DiabloMiner GPU Miner

[nmat]

You initialise the networkStates array with a length of 0 if the url command line option is not used, resulting in an ArrayIndexOutOfBoundsException on line 427:

Code:

Exception in thread "main" java.lang.ArrayIndexOutOfBoundsException: 0
        at com.diablominer.DiabloMiner.DiabloMiner.execute(DiabloMiner.java:427)
        at com.diablominer.DiabloMiner.DiabloMiner.main(DiabloMiner.java:133)

Initialising it to 1 with the command line arguments:

Code:

-dd -o "swepool.net" -r 8337 -p "mypwd" -u "myuser" -w 128

I get the following exception:

Code:

Exception in thread "DiabloMiner GetWorkAsync for swepool.net" java.lang.NullPointerException
        at java.io.StringReader.<init>(StringReader.java:33)
        at org.codehaus.jackson.JsonFactory.createJsonParser(JsonFactory.java:427)
        at org.codehaus.jackson.map.ObjectMapper.readValue(ObjectMapper.java:1285)
        at org.codehaus.jackson.map.ObjectMapper.readTree(ObjectMapper.java:1007)
        at com.diablominer.DiabloMiner.DiabloMiner$NetworkState.doJSONRPC(DiabloMiner.java:688)
        at com.diablominer.DiabloMiner.DiabloMiner$NetworkState$GetWorkAsync.run(DiabloMiner.java:815)
        at java.lang.Thread.run(Thread.java:680)

A little more testing without the url option might be in order...

I still have this error on Ubuntu 10.04. Did you fix it only in the code or are the binaries also updated?

[1714195716]

1714195716

[1714195716]

1714195716

[1714195716]

1714195716

[1714195716]

1714195716

[1714195716]

1714195716

[kaosbit]

And the argument that a JSON-RPC request has to be a POST is irrelevant because the LP request as defined by the Deepbit spec is not a JSON-RPC request just because its response looks like a JSON-RPC response.

A a protocol is mutual agreement on proper procedure, so yeah, you can pretty well do anything you want if people play along. But you should not call it HTTP or JSON-RPC then, if it merely happens to look similar.
Diablo chooses to follow the very well established HTTP standard and the somewhat established JSON-RPC standard, rather than the quite ad-hoc LP spec. This is a good thing, even if current pool practice also accommodates the JSON-RPC-ish GET LP, as you indicate:

Fortunately, it's easy enough for pools and proxies to support both by not caring if there is a POST body or not.

[twmz]

But you should not call it HTTP or JSON-RPC then, if it merely happens to look similar.

Fortunately, the deepbit.net LP spec doesn't call it JSON-RPC.  And it does happen to be perfectly compliant use of HTTP (make a GET request; get back a response).

[DiabloD3]

You initialise the networkStates array with a length of 0 if the url command line option is not used, resulting in an ArrayIndexOutOfBoundsException on line 427:

Code:

Exception in thread "main" java.lang.ArrayIndexOutOfBoundsException: 0
        at com.diablominer.DiabloMiner.DiabloMiner.execute(DiabloMiner.java:427)
        at com.diablominer.DiabloMiner.DiabloMiner.main(DiabloMiner.java:133)

Initialising it to 1 with the command line arguments:

Code:

-dd -o "swepool.net" -r 8337 -p "mypwd" -u "myuser" -w 128

I get the following exception:

Code:

Exception in thread "DiabloMiner GetWorkAsync for swepool.net" java.lang.NullPointerException
        at java.io.StringReader.<init>(StringReader.java:33)
        at org.codehaus.jackson.JsonFactory.createJsonParser(JsonFactory.java:427)
        at org.codehaus.jackson.map.ObjectMapper.readValue(ObjectMapper.java:1285)
        at org.codehaus.jackson.map.ObjectMapper.readTree(ObjectMapper.java:1007)
        at com.diablominer.DiabloMiner.DiabloMiner$NetworkState.doJSONRPC(DiabloMiner.java:688)
        at com.diablominer.DiabloMiner.DiabloMiner$NetworkState$GetWorkAsync.run(DiabloMiner.java:815)
        at java.lang.Thread.run(Thread.java:680)

A little more testing without the url option might be in order...

I still have this error on Ubuntu 10.04. Did you fix it only in the code or are the binaries also updated?

The binary is also updated. And don't put -o -p and -u args in quotes, thats nonsensical.

[nmat]

The binary is also updated. And don't put -o -p and -u args in quotes, thats nonsensical.

Thanks, I got it working and it's a little faster than poclbm. Are there stable releases or is everything just "up to date"? Not complaining, just asking

But now I have other problems. I am testing things out with a slow card (I am still thinking about investing in newer hardware) so I am running at ~15 Mhash/s with an ATI 5450 on Ubuntu 10.04. The miner runs, but after a while java pushes one of my CPU cores to 100% and stays there. My connection is a little unstable and sometimes the miner prints connection errors, but nothing else... Any idea?

[ancow]

The binary is also updated. And don't put -o -p and -u args in quotes, thats nonsensical.

