Back to Home / #uml / 2007 / 10 / Prev Day | Next Day
#uml IRC Logs for 2007-10-31

---Logopened Wed Oct 31 00:00:23 2007
02:20|-|Magotari [~karol@chello089076077166.chello.pl] has quit [Quit: Lost terminal]
04:35|-|ds2 [noinf@netblock-66-245-251-24.dslextreme.com] has quit [Read error: Operation timed out]
04:38|-|ds2 [noinf@netblock-66-245-251-24.dslextreme.com] has joined #uml
04:54|-|tuxbubling [~omauras@62.80.120.214] has joined #uml
04:54<tuxbubling-#uml->>hello there
04:58<tuxbubling-#uml->>is skas3 patch worth it?
06:34|-|Urgleflogue [~plamen@83.228.65.158] has quit [Read error: No route to host]
06:38|-|Urgleflogue [~plamen@83.228.65.158] has joined #uml
08:20|-|natacha28 [~natacha28@ANantes-257-1-29-208.w90-25.abo.wanadoo.fr] has joined #uml
08:21<tuxbubling-#uml->>i'm following this page http://uml.jfdi.org/uml/Wiki.jsp?page=Chroot, but i don't have /proc/mm on my system... any idea?
08:22|-|natacha28 [~natacha28@ANantes-257-1-29-208.w90-25.abo.wanadoo.fr] has quit []
08:24<tuxbubling-#uml->>ok sound like i can let my ./proc empty... the problem is that it's segfaulting at start :(
08:25<tuxbubling-#uml->>oh my bad
08:29<tuxbubling-#uml->>Someone has an idea? http://paste.tuxbubling.org/1684
09:48|-|arun [~arun@yktgi01e0-s5.watson.ibm.com] has joined #uml
10:09<tuxbubling-#uml->>someone awake ?
10:17<tuxbubling-#uml->>got problem making my network working under chroot :(
10:28<tuxbubling-#uml->>how am i suppose to make tun/tap network working?
10:30<caker-#uml->>http://edeca.net/articles/bridging/
10:40<tuxbubling-#uml->>is bridging better than tap?
10:46<tuxbubling-#uml->>by bridgind i don't need to have iptables stuff isn't it?
10:48|-|jdike [~jdike@pool-71-248-190-161.bstnma.fios.verizon.net] has joined #uml
10:48<jdike-#uml->>Hi guys
10:58<caker-#uml->>hello Jeff
11:01<tuxbubling-#uml->>hello
11:03<tuxbubling-#uml->>is that normal i have to mount-bind /proc and /dev/shm to have my chroot working?
11:03<jdike-#uml->>chroot == UML?
11:03<jdike-#uml->>Oh, UML within chroot
11:04<jdike-#uml->>UML should complain, but work without proc mounted
11:04<jdike-#uml->>and it should also work without /dev/shm, but be slower
11:07<tuxbubling-#uml->>without /proc i got a kernel panic
11:08<tuxbubling-#uml->>but i noticed that i can umount it right after it start booting
11:08<jdike-#uml->>hmm
11:09<jdike-#uml->>I just got rid of one reason for needing proc
11:09<jdike-#uml->>what's the panic?
11:09<tuxbubling-#uml->>gimme a second
11:10<tuxbubling-#uml->>http://paste.tuxbubling.org/1698
11:11<jdike-#uml->>yes
11:12<jdike-#uml->>with skas3, it's fairly important that UML have access to /proc/mm
11:12<jdike-#uml->>although that panic isn't about not being able to open it
11:13<jdike-#uml->>can you paste the full boot log?
11:16|-|hfb [~hfb@pool-71-106-219-180.lsanca.dsl-w.verizon.net] has joined #uml
11:23<tuxbubling-#uml->>jdike: 2s
11:24<tuxbubling-#uml->>http://paste.tuxbubling.org/1700
11:25<tuxbubling-#uml->>i'm not in skas3
11:25<jdike-#uml->>#
11:25<jdike-#uml->>Checking for the skas3 patch in the host:
11:25<jdike-#uml->>#
11:25<jdike-#uml->> - /proc/mm...found
11:25<jdike-#uml->>#
11:25<jdike-#uml->> - PTRACE_FAULTINFO...not found
11:25<jdike-#uml->>#
11:25<jdike-#uml->> - PTRACE_LDT...not found
11:26<jdike-#uml->>it seems to think you have partial skas3 support
11:26<tuxbubling-#uml->>oh
11:26<tuxbubling-#uml->>i'm using a prebuilt kernel
11:26<jdike-#uml->>host kernel?
