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

[sirky]

Did you fix the 0 byte issue?

No, it's happening on 3 machines, any of which have SDK 2.4 on them.  I'm trying to roll back to 2.1 on some of them, but they're giving me some issues.

Ah, just sounded like you may have.

Carry on!

[Meatball]

Trying to figure out that 0 byte error.  Here's what I'm seeing on one of my machines running Win7 x64 with Catalyst 11.5 and SDK 2.1 and 2 Sapphire 5830's.

Code:

Problem signature:
  Problem Event Name:	APPCRASH
  Application Name:	cgminer.exe
  Application Version:	0.0.0.0
  Application Timestamp:	4e3006d7
  Fault Module Name:	cgminer.exe
  Fault Module Version:	0.0.0.0
  Fault Module Timestamp:	4e3006d7
  Exception Code:	c0000005
  Exception Offset:	00003915
  OS Version:	6.1.7601.2.1.0.256.48
  Locale ID:	1033
  Additional Information 1:	0a9e
  Additional Information 2:	0a9e372d3b4ad19135b953a78882e789
  Additional Information 3:	0a9e
  Additional Information 4:	0a9e372d3b4ad19135b953a78882e789

[Meatball]

Think I half figured out the 0 byte crash.  It's phatk and since by default cgminer uses phatk it crashes.  Tossing "-k poclbm" in as a switch and it fires right up.

Now, why phatk would cause the crash, I don't know...

[xcooling]

code has been fixed to prevent looping. (start /WAIT

Code:

REM Grab Local Machine Name and set to variable %host%

set host=%COMPUTERNAME%

REM Check variable %host% to determine and launch correct machine subroutine.

if %host% == Miner1 goto:Miner1
if %host% == Miner2 goto:Miner2
goto:eof

REM Individual machine launch subroutines - Timout value to allow for MSI Afterburner to fully load before miners start up.

:Miner1
echo Miner 1 Batch Startup...
TIMEOUT /T 75
cd /d C:\cgminer-1.5.1
start "Miner 1 CGMiner Status" /WAIT /AFFINITY 0x1 /NORMAL cgminer -o http://mainserver:port -u userid -p password -I 8 -o http://backupserver:port -u userid -p password
goto:eof

:Miner2
echo Miner 2 Batch Startup...
TIMEOUT /T 75
cd /d C:\cgminer-1.5.1
start "Miner 2 CGMiner Status" /WAIT /AFFINITY 0x1 /NORMAL cgminer -o http://mainserver:port -u userid -p password -I 8 -o http://backupserver:port -u userid -p password
goto:eof

:eof

[JinTu]

What is wrong with  leaving in the manual override option?  What is crappy about leaving people  the option to not install more abstraction layers on things they don't want?  If I want enough abstraction layers I may as well just go use windows and autoupdates.

Did you actually read ./configure --help at some point? Is it your vax experience that did not prevent you from blaming pkg-config that is preventing you from exporting LIBCURL_LIBS and LIBCURL_CFLAGS before running ./configure if so you wish? --with-libcurl option was broken and hence removed. If you don't want to use pkg-config, fine, override the flags and there you go. I have just checked it by uninstalling pkg-config and it works fine.

Works for me as well ala:

Code:

CFLAGS=-I/usr/local/cuda/include LDFLAGS="-L/usr/local/cuda/lib -L/usr/lib/nvidia-graphics-270.41.19" LIBRARY_PATH="/usr/local/cuda/lib:/usr/lib/nvidia-graphics-270.41.19:/usr/local/lib" LIBCURL_LIBS=/usr/local/lib/libcurl.so LIBCURL_CFLAGS=-I/usr/local/include ./configure

[kano]

The point is I don't want to install pkg-config. I know exactly where say libcurl is,  as I of course build and install that from source.   If there is another dependency I would do the same.  At least with makefiles I could also just edit them easily myself quickly.  Autotools already makes that a longer cycle to fix when things go wrong.  What does it hurt to at least give the option for me to tell it exactly where to get the library it wants to link to?  Put in the README/INSTALL the version dependencies (though the library itself should offer some  way to check this, and let it abort if it is a bad version).