I'd watch myself before calling anything I don't understand nonsensical. Especially -p and -u arguments may contain special characters that should be escaped from the shell, so it's generally good practice to put them in double-quotes when using them in a shell script.

[DiabloD3]

The binary is also updated. And don't put -o -p and -u args in quotes, thats nonsensical.

I'd watch myself before calling anything I don't understand nonsensical. Especially -p and -u arguments may contain special characters that should be escaped from the shell, so it's generally good practice to put them in double-quotes when using them in a shell script.

Sure, if you're on a shell that actually handles that. Windows afaik doesn't.

[ellipsis]

Diablo, I ran into a few NPEs from commit 63d5dbb9732c6253cde2ae962aa2ae456f4ce02f. Would you like a patch sent somewhere?

[DiabloD3]

Diablo, I ran into a few NPEs from commit 63d5dbb9732c6253cde2ae962aa2ae456f4ce02f. Would you like a patch sent somewhere?

Get an account on github and send me a merge request. Its the easiest way to send me code.

[ellipsis]

Eh, too much trouble. I see you already fixed the exec java bit for Linux. For the others, replace your !"".equals(String) calls (8 instances) with a call to a method like this:

Code:

protected boolean hasLength(String str) {
    return str != null && str.length() > 0;
}

Your splitting on commas for pool support breaks values with commas in them (e.g., passwords).

Around 831, there should be a null check on getWorkParser.

[DiabloD3]

Eh, too much trouble. I see you already fixed the exec java bit for Linux. For the others, replace your !"".equals(String) calls (8 instances) with a call to a method like this:

Code:

protected boolean hasLength(String str) {
    return str != null && str.length() > 0;
}

Your splitting on commas for pool support breaks values with commas in them (e.g., passwords).

Around 831, there should be a null check on getWorkParser.

Java coders prefer checking against "" instead of checking length for some reason. Almost all the places I check also check for null as well.

And why around 831? At no point can that be null.

[ellipsis]

Eh, too much trouble. I see you already fixed the exec java bit for Linux. For the others, replace your !"".equals(String) calls (8 instances) with a call to a method like this:

Code:

protected boolean hasLength(String str) {
    return str != null && str.length() > 0;
}

Your splitting on commas for pool support breaks values with commas in them (e.g., passwords).

Around 831, there should be a null check on getWorkParser.

Java coders prefer checking against "" instead of checking length for some reason. Almost all the places I check also check for null as well.

Maybe retarded Java coders. !"".equals(null) evaluates as true (!false). You intend for it to return false, as in this case:

Code:

        if(!"".equals(url.getUserInfo())) {
          String[] userPassSplit = url.getUserInfo().split(":");

If url.getUserInfo() returns null, you will proceed into the conditional block and dereference a null.

Code:

        if(hasLength(url.getUserInfo())) {
          String[] userPassSplit = url.getUserInfo().split(":");

hasLength will return false, and skip the block. Hurrah.

And why around 831? At no point can that be null.

Really? What if you get an InterruptedException before you return from getWorkQueue.take()?

Lesson complete.

[nmat]

I want to reduce GPU usage because I use the computer where I mine. I suppose this is done with the -f option, but I can't find any documentation for it (except "target execution timing"). What sort of values should be used here?

[ihokamik]

I want to reduce GPU usage because I use the computer where I mine. I suppose this is done with the -f option, but I can't find any documentation for it (except "target execution timing"). What sort of values should be used here?

It's a matter of preference but here is the values I use.

-f 30 for browsing
-f 60 for HD video
-f 240 for games

It's also a good idea to disable flash acceleration and aero.

[DiabloD3]

Really? What if you get an InterruptedException before you return from getWorkQueue.take()?

Lesson complete.

Then I should be checking for the exception, not if it is != null.

[ellipsis]

Really? What if you get an InterruptedException before you return from getWorkQueue.take()?

Then I should be checking for the exception, not if it is != null.

Yes, a better alternative would be to move your getWorkParser variable declaration at 819 and everything below and outside of the catch (830-853) into the try block. Since I'm a stickler about these things, I would also make getWorkParser final and toss the unnecessary null assignments to get rid of another possible null dereference compiler warning. If you were aggressively nulling the variable because it was in a loop because you were worried about holding a reference in the loop, that GC weirdness workaround hasn't been necessary for a very long time. Besides, you will be reducing scope now, which is cleaner than null assignment, and the GC is non-deterministic, anyway.

[nmat]

It's a matter of preference but here is the values I use.

-f 30 for browsing
-f 60 for HD video
-f 240 for games

It's also a good idea to disable flash acceleration and aero.

Thank you

[macrum]

Hello!

I've found a bug in the current version...

I can't connect to my mining pool with the -l option anymore. -o still works. Please check this out.

