/srv/irclogs.ubuntu.com/2006/06/02/#ubuntu-ports.txt

=== fabbione [n=fabbione@217.205.109.249] has joined #ubuntu-ports
=== tmarble [n=tmarble@192.18.101.5] has joined #ubuntu-ports
tmarblefabbione: buon giorno!04:07
fabbionehey tmarble 04:07
tmarbleto answer your question, yes I do have a launchpad account: https://launchpad.net/people/tmarble04:08
fabbionetmarble: ok :) would you be so kind to send the info via email? otherwise i am going to forget in about 2 seconds :)04:09
tmarblenp04:09
tmarbleI notice you have /sparc on the wiki..... but I can't seem to link to it using an (intuitive) WikiWord..... e.g. sparc/TODO does NOT work?04:10
fabbionehttps://wiki.ubuntu.com/sparc/KnownIssues04:11
fabbionehttps://wiki.ubuntu.com/sparc/TODO04:11
fabbionethey both work for me04:12
tmarbleperhaps it's because / (root) and /sparc are considered different wikis04:12
tmarbleand on my page https://wiki.ubuntu.com/TomMarble  I have to do this [https://wiki.ubuntu.com/sparc/TODO sparc/TODO] 04:13
tmarbleas opposed to just sparc/TODO04:13
fabbionepossibly04:13
fabbionetmarble: do you have a t1000 handy?04:13
tmarblehandy?  not sitting next to me, but I can probably get to one relatively soon04:14
fabbioneok04:14
fabbionethat would be handy for some testing04:14
fabbionei am having issue with the one i got today04:14
fabbioneit doesn't even tftpboot04:14
tmarbleok04:15
tmarblehmm :-(04:15
fabbionenever mind to get to debug the driver04:15
fabbionethe more machines we get to test, the better04:15
tmarbleso do you mean you can't boot from CD either?04:15
=== doko [n=doko@dslb-088-073-101-034.pools.arcor-ip.net] has joined #ubuntu-ports
fabbionethere is no cd in the t1000 :)04:15
dokoanother channel?04:15
tmarbledoko: guten tag!04:15
tmarbledid you tftpboot the T2000?04:16
fabbioneyeps04:16
fabbionealso from cd04:16
fabbioneit works both ways04:16
fabbionebut the t1000 has only netboot04:16
fabbioneand if it can frigging tftpboot it's the machine of doom04:16
tmarblewhy isn't it tftpbooting?04:16
fabbioneit seems like the OBP is looping on the rarpd request04:17
fabbionebut it might be an issue with the server too04:17
fabbioneaccording to tcpdump rarpd is replying back04:17
fabbionebut the OBP keeps going with "Requesting Internet Address for..."04:17
tmarbledo you have dhcp3-server set with the MAC in /etc/dhcp3/dhcpd.conf?04:18
fabbionewith rarpd it should not be required04:18
fabbionebut no there is no dhcp304:18
tmarbleyou don't like this package? http://packages.ubuntu.com/dapper/net/dhcp3-server04:19
fabbioneyes i do. but i don't have root on the tftpboot server04:19
fabbioneand rarpd is supposed to give the ip to the OBP04:19
fabbioneafaik OBP never moved away from rarp04:20
fabbionewe can test that04:20
fabbionefor sure04:20
tmarbleyeah, well, maybe we can *fix* that at some point (OBP is a story all onto itself ;-)04:20
fabbioneyeah but i need to get across the OBP barrier first04:20
tmarblehave you sniffed packets to see if rarpd is doing the 'right thing'?04:21
fabbionetmarble: yeps.. as i said above04:21
tmarbleok... I got some people inside Sun to volunteer yesterday (including device driver guru's) so let me check my e-mail for potential victims04:21
fabbionei suggest you wait once i can get the sysadmin online and check with him04:22
fabbioneperhaps it is as stupid as the netswitch being retarded04:22
tmarbleare you working on the machines in Kensington?04:23
fabbioneeps04:23
fabbioneyeps04:23
tmarbleAFAIK rarpd responses should be broadcast04:23
fabbionethey happen at ethernet level04:24
fabbionethere is no ip to broadcast at that time yet04:24
tmarbleit would be really antisocial of the switch not to forward RARP packets, however04:24
fabbionewell it's HP switch :P04:26
tmarblewe like to refer to them as the House of Printers ;-)04:26
fabbioneahhaha04:26
tmarbledid you see http://ubuntuforums.org/showthread.php?p=1081276#post108127604:28
tmarbledo you have a dashboard of packages that are (or need to be) built for sparc?04:29