What is wrong with  leaving in the manual override option?  What is crappy about leaving people  the option to not install more abstraction layers on things they don't want?  If I want enough abstraction layers I may as well just go use windows and autoupdates.

Correct me if I'm wrong but it sounds like you're saying it should have two sources of that information rather than one ...
i.e. some extra copy of the configuration information that will of course be unreliable due to it being exactly that ...

If you have been building software since the VAX days (as I have also ... actually PDP11, but I digress)
then you should also realise that if that is what you are asking then you shouldn't be asking it ...

[The00Dustin]

In keeping up with this thread, I have seen two or three posts saying that shares/minute is a better indicator of performance than MHash/second.  However, I don't understand this argument.

I understand that higher MHash/s doesn't do any good if the GPU is unstable and shares aren't found or shares are rejected, but assuming a person keeps the GPU stable, doesn't the shares/minute vary with luck and difficulty while the MHash/s is what is actually being processed (regardless of luck and difficulty)?

IOW, when a person goes from 310 MHash/s to 320 MHash/s and shares/minute drops from 4.15 to 4.10, it seems logical to me to assume that the 320 MHash/s is better and the change in shares/minute is luck/difficulty related (at least assuming when comparing two relatively short runs [less than a day for sure, a couple days? I'm not sure]).

I'm not saying I don't believe shares/minute is a good thing to pay attention to as well, though.  Obviously if the same scenario involved going from 280 MHash/s to 330 MHash/s and suddenly getting 3 shares/minute instead of 4 shares/minute an issue with the miner or stability would be indicated.

That said, am I misinterpreting something, or is the suggestion that shares/minute is more important than MHash/s oversimplified?

[d3m0n1q_733rz]

In keeping up with this thread, I have seen two or three posts saying that shares/minute is a better indicator of performance than MHash/second.  However, I don't understand this argument.

I understand that higher MHash/s doesn't do any good if the GPU is unstable and shares aren't found or shares are rejected, but assuming a person keeps the GPU stable, doesn't the shares/minute vary with luck and difficulty while the MHash/s is what is actually being processed (regardless of luck and difficulty)?

IOW, when a person goes from 310 MHash/s to 320 MHash/s and shares/minute drops from 4.15 to 4.10, it seems logical to me to assume that the 320 MHash/s is better and the change in shares/minute is luck/difficulty related (at least assuming when comparing two relatively short runs [less than a day for sure, a couple days? I'm not sure]).

I'm not saying I don't believe shares/minute is a good thing to pay attention to as well, though.  Obviously if the same scenario involved going from 280 MHash/s to 330 MHash/s and suddenly getting 3 shares/minute instead of 4 shares/minute an issue with the miner or stability would be indicated.

That said, am I misinterpreting something, or is the suggestion that shares/minute is more important than MHash/s oversimplified?

The reason for taking shares/min into account rather than just MHash/s is because MHash/s is only part of the equation.  You have to take into account accuracy and efficiency.  You can compute hashes all day, but unless you're computing the right ones it's not going to do you any good.  So cgminer found a way to still maintain a decent hashing speed while also making sure that the hashes it computes are actually worth a darn.  See the full picture?

[ancow]

Correct me if I'm wrong but it sounds like you're saying it should have two sources of that information rather than one ...
i.e. some extra copy of the configuration information that will of course be unreliable due to it being exactly that ...

The way I understand it, he wants an additional, manual source of information to override the default one.
I don't see much wrong with this request...

[spiccioli]

In keeping up with this thread, I have seen two or three posts saying that shares/minute is a better indicator of performance than MHash/second.  However, I don't understand this argument.

I understand that higher MHash/s doesn't do any good if the GPU is unstable and shares aren't found or shares are rejected, but assuming a person keeps the GPU stable, doesn't the shares/minute vary with luck and difficulty while the MHash/s is what is actually being processed (regardless of luck and difficulty)?

IOW, when a person goes from 310 MHash/s to 320 MHash/s and shares/minute drops from 4.15 to 4.10, it seems logical to me to assume that the 320 MHash/s is better and the change in shares/minute is luck/difficulty related (at least assuming when comparing two relatively short runs [less than a day for sure, a couple days? I'm not sure]).

