OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.1

[tnkflx]

Interested in remote access to a 6990 rig?

A quick prod would not hurt, but I expect I'll find the same thing. Of course I can only work with linux :s

No problem, I'll see what I can see set up. Do you have a preference in Linux distro? ;-)

[1714103713]

1714103713

[1714103713]

1714103713

[1714103713]

1714103713

[1714103713]

1714103713

[1714103713]

1714103713

[-ck]

No problem, I'll see what I can see set up. Do you have a preference in Linux distro? ;-)

Thanks, ubuntu or debian are the likely culprits.

[runeks]

I'm running cgminer like this:

Code:

cgminer -c Desktop/p2pool.json

and here is the content of Desktop/p2pool.json:

Code:

{
"pools" : [
	{
		"url" : "http://127.0.0.1:9332",
		"user" : "nouser",
		"pass" : "nopass1"
	},
	{
		"url" : "http://arsbitcoin.com:8344",
		"user" : "<removed>",
		"pass" : "default"
	}
],

	"kernel" : "phatk",
   	"kernel-path" : "/usr/local/bin",
	"worksize" : "256",
	"vectors" : "2",
	"intensity" : "6",

	"donation" : "0.5"
}

as you can see I have set the donation amount to 0.5%, yet cgminer still says:

Code:

[2012-01-24 14:28:44] --donation is disabled, please consider just 0.5% :-(

how come? Is the configuration file above not correct? I'm running cgminer 2.1.2.

[-ck]

Well, after much poking and prodding and asking on forums and scouring APIs and lots of ridiculous documents, I can see there is no official way to tell that two GPUs on a dual GPU card belong to the same device. There is nothing shared between the two. However, having access to an actual device I was able to come u p with the following. This has been committed to the git tree:

Detect dual GPU cards via the indirect information of - 1st card has a fan controller. 2nd card does not have a fan controller, cards share the same device name, and are one bus ID apart. Use this information to create paired devices, displaying the fan speed on both devices and using the hottest temperature from either GPU to control the overall fanspeed.

I have confirmed this works on 5970s. Next would be the 6990s...

Now I better get some sleep 

[jjiimm_64]

Well, after much poking and prodding and asking on forums and scouring APIs and lots of ridiculous documents, I can see there is no official way to tell that two GPUs on a dual GPU card belong to the same device. There is nothing shared between the two. However, having access to an actual device I was able to come u p with the following. This has been committed to the git tree:

Detect dual GPU cards via the indirect information of - 1st card has a fan controller. 2nd card does not have a fan controller, cards share the same device name, and are one bus ID apart. Use this information to create paired devices, displaying the fan speed on both devices and using the hottest temperature from either GPU to control the overall fanspeed.

I have confirmed this works on 5970s. Next would be the 6990s...

Now I better get some sleep  

ck,  this could be problematic on windoz machines....  sometimes windows mixes up the order of the gpus. I know windoz is everyone's favorite whipping boy, but it is a necessary evil.

Code:

cgminer version 2.1.2 - Started: [1327278549]    Rig: miner14
(5s):?  (avg): 1723.83 Mh/s  |    Q:32805   A:52875   R:584   HW:0   E:?%   U:23.21/m
TQ:?   ST:137   SS:?   DW:2095   NB:228   LW:94056   GF:29   RF:447
Connected to http://us.eclipsemc.com:8337 with LP as user ?
GPU 0:	74.5C	  	  | 	361.1/360.0Mh/s | 	800Mhz 	1.05V 	A:11121	R:121	HW:0	U:4.88/m 	I: 7
GPU 1:	73.5C	  	  | 	361.3/360.1Mh/s | 	800Mhz 	1.05V 	A:10746	R:125	HW:0	U:4.72/m 	I: 7
GPU 2:	73.5C	3498RPM 	55% | 	360.4/360.1Mh/s | 	800Mhz 	1.05V 	A:11168	R:127	HW:0	U:4.90/m 	I: 7
GPU 3:	74.0C	3695RPM 	61% | 	361.5/360.1Mh/s | 	800Mhz 	1.05V 	A:11098	R:134	HW:0	U:4.87/m 	I: 7
GPU 4:	66.5C	2666RPM 	50% | 	280.4/283.4Mh/s | 	900Mhz 	1.15V 	A:8742	R:77	HW:0	U:3.84/m 	I: 7

edit:  to make matters worse.  GPU 4:  is actual in the 0 slot and has this monitor plugged into it...  just the way windows sometimes loads the drivers..

