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

[os2sam]

after checking telnet client I tried telnet us3.eclipsemc.com 3333 in the terminal
a line came up and then the screen cleared before I could see what it said.

That means you can reach that port on that server.  So it is working correctly.

Make sure your Windoze firewall isn't blocking cgminer.exe or cgminer-nogpu.exe.

The target here is to run cgminer on a Pi.
But, I guess we've further verified that the port is available on the server. 

Ah, right.

Yep verified the port on EMC.  Still weird that it won't accept it on port 3333.

[village.idiot]

Yeah.
I've tried both command line and pool management.

Thanks for the help, though

[ProfMac]

after checking telnet client I tried telnet us3.eclipsemc.com 3333 in the terminal
a line came up and then the screen cleared before I could see what it said.

That means you can reach that port on that server.  So it is working correctly.

Make sure your Windoze firewall isn't blocking cgminer.exe or cgminer-nogpu.exe.

The target here is to run cgminer on a Pi.
But, I guess we've further verified that the port is available on the server. 

I was very happy to learn about the TP-Link 703n, because it was a self contained Linux box with a MIPS cpu and networking hardware built in.  Occasionally I get excited about some other hardware that I have software tools for.

I know the raspberry pi has some nice video interfacing, and is pretty inexpensive.  However, I wonder what the particular reason is that people want to run cgminer from it?

[-ck]

I'm trying to run separate cgminer instances on my machine using "-d" and "--remove-disabled" in each batch file, but regardless, the second instance always say's "devices in use" and closes. (The devices are BFL Singles)

Is there a way to accomplish this?

Those commands are largely for managing GPUs. Use the --usb command (as per readme).

[village.idiot]

I was very happy to learn about the TP-Link 703n, because it was a self contained Linux box with a MIPS cpu and networking hardware built in.  Occasionally I get excited about some other hardware that I have software tools for.

I know the raspberry pi has some nice video interfacing, and is pretty inexpensive.  However, I wonder what the particular reason is that people want to run cgminer from it?

I can only speak for myself.
I like that it is compact and low power.
I'm not having to keep a computer running just to service the ASICs.

I've also noticed my Little Single is running more consistently.
On the Windows notebook, the ASIC fan slowing down periodically would get my attention
and the screen would show error messages until something finally got worked out and it would get back up to speed.
I'm suspecting this was a Windows issue.  I didn't run the ASICs on Linux other than the Pi.

[kano]

I'm trying to run separate cgminer instances on my machine using "-d" and "--remove-disabled" in each batch file, but regardless, the second instance always say's "devices in use" and closes. (The devices are BFL Singles)

Is there a way to accomplish this?

With --usb you can say how many devices to find or how many each driver should find

So for each cgminer
 --usb BAS:1
would grab only one BAS device.

If you use
 --usb :1
it would grab the first device of any USB mining devices that you have.
(this also has the minor advantage that it stops polling for hot-plugged devices until the current device disconnects)

[gristlelump]

When attempting to mine scrypt with the latest version of CGminer I get this error  "error -56 enquequeing: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel)".   The miner then procedes to disable the GPU.

I'm running a XFX 5870 HD 1GB VC on Windows XP SP3 with 2GB system ram.  the card is recognized as GPU 0 and the -n function shows it as ATI stream.  It's set to run on Give-Me-Coins pool in this fashion   

"cgminer --scrypt -o http://ltc.give-me-coins.com:3334 -u USER.WORKER -p PASSWORD"      via the BAT file

  In between tries I delete the bin file as well as rename the conf file which I also rename as to not interfere with the BAT file.  This error also occurs when I try and run CGminer.exe through the conf file.  Running from the cmd prompt also has the same results.   I'll also add that I have no problems mining BTC with card and CGminer.  CGminer was not compiled.  I downloaded the binaries  from the main repository.  Any suggestion to solve this problem will be appreciated.

    did you have?

    setx GPU_MAX_ALLOC_PERCENT 100
    setx GPU_USE_SYNC_OBJECTS 1

    in your batch file or before calling cgminer?


