in the mail file is [05:13:00] <dmick> Subject: Re: [tools-discuss] Re: How do I recover from mercurial onnv-gate [05:13:01] <dmick> being regenerated ? [05:13:24] <whaq> heh wth [05:13:30] <dmick> (i.e. large extra spaces between "onnv-gate" and "being", and between "being" and "regenerated" [05:13:43] <dmick> TBird replaces *both* of those with NL [05:14:01] <dmick> in the summary pane, not in the header display of the individual message [05:14:39] <richlowe> in the header of the messpage pane, I seem to get 5-ish spaces per-nl. [05:14:45] *** alanc is now known as alanc-away [05:14:48] <richlowe> the message I'm looking at has a literal \n\t in the subject. [05:14:58] <dmick> IIRC only one space of indent is required [05:15:12] <dmick> but some things do \t instead [05:15:23] <dmick> hm. which RFC was that again.... [05:15:46] <richlowe> now, why I end up with a newline in the other message, which has nothing but single-spacing in a one line subject, I have no idea. [05:16:40] <dmick> RFC821, but it was obsoleted by something [05:17:25] <dmick> 2821 I guess [05:19:43] <dmick> no, not [2]821, 2822 [05:20:23] <dmick> right. any whitespace char is fine. [05:20:52] <dmick> and the standard explicitly says that it should be removed before any further syntactic and semantic evaluation. I would assume that includes "display". [05:21:19] <richlowe> https://bugzilla.mozilla.org/show_bug.cgi?id=271312 [05:21:22] <richlowe> looks to be that. [05:22:24] <dmick> the bug isn't correct tho; the ws should not be removed, only the cr/nl [05:23:32] *** ddunham has joined #opensolaris [05:24:32] <richlowe> comment #1 seems to be on the right lines though. [05:25:50] <dmick> ya [05:25:54] <onbot> commit by Sebastien Roy: 6504039 mactype plugin unloaded when still in use and causes panic; 6505361 i_mac_impl_count race condition in mac_unregister() and mac_register(); 6505570 manually loading mactype plugin using modload makes it unloadable [05:26:35] <richlowe> look at the bug it's dependent on, it seems the summary pane removes the tab and leaves the NL, and the message pane does the reverse. [05:26:40] <richlowe> wonder who thought that was a good idea. [05:27:22] <dmick> two different people misreading 2822 in the opposite ways. neet. [05:33:21] *** jsubl2 has quit IRC [06:00:46] *** LurkinLurker has quit IRC [06:02:02] *** LurkinLurker has joined #opensolaris [06:03:53] *** aliquis has joined #opensolaris [06:05:30] <aliquis> What is Gnome/JDS doing after upgrade of Solaris when it says "Completing post install setup" or something like that? I've left it for a few minutes but it doesn't go on, does it just take long time or do it fail with whatever it does? I removed .gnome* and .gtk* but that didn't help and speeded it up, something else which I can remove to make it start faster? [06:06:47] <richlowe> that won't help. [06:06:58] <richlowe> it's doing a bunch of stuff it used to do during install, but they moved to after install to speed up install. [06:07:36] <richlowe> easy way to make it seem to go faster is to go get some coffee while it runs. :) [06:08:47] <aliquis> Ok, aslong as it will finish I'm happy, I hope I didn't made it fail or something by pressing ctrl-alt-backspace while it was doing it =P [06:09:30] *** Fetch_ is now known as Fetch [06:09:41] <aliquis> Atleast b54 boots now, jmcp gave me urls to the bug (of something crashing on socket 754 amds from build 51 and up or whatever) and later on one of the tabs I saw it just happened when you used three memory modules so with one removed it works fine =P [06:09:57] <aFlag> oh my god, not crtl-alt-backspace! Now you'll never be able to even find a girlfriend anymore :( [06:09:57] <aliquis> I'll switch window manager then, thanks [06:10:05] <aFlag> it's like a curse [06:10:16] <aliquis> aFlag: That's no problem, I couldn't before either. [06:10:21] <aFlag> hehe [06:10:43] <aliquis> Say "Yes" for more information, "No" to cancel further presentation ;) [06:10:48] *** freakazoid0223 has quit IRC [06:11:59] <richlowe> zapping X is fine. [06:12:15] <richlowe> I tend to log into a failsafe terminal, start another xterm with prstat in it, and slack until postrun stops. [06:12:23] <richlowe> (postrun being the thing causing the annoying wait on JDS login) [06:12:57] <aFlag> A friend of mine used to say that the number of girls you can get is inversely proportional to the number of times you compile the kernel. He always made me press enter for compiling the kernel during our OS class :P [06:13:31] <fedorared> yay, updated to b54 [06:13:57] *** astinus has quit IRC [06:17:41] *** freakazoid0223 has joined #opensolaris [06:18:31] *** whaq has quit IRC [06:19:41] <aliquis> aFlag: One rule I've seen is that your own "look score" from 1-10 + how much you try from 1-10 together and then divided by four tell you how good looking girls you will usually get and divided by 2 how good looking the best ones will be or something ;) [06:20:07] <aliquis> so looks 4 + try 1 = 5 / 4 = girl with look 1.25 usually, 2.5 at best ;) [06:20:34] <dmick> as though a proper nerd would be interested in evaluating mates by looks alone, anyway :) [06:20:51] <aFlag> I'm not good looking, but I try real hard :) [06:21:04] <aliquis> I'm to affraid to try ;D [06:22:36] <aFlag> I was affraid before, but I made myself go for it anyways, and now I'm not affraid at all :) [06:23:20] *** slowhog has joined #opensolaris [06:23:25] <aliquis> Also I read in some magazine a suggestion to make your choices good and chances for, uhm, what is it called when you meet someone else in english? Anyway the solution was to meet say 12 girls and rate them all, and then take the first one you find which is better than the best one of them. That would make pretty sure your selection is rather good and that you won't feel an urge to look for another one later, of cour [06:23:25] <aliquis> se it's even better if you consider say 500 girls but that will take longer time (more than life?), while testing say 0 or 2 will make your choice rather crappy and you'll end up with something you aren't happy with ;), 12 was a number they thought was reasonable good ;) [06:24:55] *** fedorared has quit IRC [06:25:21] <aFlag> i'm not sure how is it called in english, a date maybe? I'm not a native english speaker. [06:26:21] <aFlag> For me the best way to go is just to try to get any girl that you see on a night club or a party. When you try enough you're bound to find a girl that's crazy enough to go out with you [06:26:58] <aliquis> No, not date, when you are.. xxx by meeting someone else than your current/real/.. partner/wife [06:27:12] <aliquis> cheating [06:27:13] <aliquis> =P [06:27:23] <aFlag> oh, i see :) [06:27:46] <aFlag> Knuth wrote a book on stuff like that :P [06:27:54] <aFlag> It's probably a funny reading :) [06:27:59] <aFlag> I've read parts of it [06:28:09] <aliquis> Yeah, a swedish book character/cartoon figure tried that "law of big numbers" [06:28:16] <aliquis> he only had to date 9 million girls or whatever it was [06:28:19] <aliquis> ;) [06:28:43] <aliquis> Made lots of notes and had his friends leave them everywhere to raise the number count [06:29:17] <aFlag> hehe that's cool :) [06:30:10] <aFlag> if you try 9 million girls, it doesn't matter how ugly or weird you are, one of them will like you [06:32:05] <aliquis> Well if I go out I don't feel an urge to try since I don't know people anyway and it's easier to just don't try, also you have so small amount of time and it's such a noisy environment so that sucks. Party would be better since the environment is better and time isn't such an issue but to begin with I'm to old so there aren't many parties longer. Just meet someone at any time at all would probably be better but I do [06:32:06] <aliquis> n't study longer or have got a work so that doesn't work that well either, I don't meet many people at all. I know one I could ask if she would have lunch with me or whatever but I don't know how good the chances are and if she says no I don't know how that will affect other stuff and even if she says yes and eat with me I wouldn't know or understand the actual purpose and intentions she has by doing so =P [06:32:06] *** matt__ has joined #opensolaris [06:32:40] <aliquis> aFlag: Atleast one can try the well known "Will you have sex with me if I can get 5 million hits on my website?"