Windows S5R3 App 4.15 available for Beta Test

Brian Silvers
Brian Silvers
Joined: 26 Aug 05
Posts: 772
Credit: 282,700
RAC: 0

RE: Not much news. I have

Message 74897 in response to message 74896

Quote:

Not much news. I have been busy hunting Bugs that lead to more serious client errors, and for some reason I don't fully understand yet the SSE code that works fine in the MacOS Intel App give wrong results when compiled with a diferent compiler (e.g. gcc on Linux).

I'm still working on this, but the next two weeks will be very busy for basically the whole project team (for resons / deadlines unrelated to this project), I don't know how much progress will be made during these.

BM

Well, as I asked (via an @Bernd in case you are reading) in another thread, could you talk things over with Bruce and see if you can bump the deadline up to at least 18 days, preferably 21 days again, until such time as you can get the SSE optimization ported? The deadline pressure was eased quite a bit by reducing the size of the units, but the runtime didn't drop to 1/3 of the big S5R2 results, only to like around 2/5 to 1/2 (40-50%).

Brian

ErichZann
ErichZann
Joined: 11 Feb 05
Posts: 120
Credit: 81,582
RAC: 0

RE: Not much news. I have

Message 74898 in response to message 74896

Quote:

Not much news. I have been busy hunting Bugs that lead to more serious client errors, and for some reason I don't fully understand yet the SSE code that works fine in the MacOS Intel App give wrong results when compiled with a diferent compiler (e.g. gcc on Linux).

I'm still working on this, but the next two weeks will be very busy for basically the whole project team (for resons / deadlines unrelated to this project), I don't know how much progress will be made during these.

BM

Ok, thanks for the answer!

Mikie Tim T
Mikie Tim T
Joined: 22 Jan 05
Posts: 103
Credit: 207,633,459
RAC: 0

RE: Not much news. I have

Message 74899 in response to message 74896

Quote:

Not much news. I have been busy hunting Bugs that lead to more serious client errors, and for some reason I don't fully understand yet the SSE code that works fine in the MacOS Intel App give wrong results when compiled with a diferent compiler (e.g. gcc on Linux).

I'm still working on this, but the next two weeks will be very busy for basically the whole project team (for resons / deadlines unrelated to this project), I don't know how much progress will be made during these.

BM

Anything to report now? We were generous enough to give you an extra day after the 2 weeks you blocked off your schedule before starting to hound you again ;-)

Steve Hartel
Steve Hartel
Joined: 4 Mar 05
Posts: 1
Credit: 6,408,240
RAC: 23,332

My work units are finishing

My work units are finishing in about 14 seconds. Here are the details of the most recent one. I've tried resetting the project, but that didn't help. What should I try next?

Task details
Task ID 90273258
Name h1_0716.95_S5R2__267_S5R3a_0
Workunit 36152059
Created 28 Dec 2007 0:56:34 UTC
Sent 28 Dec 2007 0:56:35 UTC
Received 28 Dec 2007 1:56:58 UTC
Server state Over
Outcome Client error
Client state Compute error
Exit status 99 (0x63)
Computer ID 84427
Report deadline 11 Jan 2008 0:56:35 UTC
CPU time 5.375
stderr out 5.10.28

- exit code 99 (0x63)

2007-12-27 20:55:42.5955 [normal]: Start of BOINC application 'projects/einstein.phys.uwm.edu/einstein_S5R3_4.11_windows_intelx86.exe'.
2007-12-27 20:55:42.8142 [normal]: INFO: OPENGL32.DLL loaded
2007-12-27 20:55:42.8142 [normal]: INFO: GLU32.DLL loaded
2007-12-27 20:55:44.9080 [debug]: Reading SFTs and setting up stacks ... Ending frequency 717.267677 not contained in SFT (sequence)
(expected bin: 1291082, last bin in SFT sequence: 1291049)
Level 0: $Id: HierarchicalSearch.c,v 1.184 2007/09/28 21:26:58 reinhard Exp $
Function call `SetUpSFTs( &status, &stackMultiSFT, &stackMultiNoiseWeights, &stackMultiDetStates, &usefulParams)' failed.
file HierarchicalSearch.c, line 662
2007-12-27 20:55:50.4705 [normal]:
Level 1: $Id: HierarchicalSearch.c,v 1.184 2007/09/28 21:26:58 reinhard Exp $
2007-12-27 20:55:50.4705 [normal]: Status code -1: Recursive error
2007-12-27 20:55:50.4861 [normal]: function SetUpSFTs, file HierarchicalSearch.c, line 1428
2007-12-27 20:55:50.4861 [normal]:
Level 2: $Id: SFTfileIO.c,v 1.126 2007/08/27 14:13:17 bema Exp $
2007-12-27 20:55:50.4861 [normal]: Status code -1: Recursive error
2007-12-27 20:55:50.4861 [normal]: function LALLoadMultiSFTs, file SFTfileIO.c, line 1063
2007-12-27 20:55:50.4861 [normal]:
Level 3: $Id: SFTfileIO.c,v 1.126 2007/08/27 14:13:17 bema Exp $
2007-12-27 20:55:50.5017 [normal]: Status code 2: Error in file-IO
2007-12-27 20:55:50.5017 [normal]: function LALLoadSegmentedSFTs, file SFTfileIO.c, line 869
2007-12-27 20:55:50.5017 [CRITICAL]: BOINC_LAL_ErrHand(): now calling boinc_finish()

]]>

Validate state Invalid
Claimed credit 0.0117706041152258
Granted credit 0
application version 4.11
--------------------------------------------------------------------------------

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5,196
Credit: 41,734,612,516
RAC: 44,941,576

RE: What should I try

Message 74901 in response to message 74900

Quote:
What should I try next?

