6.2 freebsd port unable to get work


log in

Advanced search

Questions and Answers : Unix/Linux : 6.2 freebsd port unable to get work

Author Message
void
Send message
Joined: 2 Oct 99
Posts: 4
Credit: 95,920
RAC: 0
Italy
Message 743944 - Posted: 25 Apr 2008, 13:04:48 UTC

Regards all,
i been struggling against this problem for some weeks now, reading forums, howtos, documentation and lists, but to no avail ...
i installed /usr/ports/astro/boinc-setiathome-enhanced on FreeBSD 6.2-STABLE with no problem, it also installed boinc_client and boinc_cmd fine ...
started the client with --attach_project setiathome and ok ... got my account, and stuff ...
then the same error message again and again :

[SETI@home]Message from server: platform 'i386-portbld-freebsd' not found

read this thread

http://setiathome.berkeley.edu/forum_thread.php?id=44545

but stopping it all and relaunching the client with

/usr/local/bin/boinc_client -update_prefs http://setiathome.berkeley.edu

did absolutely nothing ...
the client just hangs there uselessly forever ... tried waking it up with

boinc_cmd --get_project_config http://setiathome.berkeley.edu/

and it just said

[---] Fetching configuration file from http://setiathome.berkeley.edu/get_project_config.php

doing nothing ... so i tried

boinc_cmd --project http://setiathome.berkeley.edu/ reset

and it said 'resetting project', again doing nothing more

#

last, i tried recompiling both boinc client and setiathome app with

--with-boinc-platform=i686-pc-freebsd

as found here

http://boinc.berkeley.edu/trac/wiki/BoincPlatforms

, only to see it say again

[SETI@home]Message from server: platform 'i686-pc-freebsd' not found

what ??? ok, it just doesn't bark if i recompile it all with platform switch 'i686-pc-linux-gnu', but i honestly can't run linux emu on all of my servers just to have SETI run ...

any suggestions ? i'm getting mad on this matter and not contributing my many many spare cylces here ...

thanx all !

Silvio
surt at tzone dot it
____________

Dotsch
Volunteer tester
Avatar
Send message
Joined: 9 Jun 99
Posts: 2422
Credit: 847,669
RAC: 34
Germany
Message 744104 - Posted: 25 Apr 2008, 18:41:13 UTC

Please verify if in the BOINC direcotry under projects/setiathome.berkeley.edu is a app_info.xml and setiathome-5..., if not reinstall the SETI application again.

Please start the BOINC client with "./boinc_client", from a other terminal cd to the BOINC directory and issue a "./boinc_cmd --project http://setiathome.berekeley.edu update". - Did not change something ?

Could you please post the messages from the BOINC client here.
____________

void
Send message
Joined: 2 Oct 99
Posts: 4
Credit: 95,920
RAC: 0
Italy
Message 744447 - Posted: 26 Apr 2008, 12:45:36 UTC - in response to Message 744104.

Please verify if in the BOINC direcotry under projects/setiathome.berkeley.edu is a app_info.xml and setiathome-5..., if not reinstall the SETI application again.


Thanx Dotsch, strong hint there ... in fact, it seems that the port system dependency is broken here : not just astro/boinc-setiathome-enhanced depends on net/boinc-client, but it actually dependes on boinc-client_run_once_and_attacched_to_setiathome, given that the install directory of the seti binary and app file doesn't exist previous to this ... an alternative could be to have the seti port forcibly create a $BOINCDIR/projects/setiathome.berkeley.edu/ dir where to install its essential ... great suggestion anyway, thanx again ...
#
then on to problem two ;)
here's the output of the latest run :

