Posts by Gordon Lack
log in
1) Message boards : News : Sieve in Production; Large and Solo deprecated (Message 21413)
Posted 836 days ago by Gordon Lack
Single precision, double precision, Collatz does not care because it uses integers.

To be honest, I've never seen an error like that ever before. It looks like the app was compiled with an optimization that your computer can't handle. Which specific Linux distro and version are you using?
This is on Linux Mint 17.2 (based on Ubuntu 14.04).
The fglrx driver is 14.10 (or 14.4.rev2...)
2) Message boards : News : Sieve in Production; Large and Solo deprecated (Message 21395)
Posted 837 days ago by Gordon Lack
I deprecated the solo_collatz and large_collatz applications
So I enabled sieve on my Linux system with an AMD 6670 GPU (which is a single-precision GPU).

If the sieve application fails immediately after it starts...
It did, but this being Linux it's not a MS Vidual C++ runtime issue...

This is the error log:


<core_client_version>7.2.42</core_client_version>
<![CDATA[
<message>
process exited with code 1 (0x1, -255)
</message>
<stderr_txt>
Collatz Conjecture Sieve 1.21 Linux x86_64 for OpenCL
Written by Slicker (Jon Sonntag) of team SETI.USA
Based on the AMD Brook+ kernels by Gipsel of team Planet 3DNow!
Sieve code and OpenCL optimization provided by Sosiris of team BOINC@Taiwan
LLVM ERROR: Cannot select: 0x18db810: i32 = setcc 0x1697ab0, 0x16960a0, 0x18db710 [ORD=14] [ID=50]
0x1697ab0: i64 = add 0x16960a0, 0x1696da0 [ORD=10] [ID=46]
0x16960a0: i64 = AMDILISD::VEXTRACT 0x1390ae0, 0x13901e0 [ORD=9] [ID=36]
0x1390ae0: v2i64,ch = CopyFromReg 0xf39610, 0x13909e0 [ORD=9] [ID=30]
0x13909e0: v2i64 = Register %vreg38 [ORD=9] [ID=2]
0x13901e0: i32 = TargetConstant<1> [ORD=2] [ID=21]
0x1696da0: i64 = add 0x1696aa0, 0x1696ca0 [ORD=8] [ID=45]
0x1696aa0: i64,ch = load 0xf39610, 0x16965a0, 0x16966a0<LD4[%arrayidx], zext from i32> [ORD=5] [ID=44]
0x16965a0: i32 = add 0x13904e0, 0x16964a0 [ORD=3] [ID=43]
0x13904e0: i32,ch = CopyFromReg 0xf39610, 0x13903e0 [ORD=3] [ID=29]
0x13903e0: i32 = Register %vreg35 [ORD=3] [ID=1]
0x16964a0: i32 = shl 0x18db210, 0x16963a0 [ORD=3] [ID=42]
0x18db210: i32 = AMDILISD::VEXTRACT 0x1390ee0, 0x13901e0 [ORD=2] [ID=38]
0x1390ee0: v4i32,ch = llvm.AMDIL.get.global.id 0xf39610, 0x1390de0 [ORD=1] [ID=32]
0x1390de0: i32 = TargetConstant<2580> [ORD=1] [ID=4]
0x13901e0: i32 = TargetConstant<1> [ORD=2] [ID=21]
0x16963a0: i32 = Constant<2> [ORD=3] [ID=7]
0x16966a0: i32 = undef [ORD=4] [ID=8]
0x1696ca0: i64 = shl 0x16969a0, 0x1696ba0 [ORD=7] [ID=40]
0x16969a0: i64 = zero_extend 0x1390ce0 [ORD=6] [ID=37]
0x1390ce0: i32,ch = CopyFromReg 0xf39610, 0x1390be0 [ORD=6] [ID=31]
0x1390be0: i32 = Register %vreg39 [ORD=6] [ID=3]
0x1696ba0: i32 = Constant<26> [ORD=7] [ID=9]
0x16960a0: i64 = AMDILISD::VEXTRACT 0x1390ae0, 0x13901e0 [ORD=9] [ID=36]
0x1390ae0: v2i64,ch = CopyFromReg 0xf39610, 0x13909e0 [ORD=9] [ID=30]
0x13909e0: v2i64 = Register %vreg38 [ORD=9] [ID=2]
0x13901e0: i32 = TargetConstant<1> [ORD=2] [ID=21]