-trick, but so many have already used it ;) [06:33:36] <aliquis> brb [06:33:39] *** aliquis has quit IRC [06:36:06] *** matt__ has quit IRC [06:36:10] *** whaq has joined #opensolaris [06:36:31] *** matt__ has joined #opensolaris [06:37:50] *** matt__ has quit IRC [06:38:09] *** freakazoid0224 has joined #opensolaris [06:38:34] *** freakazoid0223 has quit IRC [06:39:04] *** freakazoid0224 has quit IRC [06:39:19] *** freakazoid0223 has joined #opensolaris [06:43:29] *** halton has joined #opensolaris [06:45:36] *** lloy0076 has joined #opensolaris [06:45:40] * lloy0076 strange [06:46:04] *** aliquis has joined #opensolaris [06:46:07] <lloy0076> Ah - I have an older version of gaim. [06:47:11] *** lloy0076 has left #opensolaris [06:49:32] *** aliquis has quit IRC [06:49:53] *** aliquis has joined #opensolaris [06:53:31] *** ShadowHntr has quit IRC [07:02:40] *** dmick has left #opensolaris [07:05:11] <aliquis> A pre-police man here in Sweden which worked with underground crime related activites have quit his job and now work for a sex club in Stockholm with connections to similair people, and they say he knows who work undercover and such. Good idea to switch work, I wonder what his new salary is ;) [07:08:25] <onbot> commit by ae112802: 6493672 eeprom binary group id broken on sun4v platform [07:08:34] <aFlag> hehe I think he just like to wear a police uniform :P [07:11:44] <aliquis> What? He used to be a police, but he have switched to the other side ;) [07:11:47] *** bougie has joined #opensolaris [07:12:03] <aliquis> my post installation stuff from JDS still haven't finished, how can I tell if it do anything at all? [07:13:05] *** mercykiller has joined #opensolaris [07:13:13] *** Posixzombie has joined #opensolaris [07:13:17] <aFlag> well, I was making de suposition that now he wears a policeman costume to please his clients [07:14:57] <aliquis> ah, in the sex club you mean? [07:15:00] <aliquis> yeah maybe [07:15:55] <aFlag> I don't know about the jds, though [07:16:10] <aFlag> I haven't made much progress with opensolaris so far hehe [07:20:37] <aliquis> I just wonder if it does anything at all except showing the progress bar [07:24:39] *** ddunham has left #opensolaris [07:26:56] <bougie> hello :) [07:28:22] <aliquis> hello hey [07:30:09] *** MaGre has joined #opensolaris [07:38:02] *** halton has quit IRC [07:45:18] <Posixzombie> Does anybidy know: is there "bonding" in solaris of Nics (like there is in linux); what it does is give 2 nics the same IP (and they behave as "master-slave") [07:45:58] <syphilis> yes, look at dladm [07:47:44] <Posixzombie> syphilis, thnks ; man dladm and grepping for bonding yields no results but I assume it is add-aggr [07:48:27] *** lloy0076 has joined #opensolaris [07:49:58] <aliquis> My "Completing post install setup..." from gnome-session still haven't finished.. [07:51:16] *** lloy0076 has quit IRC [07:53:15] *** Drone has quit IRC [07:55:55] *** Drone has joined #opensolaris [07:56:21] *** halton has joined #opensolaris [08:12:13] *** BadKarma has left #OpenSolaris [08:12:13] *** BadKarma has joined #OpenSolaris [08:15:24] <onbot> commit by johansen-osdev: 4894692 caching data in heap inflates crash dump; 6499454 time to increase size of kmem default allocation caches; 6499459 vm should stop checking kvp directly [08:15:53] *** Tpenta has quit IRC [08:17:32] *** Tpenta has joined #opensolaris [08:20:00] *** Posixzombie has quit IRC [08:23:06] *** aliquis has quit IRC [08:23:53] *** ChanServ sets mode: +o Tpenta [08:32:01] *** anthony79 has joined #opensolaris [08:34:24] *** aliquis has joined #opensolaris [08:40:47] *** LordKing has joined #opensolaris [08:41:11] *** aliquis has quit IRC [08:41:43] *** aliquis has joined #opensolaris [08:45:00] <aliquis> Uhm, I ran postrun with svcadm instead and after just two minutes or so it finished, so now I'm in JDS but windowskey-r for run dialog didn't worked and now when I minimized windows I noticed they don't end up in the taskbar. I could add a window selector applet which works but is that how it's supposed to work now? It's ok thought since it look rather clean without all the window names down there I suppose. [08:47:41] *** swa_mobil has quit IRC [08:48:52] *** gustav3d has joined #opensolaris [08:52:07] *** benr has joined #opensolaris [08:52:08] *** ChanServ sets mode: +o benr [08:52:52] <benr> Tpenta, ping [08:53:23] <Tpenta> benr: ACK [08:55:08] <benr> hey Alan. ;) [08:55:18] <Tpenta> how you doing ben? [08:55:40] <benr> question for you: is it possible from a crash dump to determine which process was running on CPU that caught the panic? [08:56:09] <Tpenta> yes [08:56:10] <richlowe> ::cpuinfo [08:56:25] <Tpenta> echo ::cpuinfo -v | mdb -k unix.0 vmcore.0 [08:57:03] <benr> and is it also possible to determine what zone that process was running on? [08:57:44] <richlowe> <procaddr>::print proc_t p_zone->zone_name [08:58:12] <Doc> thread sunnary|grep PANIC [08:58:16] <Doc> summary even [08:58:28] <benr> sweet, thanks. [08:58:41] <Doc> (of course, you'll need to download SCAT first - plus i dont think it actually puts the PANIC on the same line as the actual process) [08:58:57] <richlowe> and it's possibly (likely?) that Ben is running on x86, and can't download a scat that would work. [08:59:00] <Tpenta> and there's also the fact that scat 4.2 will not do solaris 10 or newer and it only does sparc [08:59:04] <richlowe> s/possibly/possible/ [09:00:21] <benr> Yes, this is Nevada B43 on X86. [09:02:28] <aliquis> Anyone running SNV54 which can tell if minimized windows ends up in the taskbar? [09:02:33] <aliquis> in jds that is [09:02:42] <anthony79> benr: in mdb, ::ps -z will show which zone each process belongs to [09:03:16] <benr> ah, sweet. [09:03:21] <benr> didn't know about -z [09:04:21] <benr> In output like this: [09:04:23] <benr> > ::ps -z [09:04:24] <benr> S PID PPID PGID SID ZONE UID FLAGS ADDR NAME [09:04:24] <benr> R 0 0 0 0 0 0 0x00000001 fffffffffbc25900 sched [09:04:24] <benr> R 3 0 0 0 0 0 0x00020001 ffffffff9970d928 fsflush [09:04:24] <benr> R 2 0 0 0 0 0 0x00020001 ffffffff9970e558 pageout [09:04:33] <benr> The "S" column is state right? [09:04:43] <Tpenta> yes [09:04:45] <benr> Why does everything show as "R"? I assume thats "Runing" [09:05:02] <benr> in the dumps every process is showing as R [09:05:21] <Tpenta> I think that's runnable, not running [09:05:47] <Tpenta> runnable as in on a dispatch queue waiting for cpu [09:06:09] <richlowe> I have doubts it's runnable. [09:06:44] <benr> there are several hundred proccesses showing, everything that would normally be up..... I can't believe that all of them were in the run queue at the time of