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

[set]

...
 I really need you'r help. Fighting with this whole last week (

Is X running?
Do you have access to X?
export DISPLAY=:0 xlsfonts
... README GPU-README

I'm not sure why I need X on rig without any monutor attached. I have Server Ubuntu 12.04 LTS there.
Also I did DISPLAY=:0 but has no luck

[kano]

...
 I really need you'r help. Fighting with this whole last week (

Is X running?
Do you have access to X?
export DISPLAY=:0 xlsfonts
... README GPU-README

I'm not sure why I need X on rig without any monutor attached. I have Server Ubuntu 12.04 LTS there.
Also I did DISPLAY=:0 but has no luck

If you are not sure then you just didn't see it in the GPU-README

On Linux you virtually always need to export your display settings before
starting to get all the cards recognised and/or temperature+clocking working:

export DISPLAY=:0

in the FAQ:

Q: Cgminer cannot see any of my GPUs even though I have configured them all
to be enabled and installed OpenCL (+/- Xorg is running and the DISPLAY
variable is exported on linux)?
A: Check the output of 'cgminer -n', it will list what OpenCL devices your
installed SDK recognises. If it lists none, you have a problem with your
version or installation of the SDK.

and

GPUs mine with OpenCL software via the GPU device driver. This means you need
to have both an OpenCL SDK installed, and the GPU device driver RUNNING (i.e.
Xorg up and running configured for all devices that will mine on linux etc.)

Yeah oddly enough, I did say you need it for a reason ...

[The Mafia]

I have a bit of an issue, maybe a bug in cgminer.

I have two machines running, one with 2 x 7950 and 1 x 7970. The other has a single 7970.

I get this error - Which I assume is my internet dropping and reconnecting:

http://users.on.net/pacey/error.jpg

One machine (The big one) cgminer crashes. The other machine with the single card, cgminer just hangs (Doesn't crash), and does nothing (Doesn't reconnect or resume hashing).


Can anyone give me any input as to fix this, or maybe tell it to reconnect? I'm thinking on my big machine, when cgminer crashes, this could be a bit more of an issue than just a simple reconnect?

[set]

@kano
So i did install X server as described here http://askubuntu.com/questions/213678/how-to-install-x11-12-04-lts-x64
and then starded it with xinit
but now i getting this error

Code:

set@miner:~/cgminer-3.1.1-x86_64-built$ sudo aticonfig --adapter=all -f --initial
Uninitialised file found, configuring.
Using /etc/X11/xorg.conf
Saving back-up to /etc/X11/xorg.conf.fglrx-1
set@miner:~/cgminer-3.1.1-x86_64-built$ export DISPLAY=:0
set@miner:~/cgminer-3.1.1-x86_64-built$ ./cgminer -n
X Error of failed request:  BadRequest (invalid request code or no such operation)
  Major opcode of failed request:  136 ()
  Minor opcode of failed request:  19
  Serial number of failed request:  8
  Current serial number in output stream:  8

What does it mean?

[set]

Also I'm having bad feeling after this

Code:

set@miner:~$ modprobe fglrx
FATAL: Module fglrx not found.

Does it mean drivers are not installed? If so - should i install it just with `apt-get install fglrx`?

[os2sam]

thanks I will read again the gpu-temp flags section.
Just one thing, why  --temp-cutoff doesn't work?
It won't start mining again when temp drops....
The whole time it is "Device recovered to temperature below target, re-enabling"
and nothing happends, or must be there another flag for starting mining again?

From the Top Post - https://bitcointalk.org/index.php?topic=28402.msg357369#msg357369

>> --temp-cutoff <arg> Temperature where a device will be automatically disabled, one value or comma separated list (default: 95)

To me that implies that the GPU is disabled permanently until you take action to re-enable it.

Now the verbiage in your post does imply that it is *trying* to re-enable your GPU, so I don't know why it isn't.

But I would use a reasonable Temp Target around 60C to 75C, Temp Overheat of 85C to 90C and a Temp Cutoff around 95C

>> --temp-overheat <arg> Overheat temperature when automatically managing fan and GPU speeds (default: 85)
>> --temp-target <arg> Target temperature when automatically managing fan and GPU speeds (default: 75)

Why don't you post your CGMiner command line and tell us what GPU's you are using in this rig?  Maybe we can come up with something that works better?
Sam

[set]

So I did

Code:

apt-get install fglrx
apt-get install fglrx-updates

And problem above disappeared. X Server running. But still cgminer does not see my 6850.

Code:

set@miner:~/cgminer-3.1.1-x86_64-built$ export DISPLAY=:0
set@miner:~/cgminer-3.1.1-x86_64-built$ ./cgminer -n
No protocol specified
No protocol specified
 [2013-05-25 15:23:48] CL Platform 0 vendor: Advanced Micro Devices, Inc.                    
 [2013-05-25 15:23:48] CL Platform 0 name: AMD Accelerated Parallel Processing                    
 [2013-05-25 15:23:48] CL Platform 0 version: OpenCL 1.2 AMD-APP (1016.4)                    
 [2013-05-25 15:23:48] Error -1: Getting Device IDs (num)                    
 [2013-05-25 15:23:48] clDevicesNum returned error, no GPUs usable                    
 [2013-05-25 15:23:48] 0 GPU devices max detected                    
 [2013-05-25 15:23:48] USB all: found 12 devices - listing known devices                    
 [2013-05-25 15:23:48] No known USB devices    

[vapourminer]

I must doing something wrong here.