architeuthis# su - boinc -c /usr/local/bin/boinc_client
2008-04-26 03:30:03 [---] Starting BOINC client version 5.8.15 for i386-portbld-freebsd
2008-04-26 03:30:03 [---] log flags: task, file_xfer, sched_ops
2008-04-26 03:30:03 [---] Libraries: libcurl/7.16.1 zlib/1.2.3
2008-04-26 03:30:03 [---] Data directory: /var/db/boinc
2008-04-26 03:30:03 [SETI@home] Found app_info.xml; using anonymous platform
2008-04-26 03:30:03 [---] Processor: 1 i386 Intel(R) Celeron(TM) CPU 1300MHz [] [sse mmx]
2008-04-26 03:30:03 [---] Memory: 250.47 MB physical, 0 bytes virtual
2008-04-26 03:30:03 [---] Disk: 32.24 GB total, 26.84 GB free
2008-04-26 03:30:03 [SETI@home] URL: http://setiathome.berkeley.edu/; Computer ID: 2034538; location: home; project prefs: default
2008-04-26 03:30:03 [---] General prefs: from SETI@home (last modified 2005-09-01 11:21:15)
2008-04-26 03:30:03 [---] Host location: home
2008-04-26 03:30:03 [---] General prefs: no separate prefs for home; using your defaults
2008-04-26 03:30:03 [SETI@home] Starting 24mr08aj.12788.14387.16.8.57_1
SIGSYS: system call given invalid argument

now the 'anonymous platform system' sets correctly in, the workunit is there to be processed, but eventually the binary itself barks himself out to death ...
here's the environment :

architeuthis# pwd
/var/db/boinc
architeuthis# ls -la
total 52
drwxr-xr-x 4 boinc nobody 512 Apr 26 03:30 .
drwxr-xr-x 10 root wheel 512 Aug 20 2007 ..
-rw-r--r-- 1 boinc nobody 2242 Apr 26 03:29 account_setiathome.berkeley.edu.xml
-rw-r--r-- 1 boinc nobody 8377 Apr 26 03:30 client_state.xml
-rw-r--r-- 1 boinc nobody 8377 Apr 26 03:29 client_state_prev.xml
-rw-r--r-- 1 boinc nobody 722 Apr 26 03:29 global_prefs.xml
-rw------- 1 boinc nobody 32 Apr 26 02:14 gui_rpc_auth.cfg
-rw-r--r-- 1 boinc nobody 0 Apr 26 02:14 lockfile
-rw-r--r-- 1 boinc nobody 151 Apr 26 02:14 master_setiathome.berkeley.edu.xml
drwxr-xr-x 3 boinc nobody 512 Apr 26 02:14 projects
-rw-r--r-- 1 boinc nobody 7029 Apr 26 03:29 sched_reply_setiathome.berkeley.edu.xml
-rw-r--r-- 1 boinc nobody 3411 Apr 26 03:28 sched_request_setiathome.berkeley.edu.xml
drwxrwx--x 3 boinc nobody 512 Apr 26 03:29 slots
-rw-r--r-- 1 boinc nobody 422 Apr 26 03:29 statistics_setiathome.berkeley.edu.xml
architeuthis# ls -la projects/setiathome.berkeley.edu/
total 2292
drwxr-xr-x 2 boinc nobody 512 Apr 26 03:29 .
drwxr-xr-x 3 boinc nobody 512 Apr 26 02:14 ..
-rw-r--r-- 1 boinc nobody 375340 Apr 26 03:29 24mr08aj.12788.14387.16.8.57
-r--r--r-- 1 boinc nobody 462 Apr 26 02:35 app_info.xml
-rw-r--r-- 1 boinc nobody 52779 Apr 26 02:14 arecibo_181.png
-rw-r--r-- 1 boinc nobody 2536 Apr 26 02:14 sah_40.png
-rw-r--r-- 1 boinc nobody 25488 Apr 26 02:14 sah_banner_290.png
-rw-r--r-- 1 boinc nobody 35399 Apr 26 02:14 sah_ss_290.png
-r-xr-xr-x 1 boinc nobody 1786284 Apr 26 02:35 setiathome-5.13.i386-portbld-freebsd
-rw-r--r-- 1 boinc nobody 71 Apr 26 03:29 slideshow_setiathome_enhanced_00
-rw-r--r-- 1 boinc nobody 72 Apr 26 03:29 slideshow_setiathome_enhanced_01
-rw-r--r-- 1 boinc nobody 75 Apr 26 03:29 slideshow_setiathome_enhanced_02
-rw-r--r-- 1 boinc nobody 67 Apr 26 03:29 stat_icon

the binary itself is architecture-wise correctly built
architeuthis# file projects/setiathome.berkeley.edu/setiathome-5.13.i386-portbld-freebsd
projects/setiathome.berkeley.edu/setiathome-5.13.i386-portbld-freebsd: ELF 32-bit LSB executable, Intel 80386, version 1 (FreeBSD), for FreeBSD 6.2 (602107), dynamically linked (uses shared libs), stripped


the situation doesn't change whether boinc client is run as boinc or root, and i haven't found absolutely no reference to this in the threads ...

maybe a library / path issue, in your opinion ?

