Welcome to MilkyWay@home

Computational Error?


Advanced search

Message boards : Number crunching : Computational Error?
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
George Del Monte

Send message
Joined: 1 Apr 13
Posts: 2
Credit: 320
RAC: 0
1 credit badge7 year member badge
Message 58281 - Posted: 15 May 2013, 21:43:12 UTC

I'm not a geek but I'm not totally helpless, nevertheless, I have no clue what to do to to clear the fact that my Milkyway@home has computational errors. Any suggestions to get things right - or should I go to another project?
ID: 58281 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
GaryG
Avatar

Send message
Joined: 29 Aug 12
Posts: 31
Credit: 40,781,945
RAC: 0
30 million credit badge7 year member badge
Message 58284 - Posted: 15 May 2013, 23:56:39 UTC - in response to Message 58281.  

I'm not a geek but I'm not totally helpless, nevertheless, I have no clue what to do to to clear the fact that my Milkyway@home has computational errors. Any suggestions to get things right - or should I go to another project?


If you select the tasks link from your account page and then select one of the error tasks, left hand most column, it will display the results of the task. I selected a couple and in both cases the error is the same:


<core_client_version>7.0.65</core_client_version>
<![CDATA[
<message>
process exited with code 127 (0x7f, -129)
</message>
<stderr_txt>
../../projects/milkyway.cs.rpi.edu_milkyway/milkyway: error while loading shared libraries: libboinc_api.so.6: cannot open shared object file: No such file or directory

</stderr_txt>
]]>

I do not use Linux, but it could be a missing file or perhaps a permissions issue. Maybe someone who runs Linux will post with further ideas. Don't give up this is a great project.
ID: 58284 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profilemmstick
Avatar

Send message
Joined: 23 Nov 09
Posts: 29
Credit: 17,119,258
RAC: 0
10 million credit badge10 year member badge
Message 58294 - Posted: 16 May 2013, 13:35:51 UTC

I'm not sure why you'd be getting that error. It must pertain to how you installed BOINC. Using Ubuntu here, I just added a PPA that keeps up to date with the latest BOINC clients, so it will automatically upgrade to the latest BOINC over time. It worked perfectly fine after installation without needing to do anything special.

sudo add-apt-repository ppa:costamagnagianfranco/locutusofborg-ppa; sudo apt-get install boinc -y
ID: 58294 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
GregO

Send message
Joined: 29 Jun 11
Posts: 6
Credit: 620,411
RAC: 0
500 thousand credit badge8 year member badge
Message 58301 - Posted: 16 May 2013, 22:04:19 UTC - in response to Message 58281.  

Me too!
It seems since I did the upgrade from BOINC v.28 to 7.0.64(x64) I now receive a "Computation error" with every (opencl_amd_ati) GPU work unit I do.
ID: 58301 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
GregO

Send message
Joined: 29 Jun 11
Posts: 6
Credit: 620,411
RAC: 0
500 thousand credit badge8 year member badge
Message 58305 - Posted: 17 May 2013, 5:16:22 UTC - in response to Message 58301.  

Me too!
It seems since I did the upgrade from BOINC v.28 to 7.0.64(x64) I now receive a "Computation error" with every (opencl_amd_ati) GPU work unit I do.


I read another post that left me wondering if may have been the update to AMD Catalyst Version 13.4
ID: 58305 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profilemmstick
Avatar

Send message
Joined: 23 Nov 09
Posts: 29
Credit: 17,119,258
RAC: 0
10 million credit badge10 year member badge
Message 58306 - Posted: 17 May 2013, 7:08:35 UTC

I've been running Catalyst 13.4 via xorg edgers PPA with my Radeon HD 7950 without a single problem.
ID: 58306 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
ProfileSandy Sandmeyer
Avatar

Send message
Joined: 20 Aug 10
Posts: 2
Credit: 4,290,449
RAC: 0
3 million credit badge9 year member badge
Message 58315 - Posted: 17 May 2013, 19:55:19 UTC - in response to Message 58301.  

Me too!
It seems since I did the upgrade from BOINC v.28 to 7.0.64(x64) I now receive a "Computation error" with every (opencl_amd_ati) GPU work unit I do.


I'm having the same issue since updating to 7.0.64 (x86).
ID: 58315 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
grumpy

Send message
Joined: 14 Dec 07
Posts: 6
Credit: 3,146,179
RAC: 1,060
3 million credit badge10 year member badge
Message 58322 - Posted: 18 May 2013, 16:30:35 UTC
Last modified: 18 May 2013, 16:31:14 UTC

Had trouble myself when I updated to catalyst 13.4.Had to switch back to 12.8 to work with boinc 7.0.28 x64. Some say it's because openCl. Strangely I tried to switch openCl back to earlier version ( with 13.4) that did not work ( same problem). Catalyst 12.8 had and a openCl.msi installation package versus 13.4 that did not, If I remember correctly, they included (openCl) them in their catalyst installation package.

