mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   No Prime Left Behind (https://www.mersenneforum.org/forumdisplay.php?f=82)
-   -   PRPnet rally Feb. 18th-25th (https://www.mersenneforum.org/showthread.php?t=16491)

mdettweiler 2012-02-20 22:45

The max cache is 200 (it was at 50 a moment ago but I upped it to 200, since you're planning to do offline work :smile:). The deadline is 72 hours, i.e. 3 days.

(BTW: you can check the deadline on any (recent-version) PRPnet server by looking at the pending-tests list on the [url=http://noprimeleftbehind.net:2000/all.html]stats page[/url]. It shows "Age" and "Expired" columns for all pending tests.)

AMDave 2012-02-20 23:31

If anyone sees that sod Murphy, give him a black eye from me.
None of my results are showing up with that old client from any of 8 cores.
That leaves me on the sideline in the Cheer Squad.
Yayy. Go teaaam. Gimme a pom-pom.

AMDave 2012-02-21 01:15

I got 4.3.7 working on 1 box and it is reporting some work finally.

I tried all versions from 5.0.5 back to 4.3.7 on the laptops and they all fail on the T4500 chips for both 32 bit and 64 bit options. It's a bit disappointing.
3.1.7 still works on the T4500 cores, but now the server doesn't recognise the results.
I suspect someone has got trigger happy with some compiler flags. :P
A good incentive to go back to compiling my own clients again.
It's nice when things 'just work'.

So 4 out of 22 cores isn't a complete failure.

PS - someone needs to watch what end-of-line characters they use. that 4.3.7 master ini file is awful when you look at it in Notepad. Much better in N++ though.

rogue 2012-02-21 01:59

[QUOTE=AMDave;290187]I got 4.3.7 working on 1 box and it is reporting some work finally.

I tried all versions from 5.0.5 back to 4.3.7 on the laptops and they all fail on the T4500 chips for both 32 bit and 64 bit options. It's a bit disappointing.
3.1.7 still works on the T4500 cores, but now the server doesn't recognise the results.
I suspect someone has got trigger happy with some compiler flags. :P
A good incentive to go back to compiling my own clients again.
It's nice when things 'just work'.

So 4 out of 22 cores isn't a complete failure.

PS - someone needs to watch what end-of-line characters they use. that 4.3.7 master ini file is awful when you look at it in Notepad. Much better in N++ though.[/QUOTE]

It isn't an issue of the PRPNet client, but the llr that is bundled with the package released by PrimeGrid. I don't bundle any of the testing software with PRPNet, but I do provide links to d/l them.

I never use NotePad to edit text files on Windows for the reason you give amongst many others.

AMDave 2012-02-21 02:28

rogue, my apologies. I know it's not you and the llr is bundled. I'ts generally agreed that the suspect here is the LLR binary and as I posted higher up, I tried others and they also failed. However the llr that was bundled with 3.1.7 prpnetclient and earlier did work. Sometime since then, someone upstream has likely used some flags that the T4500's don't support. Unfortunately the 3.1.7 client package is not posting the results successfully with the prpnet server, which is to be expected over time. I could try pulling that old llr forward into the 5.0.5 prpnetclient, but I'm over it and now the laptops are doing something else.

Re-the text file, glad to hear it. :) but someone messed up your master ini around the start and stop options in the 4.3.7 windows 7z on Lennart's downloads page. It is fine in the more recent versions. I mention it if anyone else like me has to go back and use that version then maybe Lennart can fix it and repackage the 7z file.

It all works for everyone else. I'm just having a bad run. Ignore me.

Mark Doom 2012-02-21 04:22

So, I'm new here to NPLB, however am excited to be participating in this rally :-)

Anyways, I just wanted to throw a question out there... my first WUs just completed and stats updated for the first hour I returned WUs, and it has a blank team name for me. I noticed this for a few other of my "teammates" as well.

My .ini file properly has team name in it. Is there something else I am missing?

