Running 6.23 under wine on Ubuntu 10.10 AMD64

Moderators: slegrand, Site Moderators, PandeGroup

Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby davidcoton » Sun Jan 30, 2011 2:43 pm

I've folded lots on Windows, but this is my first attempt under Linux

PC is an Asus MoBo , Athlon II X4 640, 4GB RAM. Video is an NVidia GTS250.

I've installed Ubuntu 10.10 (AMD64) and Wine, then the Cuda 3.2 toolkit (Is this version OK?) and Folding@home-Win32-NV-GPU-systray-623.msi under wine.

SMP is folding nicely, but GPU is different...

Well, it seems nearly to work ... FAH starts, downloads a work unit and the core ... and falls over.

Code: Select all

--- Opening Log file [January 30 14:11:29 UTC]





# Windows GPU Console Edition #################################################

###############################################################################



                       Folding@Home Client Version 6.23



                          http://folding.stanford.edu



###############################################################################

###############################################################################



Launch directory: C:\Program Files\Folding@home\Folding@home-gpu

Arguments: -config



[14:15:57] - Ask before connecting: No

[14:15:57] - User name: davidcoton (Team 0)

[14:15:57] - User ID not found locally

[14:15:57] + Requesting User ID from server

[14:15:57] - Getting ID from AS:

[14:15:57] Connecting to http://assign-GPU.stanford.edu:8080/

[14:15:58] Posted data.

[14:15:58] Initial: 9206; - Received User ID =xxxxxxxxx
[14:15:58] - Machine ID: 5

[14:15:58]

[14:15:58] Work directory not found. Creating...

[14:15:58] Could not open work queue, generating new queue...

[14:15:58] Initialization complete

[14:15:58] - Preparing to get new work unit...

[14:15:58] + Attempting to get work packet

[14:15:58] - Will indicate memory of 3960 MB

[14:15:58] - Detect CPU. Vendor: AuthenticAMD, Family: 15, Model: 5, Stepping: 3

[14:15:58] - Connecting to assignment server

[14:15:58] Connecting to http://assign-GPU.stanford.edu:8080/

[14:15:58] - Autosending finished units... [January 30 14:15:58 UTC]

[14:15:58] Trying to send all finished work units

[14:15:58] + No unsent completed units remaining.

[14:15:58] - Autosend completed

[14:15:59] Posted data.

[14:15:59] Initial: 43AB; - Successful: assigned to (171.67.108.21).

[14:15:59] + News From Folding@Home: Welcome to Folding@Home

[14:15:59] Loaded queue successfully.

[14:15:59] Connecting to http://171.67.108.21:8080/

[14:16:00] Posted data.

[14:16:00] Initial: 0000; - Receiving payload (expected size: 65479)

[14:16:01] - Downloaded at ~63 kB/s

[14:16:01] - Averaged speed for that direction ~63 kB/s

[14:16:01] + Received work.

[14:16:01] + Closed connections

[14:16:01]

[14:16:01] + Processing work unit

[14:16:01] Core required: FahCore_11.exe

[14:16:01] Core not found.

[14:16:01] - Core is not present or corrupted.

[14:16:01] - Attempting to download new core...

[14:16:01] + Downloading new core: FahCore_11.exe

[14:16:01] Downloading core (/~pande/Win32/x86/NVIDIA/G80/Core_11.fah from www.stanford.edu)