tmarblefor example yesterday it *seemed* that autofs was not yet available?04:29
fabbioneno i don't read the forums04:30
fabbionethe packages are all autobuilt by buildd04:30
fabbioneautofs is in universe afaik04:30
fabbioneso it's not really checked04:30
fabbionebut it should be installable04:30
fabbionenice to know.. the U60 is tested well :) both David and I have one ;)04:31
tmarblein Debian non-free is NOT autobuilt by default... what about for Ubuntu components universe and multiverse?04:31
tmarblethere's an engineer in Boston installing on a SunBlade 2500 as we speak04:32
fabbioneoh autofs is non-free?04:33
fabbionedo you have a sun irc channel where you all guys meet up_?04:33
fabbioneFilename: pool/main/a/autofs/autofs_4.1.3-8ubuntu3_i386.deb04:33
fabbioneit's probably due to the documentation that's non-free04:33
fabbioneSB should be ok04:34
fabbionetested on 1000 and 150004:34
fabbionei assume 2500 will be good04:34
tmarblewho hooo!  I'm on a T200004:34
tmarbleLinux memleak 2.6.15-23-sparc64-smp #1 SMP Tue May 23 14:15:47 UTC 2006 sparc64 GNU/Linux04:34
fabbioneehhe04:34
fabbionehow many cores do you have there?04:34
tmarbletestme@memleak:~$ cat /proc/cpuinfo04:35
tmarblecpu             : UltraSparc T1 (Niagara)04:35
tmarblefpu             : UltraSparc T1 integrated FPU04:35
tmarbleprom            : OBP 4.19.0 2005/10/27 17:2404:35
tmarbletype            : sun4v04:35
tmarblencpus probed    : 3204:35
tmarblencpus active    : 3204:35
fabbioneHUMPF04:35
tmarbleD$ parity tl1   : 004:35
tmarbleI$ parity tl1   : 004:35
=== fabbione is jalous
tmarbleCpu0Bogo        : 2401.0804:35
tmarbleCpu0ClkTck      : 0000000047868c0004:35
fabbionei got a 6 cores one04:35
fabbioneoh you even have 1.2Ghz CPU!04:35
fabbionebastard04:35
=== fabbione envies
=== fabbione envies
tmarblebetween webmink and sabdfl I'm sure we can improve the situation over time ;-)04:35
fabbionenah i am ok with this one04:36
fabbionenext one i want a quad-T204:36
fabbioneBEEFY one04:36
tmarbleyeah, 32 GB04:36
fabbioneneat04:36
fabbionei didn't get to install the one we got here04:37
fabbionebut it should be the same04:37
tmarblethinking about how we can harness energy like recupero in the forum...04:40
tmarbleperhaps we should start a wiki page on install successes/failures/issues..... sparc/InstallReports ?04:41
fabbionetmarble: good idea04:41
fabbionego ahead :)04:41
tmarblek04:41
fabbionemake sure to add details to the table04:42
tmarbleplease elaborate?04:42
fabbionelike model, cpu, n of cpu, ram, installation methos04:42
fabbionemethod04:42
fabbioneU60 is quite generic04:42
tmarblebetter yet... let's write a "platform info" script (we do this all the time)04:42
fabbioneit doesn't tell us much other than we share the same mobo with a E450 ;)04:42
tmarblethat captures relevant stuff /proc/cpuinfo /proc/meminfo etc....04:42
fabbioneyeah04:42
tmarbleso now what can we use (should we use) as a VCS for that?04:43
tmarblefor example, is there a bzr space where we can share scripts, etc.?04:43
fabbionenot yet..04:43
fabbionei have the sysadmins online now04:44
fabbionesorry i will not pay much attention to irc04:44
fabbionehave to get the t1000 cabled properly04:44
tmarblenp04:44
fabbioneor whatever it's the issue04:44
tmarblei'll ask around about the rarpd / OBP thing04:44
fabbionedhcp doesn't help with OBP04:56
tmarblefrom Gary in Boston....04:58
tmarbleissues with sunblade 2500 2x1.2  72gb disk, xvr-600 video04:58
tmarble1.215625 tg3: Attempt to do nvram_read on Sun 570x04:58
tmarble15.122338 sda: asking for cache data failed04:58
tmarble15.134332 sda: assuming drive cache: write through04:58
tmarbleIt has a problem recognizing the scsi disk04:58
fabbioneplease direct them to launchpad.net and to file bugs04:58
tmarbleGary is a linux n00b so I may "qualify" the bug before filing it04:59
fabbionei would like to understand what makes him think that that message is an error05:00
fabbionedoes the installer recognize the disks?05:00
tmarbleis there a problem with the GUI installer?  is there a GUI for sparc?05:02