</stderr_txt>
]]>
Any ideas? the solo_collatz ran without issue.
3) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19408)
Posted 1350 days ago by Gordon Lack
If later you discover it is broke ...
FYI - it did. So I've sent the Raspberry Pi to School, and set up micro-only options for that.
4) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19388)
Posted 1351 days ago by Gordon Lack
Re your app_config you posted .... if clarify what you want that to do, I will write another version for you, it does seem a little strange on the face of it.
I'm saying that the maximum number of mini_collatzes I want to run is 0. So hopefully it will never download any, which will leave it with just the micro_collatzes to download and run.

I'm happy leaving that in place as long as it seems to be working.
5) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19386)
Posted 1351 days ago by Gordon Lack
Ah - by AMD version 14 you mean a graphics card driver.

But this is a Raspberry Pi, so AMD has nothing to do with it (it does on the system I'm typing this on, which is where I do have a v14 driver). There is no GPU involved here - just an ARM CPU.
6) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19385)
Posted 1351 days ago by Gordon Lack

Current AMD Version is 14.4 go to the AMD site to download.
BOINC Current Version is 7.2.42

What (and where) is the AMD site?
What does it have to do with the current BOINC manager version, and why is the version number so different?
7) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19383)
Posted 1352 days ago by Gordon Lack
I had thought about setting up fake locations, but then forgot about it. I also couldn't remember whether the locations were project-specific (since if they aren't that would affect other projects running in the system).

So I created a project app_config.xml file containing:
<app_config> <app> <name>mini_collatz</name> <max_concurrent>0</max_concurrent> </app> </app_config>
and I'll see how that goes.

Version 7.3.13 (as yet unreleased) of the manage has a "fraction_done_exact" option, that would help, as the remaining time would quickly become correct once the job starts to run.


Any reason why both the micro and mini jobs both have a 2-week deadline, when the micro ones will take ~10 hours while the mini ones will run for ~7 days?
8) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19382)
Posted 1352 days ago by Gordon Lack
Ah, yes. I thought about setting up different work/home/school options, but that location might obtain for all projects that the system connects to (I can't remember - perhaps not).

For the moment I've put a app_config.xml in place, containing:
<app_config>
<app>
<name>mini_collatz</name>
<max_concurrent>0</max_concurrent>
</app>
</app_config>
and I'll see how that fares. If that doesn't work I'll resort to the fake location method.
9) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19376)
Posted 1352 days ago by Gordon Lack
OK. It works.

But there is another problem.

The micro job estimated time when it started was 55 hours - it actually completed in 10. That's fine.

But I have now been sent a mini job , and that has a starting estimate of 810 hours (almost 34 days), but has a deadline of 14 days hence. Consequently it is now running at high priority, and will (I suspect) continue to do so for another ~5 days (until the remaining time estimate drops below the time to deadline), so no other project will get a look-in for that time.

How can I get only micro jobs sent to this one system?
The only preferences I can see to set are per-account, not per-system, and I have other systems running this project which I do not wish to restrict to micro jobs.
10) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19371)
Posted 1352 days ago by Gordon Lack
OK. Got it working.

It needed to be in the options section in the cc_config.xml file, so:
<cc_config>
<options>
<alt_platform>arm-unknown-linux-gnueabihf</alt_platform>
</options>
<log_flags>
...
I have a job running - I'll see how it goes...
11) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19368)
Posted 1353 days ago by Gordon Lack
That just gives:
Notice from BOINC
Unrecognized tag in cc_config.xml: <alt_platform>
in the log.
12) Message boards : Unix/Linux : Running on Raspberry Pi (Message 19357)
Posted 1354 days ago by Gordon Lack
I see that Raspberry Pi binaries were added at the end of April.

However, when I add mine to the project it gets the response:
Mon 05 May 2014 11:04:27 BST | Collatz Conjecture | Message from server: This project doesn't support computers of type armv7l-unknown-linux-gnueabihf
So how do I get it to run?




Main page · Your account · Message boards


Copyright © 2018 Jon Sonntag; All rights reserved.