I want to run only a jalapeno on my netbook, its an atom with integrated graphics. I have the zadig driver installed and the jally is detected. I fire up cgminer with -G or --disable-gpu and it still fails with a windows dialog box popup that says "no opencl.dll found." well I am not using opencl, I want the jally only to mine. what am I missing? using 3.1.1, and the command line is cgminer -G or cgminer --disable-gpu. does opencl.dll actually need to exist even if its not needed?

also, a minor display bug, when I plug the jally into my desktop with an already running cgminer, it detects the jally and it starts hashing with it but the status line for the jally is not added, it just lists the 6870 that was already hashing. shares are reported though in the scrolling part.

[-ck]

I must doing something wrong here.

I want to run only a jalapeno on my netbook, its an atom with integrated graphics. I have the zadig driver installed and the jally is detected. I fire up cgminer with -G or --disable-gpu and it still fails with a windows dialog box popup that says "no opencl.dll found." well I am not using opencl, I want the jally only to mine. what am I missing? using 3.1.1, and the command line is cgminer -G or cgminer --disable-gpu. does opencl.dll actually need to exist even if its not needed?

also, a minor display bug, when I plug the jally into my desktop with an already running cgminer, it detects the jally and it starts hashing with it but the status line for the jally is not added, it just lists the 6870 that was already hashing. shares are reported though in the scrolling part.

Use the nogpu executable file.

[Blisk]

os2sam thank you for helping.
I am looking for "universal" for ati 6850 and ati 7770 and for ati 7850
-I d --auto-gpu --auto-fan --gpu-fan 60-80 --temp-overheat 85 --temp-cutoff 95

[vapourminer]

Use the nogpu executable file.

DOH!

thanks

[os2sam]

os2sam thank you for helping.
I am looking for "universal" for ati 6850 and ati 7770 and for ati 7850
-I d --auto-gpu --auto-fan --gpu-fan 60-80 --temp-overheat 85 --temp-cutoff 95

OK, the only similar hardware I have is my 5770 and here is what I use for it.  Also I have a 5830 which is the first GPU and 5770 which is my 2nd and the 5770 is driving my display

-I 8,d --auto-gpu --gpu-engine 300-960,300-920 --gpu-memclock 250 --auto-fan --gpu-fan 25-85 --temp-target 75 --temp-overheat 85 --temp-cutoff 95

You HAVE to put an Engine Clock RANGE so that CGMiner can clock down to meet the target temp when the fans have done all they can.

Also if you want only one set of clock range's for all your GPU's you can do that, but, you will need to select the highest engine clock of your slowest GPU and then set your low engine clock range to 300 or some low number per what ckolivas said earlier.
Sam

[Blisk]

thanks I will test new settings for now with this old-new settings no PC has shut down.
I just didn't wan't to set gpu's clock range to leave it automatically, I will set and test.

[os2sam]

thanks I will test new settings for now with this old-new settings no PC has shut down.
I just didn't wan't to set gpu's clock range to leave it automatically, I will set and test.

Well if you don't believe me here it is from the CGMiner Primary author.

Unless you give it a range of engine clock speeds, auto-gpu can only do one thing: Turn it off when it hits 95 degrees. Give it a range of engines for it clock it down.

[Blisk]

thanks I will test new settings for now with this old-new settings no PC has shut down.
I just didn't wan't to set gpu's clock range to leave it automatically, I will set and test.

Well if you don't believe me here it is from the CGMiner Primary author.

Unless you give it a range of engine clock speeds, auto-gpu can only do one thing: Turn it off when it hits 95 degrees. Give it a range of engines for it clock it down.

Yes I belive you is not question of that.
One question can I set gpu clock for example on PC with 6850 and for another pc with 7850 radeon like this?
--auto-gpu --gpu-engine 300-960 --gpu-memclock 250 --auto-fan --gpu-fan 50-85 --temp-target 75 --temp-overheat 85 --temp-cutoff 95

So if I understand even if 6850 doesn't work on gpu-engine 960 it will work on slow range? Or it wil freeze and shuts down?
Or I must sets separatelly for 6850 and 7850?
Because I plan to put in mining two laptops too which have avout 120Mhash and really like to have for all pcs one settings. If you understand me.

I just tested this on my PC at home on radeon 6850
--auto-gpu --gpu-engine 200-960 --auto-fan --gpu-fan 50-85 --temp-target 75 --temp-overheat 85 --temp-cutoff 75
and device recovered to temperature below target, re enabling
Stratum from pool  0 detected new block and I get 0.000
GPU 0:   31 oC 2516RPM  | 0.000/86.27Mh/s  | A:1  R:0  HW:0  U:0.11/m I:5

and that 0.000 stays on 0.000

[os2sam]

thanks I will test new settings for now with this old-new settings no PC has shut down.
I just didn't wan't to set gpu's clock range to leave it automatically, I will set and test.