fabbionethere is no GUI installer for sparc05:02
fabbioneit was the switch05:06
fabbioneproblem solved05:06
tmarblesweet!05:06
tmarbleok... seems Gary's problem is that he has a SCSI disk -- which may not be recognized05:07
tmarbledoes the install ISO have an initrd (to bootstrap SCSI)?05:07
fabbioneok we will debug that later05:07
fabbionethe best is to do netbooting05:08
fabbionethe iso has everything05:08
fabbionescsi/ide etc..05:08
fabbionebut already the sda message is a hint that it's there05:08
tmarbleright (I was thinking that ;-)05:08
fabbioneif the disk for "scratch me hard"05:09
fabbioneask him to go console and do:05:09
fabbionedd if=/dev/zero of=/dev/sda bs=512 count=105:09
fabbioneand restart the partitioner05:09
tmarbleseems to be working better now... it apparently sees the disk (he did a OBP update AND installed a new disk)05:10
tmarblewhat do you mean: if the disk for "scratch me hard"05:10
tmarbleok.. he's partitioning now05:11
fabbionei mean if the disk has no sensible data and can e scratched05:11
fabbiones/e/be/05:11
tmarbleis 10 GB big enough for / ?05:12
fabbioneway too much05:13
fabbione2.5GB and you have space to swim05:14
tmarbleI really hate running out05:14
tmarblewhen you have a big update05:14
fabbioneif you use the Guided Partitioner will do the right thing for you05:14
fabbioneyou can even install on LVM if that's make you more happy ;)05:14
tmarblelvm not necessary now (but it is nice)05:15
fabbionethe guided partitioner can do it for you ;05:15
fabbione;)05:15
=== fabbione id happily code that
tmarblecool05:16
tmarbledo you know of our Sun video card xvr-600 is supported?  (if not I assume we will need to start with VESA)05:38
tmarbletestme@dhcp-ubur02-174-225:~$ cat /proc/cpuinfo05:43
tmarblecpu             : TI UltraSparc IIIi (Jalapeno)05:43
tmarblefpu             : UltraSparc IIIi integrated FPU05:43
tmarbleprom            : OBP 4.17.1 2005/04/11 14:2405:43
tmarbletype            : sun4u05:43
tmarblencpus probed    : 205:43
tmarblencpus active    : 205:43
fabbionetmarble: no idea.. all my machines are headless05:45
fabbioneis that a pci or sbus card?05:45
tmarbletestme@dhcp-ubur02-174-225:~$ lspci05:46
tmarble0000:00:00.0 Host bridge: Sun Microsystems Computer Corp. Tomatillo PCI Bus Module05:46
tmarble0000:00:03.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5703 Gigabit Ethernet05:46
tmarble0001:00:00.0 Host bridge: Sun Microsystems Computer Corp. Tomatillo PCI Bus Module05:46
tmarble0001:00:04.0 SCSI storage controller: LSI Logic / Symbios Logic 53c1010 66MHz  Ultra3 SCSI Adapter (rev 01)05:46
tmarble0001:00:04.1 SCSI storage controller: LSI Logic / Symbios Logic 53c1010 66MHz  Ultra3 SCSI Adapter (rev 01)05:46
tmarble0002:00:00.0 Host bridge: Sun Microsystems Computer Corp. Tomatillo PCI Bus Module05:46
tmarble0002:00:06.0 Non-VGA unclassified device: ALi Corporation M7101 Power Management Controller [PMU] 05:46
tmarble0002:00:07.0 ISA bridge: ALi Corporation M1533 PCI to ISA Bridge [Aladdin IV] 05:46
tmarble0002:00:08.0 Multimedia audio controller: ALi Corporation M5451 PCI AC-Link Controller Audio Device (rev 02)05:46
tmarble0002:00:0a.0 USB Controller: ALi Corporation USB 1.1 Controller (rev 03)05:46
tmarble0002:00:0b.0 USB Controller: ALi Corporation USB 1.1 Controller (rev 03)05:46
tmarble0002:00:0d.0 IDE interface: ALi Corporation M5229 IDE (rev c4)05:46
tmarble0003:00:00.0 Host bridge: Sun Microsystems Computer Corp. Tomatillo PCI Bus Module05:47
tmarble0003:00:02.0 Display controller: 3DLabs Sun XVR-500 Graphics Accelerator (rev 01)05:47
tmarbletestme@dhcp-ubur02-174-225:~$05:47
fabbioneARGH05:47
fabbioneplease don't flood the channel05:47
tmarblesorry05:47
fabbionecan you please give me lspci -n | grep 0300 ?05:51
tmarblelI think you want this05:52
tmarble0003:00:02.0 Display controller: 3DLabs Sun XVR-500 Graphics Accelerator (rev 01)05:52
tmarble0003:00:02.0 0380: 3d3d:07a2 (rev 01)05:52
fabbioneyes i need the lspci -n |  grep 0300 for that05:53
fabbioneuh weird05:53