Silvio
surt at tzone dot it
____________

Dotsch
Volunteer tester
Avatar
Send message
Joined: 9 Jun 99
Posts: 2422
Credit: 847,669
RAC: 34
Germany
Message 744473 - Posted: 26 Apr 2008, 14:07:34 UTC

Could you please do a "ldd setiathome-5...." to verify that all libs are installed and found.
Btw. it shows setiathome-5.13 which is a older, but still usable, version. So far I know is there also a setiathome 5.27 or 5.28 available from the FreeBSD ports. I recommend you to use 5.2x.
____________

void
Send message
Joined: 2 Oct 99
Posts: 4
Credit: 95,920
RAC: 0
Italy
Message 745769 - Posted: 29 Apr 2008, 12:56:07 UTC - in response to Message 744473.

Could you please do a "ldd setiathome-5...." to verify that all libs are installed and found.
Btw. it shows setiathome-5.13 which is a older, but still usable, version. So far I know is there also a setiathome 5.27 or 5.28 available from the FreeBSD ports. I recommend you to use 5.2x.


Hi Dotsch, and thanx for your patience ...
i been doing some more research on this SIGSYS problem, and that's what i found out :

first of all, it not setiathome's problem, but boinc's : in standalone mode, setiathome-5.27.i386-portbld-freebsd [nu version, nice suggestion ;)] works like charm

Can't open init data file - running in standalone mode
setiathome_enhanced 5.27 ${gcc_version_string}

Work Unit Info:
...............
WU true angle range is : 3.006426
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)

v_GetPowerSpectrum 0.00610 0.00000 test
v_vGetPowerSpectrum 0.00518 0.00000 test
v_vGetPowerSpectrum2 0.00757 0.00000 test
v_vGetPowerSpectrumUnrolled 0.00667 0.00000 test
v_vGetPowerSpectrumUnrolled2 0.00596 0.00000 test
v_vGetPowerSpectrum 0.00518 0.00000 choice

v_ChirpData 0.35504 0.00000 test
fpu_ChirpData 0.16324 0.00001 test
sse1_ChirpData_ak 0.12132 0.00000 test
sse1_ChirpData_ak 0.12132 0.00000 choice

v_Transpose 0.19564 0.00000 test
v_Transpose2 0.10489 0.00000 test
v_Transpose4 0.06057 0.00000 test
v_Transpose8 0.09930 0.00000 test
v_pfTranspose2 0.10574 0.00000 test
v_pfTranspose4 0.06076 0.00000 test
v_pfTranspose8 0.10321 0.00000 test
v_vTranspose4 0.06079 0.00000 test
v_vTranspose4np 0.06041 0.00000 test
v_vTranspose4ntw 0.04052 0.00000 test
v_vTranspose4x8ntw 0.03939 0.00000 test
v_vTranspose4x16ntw 0.05344 0.00000 test
v_vpfTranspose8x4ntw 0.04000 0.00000 test
v_vTranspose4x8ntw 0.03939 0.00000 choice

FPU opt folding 0.00742 0.00000 test
AK SSE folding 0.00631 0.00000 test
BH SSE folding 0.00610 0.00000 test
BH SSE folding 0.00610 0.00000 choice


second, ldd saiz all fine :

ldd /usr/bin/boinc_client
/usr/bin/boinc_client:
libssl.so.4 => /usr/lib/libssl.so.4 (0x280d7000)
libcrypto.so.4 => /lib/libcrypto.so.4 (0x28106000)
libcurl.so.4 => /usr/local/lib/libcurl.so.4 (0x281fe000)
libz.so.3 => /lib/libz.so.3 (0x2822d000)
libstdc++.so.5 => /usr/lib/libstdc++.so.5 (0x2823e000)
libpthread.so.2 => /usr/lib/libpthread.so.2 (0x28311000)
libm.so.4 => /lib/libm.so.4 (0x28335000)
libc.so.6 => /lib/libc.so.6 (0x2834e000)


The only reference i found out to the sigsys affaire is in an einsteinathome forum, but nothing conclusive emerges there, too ...
http://einstein.phys.uwm.edu/forum_thread.php?id=4969

still searching ...