11:26<tuxbubling-#uml->>2.6.23 vanilla
11:26<jdike-#uml->>but that's not prebuilt
11:26<jdike-#uml->>UML is prebuilt
11:26<tuxbubling-#uml->>yes that's what i meant
11:27<jdike-#uml->>and that was inside the chroot?
11:27<tuxbubling-#uml->>yes
11:27<tuxbubling-#uml->>outside that works flawlessly
11:27<jdike-#uml->>without /proc mounted inside?
11:27<tuxbubling-#uml->>and inside the chroot i have to mount-bind /proc
11:28<tuxbubling-#uml->>yes
11:28<jdike-#uml->>and when you do, UML boots?
11:28<tuxbubling-#uml->>boots only if i have it mount-bind
11:28<tuxbubling-#uml->>else panic
11:29<jdike-#uml->>can you do the panic again, and strace UML?
11:29<jdike-#uml->>then paste the strace log?
11:30<tuxbubling-#uml->>i can yes
11:30<jdike-#uml->>at least to the point where it says write("found")
11:30<jdike-#uml->>for /proc/mm
11:33<tuxbubling-#uml->>http://www.tuxbubling.org/public/uml_strace.txt
11:33|-|arun [~arun@yktgi01e0-s5.watson.ibm.com] has left #uml [Leaving]
11:35<tuxbubling-#uml->>i'm running in the chroot with chpst
11:37<tuxbubling-#uml->>but i don't think that makes too much differences
11:38<tuxbubling-#uml->>i'm on x86_64 too
11:40<tuxbubling-#uml->>does it help ?
11:41<jdike-#uml->>access("/proc/mm", W_OK)
11:41<jdike-#uml->> = 0
11:42<jdike-#uml->>I guess access is the wrong way to test for file existence
11:42<tuxbubling-#uml->>i don't have any /proc/mm on my host actually
11:42<tuxbubling-#uml->>ok believe me or not
11:42<tuxbubling-#uml->>i'm dumb
11:42<jdike-#uml->>that test just seems to be telling you "yeah I can make /proc/mm"
11:42<tuxbubling-#uml->>i rm chroot/proc/mm
11:43<tuxbubling-#uml->>and now it works
11:43<jdike-#uml->>not "/proc/mm is there and I can write to it"
11:43<jdike-#uml->>it was a normal file with some crap in it, then
11:43<tuxbubling-#uml->>yes
11:43<tuxbubling-#uml->>damn
11:44<jdike-#uml->>is there a UML bug here?
11:44<tuxbubling-#uml->>sorry to have bothered for thet
11:45<tuxbubling-#uml->>actually no bug... i shouldn't have created that /proc/mm
11:45<tuxbubling-#uml->>:|
11:45<tuxbubling-#uml->>would have save me some time
11:46<jdike-#uml->>did you do that by hand?
11:47<tuxbubling-#uml->>yep :(
11:47<tuxbubling-#uml->>dumbly followed the wiki
11:48<jdike-#uml->>the wiki told you to cat > /proc/mm?
11:48<jdike-#uml->>or touch it or something?
11:48<tuxbubling-#uml->>touch it
11:48<tuxbubling-#uml->>http://uml.jfdi.org/uml/Wiki.jsp?page=Chroot
11:48|-|karol [~karol@adps180.neoplus.adsl.tpnet.pl] has joined #uml
11:48|-|karol changed nick to Magotari
11:49<jdike-#uml->>that might be OK if you did the bind-mount right after that
11:49<jdike-#uml->>the mount point has to exist, probably
11:49<tuxbubling-#uml->>might if i had /proc/mm on the host ;)
11:50<jdike-#uml->>right - you should have stopped right after the mount failed
11:50<tuxbubling-#uml->>as i have not...
11:50<tuxbubling-#uml->>yes
11:51<Magotari-#uml->>jdike: When UML crashes it leaves the umid directories created. I bashed together a patch, it solved the problem for me, but it is stuff of coding nightmares... Just reporting something which might be helpful to fix in the future.
11:51<tuxbubling-#uml->>so if understood correctly, it's running by default in skas0 isn't it?
11:52<jdike-#uml->>yes
11:52<jdike-#uml->>Magotari, crashes how?