fabbione380 _?05:53
fabbionethat's not GFX class05:53
=== fabbione checks again the specs
tmarblenope05:53
tmarblewe have guys working on Xorg here so I can ping them if necessary05:54
fabbione        3d3d07a2        unknown unknown Sun XVR-500 Graphics Accelerator05:54
fabbioneit's not known to me05:54
fabbionelet me ask Alan05:54
fabbionethe worst case try vesa05:55
fabbioneor glint05:55
fabbioneglint driver seems to have tons of references to 3DLabs chipsets05:55
tmarbleah.. didn't know about glint05:55
fabbionenot sure it will work tho05:57
fabbioneit doesn't say anywhere XVR-500 05:57
tmarbleI now am logged into the T2000 and SunBlade 2500 -- need to apt-get a number of things and configure the networkign stuff (automount, etc.)05:58
tmarblethen I'll work on the graphics for the SunBlade05:58
fabbionesure take your time05:59
fabbionei am still debugging and testing a memory corruption issue with CPU < Sparc III05:59
tmarbleargh06:00
fabbioneit's not very easy to trigger06:02
fabbioneso don't be too worried06:02
fabbioneit happens only when installing or a heavy usage of dpkg06:02
fabbionethey suggest glint..06:04
fabbionetry that one06:04
tmarbleok06:06
tmarbleWhich CPU model/system has the problem?  You *might* be tripping over the e-cache bug06:13
fabbioned-cache ;)06:16
fabbioneanyway we have a possible fix and we are testing it06:16
fabbioneit affects all machines that use CPU's older than Sparc III06:17
fabbioneso II and IIi06:17
tmarbleyep... there were various fixes for Solaris but I don't know how well they would xfer to Linux06:19
fabbionethey don't because we don't get to read the solaris code06:21
fabbioneto avoid all possible IP clashing and so on06:21
fabbionesince they have difference licence it's "dangerous"06:21
tmarbleregretablly the CDDL is GPL incompatible :-(06:22
fabbione<ajax> there's a wildcat driver on 3dlabs' website if you dig REALLY HARD06:22
fabbione<ajax> very closed source06:22
fabbionethat's for your card06:22
fabbioneno idea if it builds on sparc06:22
fabbioneit's worth a shot06:22
tmarbleah, ok06:22
tmarbleI'll ask our Xorg guys (if they have contributed source to Xorg then that would be best)06:23
fabbionetmarble: i did speak with Alan Coppersmith06:23
fabbionehe works for xorg/SUN ;)06:24
tmarbleright!06:24
fabbionebut of course if you have other contacts it's even better06:24
tmarbleoh, that Alan (you spoke with just a minute ago)?06:24
fabbioneyes06:25
tmarbleis he on IRC?  what's his nick?06:25
fabbionealanc06:25
fabbionehe is on #xorg-devel06:25
tmarblethanks (found him on our internal IRC as well)06:28
fabbionehehe06:30
fabbioneso you have an internal IRC :)06:30
fabbioneis it accessable from outside?06:30
tmarbleregrettably no (I asked ;-)06:31
fabbioneeheh ok06:33
tmarblejust to confirm... is this the line I want:06:46
tmarbledeb http://us.archive.ubuntu.com/ubuntu/ dapper main restricted universe multiverse06:46
fabbionewell it depends06:46
fabbioneuniverse and multiverse are not supported06:47
fabbioneand you also want another 2 lines with dapper-security and dapper-updates06:47
fabbioneotherwise it looks good06:47
tmarblegreat.. just to confirm my understanding of "Suites"06:47
tmarbledapper = never changes06:47
tmarbledapper-security = security fixes ONLY06:47
tmarbledapper-updates = bug fixes (NO INTERSECTION or SUPERSET of dapper-security)?06:48
tmarbledapper-backports = ?06:48
fabbionedapper = right06:48
fabbionedapper-security and dapper-updates are theoretically what you say06:49
fabbionebut there are some corner cases when a package gets both security and bug fixes06:49
fabbioneso both the suites need to get different uploads06:49
tmarblebut normally no overlap?06:49
fabbionebut that's nothing you really don't need to worry about06:49
fabbionenormally no overlap06:49
fabbioneskip backports06:49
fabbionebackports is crackful stuff 06:50
fabbionethat's entirely unsupported06:50
=== Brad_ [n=braddr@bellevue.puremagic.com] has joined #ubuntu-ports

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!