Well if you don't believe me here it is from the CGMiner Primary author.

Unless you give it a range of engine clock speeds, auto-gpu can only do one thing: Turn it off when it hits 95 degrees. Give it a range of engines for it clock it down.

Yes I belive you is not question of that.
One question can I set gpu clock for example on PC with 6850 and for another pc with 7850 radeon like this?
--auto-gpu --gpu-engine 300-960 --gpu-memclock 250 --auto-fan --gpu-fan 50-85 --temp-target 75 --temp-overheat 85 --temp-cutoff 95

So if I understand even if 6850 doesn't work on gpu-engine 960 it will work on slow range? Or it wil freeze and shuts down?
Or I must sets separatelly for 6850 and 7850?
Because I plan to put in mining two laptops too which have avout 120Mhash and really like to have for all pcs one settings. If you understand me.

To be realistic your not going to be able to set all of your PC's to the same settings.  Different GPU's will need different settings or you will have a HUGE performance deficit for your higher end GPU's.  Even the same GPU's in different PC's will most likely have different settings because every card/GPU chip has it's own unique characteristics.

For your 6850 this is *probably* ok

--auto-gpu --gpu-engine 300-940 --gpu-memclock 300 --auto-fan --gpu-fan 50-85 --temp-target 75 --temp-overheat 85 --temp-cutoff 95

For your 7850

--auto-gpu --gpu-engine 300-960 --gpu-memdiff 150 --auto-fan --gpu-fan 50-85 --temp-target 75 --temp-overheat 85 --temp-cutoff 95

If you don't know what your GPU engines are running at then *before* you set --gpu-engine run with just --auto-gpu and go into CGMiner's GPU screen and see what it is running at by default and use that number for your upper limit.

You can also use this hardware comparison chart to help you optimize each GPU, if you want.

https://en.bitcoin.it/wiki/Mining_hardware_comparison

If you set your GPU clock too high it will either go dead and/or get a bunch of hardware error's.
Sam

[os2sam]

I just tested this on my PC at home on radeon 6850
--auto-gpu --gpu-engine 200-960 --auto-fan --gpu-fan 50-85 --temp-target 75 --temp-overheat 85 --temp-cutoff 75
and device recovered to temperature below target, re enabling
Stratum from pool  0 detected new block and I get 0.000
GPU 0:   31 oC 2516RPM  | 0.000/86.27Mh/s  | A:1  R:0  HW:0  U:0.11/m I:5

and that 0.000 stays on 0.000

OK, in your example above your temp target is 75, your temp overheat is 85 and your temp cutoff is 75.

Your telling CGMiner to target 75 as the temp you want to run and then telling it to react to the temps at 85 and then telling it to cut off the GPU at 75.  So you are basically telling CGMiner to turn your GPU off.

Do this on this 6850 GPU

--auto-gpu --gpu-engine 200-850 --auto-fan --gpu-fan 50-85 --temp-target 75 --temp-overheat 85 --temp-cutoff 95

850 Mhz seems to be the default for this GPU.

75C is the target temp you want to run at. The fan speeds will change to maintain this temp.
85C is the temp that it will start lowering the Engine Clock.
95C is the temp it will completely turn the GPU off to protect it.  And I think it will require manual action to restart it.

[Blisk]

soryy about the post --temp-cutoff 75
it was just for testing purpose to see when cgminer cuts off gpu and if it will starts again when temperature drops.

it will be good if cgminer can lower gpu engine by default if I setup gpu fan speed for exampe on 90%.
Maybe will be good some more auto-default setup of cgminer withouth so many parameters

[os2sam]

soryy about the post --temp-cutoff 75
it was just for testing purpose to see when cgminer cuts off gpu and if it will starts again when temperature drops.

--temp-target <arg> Target temperature when automatically managing fan and GPU speeds (default: 75)
--temp-overheat <arg> Overheat temperature when automatically managing fan and GPU speeds (default: 85)
--temp-cutoff <arg> Temperature where a device will be automatically disabled, one value or comma separated list (default: 95)

As I said I think the temp cutoff requires manual interaction whereas the temp overheat is where it recovers from.

So are you working yet?

[Blisk]

soryy about the post --temp-cutoff 75
it was just for testing purpose to see when cgminer cuts off gpu and if it will starts again when temperature drops.

--temp-target <arg> Target temperature when automatically managing fan and GPU speeds (default: 75)
--temp-overheat <arg> Overheat temperature when automatically managing fan and GPU speeds (default: 85)
--temp-cutoff <arg> Temperature where a device will be automatically disabled, one value or comma separated list (default: 95)

As I said I think the temp cutoff requires manual interaction whereas the temp overheat is where it recovers from.

So are you working yet?

yes it works ok. thank you
but as I said maybe developer will think about more auto setup cgminer which will lower gpu engine with fan settings or temperature settings

cut off doesnt restart gpu