11:52<Magotari-#uml->>Any panic.
11:52<Magotari-#uml->>The umid dir remains after a panic.
11:52<jdike-#uml->>stupid question, yes, you're right
11:52<Magotari-#uml->>This is with 23-mm1
11:53<jdike-#uml->>The thing is, when you're in panic, you have no idea what your state is, or what you can successfully do
11:53<jdike-#uml->>so just bail out hopefully with some useful information on the screen
11:54<jdike-#uml->>and when UML boots back up and sees the umid directory, it will carefully look for that UML
11:54<Magotari-#uml->>Yeah, I am aware. This is the reason why I am not submitting that one, but it is irritating to go into my .uml and find lots of virtual machine corpses.
11:54<jdike-#uml->>and if it's not there, it will take over the directory
11:54<jdike-#uml->>Oh, the random ones
11:54<Magotari-#uml->>Yes, those.
11:55<jdike-#uml->>yeah
11:55<Magotari-#uml->>I am aware of the nonrandom ones.
11:55<Magotari-#uml->>No problem with those in my book.
11:57<jdike-#uml->>BTW, I fiddled your cmov stuff a little
11:57<jdike-#uml->>http://rafb.net/p/fnQ1Me36.html
11:58<jdike-#uml->>also test_for_host_cmov doesn't return anything - that was kind of pointless
12:00<Magotari-#uml->>Hmmm...
12:00<Magotari-#uml->>I thought cmovz reg reg would be a nop. Ah well.
12:01<jdike-#uml->>ha
12:01<Magotari-#uml->>Especially if it is the same reg.
12:01<jdike-#uml->>I was concerned about it mangling an in-use register
12:01<jdike-#uml->>I didn't think about whether that is a no-op
12:01<Magotari-#uml->>Yeah, I see what you mean. I chose ecx because it seemed safer than eax.
12:01<jdike-#uml->>but this way makes it irrelevant whether it's a no-p
12:02<jdike-#uml->>no-op
12:02<Magotari-#uml->>But I agree, better to be on the safe side.
12:02<Magotari-#uml->>Just one more question. Why long? Won't an int suffice?
12:04<jdike-#uml->>yes
12:04<jdike-#uml->>but on all architectures that I know of, long is the size of a register
12:04<jdike-#uml->>and on some, int is shorter
12:06<Magotari-#uml->>Erm... But this file is explicitly for 32 bits, isn't it? arch/um/sys-i386/bugs.c
12:06<jdike-#uml->>yes
12:07<jdike-#uml->>but int and long are the same size, so it doesn't matter
12:07<Magotari-#uml->>*shame on me* :/
12:07<Magotari-#uml->>I always thought long was longer.
12:07<Magotari-#uml->>Ok, gotcha.
12:08<Magotari-#uml->>Sorry to be a bother again. Back to crawling the code for problems.
12:09<tuxbubling-#uml->>thx for help
12:09<tuxbubling-#uml->>cya
12:09|-|tuxbubling [~omauras@62.80.120.214] has left #uml []
12:25|-|ram [~ram@pool-71-117-247-93.ptldor.fios.verizon.net] has joined #uml
12:33<jdike-#uml->>bugs.c down to 76 lines
12:34<Magotari-#uml->>Yeah, I removed something like 120 lines from it with my patch. Most files which make up UML are really small, no more than 400 lines most of the time.
12:35<Magotari-#uml->>Really nice to read for a newbie like me. Not too complicated at all.
12:38<jdike-#uml->>The low-level stuff is pretty easy for the most part
12:39<jdike-#uml->>For most arches, that's where the hairy stuff is
12:41<Magotari-#uml->>Hmm... I just built UML without support for the config file embedded into it. However, --showconfig works right as rain.
12:41<jdike-#uml->>there are two copies of the config file
12:42<jdike-#uml->>the one that shows up in /proc/config.gz, which you just turned off
12:42<jdike-#uml->>and the one that UML saves away for --showconfig, which is done always
12:45<Magotari-#uml->>Yeah, I see that. I could try to get that 'fixed' but probably it would bite us both when helping someone with uml one day... Not worth it to save 10K.