Funny thing about this is that I have always included this in the BAT file.  Initially the miner was just instantly shutting down no matter what I did.  I then got some help from the MOD at my mining pool and he did some research and found out that Windows XP does not natively support the SETX command.  A little bit of info that all the guides LEAVE OUT!  I had to download Windows XP Service Pack 2 Support Tools to get the SETX function into XP which was a whole nightmare in itself (but I won't go there).  Once I got that setup incorporated the setx GPU_MAX_ALLOC_PERCENT 100 into the BAT file but not the setx GPU_USE_SYNC_OBJECTS 1.  There seems to be a lot of confusion about how to run CGminer on Windows Xp as opposed to Windows 7 or Linux systems.  If you could explain what the setx commands do it would be much appreciated!

[Aurum]

When attempting to mine scrypt with the latest version of CGminer I get this error  "error -56 enquequeing: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel)".   The miner then procedes to disable the GPU.

I'm running a XFX 5870 HD 1GB VC on Windows XP SP3 with 2GB system ram.  the card is recognized as GPU 0 and the -n function shows it as ATI stream.  It's set to run on Give-Me-Coins pool in this fashion  

"cgminer --scrypt -o http://ltc.give-me-coins.com:3334 -u USER.WORKER -p PASSWORD"      via the BAT file

  In between tries I delete the bin file as well as rename the conf file which I also rename as to not interfere with the BAT file.  This error also occurs when I try and run CGminer.exe through the conf file.  Running from the cmd prompt also has the same results.   I'll also add that I have no problems mining BTC with card and CGminer.  CGminer was not compiled.  I downloaded the binaries  from the main repository.  Any suggestion to solve this problem will be appreciated.

    did you have?

    setx GPU_MAX_ALLOC_PERCENT 100
    setx GPU_USE_SYNC_OBJECTS 1

    in your batch file or before calling cgminer?


Funny thing about this is that I have always included this in the BAT file.  Initially the miner was just instantly shutting down no matter what I did.  I then got some help from the MOD at my mining pool and he did some research and found out that Windows XP does not natively support the SETX command.  A little bit of info that all the guides LEAVE OUT!  I had to download Windows XP Service Pack 2 Support Tools to get the SETX function into XP which was a whole nightmare in itself (but I won't go there).  Once I got that setup incorporated the setx GPU_MAX_ALLOC_PERCENT 100 into the BAT file but not the setx GPU_USE_SYNC_OBJECTS 1.  There seems to be a lot of confusion about how to run CGminer on Windows Xp as opposed to Windows 7 or Linux systems.  If you could explain what the setx commands do it would be much appreciated!

   setx GPU_MAX_ALLOC_PERCENT 100
    setx GPU_USE_SYNC_OBJECTS 1
are useless in Windows, eliminate them and then go System Variables/Environment Variables and delete them. They're not your problem, they just do nothing for you in Windows.

Are you using a conf file??? Here's what I use for a 5830:
{
"scrypt" : true,
"intensity" : "18",
"worksize" : "256",
"lookup-gap" : "2",
"thread-concurrency" : "5632",
"gpu-engine" : "800",
"gpu-memclock" : "1000",
"gpu-vddc" : "0",
"gpu-fan" : "0-85",
"auto-fan" : true,
"gpu-reorder" : true,
"gpu-threads" : "1",
"temp-cutoff" : "95",
"temp-overheat" : "88",
"temp-target" : "75",
"temp-hysteresis" : "9",
"api-port" : "4028",
"hotplug" : "0",
"log" : "5",
"shares" : "0",
"kernel-path" : "/usr/local/bin"
}

[Chalkbot]

[Those commands are largely for managing GPUs. Use the --usb command (as per readme).

With --usb you can say how many devices to find or how many each driver should find

So for each cgminer
 --usb BAS:1
would grab only one BAS device.

If you use
 --usb :1
it would grab the first device of any USB mining devices that you have.
(this also has the minor advantage that it stops polling for hot-plugged devices until the current device disconnects)

Thank you guys, this is working just fine.