panic. [09:06:49] <richlowe> Hm, or something is screwy. [09:08:16] <richlowe> on here *everything* shows as runnable, when it seems reasonable that at the very least one should be onproc. [09:08:20] <richlowe> though this is live. [09:09:54] *** mnowak_ has quit IRC [09:10:06] *** mnowak_ has joined #opensolaris [09:10:54] <benr> hmmm. [09:10:56] * benr digs. [09:11:03] <benr> Thanks Alan and Rich. :) [09:11:14] <benr> and Anthony too. ;) [09:11:26] <richlowe> you'd get a (seemingly) better indication of that from ::thread [09:11:45] <benr> ahhhhhhhhhhh, much better. [09:14:26] <richlowe> heh, because I confused process states and thread states. [09:14:33] <benr> but then, how do I determine what process each thread belongs too? [09:16:45] <richlowe> ::ps -zt seems like it would do everything you want. [09:16:50] *** mazon is now known as Mazon [09:17:58] *** gustav3d has quit IRC [09:19:05] <benr> perfect. [09:19:06] <benr> thanks. [09:23:13] *** lloy0076 has joined #opensolaris [09:24:28] <lloy0076> I suppose I should open the wrapping. [09:28:21] *** salamanders has joined #opensolaris [09:31:05] *** LurkinLurker has quit IRC [09:33:57] *** dunc has quit IRC [09:34:00] *** simfordWFH has joined #OpenSolaris [09:34:42] *** LurkinLurker has joined #opensolaris [09:34:53] *** tsoome has quit IRC [09:36:01] <cmihai> aliquis: what? [09:36:08] <cmihai> aliquis: about minimised windows? [09:39:57] *** slowhog has left #opensolaris [09:40:08] *** simford has quit IRC [09:40:48] *** sniffy has quit IRC [09:45:48] *** giant| has quit IRC [09:45:55] <aliquis> cmihai: Just when I minimize them in JDS they don't turn up in the task bar (start menu) [09:46:39] <ofu> prtdiag -v is still broken on x4600 with sol10u3 [09:46:51] <Doc> ofu: how so? [09:47:04] <Doc> worked fine for me a few hours ago on s10u2 [09:47:04] <cmihai> aliquis: odd, did you try to restart / recrate them? The bars.. [09:47:09] <ofu> the memory sockets and pci-slots are wrong [09:47:50] <Tpenta> i believe that was fixed in b56 today [09:48:17] <ofu> http://www.fuckner.net/~ofu/prtdiag-x4600-sol10u3 [09:48:31] <Tpenta> nah, that was prtconf -v [09:48:39] <Tpenta> 6504229 prtconf -v core dumps with snv_54 [09:48:40] <ofu> i only use on pci-x slot and memory on all 4 cpus [09:49:15] <richlowe> http://bugs.opensolaris.org/view_bug.do?bug_id=6430609 [09:49:18] * richlowe mutters [09:49:59] <ofu> ah, thx [09:51:22] <richlowe> that wasn't an answer to you. [09:51:23] <Doc> that's a pretty seriously fucked up CR [09:51:30] <richlowe> it was the stupid bloody comment. [09:51:34] <richlowe> (in the CR) [09:51:44] <Doc> well, it is for ofu too [09:52:02] <Doc> oh well.. i just installed windows on my X4600 anyway [09:52:06] <richlowe> if it *is*, I just got lucky, since all I can see is the wittisism. [09:52:13] <Doc> lets see how fast autopano runs on a 16 core machine [09:52:22] [09:52:34] <Doc> i've looked at the whole CR, and it does appear to be it [09:53:11] *** dlynes_laptop has joined #opensolaris [09:53:39] <Doc> thhe short version of the workaround is to use ipmitool [09:54:37] <Doc> the long version.. well.. believe me, you dont wanna know [09:54:52] *** tsoome has joined #opensolaris [09:57:10] *** Cyrille has joined #opensolaris [09:58:30] *** MaGre has quit IRC [10:00:37] *** _357181_ is now known as estibi [10:02:11] <richlowe> benr: hg log -pk 6459866 [10:04:56] *** dunc has joined #opensolaris [10:06:08] *** bnitz has quit IRC [10:10:03] *** alanc_away has quit IRC [10:10:52] *** alanc has joined #opensolaris [10:11:38] <benr> richlowe, ah, ty! :) [10:12:13] *** udos has quit IRC [10:14:10] <benr> richlowe, I didn't realize that older commits were in hg as well. [10:14:34] <richlowe> everything since launch. [10:16:16] <benr> sweet. [10:16:43] <benr> richlowe, do you know how I'd go about pulling an old build (43)? [10:17:07] <benr> I was going to look for the old tarballs, but I'd love to know how to do it in hg. [10:17:16] <benr> not sure if each build is tagged. [10:17:53] <richlowe> http://dlc.sun.com/osol/on/downloads/hg-build-snapshots/ [10:18:28] <richlowe> though if you already have a tree locally, you can hg clone -r ... <that tree>, rather than pull it from hg.opensolaris.org again. [10:19:01] <richlowe> but 43 doesn't look to have changed, so just cloning it at onnv_43 should work. [10:19:03] <benr> sweet, thanks. [10:38:50] *** uebayasi has quit IRC [10:40:29] *** MattMan has joined #opensolaris [10:49:51] <raph_ael> hello [10:54:07] *** dunc has quit IRC [10:54:15] *** alobbs has joined #opensolaris [10:54:46] *** Gman has joined #opensolaris [10:54:58] <Tpenta> Media Training Questionnaire I need to think about this one a little more. hi glynn [10:55:05] <Tpenta> hi glyn [10:55:09] <Gman> hi Tpenta [10:55:13] <Tpenta> oops dropped the mouse and hit the paste buffer [10:55:21] <asyd> questionnaire is an english word too ? [10:55:39] <Tpenta> yes [10:55:48] <Doc> i think what tpenta needs is mouse training rather than media training :) [10:55:55] <Tpenta> :) [10:56:04] <Doc> tpenta: we pulled rorys office down [10:56:09] <Tpenta> glynn, heard any more on that python/gnome coredump issue? [10:56:13] <Tpenta> awwwwww [10:56:21] <Tpenta> i hope you have photos [10:56:23] <Doc> he was devistated [10:56:42] <Gman> Tpenta, nope, haven't really investigated it to be honest [10:56:45] <Doc> he went out for a meeting, and came back half an hour later to no office [10:57:12] <Tpenta> why? [10:58:03] <Doc> why not? [10:58:46] <Tpenta> it looks like it has been fixed in the root tree but we dont have it yet [10:58:48] *** aliquis has quit IRC [11:00:31] <Tpenta> I started looking at pulling the individual bits over to test but it started looking very much like I was going to have to build the entirity of the gnome support [11:01:16] <lloy0076> Anyone know if distcc works with Sun Studio compilers...?? [11:04:30] *** timelyx has quit IRC [11:04:41] *** dunc has joined #opensolaris [11:04:56] *** phus has quit IRC [11:05:05] *** phus has joined #opensolaris [11:06:03] *** benr has quit IRC [11:06:26] *** timelyx has joined #opensolaris [11:06:29] *** simfordWFH has quit IRC [11:07:42] *** lloy0076 has quit IRC [11:08:08] *** anthony79 has quit IRC [11:28:58] *** Snake007uk has joined #opensolaris [11:30:03] *** Gr|ffous has left #opensolaris [11:30:21] *** Gr|ffous has joined #opensolaris [11:33:51] <tsoome> hm, upgrade to b53 suggests there is not enough space (19GB free), some kind of known bug? [11:36:40] *** halton has left #opensolaris [11:44:55] <jteo> re. [11:45:21] <timeless> heh [11:48:01] *** bobbyz is now known as bobbyZzz [11:48:14] *** yongsun has quit IRC [11:48:18] <Stric> I installed b53 on an 8G partition (18G disk) with half to spare when it was done.. [11:49:00] <timeless> i always install into a 4g ufs / partition [11:49:07] * timeless is running 53 [11:49:43] <tsoome> well, I have just found b54 is available, so, it will takse some time now... [11:50:57] <tsoome> upgrade is from snv_50 btw [11:53:01] *** aFlag has left #opensolaris [12:10:12] *** Cyrille has quit IRC [12:10:31] *** Cyrille has joined #opensolaris [12:18:17] *** Tpenta has quit IRC [12:20:48] <Gman> woo, for website bugs :) [12:21:06] <jteo> hmm? [12:21:55] <cmihai> http://img501.imageshack.us/img501/8871/biggunsunrw9.jpg - svn_54 TwinView ;] [12:22:24] <Gman> jteo, because if the website was OS'd, i'd fix them myself ;) [12:29:01] *** MattMan is now known as MattAFC [12:35:36] *** Posixzombie has joined #opensolaris [12:48:03] <Auralis> nice backdrop [12:49:18] *** MattAFC is now known as MattMan [12:55:26] <onbot> commit by danmcd: 6504195 IPv6 doesn't send up self-generated ICMP6_PACKET_TOO_BIG to tunnels.; 6504199 tun does a lot of freeb(ipsec_mp) where ipsec_mp is 0xdeadbeef. [13:01:03] *** MattMan has quit IRC [13:01:11] *** Dar has quit IRC [13:01:17] *** Gman has quit IRC [13:07:00] *** kimc has joined #opensolaris [13:08:40] *** dunc_ has joined #opensolaris [13:09:25] *** Posixzombie has quit IRC [13:09:49] *** adam_z has joined #opensolaris [13:10:18] <kimc> good morning [13:12:07] <adam_z> morning [13:13:38] <kimc> i'd like to create a partition out of unused disk space, then create some slices in that ..is this possible ? [13:13:56] <Doc> yes, but you need to use Solaris Volume Manager (SVM) and soft partitions [13:14:08] *** sparc-kly has joined #opensolaris [13:14:35] <kimc> ah yes, i was reading about soft partitions and using those to contain zones.. [13:14:55] <kimc> root zones.. [13:15:19] <kimc> so you can't have more than one Solaris partition ? [13:19:05] *** dunc has quit IRC [13:38:22] *** vmhobbes has joined #OpenSolaris [13:38:24] *** gisburn has quit IRC [13:39:17] *** Posixzombie has joined #opensolaris [13:45:09] <tsoome> kimc nope [13:51:40] <kimc> ok so the only way to go is to figure out how to create soft partitions with SVM.. but i don't see my newly created disk slice [13:53:11] <kimc> i assume you can use SVM to create soft partitions out of unsed disk space, right ? [13:53:17] *** hile_ has joined #opensolaris [13:59:25] <kimc> i'm going to reinstall sxcr for a fresh start.. should i create a Solaris partition at this point and an leave the other 4 partitions Unused ? [13:59:39] <kimc> other 3 partitions that is [14:00:17] <kimc> then use SVM to create soft partitions in the unused space ? [14:00:23] <phus> yes [14:00:32] <phus> no actually [14:00:41] <phus> you create metadevices on the free partitions [14:00:57] <phus> s3/s4 you need about 10mb partitions [14:01:14] <phus> then once you have svm configured you can create many soft partitions and mount them [14:01:59] <kimc> do you use SVM to create the soft partitions within the 10 mb partitions you referred to ? [14:02:11] <kimc> what is s3/s4 ? [14:02:53] <hile_> i always used s7 for mddb replicas [14:03:04] <hile_> s3/4 for pub/priv regions on VRTS discs [14:06:47] <kimc> so.. since i'm reinstalling at this point should i leave the other 3 partitions unused ? [14:06:48] *** bengt_ has joined #opensolaris [14:06:57] *** bengt_ is now known as bengtf [14:10:34] *** Vanuatoo has quit IRC [14:10:39] *** Vanuato1 has joined #opensolaris [14:12:35] <kimc> ok reinstalling ..after a ehh.. SVM mishap :( [14:12:37] <PerterB> you can't use soft partitions (shouldn't they be called soft slices? ;) on your root partition because they cause the partition to get re-sliced so s7 contains state replicas and s0 contains the rest of the space [14:12:49] <PerterB> and I didn't think you could have multiple Solaris partitions on the same disk [14:13:17] <kimc> yes right only 1 Solaris partition [14:13:50] <PerterB> ok, so you can't use soft partitions on your system disks then [14:14:53] <kimc> no soft partitions on the unused space on the system disk ? [14:15:28] <PerterB> I can't see how (because of the re-slicing I mentioned above) [14:15:28] *** jamesd has quit IRC [14:15:35] <Stric> I think the discussion has a great name confusion wrt partitions/slices [14:15:42] <PerterB> yes [14:15:45] *** jamesd has joined #opensolaris [14:15:45] *** ChanServ sets mode: +o jamesd [14:16:10] <Stric> you can have / on s0 and make s1 a "host" for a gazillion soft partitions on the same physical disk [14:18:01] <PerterB> oh, actually my bad, I misread the manpage and the last time I used soft partitions was using the dedicated disk method [14:19:37] <kimc> Ah Ok.. so can SVM create the s1 or fdisk ? [14:21:08] <PerterB> no fdisk is still different... I was confusing what happens when you use the -e option creating soft partitions (scroll down to Soft Partition Options in the metainit manpage), sorry [14:22:37] <PerterB> so fdisk creates partitions on the disk, format/fmthard create slices within the Solaris partition on a disk and soft partitions are divisions within a single slice [14:23:42] <hile_> s/within a single slice/within a metadevice/ [14:24:03] <kimc> so in that case it sounds like i need to make the entire disk Solaris since you can have only 1 Solaris partition [14:24:34] <PerterB> hile_: yes :) [14:27:46] *** laca has joined #opensolaris [14:28:38] <kimc> the whole disk must be Solaris right ? [14:29:05] <PerterB> if you want to use the space from Solaris, yes [14:29:28] <kimc> yes i want to use the space for Solaris zones [14:29:53] <bengtf> when will ati be supported ? [14:30:38] <cmihai> What do you mean? [14:31:14] <cmihai> Go ask ATI when they feel like releasing a Solaris driver? But apart from that, Ati cards should work with Xorg drivers. [14:35:18] <PerterB> Oh bollocks... I just twigged what my "disk appearing not to be a disk" weird error was (http://pastebin.ca/285027)... [14:35:21] <bengtf> cmihai: well only in standard 1024 resolution, would like to have as nvidia full resolution and opengl ... [14:35:49] <PerterB> there happened to be an old entry in vfstab which had a block device in the raw device column by mistake, and metainit used that [14:36:20] <bengtf> x1600 isnt supported with xorg either [14:41:36] <cmihai> bengtf: go ask ATI [14:41:42] <cmihai> It's really up to them. [14:41:59] <cmihai> Or help make better open source drivers. There's really no way around it. [14:42:10] <bengtf> hmm, linux has it so opensource or something already ... [14:42:17] <cmihai> No [14:42:24] <cmihai> Linux has propriotary binary drivers from Ati. [14:42:27] <sickness> http://flickr.com/photos/64856052@N00/323290914/ <- lol [14:42:37] <cmihai> Like Solaris has for NVIDIA. [14:45:37] <PerterB> sickness: cute [14:46:07] *** LurkinLurker has quit IRC [14:47:47] *** LurkinLurker has joined #opensolaris [14:48:59] *** mercykiller has quit IRC [14:49:00] <bengtf> next project replace ati in portable with nvidia then ... ;( [14:50:02] <LurkinLurker> FYI, the chanserv message for #opensolaris lists a variety of URLS, one of them for the genunix wiki. Clicking on it (in xchat, at least) sends you to http://www.genunix.org/wiki, ...not http://www.genunix.org/wiki as it should. [14:53:36] *** eboutilier has joined #opensolaris [14:56:38] *** gm152 has joined #opensolaris [14:59:13] *** bunker_work has joined #opensolaris [15:00:28] *** adam_z is now known as adam_bin [15:01:44] <tsoome> not nice at all. upgrade from b50 to b54 does not work as well. [15:02:20] <tsoome> disk space reallocation is required, but there is 16GB of free space.... [15:05:20] *** dunc_ has quit IRC [15:08:04] *** laca is now known as lacaAFK [15:09:38] <kimc> SMC seems to have some problems on b54 [15:10:36] <kimc> like showing -26.71% free when there is actually df -k [15:11:04] <tsoome> SMC has always had this kind of issues [15:11:07] <kimc> about 11 gigs free [15:11:36] <tsoome> in SA trainings I always have to explain this... [15:11:57] <kimc> ahh very good.. [15:12:22] <tsoome> use format instead;) [15:12:38] <tsoome> and never both at the same time [15:12:45] <kimc> then if i double-click c1d0 it opens a box which says something like: Warning: No Solaris Fdisk Partition [15:13:38] *** LordKing has quit IRC [15:13:58] *** adam_bin has left #opensolaris [15:14:05] <tsoome> hm, it looks like pfinstall