13:03|-|ram [~ram@pool-71-117-247-93.ptldor.fios.verizon.net] has quit [Ping timeout: 480 seconds]
13:16|-|ram [~ram@pool-72-90-125-50.ptldor.fios.verizon.net] has joined #uml
13:17|-|Baltam [~WIKIMOKI@bravo839.server4you.de] has quit [Remote host closed the connection]
13:21|-|Baltam [~WIKIMOKI@tor-irc.dnsbl.oftc.net] has joined #uml
13:38|-|kos_tom [~thomas@humanoidz.org] has joined #uml
13:56|-|tyler29 [~tyler@ARennes-257-1-117-55.w86-210.abo.wanadoo.fr] has joined #uml
14:48<jdike-#uml->>rjbell4, there?
15:09<rjbell4-#uml->>jdike: Yes
15:09<jdike-#uml->>what version of UML do you have?
15:09<rjbell4-#uml->>A pull of kernel.org from around October 2nd.
15:10<jdike-#uml->>hmm
15:10<jdike-#uml->>can you try 2.6.23?
15:10<jdike-#uml->>I reproduce your ptrace problems here
15:10<jdike-#uml->>can't
15:11<rjbell4-#uml->>jdike: Okay, maybe not immediately, though
15:11<jdike-#uml->>OK
15:12<jdike-#uml->>There have been fixes in this area
15:14<jdike-#uml->>just nothing specifically aimed at what you are seeing
15:25|-|kokoko1 [~Slacker@203.148.65.19] has joined #uml
15:25<kokoko1-#uml->>Hi
15:26<jdike-#uml->>Hey
15:26<kokoko1-#uml->>jeff how are you doing? :)
15:26<jdike-#uml->>not bad
15:27<kokoko1-#uml->>how's weather these days in your part of the world?
15:28<kokoko1-#uml->>its turning cold here but still dry , no rain from quite a long time.
15:28<jdike-#uml->>getting cooler - winter is on its way
15:29<kokoko1-#uml->>Anything new for UML?
15:29<rjbell4-#uml->>jdike: You're up here in either MA or NH, right?
15:29<jdike-#uml->>MA now, used to be NH
15:30<kokoko1-#uml->>running 2.6.21.4 UML
15:30<jdike-#uml->>and I meant to ask, you're an ex-DECcie?
15:30<rjbell4-#uml->>jdike: Well, by the time I got here, DEC was Compaq, then became HP.
15:31<rjbell4-#uml->>jdike: You remember Ed Cande or Mary Walker?
15:31<jdike-#uml->>hm
15:31<jdike-#uml->>both
15:31<rjbell4-#uml->>My co-workers...
15:31<jdike-#uml->>I seem to recall a Bob Bell before I left, and that was before Compaq
15:32<rjbell4-#uml->>jdike: Lots of Bob Bell's around. I get a lot of their mail, too. :)
15:32<jdike-#uml->>hehe
15:32<jdike-#uml->>OK
15:33<kokoko1-#uml->>heh, rjbell4 lucky you still working there even after the merging ;)
15:33<jdike-#uml->>Andy Grover (former ACPI maintainer, used to be at Intel) ocassionally got mail for Andy Grove (former Intel CEO) that he absolutely should not have seen :-)
15:33<kokoko1-#uml->>heh
15:36<rjbell4-#uml->>hehe
15:37<rjbell4-#uml->>kokoko1: I got laid off as part of a very large group when HP cancelled the port of TruClusters to HP-UX. I landed at EMC almost immediately, though.
15:41<kokoko1-#uml->>right, jdike looks interesting to me http://ozlabs.org/pipermail/lguest/2007-October/000573.html
15:44<kokoko1-#uml->>its always nice to see ppl using uml like Openswan, but its annoying when watching them for looking for alternatives :-s
15:46<jdike-#uml->>hehe
15:46<jdike-#uml->>I wonder if they're still stuck on 2.4
15:49|-|tyler29 [~tyler@ARennes-257-1-117-55.w86-210.abo.wanadoo.fr] has quit [Ping timeout: 480 seconds]
15:53<kokoko1-#uml->>jdike, how is the traffic 'volume' at uml mailing list?