[14:16:02] Initial: AFDE; + 10240 bytes downloaded

[14:16:02] Initial: E027; + 20480 bytes downloaded

[14:16:02] Initial: 8EE5; + 30720 bytes downloaded

[14:16:02] Initial: 9838; + 40960 bytes downloaded

[14:16:03] Initial: D615; + 51200 bytes downloaded

[14:16:03] Initial: CAA5; + 61440 bytes downloaded

[14:16:03] Initial: 3354; + 71680 bytes downloaded

[14:16:03] Initial: 6FF7; + 81920 bytes downloaded

[14:16:03] Initial: 5BA3; + 92160 bytes downloaded

[14:16:03] Initial: 23F0; + 102400 bytes downloaded

[14:16:03] Initial: C043; + 112640 bytes downloaded

[14:16:04] Initial: 2DB2; + 122880 bytes downloaded

[14:16:04] Initial: E434; + 133120 bytes downloaded

[14:16:04] Initial: 321F; + 143360 bytes downloaded

[14:16:04] Initial: 93B0; + 153600 bytes downloaded

[14:16:04] Initial: 245F; + 163840 bytes downloaded

[14:16:04] Initial: 6150; + 174080 bytes downloaded

[14:16:04] Initial: A664; + 184320 bytes downloaded

[14:16:05] Initial: 6EF8; + 194560 bytes downloaded

[14:16:05] Initial: 65A4; + 204800 bytes downloaded

[14:16:05] Initial: F31D; + 215040 bytes downloaded

[14:16:06] Initial: AD3D; + 225280 bytes downloaded

[14:16:06] Initial: FEF9; + 235520 bytes downloaded

[14:16:07] Initial: 9A18; + 245760 bytes downloaded

[14:16:08] Initial: F515; + 256000 bytes downloaded

[14:16:08] Initial: 7CD9; + 266240 bytes downloaded

[14:16:09] Initial: 0905; + 276480 bytes downloaded

[14:16:09] Initial: D379; + 286720 bytes downloaded

[14:16:09] Initial: 4E51; + 296960 bytes downloaded

[14:16:10] Initial: B099; + 307200 bytes downloaded

[14:16:10] Initial: 4645; + 317440 bytes downloaded

[14:16:10] Initial: D03A; + 327680 bytes downloaded

[14:16:10] Initial: 1144; + 337920 bytes downloaded

[14:16:10] Initial: 9388; + 348160 bytes downloaded

[14:16:10] Initial: 34A4; + 358400 bytes downloaded

[14:16:10] Initial: BD67; + 368640 bytes downloaded

[14:16:11] Initial: 0594; + 378880 bytes downloaded

[14:16:11] Initial: AFD1; + 389120 bytes downloaded

[14:16:11] Initial: 1A2A; + 399360 bytes downloaded

[14:16:11] Initial: 59F1; + 409600 bytes downloaded

[14:16:11] Initial: 2512; + 419840 bytes downloaded

[14:16:11] Initial: 74B1; + 430080 bytes downloaded

[14:16:12] Initial: 94F2; + 440320 bytes downloaded

[14:16:12] Initial: D8E0; + 450560 bytes downloaded

[14:16:12] Initial: 01A4; + 460800 bytes downloaded

[14:16:13] Initial: BCAF; + 471040 bytes downloaded

[14:16:13] Initial: B064; + 481280 bytes downloaded

[14:16:13] Initial: 3236; + 491520 bytes downloaded

[14:16:13] Initial: 3291; + 501760 bytes downloaded

[14:16:14] Initial: 6A7F; + 512000 bytes downloaded

[14:16:14] Initial: 61F7; + 522240 bytes downloaded

[14:16:14] Initial: F17C; + 532480 bytes downloaded

[14:16:14] Initial: 4E88; + 542720 bytes downloaded

[14:16:14] Initial: 7111; + 552960 bytes downloaded

[14:16:14] Initial: 1568; + 563200 bytes downloaded

[14:16:14] Initial: 9B04; + 573440 bytes downloaded

[14:16:15] Initial: 783B; + 583680 bytes downloaded

[14:16:15] Initial: 2FC8; + 593920 bytes downloaded

[14:16:15] Initial: E6E7; + 604160 bytes downloaded

[14:16:15] Initial: E7AB; + 614400 bytes downloaded

[14:16:15] Initial: D677; + 624640 bytes downloaded

[14:16:15] Initial: 5C98; + 634880 bytes downloaded

[14:16:15] Initial: 7BEE; + 645120 bytes downloaded

[14:16:16] Initial: D127; + 655360 bytes downloaded

[14:16:16] Initial: CADE; + 665067 bytes downloaded

[14:16:16] Verifying core Core_11.fah...

[14:16:16] Signature is VALID

[14:16:16]

[14:16:16] Trying to unzip core FahCore_11.exe

[14:16:16] Decompressed FahCore_11.exe (1908736 bytes) successfully

[14:16:21] + Core successfully engaged

[14:16:26]

[14:16:26] + Processing work unit

[14:16:26] Core required: FahCore_11.exe

[14:16:26] Core found.

[14:16:26] Working on queue slot 01 [January 30 14:16:26 UTC]

[14:16:26] + Working ...

[14:16:26] - Calling '.\FahCore_11.exe -dir work/ -suffix 01 -priority 96 -checkpoint 15 -verbose -lifeline 8 -version 623'


[14:16:30] CoreStatus = C0000135 (-1073741515)

[14:16:30] Client-core communications error: ERROR 0xc0000135

[14:16:30] This is a sign of more serious problems, shutting down.





--- Opening Log file [January 30 14:18:13 UTC]





# Windows GPU Console Edition #################################################

###############################################################################



                       Folding@Home Client Version 6.23



                          http://folding.stanford.edu



###############################################################################

###############################################################################



Launch directory: C:\Program Files\Folding@home\Folding@home-gpu

Arguments: -local -verbosity 9 -advmethods



[14:18:13] - Ask before connecting: No

[14:18:13] - User name: davidcoton (Team 0)

[14:18:13] - User ID: xxxxxxxxxxx
[14:18:13] - Machine ID: 5

[14:18:13]

[14:18:13] Loaded queue successfully.

