Re: [Openmosix-devel] What's the status of the Userland tools för OM 2.6 ?
Ralf Oelschlaegel <[email protected]>
| Newsgroups | gmane.linux.cluster.openmosix.devel |
|---|---|
| Message-ID | <[email protected]> |
Evan Hisey wrote:
> Anders-
> As much as hate to admit it, there has been a stall in the userland
> tools. The branch we were working on ran in to a snag and have not yet
> recovered. I am not sure of the exact status of the other 2 attempts.
> Right now migration is having to be done manually on teh 2.6 tree.
All my tests with om2.6 dosn't work for manually migration (s. test with 2.6.13
/ posting from 25.10.2005, also a test with 2.6.14 and a mail to Vincent).
Bruce announced another test with 2.6.14 at 23.11.2005 for Kris. He write
"cnode2 should be live tomorrow :)" but no message about succes.
My question: can anybody write down step by step to build a working om2.6 kernel?
Here is my last mail to Vincent:
Vincent write
>
> There's a 2.6.14 now btw.
>
update is done, compilation and boot seems to be good
> looks like some stuff happen during the migration process, and the
> kernel decide to kill the task after starting the process.
> I'll have a look later
I changed my distri from SuSE9.3 to 10.0 (with om-2.6.14-kernel) but the same
problems, if migration from sendpc to recvpc with a pure-cpu-job:
sendpc:~ # echo "192.168.0.2" > /proc/6873/om/where
Nov 17 13:26:01 sendpc kernel: [OMBUG] comm_connect: failed to connect
Nov 17 13:26:01 sendpc kernel: [OMBUG] task_local_send: error
Nov 17 13:26:01 sendpc kernel: [OMBUG] task_local_send: failed
Nov 17 13:26:43 sendpc kernel: [OMBUG] comm_recv: got -32 of 8
Nov 17 13:26:43 sendpc kernel: deputy communication died.
Nov 17 17:20:34 recvpc kernel: [OM] receiving new process
Nov 17 17:20:34 recvpc kernel: [OM] starting process(6379)
Nov 17 17:20:34 recvpc kernel: ------------[ cut here ]------------
Nov 17 17:20:34 recvpc kernel: kernel BUG at mm/mmap.c:2005!
Nov 17 17:20:34 recvpc kernel: invalid operand: 0000 [#1]
Nov 17 17:20:34 recvpc kernel: Modules linked in: ipt_pkttype ipt_LOG ipt_limit
speedstep_lib autofs4 freq_table thermal processor fan button battery ac usbhid
ide_cd cdrom uhci_hcd usbcore i2c_viapro via686a i2c_isa i2c_core via_agp
agpgart generic ipt_REJECT ipt_state iptable_mangle iptable_nat ip_nat
iptable_filter ip6table_mangle ip_conntrack ip_tables ip6table_filter ip6_tables
ipv6 dm_mod
Nov 17 17:20:34 recvpc kernel: CPU: 0
Nov 17 17:20:34 recvpc kernel: EIP: 0060:[<c015818a>] Not tainted VLI
Nov 17 17:20:34 recvpc kernel: EFLAGS: 00010282 (2.6.14-om-default)
Nov 17 17:20:34 recvpc kernel: EIP is at exit_mmap+0x13a/0x150
Nov 17 17:20:34 recvpc kernel: eax: fffffffd ebx: 00000000 ecx: dc71efbc
edx: dffebcc0
Nov 17 17:20:34 recvpc kernel: esi: 00000000 edi: dffca040 ebp: 0000000b
esp: d53ede64
Nov 17 17:20:34 recvpc kernel: ds: 007b es: 007b ss: 0068
Nov 17 17:20:34 recvpc kernel: Process omkmigd (pid: 6379, threadinfo=d53ec000
task=dc79baa0)
Nov 17 17:20:34 recvpc kernel: Stack: dc71ea14 dc71eb1c 00000000 00000000
ffffffff d53ede80 00000000 00000045
Nov 17 17:20:34 recvpc kernel: c03dd1ec dffca040 dc79baa0 00000001
c011d386 dffca040 d53ec000 c0121ed5
Nov 17 17:20:34 recvpc kernel: dffca040 d53edf14 dc79bf04 dc79baa0
d53ec000 0000000b d53ec000 d53edf14
Nov 17 17:20:34 recvpc kernel: Call Trace:
Nov 17 17:20:34 recvpc kernel: [<c011d386>] mmput+0x26/0x70
Nov 17 17:20:34 recvpc kernel: [<c0121ed5>] do_exit+0xe5/0x3a0
Nov 17 17:20:34 recvpc kernel: [<c0122204>] do_group_exit+0x34/0x70
Nov 17 17:20:34 recvpc kernel: [<c012c1bd>] get_signal_to_deliver+0x1cd/0x2c0
Nov 17 17:20:34 recvpc kernel: [<c010425f>] do_signal+0x6f/0x180
Nov 17 17:20:34 recvpc kernel: [<c0104542>] restore_nocheck+0x7/0xd
Nov 17 17:20:34 recvpc kernel: [<c0104543>] restore_nocheck+0x8/0xd
Nov 17 17:20:34 recvpc kernel: [<c0104547>] restore_nocheck+0xc/0xd
Nov 17 17:20:34 recvpc kernel: [<c01321e0>] search_exception_tables+0x20/0x40
Nov 17 17:20:34 recvpc kernel: [<c0104547>] restore_nocheck+0xc/0xd
Nov 17 17:20:34 recvpc kernel: [<c011a6e3>] fixup_exception+0x13/0x30
Nov 17 17:20:34 recvpc kernel: [<c0104547>] restore_nocheck+0xc/0xd
Nov 17 17:20:34 recvpc kernel: [<c0307bd8>] do_general_protection+0xa8/0x1e0
Nov 17 17:20:34 recvpc kernel: [<c01fe633>] arch_kickstart+0xa3/0xb0
Nov 17 17:20:34 recvpc kernel: [<c0105ea0>] do_iret_error+0x0/0xc0
Nov 17 17:20:34 recvpc kernel: [<c0104397>] do_notify_resume+0x27/0x38
Nov 17 17:20:34 recvpc kernel: [<c010459e>] work_notifysig+0x13/0x19
Nov 17 17:20:34 recvpc kernel: Code: 00 8b 5e 0c 89 34 24 e8 05 da ff ff 85 db
89 de 75 ef 8b 87 8c 00 00 00 85 c0 75 0f 83 c4 24 5b 5e 5f c3 0f 20 d8 0f 22 d8
eb a1 <0f> 0b d5 07 c5 ea 31 c0 eb e7 8d b6 00 00 00 00 8d bf 00 00 00
Nov 17 17:20:34 recvpc kernel: <1>Fixing recursive fault but reboot is needed!
Don't know, what to do - any help?
Ralf
--
-------------------------------------------------------------------------------
Ralf Oelschlägel Tel.: (030) 2093 3199
Fax.: (030) 2093 3191
Humboldt-Universität zu Berlin e-mail: [email protected]
Institut für Informatik http://www.informatik.hu-berlin.de/~oelschla
D 10099 Berlin, Unter den Linden 6 Sitz: Rudower Chausee 25 (JvNH), Raum 3.3.15
-------------------------------------------------------------------------------
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click