is broken and prevents my upgrade:( [15:16:53] *** dunc has joined #opensolaris [15:18:04] *** LordKing has joined #opensolaris [15:26:41] *** adam_bin has joined #opensolaris [15:31:02] <hile_> using smc is likely your first mistake [15:32:51] *** sniffy has joined #opensolaris [15:52:26] <onbot> commit by kp158701: 6486273 bad mutex panic within find_tpc() [15:52:39] *** lacaAFK is now known as laca [15:53:10] * Stric just got a panic from a snv_53 by setting zfs_vdev_cache_bshift/W 0t14 [15:54:53] *** Mazon is now known as mazon [15:57:12] <Stric> looking at the source, it seems like "don't do this on active filesystems" [15:58:40] <kimc> hile_: Ok on SMC being the first mistake [15:59:12] *** bougie has quit IRC [16:00:29] *** salamanders has quit IRC [16:01:41] *** bunker has joined #opensolaris [16:01:58] *** aliquis has joined #opensolaris [16:02:06] *** bougie has joined #opensolaris [16:02:27] <aliquis> Are there still any problems with ZFS on x86 (not AMD64) or have whatever it was earlier been fixed? What was/is the problem? [16:03:08] <Stric> virtual memory range problems, I think [16:03:13] <sickness> only "problem" is that the maximum memory usage is limited to 512mb of ram [16:03:34] <sickness> whichever is the amount of ram that you have installed [16:03:51] <sickness> so if you have 2gb, for example, zfs will still use at most 512 of it [16:03:59] <tsoome> okay, I got why my upgrade is failing [16:04:02] <aliquis> k [16:04:36] <sickness> I, for example, use zfs on a 64bit machine but with only 512mb ram, so I don't really think that problem is a so big showstopper at least out of production use on "big iron" ... :P [16:05:08] <aliquis> sickness: And you are sure that is the only problem? In that case I guess I don't see it as a problem either since I don't have that much memory to use anyway. [16:05:10] <tsoome> or then again... [16:05:11] <sickness> I even managed to run all the dclarke's stress tests, withount a glitch [16:05:22] <sickness> aliquis: yeah, that's the point :) [16:05:29] *** eboutilier has quit IRC [16:05:38] <Stric> sickness: I've had big trouble with running a zfs/nfs server on only 1GB.. got hold of pretty cheap blade1k memory so now it has 2GB and is _much_ happier [16:05:49] *** bunker_work has quit IRC [16:05:57] <sickness> aliquis: if you don't have more than 512mb of ram, you shouldn't see the difference anyway... because you really don't want to swap so much on disk, I hope... [16:05:58] <Stric> too aggressive caching being the problem [16:06:29] <tsoome> Stric: memory for file server is always a good thing - to let it cache data and reduce a physical io [16:06:42] <sickness> Stric: which release? I have this zfs pr0n fileserver at home with only 512mb ram, I use it since snv47 and I'm so happy, 0 problems (nfs exported of course) [16:07:03] <Stric> tsoome: yeah, but there's a difference between "more memory giving higher performance" and "too little memory giving close to no performance at all" [16:07:41] <Stric> tsoome: I got into the situation where it constantly read like 60-70MB/s from disk, and threw away 90-95% of it, sending the rest over nfs [16:07:48] <sickness> bbl [16:07:59] <Stric> sickness: u2 up to snv53 [16:08:36] <Stric> tsoome: zfs has two layers of prefetching.. and after it had prefetched, it got into memory pressure and threw all the cache away [16:08:44] <Stric> goto 10 [16:09:10] <aliquis> sickness: Will it swap a lot with 512 or 768MB ram if i don't use it that much then? [16:09:12] <Stric> (this is an nfs backend for cdimage.debian.org, (one of) releases.mozilla.org, ftp.gnome.org etc) [16:09:28] <aliquis> in my case "file server" = machine in my closet [16:11:48] <aliquis> Stric: The reason I ask is because I don't know if some old machine (athlon thunderbird 1GHz or so) will work or if I should get the cheapest socket 939 + athlon64 cpu I can get and use that instead. [16:12:58] <jamesd> my blade 1500 with 512MB of ram often is non-responsive after big file moves because it swaps out all programs in the process. [16:13:44] <Stric> aliquis: my usage case is quite different from yours.. I'm getting into trouble with the prefetch when having for instance 50:ish parallel reads of various dvd images [16:13:47] <Stric> over nfs [16:14:03] * cmihai wonders who's idea was it to replace Alt F2 as as a default shortcut for run in JDS with Winkey R [16:15:11] <aliquis> It's winkey + r in windows, therefor it's the standard. [16:16:57] * andersmo likes using the windows key for window manager shortcuts. No applications want to use it -> no conflicts. [16:18:08] <jteo> ARC is...broken. [16:26:53] <cmihai> aliquis: it's not standard [16:26:59] <cmihai> My keyboard doesn't have a win key [16:27:07] <cmihai> It does have Meta keys though :P [16:29:01] *** alobbs has quit IRC [16:32:55] *** wraith has joined #opensolaris [16:33:36] <wraith> where I can read about zones when upgrading opensolaris? [16:33:46] <tsoome> docs.sun.com [16:34:17] <whaq> pl [16:34:26] <whaq> moin ppl [16:38:52] *** kimc has quit IRC [16:41:19] *** adam_bin has quit IRC [16:56:41] *** carbon60 has joined #opensolaris [16:57:53] *** LordKing has quit IRC [17:00:46] <tsoome> pfinstall sucks [17:03:52] <Sporq> man [17:03:57] <Sporq> sunsolve sucks. [17:04:06] <tsoome> pfinstall sucks more [17:04:49] <Sporq> dumbass site. search, find results, click results, stupid site asks you to login, then instead of returning you to the result, it dumps you back to the main search page. [17:04:59] <tsoome> it's suggesting I should have ~144GB / for b54 [17:05:19] *** laca is now known as lacaAFK [17:06:54] <Sporq> sun's really starting to annoy me. [17:06:56] *** bougie has quit IRC [17:07:58] <phus> Sporq, just get in the habbit of logging in [17:08:01] <phus> before you do anything [17:08:05] <phus> I know its really annoying [17:08:21] <Sporq> I do, sometimes though I'll return to the page after an hour or so and the session expires. [17:08:26] <Sporq> like just now. [17:08:41] <Sporq> and now that i've been watiing to login for like 3 minutes ....... [17:09:19] *** bougie has joined #opensolaris [17:10:12] <tsoome> sunsolve search is mostly just meaningless, bug descriptions are cross referencing without real information... [17:15:36] <carbon60> Any ideas on what hardware to use in order to create a Sun Cluster setup to play with? I'm looking for cheap-as-possible? [17:15:45] <carbon60> I'm talking about the storage? [17:16:15] <Sporq> you can use 64-bit amd [17:16:20] <tsoome> depends what is your target [17:16:27] <tsoome> or goal [17:16:48] <andersmo> Some people have used firewire drives for shared storage in linux/oracle clusters. [17:16:53] <tsoome> simplest storage is multi hosted JBOD SCSI [17:16:58] <andersmo> Does solaris do firewire? [17:17:02] <Sporq> yup [17:17:05] <Sporq> 10 [17:17:17] <Sporq> not sure about 9 or 8 [17:17:43] <tsoome> SC supports SCSI and FC [17:26:57] <carbon60> Multi-hosted JBOD SCSI should work, just need to find something. [17:32:26] <onbot> commit by apersson: PSARC 2006/314 Updated MIBs; 6501024 Add support for new IP/TCP/UDP MIBs [17:40:54] <sickness> aliquis: nope, I use it a lot, I have an mlnet running on it locally, plus my entire mame/uae roms remotely mounted via samba, plus my vmware images remotely mounted, no problems :) [17:42:26] <onbot> commit by alan maguire: 6425112 network/service shouldn't write to filesystems it doesn't depend on; 6499493 machine installed with snv_53 can not find default router automatically (only