[14:18:13] Initialization complete

[14:18:13]

[14:18:13] + Processing work unit

[14:18:13] - Autosending finished units... [January 30 14:18:13 UTC]

[14:18:13] Trying to send all finished work units

[14:18:13] + No unsent completed units remaining.

[14:18:13] - Autosend completed

[14:18:13] Core required: FahCore_11.exe

[14:18:13] Core found.

[14:18:13] Working on queue slot 01 [January 30 14:18:13 UTC]

[14:18:13] + Working ...

[14:18:13] - Calling '.\FahCore_11.exe -dir work/ -suffix 01 -priority 96 -checkpoint 15 -verbose -lifeline 8 -version 623'


[14:18:18] CoreStatus = C0000135 (-1073741515)

[14:18:18] Client-core communications error: ERROR 0xc0000135

[14:18:18] This is a sign of more serious problems, shutting down.


What have I not done .. or done wrong?

David
Image
davidcoton
 
Posts: 947
Joined: Wed Nov 05, 2008 3:19 pm
Location: Cambridge, UK

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby toTOW » Sun Jan 30, 2011 2:47 pm

Did you follow the Linux install guide : viewtopic.php?f=54&t=6793 ?
Folding@Home beta tester since 2002. Folding Forum moderator since July 2008.

FAH-Addict : latest news, tests and reviews about Folding@Home project.

Image
User avatar
toTOW
Super Moderator
 
Posts: 8848
Joined: Sun Dec 02, 2007 10:38 am
Location: Bordeaux, France

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby davidcoton » Sun Jan 30, 2011 3:57 pm

Thanks toTOW.

I saw the install guide, but it didn't appear immediately relevant. I'm running a later version of Ubuntu, where there is far more automation of installs, and a later version of the CUDA toolkit. Also I'm not running headless. So I'm trying to figure out what is NOT set up by the installation route I took. Being in a hurry, I hopped someone might recognize the problem. Just seen the 18 pages (!) of the thread following the install guide post, guess I'll just have to wade through all that and try each suggestion. Or give up and reorganize my hardware to put the decent graphics card in a Windows desktop instead of a purpose-built Linux folding PC.

Big sigh.

David
davidcoton
 
Posts: 947
Joined: Wed Nov 05, 2008 3:19 pm
Location: Cambridge, UK

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby bollix47 » Sun Jan 30, 2011 4:35 pm

The error code is described at :

http://fahwiki.net/index.php/CoreStatus_codes#C0000135

Because you're running X you can ignore steps 14,15 and 16.

You want to ensure your links are correct by running 12 & 13. If 13 doesn't show that you've done the correct links then you'll have to determine why.

gl
Last edited by bollix47 on Sun Jan 30, 2011 4:44 pm, edited 1 time in total.
Image
bollix47
 
Posts: 3363
Joined: Sun Dec 02, 2007 5:04 am
Location: Canada

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby dschief » Sun Jan 30, 2011 4:42 pm

What driver is installed? And is the tool kit the 32 bit version, I'm pretty sure it's still required to run the 32 bit toolkit

have you tried the following flags in your config file -gpu 0 and -forcegpu nvidia_g80
User avatar
dschief
 
Posts: 259
Joined: Tue Dec 04, 2007 5:56 am

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby davidcoton » Sun Jan 30, 2011 5:35 pm

Thanks for both replies

It looks like the automatic install of the CUDA 3.2 toolkit under wine has not put the right dlls in the right place, because step 13 fails.
I've got these dlls (copied from /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/bin)
    /home/david/.wine/drive_c/windows/system32/cublas32_32_16.dll
    /home/david/.wine/drive_c/windows/system32/cudart32_32_16.dll
    /home/david/.wine/drive_c/windows/system32/cufft32_32_16.dll
    /home/david/.wine/drive_c/windows/system32/curand32_32_16.dll
    /home/david/.wine/drive_c/windows/system32/cusparse32_32_16.dll

WHAT should they be called? WHERE should they be?

Drivers are 260.19.06.
Toolkit is 32bit version.
I've used -forcegpu and -gpu without any effect.

David
davidcoton
 
Posts: 947
Joined: Wed Nov 05, 2008 3:19 pm
Location: Cambridge, UK

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby bollix47 » Sun Jan 30, 2011 5:43 pm

WHAT should they be called? WHERE should they be?


Step 12 should give you that info. Did you rerun steps 10,11 & 12?

You could copy/paste your terminal output and post it here.
bollix47
 
Posts: 3363
Joined: Sun Dec 02, 2007 5:04 am
Location: Canada

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby davidcoton » Sun Jan 30, 2011 7:50 pm

Thanks again. I was trying to identify what parts of 10, 11, and 12 are necessary -- not easy when my Linux knowledge is almost nil.

