LINUX Beta Test App 4.55 available

Michael Karlinsky
Michael Karlinsky
Joined: 22 Jan 05
Posts: 888
Credit: 23502182
RAC: 0

I get: 5.2.8 process

I get:

5.2.8
process exited with code 22 (0x16)

2006-04-11 19:44:08.2175 [normal]: Start of BOINC application 'albert_4.55_i686-pc-linux-gnu'.
2006-04-11 19:44:08.2617 [normal]: Started search at lalDebugLevel = 0
2006-04-11 19:44:10.1300 [normal]: Checkpoint-file 'Fstat.out.ckp' not found.
2006-04-11 19:44:10.1302 [normal]: No usable checkpoint found, starting from beginning.
Detected CPU type 1

2006-04-11 19:44:10.1325 [CRITICAL]: APP DEBUG: Application caught signal 11.

2006-04-11 19:44:10.1325 [CRITICAL]: Stack trace of LAL functions in worker thread:
2006-04-11 19:44:10.1325 [CRITICAL]: TestLALDemod at line 44 of file /home/bema/einsteinathome/CFS/EaH_build_release_albert_4.55/extra_sources/lalapps-CVS/src/pulsar/FDS_isolated/CFSLALDemod_ndp_vect.c
2006-04-11 19:44:10.1326 [CRITICAL]: At lowest level status code = 0, description: NO LAL ERROR REGISTERED
2006-04-11 19:44:10.1326 [CRITICAL]: Obtained 8 stack frames for this thread.
2006-04-11 19:44:10.1326 [CRITICAL]: Use gdb command: 'info line *0xADDRESS' to print corresponding line numbers.
albert_4.55_i686-pc-linux-gnu(sighandler+0x95)[0x808abcd]
/lib/libpthread.so.0[0x4005312b]
/lib/libc.so.6[0x40088d68]
albert_4.55_i686-pc-linux-gnu(boincmain+0xc8c)[0x807f8cc]
albert_4.55_i686-pc-linux-gnu(worker+0x67)[0x808a6b3]
../../projects/einstein.phys.uwm.edu/albert_4.55_i686-pc-linux-gnu.so(_Z6foobarPv+0x22)[0x401ca6ea]
/lib/libpthread.so.0[0x400501b0]
/lib/libc.so.6(__clone+0x3a)[0x4012ae4a]

hawkeye@marge:~/BOINC/projects/einstein.phys.uwm.edu> uname -a
Linux marge 2.4.19-4GB #1 Fri Sep 13 13:14:56 UTC 2002 i686 unknown

This is SUSE 8.1 (out of the box) on PIV 1700.

Michael

Gray Handcock
Gray Handcock
Joined: 11 Mar 05
Posts: 211
Credit: 135567
RAC: 0

Hiya All Well I'm back in

Hiya All

Well I'm back in XP for the moment, to crunch those WUs in BOINC there. I am impressed with the speedup in Linux but have to say that there is still work to be done. I run a dual-boot here, so can make direct comparisons.

XP with AVG antivirus and Zone-alarm running still out-crunches my Gentoo Linux side by almost a third.

Average figures in XP are around 2 hours and 2 - 5 minutes for the big WU.

Average figures for a heavily tweaked Gentoo Linux kernel and Fluxbox (a VERY light window manager) and using Einstein 4.55 are around 3 hours and 33 - 35 minutes for same WUs (yes I have sse2 and dma turned on along with other speedups in Gentoo).

Please note I am grateful for any speedup from Linux Einestein - but if I was after pure credits I'd stick with Windows XP client.

Gray

PS - graphics worked fine both in Linux and in Windows XP - nvidia 5500FX card

tullio
tullio
Joined: 22 Jan 05
Posts: 2118
Credit: 61407735
RAC: 0

RE: hawkeye@marge:~/BOINC/p

Message 27864 in response to message 27862

Quote:

hawkeye@marge:~/BOINC/projects/einstein.phys.uwm.edu> uname -a
Linux marge 2.4.19-4GB #1 Fri Sep 13 13:14:56 UTC 2002 i686 unknown

This is SUSE 8.1 (out of the box) on PIV 1700.

Michael


Why don't you update your SuSE? You can download a 10.0 from SuSE.com. I am running albert 4.51 on an old Pentium II, SuSE 9.3, with a 45% speed increase on albert 4.40 and no problems. CPU type is 0.
Tullio
http://en.opensuse.org/Download

tekwyzrd
tekwyzrd
Joined: 25 Feb 05
Posts: 49
Credit: 2922090
RAC: 0

On a very up-to-date install

On a very up-to-date install of SuSE 10.0 I had the following on the stderr_txt info for a result:

2006-04-11 00:49:56.8056 [CRITICAL]: APP DEBUG: Application caught signal 11.
2006-04-11 00:49:56.8057 [CRITICAL]: Stack trace of LAL functions in worker thread:
2006-04-11 00:49:56.8058 [CRITICAL]: getCheckpointCounters at line 3878 of file /home/bema/einsteinathome/CFS/EaH_build_release_albert_4.55/extra_sources/lalapps-CVS/src/pulsar/FDS_isolated/ComputeFStatistic.c
2006-04-11 00:49:56.8058 [CRITICAL]: At lowest level status code = 0, description: NO LAL ERROR REGISTERED
Caught SIGABRT in graphics thread

Due to the time of it's occurrance I suspect it was the result of hitting the "show graphics" button while testing the latest boinc alpha build. Of four tasks completed with v4.55 this is the only one for which I used the "show graphics" option and the only one reporting this error.