[LogicalUnit]

Does anyone know if/when cgminer will support the new Blue Fury ASICs?

[-ck]

Does anyone know if/when cgminer will support the new Blue Fury ASICs?

I'm working on it for BF1 devices (red and blue are the same) since I received one. Should be in next version.

[gristlelump]

When attempting to mine scrypt with the latest version of CGminer I get this error  "error -56 enquequeing: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel)".   The miner then procedes to disable the GPU.

I'm running a XFX 5870 HD 1GB VC on Windows XP SP3 with 2GB system ram.  the card is recognized as GPU 0 and the -n function shows it as ATI stream.  It's set to run on Give-Me-Coins pool in this fashion  

"cgminer --scrypt -o http://ltc.give-me-coins.com:3334 -u USER.WORKER -p PASSWORD"      via the BAT file

  In between tries I delete the bin file as well as rename the conf file which I also rename as to not interfere with the BAT file.  This error also occurs when I try and run CGminer.exe through the conf file.  Running from the cmd prompt also has the same results.   I'll also add that I have no problems mining BTC with card and CGminer.  CGminer was not compiled.  I downloaded the binaries  from the main repository.  Any suggestion to solve this problem will be appreciated.

    did you have?

    setx GPU_MAX_ALLOC_PERCENT 100
    setx GPU_USE_SYNC_OBJECTS 1

    in your batch file or before calling cgminer?


Funny thing about this is that I have always included this in the BAT file.  Initially the miner was just instantly shutting down no matter what I did.  I then got some help from the MOD at my mining pool and he did some research and found out that Windows XP does not natively support the SETX command.  A little bit of info that all the guides LEAVE OUT!  I had to download Windows XP Service Pack 2 Support Tools to get the SETX function into XP which was a whole nightmare in itself (but I won't go there).  Once I got that setup incorporated the setx GPU_MAX_ALLOC_PERCENT 100 into the BAT file but not the setx GPU_USE_SYNC_OBJECTS 1.  There seems to be a lot of confusion about how to run CGminer on Windows Xp as opposed to Windows 7 or Linux systems.  If you could explain what the setx commands do it would be much appreciated!

   setx GPU_MAX_ALLOC_PERCENT 100
    setx GPU_USE_SYNC_OBJECTS 1
are useless in Windows, eliminate them and then go System Variables/Environment Variables and delete them. They're not your problem, they just do nothing for you in Windows.

Are you using a conf file??? Here's what I use for a 5830:
{
"scrypt" : true,
"intensity" : "18",
"worksize" : "256",
"lookup-gap" : "2",
"thread-concurrency" : "5632",
"gpu-engine" : "800",
"gpu-memclock" : "1000",
"gpu-vddc" : "0",
"gpu-fan" : "0-85",
"auto-fan" : true,
"gpu-reorder" : true,
"gpu-threads" : "1",
"temp-cutoff" : "95",
"temp-overheat" : "88",
"temp-target" : "75",
"temp-hysteresis" : "9",
"api-port" : "4028",
"hotplug" : "0",
"log" : "5",
"shares" : "0",
"kernel-path" : "/usr/local/bin"
}

tried your conf file and got this error.
 
Maximum buffer memory device 0 supports says 268435456
your scrypt setting come to 369098752
error -61: clCreateBuffer (padbuffer8), decrease TC or increase LG
failed to init GPU thread 0, disabling device 0
Restarting the GPU from the menu will not fix this.
Try restarting cgminer.

tried decreasing thread concurrency by cutting it in half and increased Lookup gap to 3 bu got the same -51 error mentioned above.

[kentt]

Why do I get "Pool 0 not providing work fast enough".  I get this when I'm solo mining.  I'm solving block and so it seems like it's working, but when I get this message the block I solve per week is lower that expected (by more that 10%).  Is this a problem, that I can debug?