Also i have this problem here with my second card i installed today:

[7/18/11 9:01:36 PM] Started                                                 
[7/18/11 9:01:36 PM] Connecting to: http://174.133.197.173:8332/             
[7/18/11 9:01:37 PM] Using AMD Accelerated Parallel Processing OpenCL 1.1 AMD-APP-SDK-v2.4 (595.10)
[7/18/11 9:01:40 PM] BFI_INT patching enabled, disabling hardware checking   
[7/18/11 9:01:40 PM] Added Cayman (#1) (24 CU, local work size of 256)       
[7/18/11 9:01:44 PM] BFI_INT patching enabled, disabling hardware checking   
[7/18/11 9:01:44 PM] Added Cayman (#2) (24 CU, local work size of 256)       
[7/18/11 9:01:46 PM] 174.133.197.173 accepted block 1 from Cayman (#1)       
[7/18/11 9:02:00 PM] 174.133.197.173 accepted block 2 from Cayman (#2)       
[7/18/11 9:02:12 PM] 174.133.197.173 accepted block 3 from Cayman (#1)       
[7/18/11 9:02:14 PM] 174.133.197.173 accepted block 4 from Cayman (#2)       
[7/18/11 9:02:22 PM] 174.133.197.173 accepted block 5 from Cayman (#1)       
mhash: 362.1/412.3 | accept: 5 | reject: 0 | hw error: 40718

Please help me, i really don't like those python miners...

I am running on CentOS 6 with two 6950's
If you need more info please tell!

[DiabloD3]

Hello!

I've found a bug in the current version...

I can't connect to my mining pool with the -l option anymore. -o still works. Please check this out.

Also i have this problem here with my second card i installed today:

[7/18/11 9:01:36 PM] Started                                                 
[7/18/11 9:01:36 PM] Connecting to: http://174.133.197.173:8332/             
[7/18/11 9:01:37 PM] Using AMD Accelerated Parallel Processing OpenCL 1.1 AMD-APP-SDK-v2.4 (595.10)
[7/18/11 9:01:40 PM] BFI_INT patching enabled, disabling hardware checking   
[7/18/11 9:01:40 PM] Added Cayman (#1) (24 CU, local work size of 256)       
[7/18/11 9:01:44 PM] BFI_INT patching enabled, disabling hardware checking   
[7/18/11 9:01:44 PM] Added Cayman (#2) (24 CU, local work size of 256)       
[7/18/11 9:01:46 PM] 174.133.197.173 accepted block 1 from Cayman (#1)       
[7/18/11 9:02:00 PM] 174.133.197.173 accepted block 2 from Cayman (#2)       
[7/18/11 9:02:12 PM] 174.133.197.173 accepted block 3 from Cayman (#1)       
[7/18/11 9:02:14 PM] 174.133.197.173 accepted block 4 from Cayman (#2)       
[7/18/11 9:02:22 PM] 174.133.197.173 accepted block 5 from Cayman (#1)       
mhash: 362.1/412.3 | accept: 5 | reject: 0 | hw error: 40718

Please help me, i really don't like those python miners...

I am running on CentOS 6 with two 6950's
If you need more info please tell!

Holy crap dude. What is the entire arguments you're feeding it?

Also, what did you try feeding it with -l? It should work fine if you're using it right.

[macrum]

Sorry for that little info, i was late.

I am feeding it with this:
./DiabloMiner-Linux.sh -u workername -p password -l nofeemining.com

This worked fine with the release from Wed 06 Jul 2011 (modify date of DiabloMiner-Linux.sh)

Additional Info i am using SDK 2.4 and Catalyst 11.6


Old release:

Code:

[robert@RIG02 DiabloMiner.old]$ ./DiabloMiner-Linux.sh -u username -p password -l nofeemining.com
[7/18/11 10:54:34 PM] Started                                                
[7/18/11 10:54:34 PM] Connecting to: http://nofeemining.com:8332/            
[7/18/11 10:54:34 PM] Using AMD Accelerated Parallel Processing OpenCL 1.1 AMD-APP-SDK-v2.4 (595.10)
[7/18/11 10:54:37 PM] BFI_INT patching enabled, disabling hardware checking  
[7/18/11 10:54:37 PM] Added Cayman (#1) (24 CU, local work size of 256)      
[7/18/11 10:54:42 PM] BFI_INT patching enabled, disabling hardware checking  
[7/18/11 10:54:42 PM] Added Cayman (#2) (24 CU, local work size of 256)      
mhash: 407.6/459.3 | accept: 0 | reject: 0 | hw error: 21104

Newest Binary:

Code:

[robert@RIG02 DiabloMiner]$ ./DiabloMiner-Linux.sh -u username -p password -l nofeemining.com
Exception in thread "main" java.net.MalformedURLException: no protocol: nofeemining.com
    at java.net.URL.<init>(URL.java:583)
    at java.net.URL.<init>(URL.java:480)
    at java.net.URL.<init>(URL.java:429)
    at com.diablominer.DiabloMiner.DiabloMiner.execute(DiabloMiner.java:307)
    at com.diablominer.DiabloMiner.DiabloMiner.main(DiabloMiner.java:133)

Any Ideas where this hardware errors come from(I asume hw = hardware?)? When I install only one card into my rig than there is no problem (tested both of them).