Neo 2012-02-21 05:06

Back to back primes!

Nice job PCZ. :smile:

Neo
AtP

MyDogBuster 2012-02-21 05:37

[QUOTE]So, I'm new here to NPLB, however am excited to be participating in this rally :-)

Anyways, I just wanted to throw a question out there... my first WUs just completed and stats updated for the first hour I returned WUs, and it has a blank team name for me. I noticed this for a few other of my "teammates" as well.

My .ini file properly has team name in it. Is there something else I am missing?[/QUOTE]

We have a separate stats DB other than what is in PRPNet. I have added the following to Aggie The Pew based on their PRPNet team choice. The stats should reflect this at the next update.

Caravaggio
KD7LRJ
Mark_Doom
Gary_Craig
meiligo

AMDave 2012-02-21 05:55

[QUOTE=Neo;290215]Back to back primes!
Nice job PCZ. :smile:
Neo
AtP[/QUOTE]
Absolutely. Great to see a Free-DC representative again.
And we have some new names and teams this time.
Excellent. Crunch On!

mdettweiler 2012-02-21 06:18

[QUOTE=AMDave;290195]rogue, my apologies. I know it's not you and the llr is bundled. I'ts generally agreed that the suspect here is the LLR binary and as I posted higher up, I tried others and they also failed. However the llr that was bundled with 3.1.7 prpnetclient and earlier did work. Sometime since then, someone upstream has likely used some flags that the T4500's don't support. Unfortunately the 3.1.7 client package is not posting the results successfully with the prpnet server, which is to be expected over time. I could try pulling that old llr forward into the 5.0.5 prpnetclient, but I'm over it and now the laptops are doing something else.

Re-the text file, glad to hear it. :) but someone messed up your master ini around the start and stop options in the 4.3.7 windows 7z on Lennart's downloads page. It is fine in the more recent versions. I mention it if anyone else like me has to go back and use that version then maybe Lennart can fix it and repackage the 7z file.

It all works for everyone else. I'm just having a bad run. Ignore me.[/QUOTE]
Dave, I'm not sure what's up with this but the problem definitely seems to be with the LLR included with the latest clients (perhaps the particular build, or particular version)--as you discovered yourself. Since the LLR bundled with 3.1.7 worked before (presumably on these same computers?), that one should definitely work again with the latest client--all PRPnet versions are fully cross-compatible with all command-line LLR versions (Linux llr and Windows cllr.exe).

That particular error you got, "Number sent to gwsetup is too large...", is, interestingly enough, the exact same error some fellows over at PrimeGrid were reporting with AMD Bulldozer machines on the latest development version of LLR--though in their case it was due to the latest development verison adding AVX support for a speed boost on Sandy Bridge and Bulldozer CPUs, but not (yet) properly taking into account differences in how Bulldozer implements AVX. I don't think I've ever heard of this error being thrown up by a pre-AVX machine, least of all with a non-development LLR build, but nonetheless it is what you're seeing. It would almost seem to indicate that somehow LLR doesn't know how to handle your CPUs, which is strange because a T4500 is a pretty standard laptop-spec Core 2 Duo, which [I]should[/I] be a no-brainer for LLR.

AMDave 2012-02-21 06:24

[QUOTE=MyDogBuster;290218]We have a separate stats DB other than what is in PRPNet. I have added the following to Aggie The Pew based on their PRPNet team choice. The stats should reflect this at the next update.

Caravaggio
KD7LRJ
Mark_Doom[/QUOTE]
I saw the accidental '-' at the end of the team name for KD7LRJ but someone fixed it faster than I could get to it. (Good shooting whoever that was)
The team total will be corrected automatically by the database on the next refresh. :smile:

@Max. Thanks. Yup. Used to work on same hardware. As above I suspect upstream LLR sources have made an error. Never mind. There is always next time. :)


All times are UTC. The time now is 10:09.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.