You could do a test with another project that uses the ati GPU.

Ver 13.4 was removed with the ati catalyst uninstaller.
ID: 58322 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
techbird

Send message
Joined: 15 Jan 11
Posts: 16
Credit: 40,731,537
RAC: 0
30 million credit badge9 year member badge
Message 58327 - Posted: 18 May 2013, 22:02:29 UTC

I Have the same issue. However the errors only occur on milkyway@home. Both Einstenm@home and SETI@home run fine with the 13.4 catalyst driver and boinc 7.0.28 (x64) combination. may be its a bad batch of work units, or the gpu apps for milkyway are not compatiple with this combination.
ID: 58327 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Richard Haselgrove

Send message
Joined: 4 Sep 12
Posts: 219
Credit: 448,778
RAC: 0
100 thousand credit badge7 year member badge
Message 58328 - Posted: 18 May 2013, 22:12:44 UTC - in response to Message 58327.  

I Have the same issue. However the errors only occur on milkyway@home. Both Einstenm@home and SETI@home run fine with the 13.4 catalyst driver and boinc 7.0.28 (x64) combination. may be its a bad batch of work units, or the gpu apps for milkyway are not compatiple with this combination.

It depends what the actual reported error is. SETI was plagued with "internal error: Compilation failed" from a number of drivers in the 13.2 - 13.4 range until the application was re-compiled without the "-fno-bin-llvmir" build option.
ID: 58328 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
GregO

Send message
Joined: 29 Jun 11
Posts: 6
Credit: 620,411
RAC: 0
500 thousand credit badge8 year member badge
Message 58369 - Posted: 22 May 2013, 13:43:49 UTC - in response to Message 58327.  

I allowed a batch of Seti AstroPulse v6.04 and Einstein BRP v1.34 to run and all of them were "Completed and validated" using BOINC 7.0.64 and AMD 13.4. So the problem is specific with MilkyWay. Once I read something has been addressed in this regards I'll go back in and reopen my preference to "Use AMD (ATI) GPU".
ID: 58369 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
George Del Monte

Send message
Joined: 1 Apr 13
Posts: 2
Credit: 320
RAC: 0
1 credit badge7 year member badge
Message 58401 - Posted: 25 May 2013, 16:40:57 UTC - in response to Message 58284.  

You're correct about the error messages being the same. It seems the three "computational errors" are simply the results of three sequential attempts. It means nothing to me, but attempts 1 and 2 died immediately; attempt 3 ran 0.01 seconds. The other two projects are happily buzzing along. Maybe the gods of milkyway@home are busily solving the problem, which clearly is not happening just on my computer.
ID: 58401 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Len LE/GE

Send message
Joined: 8 Feb 08
Posts: 261
Credit: 104,050,322
RAC: 0
100 million credit badge10 year member badge
Message 58402 - Posted: 25 May 2013, 19:19:19 UTC - in response to Message 58401.  
Last modified: 25 May 2013, 19:20:09 UTC

@George Del Monte

Your mw error logs are showing the following reason for your 'Computation error':

../../projects/milkyway.cs.rpi.edu_milkyway/milkyway: error while loading shared libraries: libboinc_api.so.6: cannot open shared object file: No such file or directory
ID: 58402 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
TLSI2000

Send message
Joined: 15 Mar 10
Posts: 17
Credit: 659,170,172
RAC: 328,955
500 million credit badge10 year member badgeextraordinary contributions badge
Message 58407 - Posted: 26 May 2013, 5:01:23 UTC

As I updated from AMD Catalyst 13.1 to 13.4 today, the MilkyWay work units are all failing.

Something that I did not expect, so searching the forum brought up this thread.
Running BOINC 7.0.64 on Win7 64bit AMD 69xx Cayman GPU


The error log indicates an exception, and dumps a BOINC debug trace to the log.
The error seems to occur at the end of processing.