[sexjam]

with the 3.4.3 version my accepted and rejected shares count is wrong.
Ex. in the log i see 5 or 6 shares, while the program report A:1920 R:230 ecc.ecc.

i see that the api works fine

now, i'm using the 3.1.1 version and works fine.

[twmz]

with the 3.4.3 version my accepted and rejected shares count is wrong.
Ex. in the log i see 5 or 6 shares, while the program report A:1920 R:230 ecc.ecc.

i see that the api works fine

now, i'm using the 3.1.1 version and works fine.

Because of vardiff, the UI now shows "Accepted Difficulty" instead of simple count of accepted shares (because not all shares are equal in a vardiff world).  So if you find a share with difficulty 16, "A" will increase by 16 and not by 1.

[sexjam]

 
it's great.
thank you very much.
i switch faster to 3.4.3 version 

[Aurum]

tried your conf file and got this error.
 
Maximum buffer memory device 0 supports says 268435456
your scrypt setting come to 369098752
error -61: clCreateBuffer (padbuffer8), decrease TC or increase LG
failed to init GPU thread 0, disabling device 0
Restarting the GPU from the menu will not fix this.
Try restarting cgminer.

tried decreasing thread concurrency by cutting it in half and increased Lookup gap to 3 bu got the same -51 error mentioned above.

I just looked up a 5870 and see it's a single GPU like a 5970. Here's the conf I run on my 5970 miners:
{
"scrypt" : true,
"gpu-reorder" : true,
"intensity" : "18",
"worksize" : "256",
"lookup-gap" : "2",
"thread-concurrency" : "8192",
"gpu-engine" : "725",
"gpu-memclock" : "1000",
"gpu-vddc" : "0",
"temp-cutoff" : "95",
"temp-overheat" : "88",
"temp-target" : "75",
"temp-hysteresis" : "9",
"gpu-fan" : "0-85",
"auto-fan" : true,
"gpu-threads" : "1",
"log" : "5",
"shares" : "0",
"kernel-path" : "/usr/local/bin"
}
How much RAM do you have installed? I can run 4x5970s on 4GB RAM with a Sempron 140.

[Aurum]

I just installed 10 block erupters on a powered USB hub. (To amuse myself while I wait for my BuFfaLo Chips to arrive from Kansas )  I installed the driver for them and the hub. I can mine at bitminter no problem. But, I cannot initialize them using cgminer. What am I doing wrong? TIA
cgminer-nogpu.exe -c Block_Erupters.conf

{
"pools" : [
   {
      "url" : "stratum+tcp://pool1.us.multipool.us:3350",
      "user" : "aurum.be1",
      "pass" : "pw"
   },
   {
      "url" : "stratum+tcp://pool1.us.multipool.us:3333",
      "user" : "aurum.be1",
      "pass" : "pw"
   }
]
,
"api-mcast-port" : "4028",
"api-port" : "4028",
"hotplug" : "5",
"log" : "5",
"shares" : "0",
"kernel-path" : "/usr/local/bin",
"icarus-options" : "115200:1:1",
"icarus-timing" : "3.0=100",
"usb" : ":10"
}

[kano]

I can mine at bitminter no problem. But, I cannot initialize them using cgminer. What am I doing wrong? TIA
...

i.e. you didn't use the correct drivers for cgminer.

... https://bitcointalk.org/index.php?topic=28402.msg3231722#msg3231722

[Aurum]

I can mine at bitminter no problem. But, I cannot initialize them using cgminer. What am I doing wrong? TIA
...

i.e. you didn't use the correct drivers for cgminer.

... https://bitcointalk.org/index.php?topic=28402.msg3231722#msg3231722

Thx Kano. I assumed the SiLabs driver it wanted to install when I plugged the first one in was the best.
So Zadig replaced all 10 Eruptor drivers with the WinUSB driver. But, when I start cgminer it quickly discovers 6 but cannot see the other 4. I have 6 plugged into a powered USB hub and 4 plugged into MB USB ports. I don't know how to tell which is which. I guess I'll unplug them all and plug them in one by one and see what happens.