Nothing travels faster than the speed of light with the possible exception of bad news, which obeys its own special laws.
Douglas Adams (1952 - 2001)

kami4ligo
kami4ligo
Joined: 15 Mar 05
Posts: 48
Credit: 16105651
RAC: 0

To TruX @ TruXoft : I

To TruX @ TruXoft :

I downloaded and compiled the truxoft client (tx36) from source, and found a couple of problems. I worked around all of them, but I wanted to make you aware of the problems.

I failed to find a usable mail address on your site - here my comments :
------------------------------------------------------
0.
I'm running a Debian GNU/Linux platform.
1.
The zip file expands to a boinc-5.3.12.tx35.src directory. Confusing, but harmless.
2.
The file truxoft_prefs.xml is missing from the source and the linux archives.
3.
After I got the boinc_core_release_5_3_12 form CVS and copied the tx36 files in the client/ and lib/ directories, I ran thru the compilation procedures for the boinc client.
I found two problems :
3.1
client/client_types.C failed to compiled because the function stricmp(*, *) is not defined.
I copied the function (very dumb, I know) from zip/zip/macos/source/helpers.c, into the client_types.C file, and that problem disappeared.
3.2
In client/cs_apps.C, CHAR was not defined and I changed the code to read:
.....
bool CLIENT_STATE::schedule_priority_projects() {
PROJECT *best_project = NULL;
char *best_priority = NULL, *current_priority;
....
This solved the problem.

After these changes, I ran thru the configuration / compilation routine for the boinc client again, and found no more (compilation/linkage) errors. I assume the code will run fine, and I have started to use it on one of my systems.
------------------------------------------------------
Thanks for your very valuable efforts - I'll keep you informed if anyting bad happens in the next days.

Regards.

-rg-

BenJoe
BenJoe
Joined: 16 Mar 05
Posts: 1
Credit: 1286009
RAC: 0

I have also a little

I have also a little problem:

2006. ápr. 11., kedd, 20.37.13 CEST||Suspending computation and network activity - running CPU benchmarks
2006. ápr. 11., kedd, 20.37.13 CEST|Einstein@Home|Pausing result z1_0981.5__139_S4R2a_1 (removed from memory)
2006. ápr. 11., kedd, 20.37.14 CEST||Running CPU benchmarks
2006. ápr. 11., kedd, 20.37.23 CEST||Failed to stop applications; aborting CPU benchmarks
2006. ápr. 11., kedd, 20.37.24 CEST||Resuming computation and network activity
2006. ápr. 11., kedd, 20.37.24 CEST||request_reschedule_cpus: Resuming activities
2006. ápr. 11., kedd, 20.37.24 CEST||request_reschedule_cpus: process exited
2006. ápr. 11., kedd, 20.37.24 CEST|Einstein@Home|Restarting result z1_0981.5__139_S4R2a_1 using albert version 455
2006. ápr. 11., kedd, 20.37.25 CEST||ACTIVE_TASK_SET::check_app_exited(): pid 10016 not found
2006. ápr. 12., szerda, 05.58.41 CEST||request_reschedule_cpus: process exited
2006. ápr. 12., szerda, 05.58.41 CEST|Einstein@Home|Computation for result z1_0981.5__139_S4R2a_1 finished

Boinc couldn't stop teh client to do the benchmark. This machine:
http://einsteinathome.org/host/505092

Xcamel
Xcamel
Joined: 21 Mar 05
Posts: 5
Credit: 749481774
RAC: 367931

fwiw: 4.55 is about 50%

fwiw: 4.55 is about 50% slower than 4.51 on both intel p4 & xeon, along with k6 and k8, including 1 dual-dual opteron 280. All fedora linux, > 2.6.15-xxx kernels IA32 mode (no 64 bit)

:Xcamel:

Trog Dog
Trog Dog
Joined: 25 Nov 05
Posts: 191
Credit: 541562
RAC: 0

So far one of my farm has

So far one of my farm has problems with both 4.51 and 4.55 catching signal 11. Its a P4 2Ghz running Damn Small Linux 2.0 (Dell Optiplex SX260) - following the instructions for running without graphics seems to have fixed the problem. As this is a headless system losing the graphics was no biggie.

bloed_brot
bloed_brot
Joined: 5 Apr 05
Posts: 70
Credit: 91124558
RAC: 0

RE: RE: Hi ! So far, all

Message 27870 in response to message 27860

Quote:
Quote:

Hi !

So far, all my results computed with albert_455 excepted 24588564 were successfully validated.
The message that failed reports a problem with graphics (I tried to startup the graphics, but my boincmgr doesn't correctly support graphics).


Hi,
This is no client error (outcome=success), just the validator threw your result away, because he thought it wasn't exact enough.
I got such a "0" result too today and I think the app must be modified again.

cu,
Michael

:::NB:::

My machines running on 4.55 ALL return "0" because the application is not precise enough. I do get "no client error; outcome success" but results get binned cos the validator takes them to be crapped up.

"Jamessir Bensonmam - Strange was the name of his father"

:
your thoughts - the ways :: the knowledge - your space
:

bloed_brot
bloed_brot
Joined: 5 Apr 05
Posts: 70
Credit: 91124558
RAC: 0

RE: RE: This is no client

Message 27871 in response to message 27861

Quote:
Quote:
This is no client error (outcome=success), just the validator threw your result away, because he thought it wasn't exact enough.
I got such a "0" result too today and I think the app must be modified again.
I think the reason if this problem the low-precision sin/cos lookup table.
I will suggest to Brend the usage of my interpolation algorithm.
It gives about ~2300000 times more accurate sin/cos values an the speed is about the same.

akosf, you ARE the man! So cool to have you sticking around. Thanks a million for all your hard work and contributions. You make this project rock!

:
your thoughts - the ways :: the knowledge - your space
:

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.