--------------------------------------------
Example for Task: 481431579 WorkUnit: 368827958 computer: 366518


Using AMD IL kernel
Binary status (0): CL_SUCCESS
Estimated AMD GPU GFLOP/s: 2765 SP GFLOP/s, 691 DP FLOP/s
Using a target frequency of 30.0
Using a block size of 6144 with 121 blocks/chunk
Using clWaitForEvents() for polling (mode -1)
Range: { nu_steps = 320, mu_steps = 1600, r_steps = 1400 }
Iteration area: 2240000
Chunk estimate: 3
Num chunks: 4
Chunk size: 743424
Added area: 733696
Effective area: 2973696
Initial wait: 27 ms
Integration time: 37.905295 s. Average time per iteration = 118.454047 ms


Unhandled Exception Detected...

- Unhandled Exception Record -
Reason: Access Violation (0xc0000005) at address 0x000007FEDE811DCD read attempt to address 0x00000010

Engaging BOINC Windows Runtime Debugger...
ID: 58407 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
TJ

Send message
Joined: 12 Aug 09
Posts: 262
Credit: 92,574,414
RAC: 0
50 million credit badge10 year member badge
Message 58436 - Posted: 29 May 2013, 21:17:51 UTC

I have Catalyst 13.3 for a while on my rig and all MilkyWay's will error out.
Going lower got problems with installing, de-istalling and a lot of pain. Eventually the one proper installed had an app missing according to BOINC and wouldn't crunch Einstein and Albert at home.
They run smooth now, but MW don't. So I have to wait for a new good driver or an update of the MW-app.
Greetings from,
TJ
ID: 58436 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profilemikey
Avatar

Send message
Joined: 8 May 09
Posts: 2310
Credit: 396,910,366
RAC: 114,789
300 million credit badge10 year member badgeextraordinary contributions badge
Message 58441 - Posted: 30 May 2013, 11:37:37 UTC - in response to Message 58407.  

As I updated from AMD Catalyst 13.1 to 13.4 today, the MilkyWay work units are all failing.

Something that I did not expect, so searching the forum brought up this thread.
Running BOINC 7.0.64 on Win7 64bit AMD 69xx Cayman GPU


AMD is again changing their software to favor gamers and not crunchers, they are working on it but don't hold your breath! If you ever find a version that is stable and just works AND you are not a gamer, then don't upgrade unless you hear the new one is MUCH faster and has been tested by others with your configuration. I am still running version 11.9 on one of my pc's and it works here just fine.
ID: 58441 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
P. Skingle

Send message
Joined: 19 Dec 12
Posts: 36
Credit: 387,443,191
RAC: 0
300 million credit badge7 year member badge
Message 58453 - Posted: 31 May 2013, 6:01:04 UTC - in response to Message 58441.  

AMD is again changing their software to favor gamers and not crunchers, they are working on it but don't hold your breath! If you ever find a version that is stable and just works AND you are not a gamer, then don't upgrade unless you hear the new one is MUCH faster and has been tested by others with your configuration. I am still running version 11.9 on one of my pc's and it works here just fine.


I agree completely!! Updating GPU software/drivers is almost always in-favour of the gamer. 13.4 is currently on my "thou shall not pass" software list.
ID: 58453 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
TJ

Send message
Joined: 12 Aug 09
Posts: 262
Credit: 92,574,414
RAC: 0
50 million credit badge10 year member badge
Message 58462 - Posted: 31 May 2013, 21:57:27 UTC - in response to Message 58453.  

AMD is again changing their software to favor gamers and not crunchers, they are working on it but don't hold your breath! If you ever find a version that is stable and just works AND you are not a gamer, then don't upgrade unless you hear the new one is MUCH faster and has been tested by others with your configuration. I am still running version 11.9 on one of my pc's and it works here just fine.


I agree completely!! Updating GPU software/drivers is almost always in-favour of the gamer. 13.4 is currently on my "thou shall not pass" software list.

I agree too, but there was an issue with my 12.x version and Einstein@home.
You have 13.4 drivers and MilkyWay is working with no errors, do I understand that from your post correctly?
Then I must update at once.
Thanks.
Greetings from,
TJ
ID: 58462 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
swiftmallard
Avatar

Send message
Joined: 18 Jul 09
Posts: 299
Credit: 303,442,173
RAC: 1
300 million credit badge10 year member badgeextraordinary contributions badge
Message 58463 - Posted: 31 May 2013, 22:49:55 UTC

I think he means that he refuses to use 13.4. He appears to be using 12.8
ID: 58463 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
P. Skingle

Send message
Joined: 19 Dec 12
Posts: 36
Credit: 387,443,191
RAC: 0
300 million credit badge7 year member badge
Message 58467 - Posted: 1 Jun 2013, 7:09:32 UTC

<--- 13.1
ID: 58467 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
1 · 2 · Next

Message boards : Number crunching : Computational Error?

©2020 Astroinformatics Group