I installed the v3.2 toolkit as an MSI package into Wine, so I hoped most of the work was done. However, the dll name do not match the expectations. I've now done steps 10 and 12 (using wrapper-3,2 instead of wrapper-3.0) and that gives something more reasonable in terms of dll (links) in ...windows/system32.
Step 11 has happened (directories exist).
Then I did step 7 (missed earlier), and repeated 10 and the LN commands from 12 just in case.

However, the error persists when running FAH. Looking again at the output from step 13, I've got:
Code: Select all
david@david-ubuntu:~/.wine/drive_c/windows/system32$ ldd ~/.wine/drive_c/windows/system32/cudart.dll
   linux-gate.so.1 =>  (0xf7717000)
   libcudart.so.3 => not found
   libwine.so.1 => /usr/lib32/libwine.so.1 (0xf759d000)
   libm.so.6 => /lib32/libm.so.6 (0xf7576000)
   libc.so.6 => /lib32/libc.so.6 (0xf741b000)
   libdl.so.2 => /lib32/libdl.so.2 (0xf7417000)
   /lib/ld-linux.so.2 (0xf7718000)


So one "Not Found" error, and four missing entries compared to the Guide.
Code: Select all
        libpthread.so.0 => /lib32/libpthread.so.0 (0xf73af000)
        librt.so.1 => /lib32/librt.so.1 (0xf73a6000)
        libstdc++.so.6 => /usr/lib32/libstdc++.so.6 (0xf72b7000)
        libgcc_s.so.1 => /usr/lib32/libgcc_s.so.1 (0xf72a7000)



EDIT:
It seems /usr/local/cuda/lib does not exist. What should be there? The toolkit has these lib files:
    /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/lib/Win32/cublas.lib
    /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/lib/Win32/cuda.lib
    /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/lib/Win32/cudart.lib
    /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/lib/Win32/cufft.lib
    /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/lib/Win32/curand.lib
    /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/lib/Win32/cusparse.lib
    /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/lib/Win32/nvcuvenc.lib
    /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/lib/Win32/nvcuvid.lib
    /home/david/.wine/drive_c/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v3.2/lib/Win32/OpenCL.lib

Should step 10 use this directory instead?
What can I try next?

David
Last edited by davidcoton on Sun Jan 30, 2011 8:28 pm, edited 1 time in total.
davidcoton
 
Posts: 947
Joined: Wed Nov 05, 2008 3:19 pm
Location: Cambridge, UK

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby bollix47 » Sun Jan 30, 2011 8:14 pm

I installed the v3.2 toolkit as an MSI package into Wine


Sounds like you installed the Windows version of the toolkit. You need to install the linux version.

Here's another guide that might be helpful:

http://linuxfah.info/index.php?title=Main_Page
bollix47
 
Posts: 3363
Joined: Sun Dec 02, 2007 5:04 am
Location: Canada

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby davidcoton » Sun Jan 30, 2011 8:48 pm

Looks from the linuxfah guide that the version is critical -- I'll try that and see what happens. Might take a while...

David
davidcoton
 
Posts: 947
Joined: Wed Nov 05, 2008 3:19 pm
Location: Cambridge, UK

Re: Running 6.23 under wine on Ubuntu 10.10 AMD64

Postby Sidicas » Wed Feb 16, 2011 5:24 pm

bollix47 wrote:
I installed the v3.2 toolkit as an MSI package into Wine


Sounds like you installed the Windows version of the toolkit. You need to install the linux version.

Here's another guide that might be helpful:

http://linuxfah.info/index.php?title=Main_Page


That's exactly what his problem is.. Don't run the CUDA Toolkit itself through WINE, it runs naively on Linux so there's no reason to.

Further, you should be running the *EXACT* version of the CUDA Toolkit as recommended by the guides.. The reason is because it's the same toolkit that the folding@home cores are compiled against. There might be "regressions" in the newer versions of the nvidia toolkit that can cause you headaches in trying to get it to run.. It's very rare, but it's possible.

I'd suggest you follow the guide exactly. Regardless of what distro you're using or what version of the distro you're using to be sure you don't miss any steps. It will also make any problems a lot easier to resolve when we're all on the same page.

You need to make sure your CUDA libraries get installed and linked into the locations where they're needed, the Linux installer for the CUDA Toolkit does this automatically.
Sidicas
 
Posts: 233
Joined: Sun Feb 17, 2008 4:46 pm


Return to unOfficial Linux GPU (WINE wrapper) (3rd party support)

Who is online

Users browsing this forum: No registered users and 1 guest

cron