Here are some suggestions based on the information you have already seen and reproduced in your message. Note that I've selected small bits of the stderr output which relate to the suggested actions.

Quote:

Task ID 90273258

- exit code 99 (0x63)

Your most recently failed task and the error code it returned. When you examine your full results list you will see that your computer seems to be failing in the same way on all tasks sent to it. You need to consult the first post in the sticky thread in the Problems and Bug Reports forum which gives more information about particular error codes - this one included.

When you read through the details about exit code 99, notice in particular the reference to "[CRITICAL]: Required frequency-bins [-8, 8] not covered by SFT-interval [...]". Your error message is "Ending frequency 717.267677 not contained in SFT (sequence) (expected bin: 1291082, last bin in SFT sequence: 1291049)" which is not the same but appears to be at least somewhat related. It's quite likely that error messages have changed with later beta versions from what was in place at the time Bernd wrote the original sticky message. Notice also that the file "SFTfileIO.c" is named in both places as well.

We could guess that Bernd may well be interested in these failures so thanks for posting this error report. Bernd is likely to see it as it's in a thread he started but perhaps for future reference you might note that the sticky thread about client errors might be a more reliable place to grab his attention.

Quote:
2007-12-27 20:55:42.5955 [normal]: Start of BOINC application 'projects/einstein.phys.uwm.edu/einstein_S5R3_4.11_windows_intelx86.exe'.

You appear to be participating in the beta test program as you are running an old version of the app. I would guess that you have an app_info.xml file which is preventing you from receiving the current recommended version. You should stop BOINC and delete the app_info.xml file. When you restart BOINC the correct version can then be downloaded. It is possible that this action on its own may get rid of your problems as the problem may have been fixed between version 4.11 and 4.15. At least it's worth a try. In any case, if you are going to use beta apps you should make sure to keep an eye out for when new versions are released and then upgrade as soon as convenient.

Quote:
2007-12-27 20:55:44.9080 [debug]: Reading SFTs and setting up stacks ... Ending frequency 717.267677 not contained in SFT (sequence)
(expected bin: 1291082, last bin in SFT sequence: 1291049)
Level 0: $Id: HierarchicalSearch.c,v 1.184 2007/09/28 21:26:58 reinhard Exp $
Function call `SetUpSFTs( &status, &stackMultiSFT, &stackMultiNoiseWeights, &stackMultiDetStates, &usefulParams)' failed.
file HierarchicalSearch.c, line 662

This would seem to be the point of failure and is obviously something that Bernd will have to look into as to why it is occurring.

In the meantime, since you have already reset the project, your best action would be to get a later app than 4.11 and see if that solves your problem.

Cheers,
Gary.

Jord
Joined: 26 Jan 05
Posts: 2,952
Credit: 5,669,412
RAC: 388

RE: It looks like a disk

Message 74902 in response to message 74893

Quote:

It looks like a disk error. It's not a corrupt data file, it's actually the checkpoint file that's broken. The zip messages are misleading, they have nothing to do with the actual error.

BM


In fact, it was my motherboard that was broken. Since the day before Christmas I have a new motherboard, same CPU, same memory, same drives (although now on the motherboard, not Promise card) and everything is very sturdy and staying up&running. I'm even not running with CPU throttling anymore and can play my games again while BOINC stays in memory. No hangs, no hiccups, no crashes. :-)

Brian Silvers
Brian Silvers
Joined: 26 Aug 05
Posts: 772
Credit: 282,700
RAC: 0

Since 4.15 is now the

Since 4.15 is now the official app, should this thread be unstickied?

Bikeman (Heinz-Bernd Eggenstein)
Bikeman (Heinz-...
Moderator
Joined: 28 Aug 06
Posts: 3,516
Credit: 454,089,539
RAC: 44,102

Indeed, unstickified in order

Indeed, unstickified in order not to confuse users :-)

Akos Fekete
Akos Fekete
Joined: 13 Nov 05
Posts: 560
Credit: 4,510,638
RAC: 0

I hope the next Win app will

I hope the next Win app will use the linear sin/cos code. :-)

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 3,935
Credit: 198,571,132
RAC: 52,935

RE: I hope the next Win app

Message 74906 in response to message 74905

Quote:
I hope the next Win app will use the linear sin/cos code. :-)


I just found a portable way to make this code work correctly. The reason why the 4.24 that features this code isn't showing any speedup is that the compiler is messing up the "hot loop" with a variable on the ordinary stack then, as he thinks he has only four registers left in the FPU. I really like to solve this problem on the source-code level, i.e. without any manually inserted assembler code, at least for generic Apps that should run on all CPUs.

BM

BM

Comment viewing options

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