15:53<kokoko1-#uml->>heh, i never posted to the list, i always get support from you in the # :)
15:53|-|apic [~andi@paranerd-geekophren.name] has quit [Server closed connection]
15:53|-|apic [~andi@paranerd-geekophren.name] has joined #uml
15:53<jdike-#uml->>~60 over the last week
15:53<jdike-#uml->>so 6-7/day
15:59|-|SNy_ [882b54dd25@bmx-chemnitz.de] has joined #uml
15:59|-|Netsplit cation.oftc.net <-> kilo.oftc.net quits: SNy, krau
15:59|-|SNy_ changed nick to SNy
16:00<jdike-#uml->>Four more patches off to akpm
16:01|-|krau [~cktakahas@200.184.118.132] has joined #uml
16:02|-|tyler29 [~tyler@ARennes-257-1-124-197.w86-210.abo.wanadoo.fr] has joined #uml
16:04<kokoko1-#uml->>!kernel
16:04<linbot-#uml->>kokoko1: The latest stable kernel is 2.6.23.1; the latest snapshot of the stable kernel is 2.6.24-rc1-git9; the latest beta kernel is 2.6.23-mm1.
16:05<kokoko1-#uml->>right will be merged to .23.x?
16:05<rjbell4-#uml->>jdike: I'm headed out, but thought I'd let you know that I get the same error with 2.6.23 (from gdb at least; I haven't re-straced it)
16:05<jdike-#uml->>OK
16:06<jdike-#uml->>I might need help reproducing it then
16:15|-|tyler29 [~tyler@ARennes-257-1-124-197.w86-210.abo.wanadoo.fr] has quit [Ping timeout: 480 seconds]
16:23<Magotari-#uml->>Gah. I was asleep like a rock for so long...
16:23<Magotari-#uml->>jdike: What is the ptrace problem? Maybe I could try to reproduce?
16:28<jdike-#uml->>on x86_64, gdb doesn't work for him
16:28<jdike-#uml->>works OK for me though
16:29<Magotari-#uml->>:/ I am going to post an ad. Maybe someone wants to give me a computer. Sadly, this is a blocking operation, so I cannot test now.
16:29<Magotari-#uml->>Craigslist, maybe...
16:31|-|tyler29 [~tyler@ARennes-257-1-22-150.w81-53.abo.wanadoo.fr] has joined #uml
16:31<kokoko1-#uml->>hehe
16:42|-|dang [~dang@nemesis.fprintf.net] has joined #uml
16:43|-|dang [~dang@nemesis.fprintf.net] has left #uml []
16:59|-|tyler29 [~tyler@ARennes-257-1-22-150.w81-53.abo.wanadoo.fr] has quit [Ping timeout: 480 seconds]
17:16|-|tyler29 [~tyler@ARennes-257-1-157-84.w86-214.abo.wanadoo.fr] has joined #uml
17:28|-|tchan [~tchan@c-24-13-84-219.hsd1.il.comcast.net] has quit [Remote host closed the connection]
17:29|-|tchan [~tchan@c-24-13-84-219.hsd1.il.comcast.net] has joined #uml
18:01|-|aroscha [~aroscha@chello213047053193.30.11.tuwien.teleweb.at] has joined #uml
18:09|-|hfb [~hfb@pool-71-106-219-180.lsanca.dsl-w.verizon.net] has quit [Quit: Leaving]
18:30|-|silug [~steve@38.104.162.2] has quit [Read error: Connection reset by peer]
19:24|-|peterz [~peterz@f237116.upc-f.chello.nl] has quit [Server closed connection]
19:24|-|peterz [~peterz@f237116.upc-f.chello.nl] has joined #uml
19:24|-|jdike [~jdike@pool-71-248-190-161.bstnma.fios.verizon.net] has quit [Quit: Leaving]
19:28|-|tyler29 [~tyler@ARennes-257-1-157-84.w86-214.abo.wanadoo.fr] has quit [Remote host closed the connection]
19:51|-|kos_tom [~thomas@humanoidz.org] has quit [Quit: I like core dumps]
20:09|-|aroscha [~aroscha@chello213047053193.30.11.tuwien.teleweb.at] has quit [Quit: aroscha]
20:26|-|silug [~steve@38.104.162.2] has joined #uml
20:34|-|Hunger [Hunger.hu@Hunger.hu] has joined #uml
23:08|-|Nem^ [~Nem@dslb-084-056-250-084.pools.arcor-ip.net] has quit [Ping timeout: 480 seconds]
23:18|-|Nem^ [~Nem@dslb-084-056-253-078.pools.arcor-ip.net] has joined #uml
23:59|-|VS_ChanLog [~stats@ns.theshore.net] has left #uml [Rotating Logs]
23:59|-|VS_ChanLog [~stats@ns.theshore.net] has joined #uml
---Logclosed Thu Nov 01 00:00:56 2007