Silvio
surt at tzone dot it
____________

void
Send message
Joined: 2 Oct 99
Posts: 4
Credit: 95,920
RAC: 0
Italy
Message 745777 - Posted: 29 Apr 2008, 13:30:28 UTC - in response to Message 745769.


me again ...
i switched on every possible debug flag in cc_config, and here's where it dies

29-Apr-2008 15:19:04 [---] Resuming computation
29-Apr-2008 15:19:04 [---] [cpu_sched_debug] Request CPU reschedule: Resuming computation
29-Apr-2008 15:19:04 [---] [cpu_sched_debug] Request CPU reschedule: Scheduling period elapsed.
29-Apr-2008 15:19:04 [---] [cpu_sched_debug] schedule_cpus(): start
29-Apr-2008 15:19:04 [---] [rr_sim] rr_sim start: work_buf_total 21600.864000 rrs 100.000000 trs 100.000000 ncpus 1
29-Apr-2008 15:19:04 [SETI@home] [rr_sim] set_rrsim_proc_rate: 0.999965 (rrs 100.000000, rs 100.000000, nactive 1, ocf 0.999965
29-Apr-2008 15:19:04 [SETI@home] [rr_sim] result 25mr08aa.23189.105748.8.8.76_0 finishes after 78013.692985 (78010.926118/0.999965)
29-Apr-2008 15:19:04 [SETI@home] [rr_sim] decr rrs by 100.000000, new value 0.000000
29-Apr-2008 15:19:04 [---] [rr_sim] total: idle cpus 0, last active 1, active 0, shortfall 0.000000
29-Apr-2008 15:19:04 [---] [rr_sim] proj SETI@home: last active 1, active 0, shortfall 0.000000
29-Apr-2008 15:19:04 [---] [rr_sim] done; total shortfall 0.000000
29-Apr-2008 15:19:04 [SETI@home] [cpu_sched_debug] highest debt: 0.000000 25mr08aa.23189.105748.8.8.76_0
29-Apr-2008 15:19:04 [---] [cpu_sched_debug] scheduling (regular) 25mr08aa.23189.105748.8.8.76_0
29-Apr-2008 15:19:04 [---] [cpu_sched_debug] Request enforce CPU schedule: schedule_cpus
29-Apr-2008 15:19:04 [---] [state_debug] set dirty: schedule_cpus
29-Apr-2008 15:19:04 [---] [poll_debug] CLIENT_STATE::poll_slow_events(): possibly_schedule_cpus
29-Apr-2008 15:19:04 [---] [cpu_sched_debug] enforce_schedule(): start
29-Apr-2008 15:19:04 [SETI@home] [cpu_sched_debug] want to run: 25mr08aa.23189.105748.8.8.76_0
29-Apr-2008 15:19:04 [---] [mem_usage_debug] enforce: available RAM 225.43MB
29-Apr-2008 15:19:04 [SETI@home] [cpu_sched_debug] processing 25mr08aa.23189.105748.8.8.76_0
29-Apr-2008 15:19:04 [---] [cpu_sched_debug] finished preempt loop, nrunning 1
29-Apr-2008 15:19:04 [SETI@home] [cpu_sched_debug] 25mr08aa.23189.105748.8.8.76_0 sched state 0 next 2 task state 0
29-Apr-2008 15:19:04 [SETI@home] Starting 25mr08aa.23189.105748.8.8.76_0
29-Apr-2008 15:19:04 [SETI@home] [cpu_sched] Starting 25mr08aa.23189.105748.8.8.76_0 (initial)
SIGSYS: system call given invalid argument

Exiting...


not much yet, but still searching ...

Silvio
silvio at tzone dot it
____________

Dotsch
Volunteer tester
Avatar
Send message
Joined: 9 Jun 99
Posts: 2422
Credit: 847,669
RAC: 34
Germany
Message 746486 - Posted: 1 May 2008, 11:37:04 UTC

I have no idea in the moment, what cause this problem. Evtentualy a ktrace would be helpfull. Could you please start the BOINC Client with "ktrace ./boinc_client" an mail me the file kdump.out at seti underscore boinc at dotsch dot de
I will notify the FreeBSD SETI and BOINC port maintainers about this problem, too.
____________

Questions and Answers : Unix/Linux : 6.2 freebsd port unable to get work

Copyright © 2014 University of California