after sys-unconfig?); 6503806 md_monitord Makefile install target needs a dependency on SVCMETHOD [17:42:50] *** bunker has quit IRC [17:46:16] *** hile_ has quit IRC [17:48:33] <Darwin> hi [17:52:57] *** ndroux has joined #opensolaris [17:55:57] *** carbon60 has quit IRC [18:00:16] *** Torkel9 has joined #opensolaris [18:02:41] <Torkel9> Hi! What about Solaris Express 12/06? Is it a further development of Solaris 10 11/06 or is it a more tested version of OpenSolaris? [18:03:47] <jamesd> its the latest version of solaris express and has no relation to 11/06 [18:03:55] <sickness> more tested version of opensolaris [18:04:16] <alanc> Solaris Express is the releases of the development tree of what will probably become Solaris 11 some day [18:04:17] <sickness> that's what every solaris express release is... [18:04:29] <sickness> yeah :) [18:06:37] <tsoome> it's just not cool to say "it's beta" ;) [18:06:52] *** sommerfeld has joined #opensolaris [18:06:52] *** ChanServ sets mode: +o sommerfeld [18:15:58] *** fedorared has joined #opensolaris [18:18:31] *** kimc has joined #opensolaris [18:21:48] *** aiki_kid has joined #opensolaris [18:22:53] <aiki_kid> what's the cmd for listing the dev connected? [18:23:06] *** lopa has joined #opensolaris [18:25:30] *** dlynes_laptop has quit IRC [18:29:06] *** lopa has left #opensolaris [18:32:06] *** Snake007uk has quit IRC [18:39:25] *** slowhog has joined #opensolaris [18:42:13] *** jgilje_ has joined #opensolaris [18:43:06] *** jgilje has quit IRC [18:43:26] *** razrX has quit IRC [18:45:03] *** aiki_kid has quit IRC [18:46:33] *** razrX has joined #opensolaris [18:47:43] *** dunc_ has joined #opensolaris [18:50:45] *** dunc has quit IRC [18:52:40] *** pikapika has joined #opensolaris [18:53:26] <pikapika> hello [18:56:16] *** tsoome has quit IRC [18:58:54] *** stevel has joined #opensolaris [18:58:55] *** ChanServ sets mode: +o stevel [18:59:21] <richlowe> stevel: mornin'e [18:59:58] <stevel> morning... [19:00:06] * stevel is still struggling with some network issues [19:00:09] *** tsoome has joined #opensolaris [19:01:04] *** PerterB has quit IRC [19:04:11] <cmihai> Damn, who made Jaba fast? :P [19:07:23] *** PerterB has joined #opensolaris [19:11:06] *** jgilje_ has quit IRC [19:11:07] *** jgilje has joined #opensolaris [19:12:44] *** dlg has quit IRC [19:12:56] *** |tsoome| has joined #opensolaris [19:18:15] *** dunc_ has quit IRC [19:18:39] *** mazon is now known as Mazon [19:21:56] *** dlg has joined #opensolaris [19:25:29] *** tsoome has quit IRC [19:29:11] *** swa_mobil has joined #opensolaris [19:38:23] <kimc> pwd [19:38:56] <cmihai> ~ [19:40:13] <asyd> / [19:45:30] <stevel> !! [19:46:04] <dwc-> /home/stevel/porn/goats [19:46:19] <stevel> phbt [19:46:21] <stevel> as if [19:46:24] <stevel> goats is totally not-executable [19:46:31] *** bengtf has quit IRC [19:46:56] <dwc-> as you found out the first time you tried to exec it [19:48:28] <richlowe> setopt autocd [19:49:23] *** pizdec has joined #opensolaris [19:51:27] <stevel> don't use zsh ;) [19:51:40] <pizdec> you mean zfs? [19:51:41] *** dlg has quit IRC [19:51:54] <jamesd> no he means zsh [19:51:59] *** dlg has joined #opensolaris [19:52:08] <pizdec> hehe [19:52:19] <whaq> zsh vs bash? [19:52:36] <dwc-> they both fall to the might of ksh93's mighty komodo dragon army [19:52:37] <cmihai> Bash is for gentoobies :P [19:52:44] <dwc-> blah, that was redundant [19:52:51] * dwc- reaches for the caffeine [19:52:54] <whaq> haha [19:53:11] <jamesd> cmihai, don't degrade bash like that.. thats uncalled for [19:53:26] <whaq> komodoes are smelly [19:53:45] <rydis> So are linux users. What's your point? ;) [19:53:52] <jamesd> whaq almost as bad as whaq-toes [19:54:06] <cmihai> Meh, bashing bash isn't even fun anymore [19:54:29] <whaq> just an observation ;) [19:54:40] *** deather_ has joined #opensolaris [19:57:08] * dwc- wonders why jamesd was smelling whaq-toes [19:57:37] * cmihai wonders what the hell a whaq-toe is [19:57:44] <jamesd> dwc-, they radiate through out the whole room, surprised you haven't noticed. [19:57:47] <whaq> i wear socks [19:58:07] <dwc-> I never installed the irc smell-plugin [19:58:31] <dwc-> ... for a reason [19:58:36] <cmihai> IRC stinks? [19:59:56] *** alanc is now known as alanc_away [20:01:29] <whaq> heh [20:02:16] *** pizdec has quit IRC [20:02:18] *** LordKing has joined #opensolaris [20:02:30] *** pizdec has joined #opensolaris [20:11:29] *** deather has quit IRC [20:15:24] *** mega has joined #opensolaris [20:15:30] <Torkel9> Is 384 MB RAM enough for Opensolaris x86? [20:16:23] <jamesd> yes [20:16:26] <jamesd> but more is always better [20:16:34] <jamesd> 256 is the min. [20:17:26] <Torkel9> Ok, Thank You. [20:17:50] <quasi> doesn't it still need 512 for the gui installer? [20:18:13] <jamesd> not sure.. just said the min [20:18:28] <onbot> commit by dp: 6505241 webrev misses ps2pdf, but carries on like it's still there; 6506433 new webrev botches manually supplied file lists [20:22:41] <Torkel9> http://docs.sun.com/app/docs/doc/819-6394/6n8dmvnku?a=view Perhaps it is the same as Solaris 10 11/06? [20:24:39] *** estibi is now known as _357181_ [20:28:45] *** kb7sqi has quit IRC [20:35:26] *** hile_ has joined #opensolaris [20:42:15] *** hell` has joined #opensolaris [20:42:38] *** alfism has quit IRC [20:45:25] *** mega has quit IRC [20:46:37] *** alfism has joined #opensolaris [20:51:06] *** Torkel9 has quit IRC [20:51:25] *** Torkel9 has joined #opensolaris [20:55:09] <fedorared> the installer said 400 MB RAM was needed for GUI install [20:55:37] <fedorared> and I can say that 256 MB is really as low as you'd want to go [20:57:47] <dwc-> I've done s10 on 128mb [20:57:49] <dwc-> (sparc) [20:58:07] <dwc-> just don't load up JDS and turn off all the unnecessary services [20:59:39] * delewis cringes at the thought of JDS and 128MB of physical memory [20:59:56] <dwc-> yea, that was a bit slowe [20:59:57] <dwc-> *slow [21:00:07] <dwc-> replaced gdm with dtlogin, and fired up cde [21:03:15] <LurkinLurker> Is the memory requirements the same for Solaris Express, or are you guys talking about regular solaris 10? [21:03:34] <delewis> they're the same in Solaris 10. [21:03:52] <delewis> after 1/06, that is. [21:04:13] * |tsoome| just installed recently s10 11/06 on U10 with 256MB of RAM [21:04:24] *** LordKing has quit IRC [21:04:40] <delewis> heh, it's hard to believe Solaris 10 is approaching the 2 year mark. [21:04:44] <|tsoome|> it does not run any X, of course [21:06:33] *** |tsoome| is now known as tsoome [21:07:51] <tsoome> anyone done upgrade from b50 to b53/b54? [21:08:56] *** mrdeviant has joined #opensolaris [21:09:19] *** lloy0076 has joined #opensolaris [21:09:24] <lloy0076> heh [21:09:26] <onbot> commit by Jeffrey Rutt: 6324053 ETM should avoid alloc/free of event bodies on duplicate msgs; 6343294 fmd.module ereports are generated for numerous error types; 6502549 ETM race condition when sending responses and fma events [21:09:45] <lloy0076> Well, this new AMD 64 X2 3800+ is faster than the Intel D945 at work. [21:09:56] *** jgilje has quit IRC [21:10:00] <tsoome> I've failed with it... [21:10:15] <tsoome> pfinstall seems not to like me... [21:10:41] <richlowe> what's it doing to you? [21:10:50] <richlowe> and are you using liveupgrade? [21:11:01] <tsoome> no netboot and regular one [21:11:16] <tsoome> it detects the required minimal free space frong [21:11:19] <tsoome> wrong [21:11:37] <tsoome> thats in kB: /dev/dsk/c1t0d0s0 141395904 151419533 [21:12:13] <tsoome> last