[tnkflx]

<snip>

I have confirmed this works on 5970s. Next would be the 6990s...

<snip>

Yeah yeah, working on that ;-)

[jake262144]

ck,  this could be problematic on windoz machines....  sometimes windows mixes up the order of the gpus. I know windoz is everyone's favorite whipping boy, but it is a necessary evil.

I'm not sure Windows can really be blamed about that. I clearly remember Conman bitching about crazy device enumeration being done in AMD's drivers.
So, no Windows jokes this time.

[JWU42]

Will update and rebuild and check the 5970 fan info to see how it is working...

[-ck]

False positive for picking up twin cards is a problem. False negative is not. Anyway I need access to more hardware to see what else I need to do.

[Fiyasko]

Yaaaaaaaaaaay! i blew a fan totally by my own fault of using cgminers "rapingly powerful fan control"
Time for me to not use that!, xD had my fan running at 100% for three days without knowing it xD Now it turns at 1/5th the RPM.

Lol, Never believed it before, BUT FANS ARE NOT MEANT TO RUN AT 100%!!!!

Hey guys can a fan "break" Out And Into place? All of a sudden the fan is spinning at normal speeds again, And the bearing seems just fine.

Then about 1hr later (while mining), It's slows down like a bitch and overheats till shutdown, I restart, Fan not turning, Give it a little push, and VROOOM proper speed...

Lol.. like wtf...

[rjk]

Yaaaaaaaaaaay! i blew a fan totally by my own fault of using cgminers "rapingly powerful fan control"
Time for me to not use that!, xD had my fan running at 100% for three days without knowing it xD Now it turns at 1/5th the RPM.

Lol, Never believed it before, BUT FANS ARE NOT MEANT TO RUN AT 100%!!!!

Hey guys can a fan "break" Out And Into place? All of a sudden the fan is spinning at normal speeds again, And the bearing seems just fine.

Then about 1hr later (while mining), It's slows down like a bitch and overheats till shutdown, I restart, Fan not turning, Give it a little push, and VROOOM proper speed...

Lol.. like wtf...

That's what happens when you run a fan at 100%...

[os2sam]

Yaaaaaaaaaaay! i blew a fan totally by my own fault of using cgminers "rapingly powerful fan control"
Time for me to not use that!, xD had my fan running at 100% for three days without knowing it xD Now it turns at 1/5th the RPM.

Lol, Never believed it before, BUT FANS ARE NOT MEANT TO RUN AT 100%!!!!

Hey guys can a fan "break" Out And Into place? All of a sudden the fan is spinning at normal speeds again, And the bearing seems just fine.

Then about 1hr later (while mining), It's slows down like a bitch and overheats till shutdown, I restart, Fan not turning, Give it a little push, and VROOOM proper speed...

Lol.. like wtf...

I would replace it ASAP.

[-ck]

Yaaaaaaaaaaay! i blew a fan totally by my own fault of using cgminers "rapingly powerful fan control"
Time for me to not use that!, xD had my fan running at 100% for three days without knowing it xD Now it turns at 1/5th the RPM.

Lol, Never believed it before, BUT FANS ARE NOT MEANT TO RUN AT 100%!!!!

Hey guys can a fan "break" Out And Into place? All of a sudden the fan is spinning at normal speeds again, And the bearing seems just fine.

Then about 1hr later (while mining), It's slows down like a bitch and overheats till shutdown, I restart, Fan not turning, Give it a little push, and VROOOM proper speed...

Lol.. like wtf...

That's what happens when you run a fan at 100%...

And this is why cgminer peaks autofan at 85% and reserves 100% only for overheat situations.

[Fiyasko]

*snickers* Yeah a replacement is on the way!, Sorrrrie for my minihijack~ ty for the comments

[TheHarbinger]

You sir, are a king among men.   