number is the minimum suggested 1kb blocks [21:12:22] *** dunc has joined #opensolaris [21:12:23] <richlowe> isn't that somewhere on the order of 140G? [21:12:34] <tsoome> 144GB, yes [21:12:51] <tsoome> it's W2100z with 146GB HDD [21:13:49] <richlowe> haven't encountered that, no. [21:13:54] <richlowe> but Ow. [21:14:00] <tsoome> upgrade to b53 and to b54 both failed [21:14:38] <tsoome> I remember to have this kind of issue when I had lofs mount in vfstab [21:15:24] *** Error_404 has joined #opensolaris [21:16:09] <tsoome> and running pfinstall -D testProfile by hand gave the same result [21:16:24] <tsoome> so it must be the reason:) [21:16:39] <lloy0076> How do I get my X server to either: a) drop authentication to my local network or b) use xauth when xauth isn't on my path on ON 54? [21:16:49] <lloy0076> xhosts doesn't seem to exist. [21:17:34] <tsoome> then set the path? [21:17:57] <lloy0076> tsoome: I would but WHERE are they... [21:18:03] <tsoome> man find? [21:18:32] <richlowe> /usr/openwin/bin [21:18:40] <lloy0076> richlowe: Thanks. [21:19:05] <tsoome> .oO don't give a fish..... :P [21:20:02] <lloy0076> tsoome: I've just reinstalled an OpenSolaris box. My previous box and usual setup has a PATH that has about 10-15 directories within it. [21:20:09] *** pizdec has quit IRC [21:20:20] <lloy0076> tsoome: I find that making my life with OpenSolaris on the command line takes a LOT of paths. [21:20:43] <tsoome> man find is still valid;) [21:20:54] <lloy0076> tsoome: It's valid and it's annoying. [21:21:23] <lloy0076> tsoome: Given that I've used Linux and other Unixes for about 10 years -- which you probably didn't know -- I already knew how to do that. [21:21:28] <lloy0076> tsoome: man telepathy [21:21:29] <lloy0076> :( [21:22:10] <lasseoe> 10-15 dirs? that IS a lot, what the heck for? [21:22:15] <tsoome> 10 years and you have no idea that solaris has always had a X clients in /usr/openwin/bin ? [21:22:34] <lloy0076> tsoome: I've only used Solaris for about 3 months. [21:23:18] <lloy0076> lasseoe: Basically if it's got a "/bin" I tend to add it to my path. [21:23:31] <lasseoe> lloy0076: that's a bit silly [21:23:40] <lasseoe> add it if you frequently need it [21:23:46] <tsoome> it is, quite;) [21:23:53] * lloy0076 sigh [21:23:58] <lloy0076> Actually, it works for me. [21:24:06] <lloy0076> I really don't care if you think it's silly :P [21:24:18] <tsoome> well [21:24:39] <lloy0076> I'll rephrase. [21:24:48] <lloy0076> I don't care if you think it's silly if you won't actually say why. [21:24:49] <tsoome> what I mean - there are /usr/xpg?/bin dirs, same commands, different semantics... [21:25:16] <tsoome> now it depends how you set up your path.... [21:25:30] <lloy0076> I set my path to find programsthat work for me. [21:25:58] <lloy0076> Maybe that was a better way to say what I wanted to say. That does actually take about 10 entries and possibly more. I don't have it set that way atm. [21:28:27] <tsoome> lots of dirs from /opt ? [21:29:05] <lloy0076> netbeans, eclips and /opt/bin. I suppose I could symlink some of it. [21:29:33] <tsoome> thats the way how you can make the path shorter, yes [21:30:24] <lloy0076> heh - netbeans actually is responsive now. [21:30:27] *** gammon has joined #opensolaris [21:30:40] <lloy0076> Great program, but it does take resources. [21:31:18] *** jamesd has quit IRC [21:31:58] <gammon> hello folks - anyone else having trouble with firefox 2.0 on sxcr build 54 ? [21:32:14] <LurkinLurker> What did you do to make it more resonsive, lloy0076 ? [21:32:28] <LurkinLurker> *responsive, too [21:32:31] <lloy0076> LurkinLurker: I bought a new computer (!!!) [21:32:34] <fedorared> gammon, no, what kind of trouble? [21:33:09] <gammon> firefox starts 1st time, the segfaults 2nd time [21:33:22] <gammon> clear ~/.mozilla & try again: starts ok, then segfaults 2nd time [21:33:22] <LurkinLurker> lol that'll certainly work [21:34:10] * gammon considers blaming live upgrade... again [21:34:57] *** bengtf has joined #opensolaris [21:35:54] <gammon> fedorared: did you do fresh install of snv_54 , live upgrade or bfu ? [21:36:17] <tsoome> hm, if pfinstall is broken, should I replort it from opensolaris or from osc.... [21:36:20] <fedorared> I threw a DVD in and upgraded. [21:36:51] <gammon> k [21:36:55] <tsoome> fedorared: what was the previous version? [21:37:02] <fedorared> b50 [21:37:10] <gammon> hmm [21:37:17] <tsoome> and x86 or sparc? [21:37:21] <fedorared> x86 [21:37:22] <gammon> x86 here [21:37:23] <lloy0076> Damn! The Sparc version of Eclipse doesn't work on x86 :P [21:37:32] <gammon> lol @ lloy0076 [21:37:40] <tsoome> damit.... then why my upgrade failed.... [21:38:01] <tsoome> perjaps I should try with dvd... [21:40:02] * lloy0076 sigh [21:40:22] <lloy0076> They haven't made a 3.2.1 Solaris x86 version yet. This is painful. [21:41:32] *** alanc-away is now known as alanc [21:44:42] <elektronkind> so, I'm trying to write some sdt probes for a kernel module [21:45:02] <elektronkind> and for the life of me, I can't figure out how to feed a struct to a DTRACE_PROBE [21:49:45] *** Tpenta has joined #opensolaris [21:54:21] *** anthony79 has joined #opensolaris [21:59:21] *** yippi has joined #opensolaris [22:00:20] *** gm152 has quit IRC [22:03:45] <lloy0076> bash-3.00$ java --help [22:03:46] <lloy0076> Unrecognized option: --help [22:03:54] <lloy0076> That is so not useful.. [22:04:02] <Error_404> it's perfectly useful [22:04:19] <Error_404> it's telling you not to use GNU options [22:04:33] <Error_404> try java -help [22:04:44] <lloy0076> heh [22:04:58] *** phus has quit IRC [22:04:59] *** nbkk6fo_ has joined #opensolaris [22:05:02] <jmcp> just because java has been gpl' [22:05:10] <lloy0076> EEP [22:05:19] <lloy0076> I suppose I could hack its source :P [22:05:19] *** crib4 has joined #opensolaris [22:05:20] <jmcp> gpl'd doesn mean it's been converted to use gnu long options [22:05:53] <hile_> jmcp: you surprised I could give you those directions from memory last night? [22:06:08] <jmcp> they were from memory? heck! [22:06:17] <jmcp> very much appreciated too [22:07:47] <hile_> most of it was [22:08:49] *** crib has quit IRC [22:10:23] * lloy0076 java is a pile of dog@#@$@c@#$#@p [22:10:33] <trygvis> pft [22:10:46] <lloy0076> It's now decided to say that "There is not enough space." when installing glassfish. [22:11:13] <lloy0076> But the installer won't say *where* there is not enough space. Just that there isn't. [22:12:23] *** vmhobbes has quit IRC [22:12:32] *** vmhobbes has joined #OpenSolaris [22:13:01] <Error_404> probably wherever you're unpacking glassfish [22:13:13] <Error_404> because it doesn't write outside of the directory you run the installer in [22:17:38] <lloy0076> Error_404: If glassfish takes more than 90GB of disk space, it's got a problem. [22:17:46] *** Odin- has joined #opensolaris [22:18:05] <trygvis> did you run the installer from /tmp? [22:18:14] *** fedorared has quit IRC [22:18:15] <lloy0076> It's in /export/home/opt [22:18:29] <lloy0076> ... /dev/dsk/c1d0s7 90G 1.5G 88G 2% /export/home [22:18:56] <lloy0076> If it's using /tmp it might be running out of space. But it didn't on ON53 with BFUs on my older machine. [22:19:25] <lloy0076> But, as I was saying, it's not saying *why* or *where* it's running out of space. Just that it is. [22:21:06] <Error_404> yeah, it doesn't take that much [22:21:33] <Error_404> IIRC it takes up just over half a gig [22:22:03] <lloy0076> Error_404: If that. I've found glassfish's installer to be bizarre. [22:22:27] <lloy0076> Error_404: I've always had to set its heapsize max to higher than whatever the default is. [22:23:31] *** pikapika has quit IRC [22:24:22] *** ChanServ sets mode: +o Tpenta [22:24:28] *** pikapika has joined #opensolaris [22:25:08] *** andrei has joined #opensolaris [22:25:35] <Error_404> okay, I give up [22:25:44] <Error_404> i don't understand struts tiles [22:26:43] <hali> hm, whats the boot option to force 32-bit mode? [22:27:06] <trygvis> it's the default AFAIK [22:27:13] <richlowe> it isn't. [22:27:16] <trygvis> use -d64 to force 64 bit mode [22:27:17] <trygvis> eh? [22:27:22] *** lloy0076 has quit IRC [22:27:30] <hali> i get 64-bit per default [22:27:34] <jmcp> hali: a *boot* option is to specify "kernel/unix" on the grub kernel line [22:27:54] <jmcp> eg kernel /platform/i86pc/multiboot kernel/unix [22:28:04] <trygvis> heh, I was still taking about java %) [22:28:42] *** lloy0076 has joined #opensolaris [22:29:22] <hali> jmcp: thanks, seems to work [22:29:31] <hali> vmware on a core2 didn't like 64-bit mode [22:29:48] *** hile_ has quit IRC [22:30:03] <dlg> hali: is the host os 64bit? [22:30:25] <hali> dlg: it's a 64-bit cpu yes, the host os is 32-bit though [22:30:31] <hali> 64-bit centos runs fine [22:30:54] *** Error_404 has quit IRC [22:31:01] <dlg> weird [22:31:23] <hali> dlg: solaris 64-bit is listed as "experimental" in the settings... so im not supprised [22:31:46] <andrei> hali: is this on vmware workstation version x.x.x? [22:32:11] <hali> andrei: 5.5.3 [22:32:20] <hali> i'll try in vmware server later [22:33:00] <andrei> hali: sorry, I joined late... what is the problem again exactly? [22:33:23] <hali> andrei: the whole vm locks up halfway through boot when running 64-bit mode [22:33:28] <andrei> hali: you're booting solaris on core2 cpu in a VM inside workstation and it only boots in 32-bit mode? [22:33:45] <andrei> hali: what build are you using? [22:33:50] <hali> b54 [22:34:10] *** wraith has quit IRC [22:34:11] <andrei> hali: I just downloaded it last night, but haven't tried it yet... let me try [22:34:22] <hali> 34685 of vmware workst. [22:34:38] <hali> 32-bit mode seems to work perfectly (so far) [22:35:10] <andrei> hali: good, I have the same build of workstation installed [22:35:31] <andrei> hali: it installs fine, right? (installer is 32-bit) [22:35:36] <hali> andrei: yes [22:35:40] <hali> andrei: or no actually [22:35:51] <hali> andrei: the graphical installer failed, textmode was fine [22:36:09] <hali> it may be a problem with running X in 64-bit actually... [22:36:31] <hali> i never booted solaris 64-bit in verbose mode... i would guess it locks up on starting X [22:36:39] <hali> i'll try that in a second ... adding myself a user now [22:36:48] <andrei> hali: I'm running on core2 duo processor also... installing right now [22:37:08] <andrei> hali: well, doing an upgrade from snv48 actually [22:37:19] <hali> bfu? [22:37:33] <andrei> hali: nope... full install [22:37:43] <stevel> 'afternoon andrei [22:37:46] <hali> ok, good.. then we'll have the same X releease as well [22:37:53] <andrei> hali: lots of F2 pressing :-) [22:37:54] *** ShadowHntr has joined #opensolaris [22:37:55] <hali> im no a t7200 laptop proc [22:37:59] <hali> on* [22:38:12] <andrei> hey, stevel! [22:38:13] <hali> andrei: yes, tell me about it :) [22:39:04] <andrei> hali: it's going to take awhile... about an hour I think [22:39:30] *** hile_ has joined #opensolaris [22:39:43] <andrei> stevel: how are things? [22:40:24] <stevel> andrei: not bad. just moved into rab's old seat, so i'm esaxe's roomie now ;) [22:40:49] <andrei> stevel: that's cool.. is coffee on? :-) [22:41:08] <stevel> definitely. it's nice to be within arms-reach of the coffee pot [22:41:29] <stevel> i'm wagering it'll increase my productivity at least 2 fold [22:41:30] <andrei> stevel: that reminds me... I have to go make myself one :-) [22:41:31] <esaxe> hey andrei [22:41:42] <andrei> esaxe: hey eric! [22:42:01] <esaxe> how's things? :) [22:42:02] * andrei brb... making latte [22:44:30] <kFuQ> http://tinyurl.com/y2h4gs this is fuctup [22:45:13] *** lloy0076 has quit IRC [22:48:50] * andrei is back [22:49:13] <andrei> esaxe: everything is great now that I have a cup of coffee w/ some desert :-) [22:49:18] <andrei> esaxe: and you? [22:50:18] *** mrdeviant has left #opensolaris [22:50:22] *** LurkingLurker has joined #opensolaris [22:59:24] <andrei> hali: my upgrade is about 50% done... [22:59:35] <andrei> hali: these core2's are pretty fast CPUs [22:59:54] <hali> andrei: centos is faster under vmware on my core2 than it is native on my amd 3200+ :) [23:00:37] <andrei> hali: using what benchmark? [23:00:46] <hali> andrei: day to day use [23:01:06] <hali> andrei: i do most my work running oracle under centos [23:01:19] <andrei> hali: that's cool [23:01:21] <hali> even with a laptop drive compared to a normal sata drive it "feels" faster [23:01:43] <hali> im sure it is faster actually.. bulding indexes is without a doubt faster [23:08:20] *** hile_ has quit IRC [23:08:22] *** bougie has quit IRC [23:11:16] <andrei> hali: quick question -- are you using e1000g virtual nic when booting in 64-bit mode? [23:12:39] *** yarihm has joined #OpenSolaris [23:14:05] *** hile_ has joined #opensolaris [23:14:16] *** LurkinLurker has quit IRC [23:14:31] *** karrotx has quit IRC [23:14:31] *** karrotx_ has quit IRC [23:15:13] <hali> andrei: i thought the nic wasn't supported at all? [23:15:16] <hali> in 64-bit mode [23:15:19] <hali> i got some big warning [23:16:26] <andrei> hali: default virtual nic is not supported in 64-bit mode [23:17:07] <andrei> add ethernet0.virtualDev = "e1000" to your .vmx file [23:17:28] <hali> andrei: ok [23:22:39] <hali> ok, got 64-bit booted now.. with the e1000 nic and console.. lets see if i can start X [23:22:42] *** lloy0076 has joined #opensolaris [23:22:55] <Tpenta> kfuq, i've only read one or two stories on that, but this idiot has taken the bits of conspiricy that he wants (ignoring what the judge actually said) and run off screaming into tin foil hat land [23:23:24] <Tpenta> good morning all [23:23:37] <hile_> morning alan [23:23:38] <jmcp> hi Tpenta [23:24:29] <lloy0076> Can someone download the Internet for me and send it to me on some floppies? [23:25:03] <trygvis> if you send me the floppies first [23:25:12] *** gammon has quit IRC [23:26:02] <lloy0076> trygvis: That would be quite expensive to do :) [23:28:02] *** yippi has quit IRC [23:28:45] <hali> andrei: ok, X failed to start when in 64-bit mode... works fine in 32-bit [23:30:23] <andrei> hile: I just finished upgrading [23:31:06] <hali> nothing usefull in Xorg.0.log [23:32:41] <andrei> hali: hmm... xorg started for me just fine in 64-bit mode.. [23:32:57] <andrei> hali: try upgrading vmware tools from the command line [23:33:49] *** pikapika has quit IRC [23:34:42] *** pikapika has joined #opensolaris [23:39:10] *** nbkk6fo_ is now known as phus [23:41:50] *** freakazoid0223 has quit IRC [23:42:41] *** yippi has joined #opensolaris [23:42:57] <onbot> commit by sb155480: 6464148 rename in-band descriptor in vnet to use the right prefix.; 6496374 vsw: "tunrnstile_block: unowned mutex" panic on a diskless-clients test bed; 6501505 Assertion panic in vcc on ldom creation.; 6501588 Adding a 2nd VDS to domain0 renders the 1st VDS unreachable, plus false failures on VDS1 disk [23:44:44] *** freakazoid0223 has joined #opensolaris [23:52:12] *** Mazon is now known as mazon [23:53:23] *** lloy0076 has quit IRC [23:55:08] *** sommerfeld has quit IRC