Donation % bumped up to 2.  (Until I ditch all my GPUs and go pure FPGA anyway. ) 

[1onevvolf]

I've just committed the first patch to the git tree that should enable 79xx cards to work properly under cgminer (delete any dodgy .bins created before this attempt). I have no idea if it performs even remotely well, but in theory at least, it won't behave stupidly as the current version does. If someone does run it, could you please dump the output of -D -T the first time you run it? Also it may be worth starting it with -v 1, and/or experiment with various combinations of -v and -w (default being -v 2 -w 128 if you set nothing).

I tried to test your 79xx patch on the git tree following the instructions in the README but after several hours of trying I was unable to get it to actually build on my windows machine.

But it might help you to know that I've found that using the poclbm kernel, setting vector size to 1 and hard-coding the poclbm kernel's BFI_INT to false gives me ~600MH/s (which is much better than the ~540MH/s v2.1.2 kernel with the same settings).

Further gains can be had by changing the definition of ch() and Ma() in the poclbm kernel to use bitselect() instead of boolean operations. Changing this gets cgminer's Radeon HD 7970 performance up to within 3% of that of Diablominer's  (diablo gives me 670MH/s while cgminer gives 650MH/s @ 1125MHz).

[-ck]

Thanks. The GCN compatible phatk kernel that's in the tree should be even faster than that hardcoded poclbm which is reassuring... assuming it works as planned  Alas building on windows is an absolutely pain in the arse and as I've said before even I can't remember what I installed in the end to get it building  I'll see if I can build one maybe tomorrow.

[smurfix]

Hi,

I just had an annoying multi-hour debugging session, trying to figure out why cgminer thinks there's only one GPU in my (Debian Testing) system.

$ aticonfig --list-ad
* 0. 06:00.0 ATI Radeon HD 5800 Series
  1. 07:00.0 ATI Radeon HD 5800 Series
  2. 08:00.0 ATI Radeon HD 5450

Turns out that AMD is being stupid. Again.

OpenCL 1.1 AMD-APP-SDK-v2.4 (595.10) (don't remember where I downloaded it from): library works.
OpenCL 1.2 AMD-APP (831.4) (available from http://developer.amd.com/sdks/AMDAPPSDK/Pages/default.aspx ) does NOT see more than one GPU.

You can extract the build number etc. from the library with
$ strings -a /pth/to/the/library.so | grep "^OpenCL.*AMD-APP"
 
Sorry if this has been mentioned before, but this thread is too long to scan everything.

[-ck]

OpenCL 1.2 AMD-APP (831.4) (available from http://developer.amd.com/sdks/AMDAPPSDK/Pages/default.aspx ) does NOT see more than one GPU.

You can extract the build number etc. from the library with
$ strings -a /pth/to/the/library.so | grep "^OpenCL.*AMD-APP"
 
Sorry if this has been mentioned before, but this thread is too long to scan everything.

Not at all. I think you may have discovered the cause of yet another random unexplained failure that has yet to be mentioned on this thread.

[Proofer]

Trying to understand what happened here...

Code:

[2012-01-25 09:01:07] Accepted 00000000.0292c640.b44ff0df GPU 1 thread 3 pool 0
......... omitted 15 "Accepted" lines not mentioning thread 3 ........
[2012-01-25 09:01:29] Thread 3 idle for more than 60 seconds, GPU 3 declared SICK!
[2012-01-25 09:01:29] Attempting to restart GPU
[2012-01-25 09:01:29] Thread 6 still exists, killing it off
[2012-01-25 09:01:29] Thread 7 still exists, killing it off
[2012-01-25 09:01:33] Thread 6 restarted
[2012-01-25 09:01:34] Thread 7 restarted
[2012-01-25 09:01:35] Thread 7 being disabled
[2012-01-25 09:01:35] Thread 6 being disabled

Is the "Thread 3 idle ..." correct?  Thread 3 is associated with GPU 1, not GPU 3.

While I have had occasional "SICK!"s -- 0 to a half dozen a day, but none since compiling from github a couple of days ago  -- this is the first disabling I've seen.