I'm not saying I don't believe shares/minute is a good thing to pay attention to as well, though.  Obviously if the same scenario involved going from 280 MHash/s to 330 MHash/s and suddenly getting 3 shares/minute instead of 4 shares/minute an issue with the miner or stability would be indicated.

That said, am I misinterpreting something, or is the suggestion that shares/minute is more important than MHash/s oversimplified?

The reason for taking shares/min into account rather than just MHash/s is because MHash/s is only part of the equation.  You have to take into account accuracy and efficiency.  You can compute hashes all day, but unless you're computing the right ones it's not going to do you any good.  So cgminer found a way to still maintain a decent hashing speed while also making sure that the hashes it computes are actually worth a darn.  See the full picture?

Sorry, but I don't see the full picture

What does it means when you say: "while also making sure that the hashes it computes are actually worth a darn"?

The shares it computes depends on the work the pool sent you (if you're in a pool), so they should be the same even using a different miner, or am I wrong?

spiccioli.

[The00Dustin]

In keeping up with this thread, I have seen two or three posts saying that shares/minute is a better indicator of performance than MHash/second.  However, I don't understand this argument.

I understand that higher MHash/s doesn't do any good if the GPU is unstable and shares aren't found or shares are rejected, but assuming a person keeps the GPU stable, doesn't the shares/minute vary with luck and difficulty while the MHash/s is what is actually being processed (regardless of luck and difficulty)?

IOW, when a person goes from 310 MHash/s to 320 MHash/s and shares/minute drops from 4.15 to 4.10, it seems logical to me to assume that the 320 MHash/s is better and the change in shares/minute is luck/difficulty related (at least assuming when comparing two relatively short runs [less than a day for sure, a couple days? I'm not sure]).

I'm not saying I don't believe shares/minute is a good thing to pay attention to as well, though.  Obviously if the same scenario involved going from 280 MHash/s to 330 MHash/s and suddenly getting 3 shares/minute instead of 4 shares/minute an issue with the miner or stability would be indicated.

That said, am I misinterpreting something, or is the suggestion that shares/minute is more important than MHash/s oversimplified?

The reason for taking shares/min into account rather than just MHash/s is because MHash/s is only part of the equation.  You have to take into account accuracy and efficiency.  You can compute hashes all day, but unless you're computing the right ones it's not going to do you any good.  So cgminer found a way to still maintain a decent hashing speed while also making sure that the hashes it computes are actually worth a darn.  See the full picture?

Sorry, but I don't see the full picture

What does it means when you say: "while also making sure that the hashes it computes are actually worth a darn"?

The shares it computes depends on the work the pool sent you (if you're in a pool), so they should be the same even using a different miner, or am I wrong?

spiccioli.

In addition to this, wouldn't that mean that a higher MHash/second is automatically better on cgminer?  The posts I am referring to don't seem to be scolding people for comparing MHash/second between another miner and cgminer, they seem to be scolding people for paying attention to the MHash/second in different runs of cgminer.  If cgminer was already dealing with accuracy and efficiency, wouldn't suggesting that a higher MHash/second in cgminer wasn't necessarily a good theng be spurious?

[plantucha]

In keeping up with this thread, I have seen two or three posts saying that shares/minute is a better indicator of performance than MHash/second.  However, I don't understand this argument.

I understand that higher MHash/s doesn't do any good if the GPU is unstable and shares aren't found or shares are rejected, but assuming a person keeps the GPU stable, doesn't the shares/minute vary with luck and difficulty while the MHash/s is what is actually being processed (regardless of luck and difficulty)?

IOW, when a person goes from 310 MHash/s to 320 MHash/s and shares/minute drops from 4.15 to 4.10, it seems logical to me to assume that the 320 MHash/s is better and the change in shares/minute is luck/difficulty related (at least assuming when comparing two relatively short runs [less than a day for sure, a couple days? I'm not sure]).

I'm not saying I don't believe shares/minute is a good thing to pay attention to as well, though.  Obviously if the same scenario involved going from 280 MHash/s to 330 MHash/s and suddenly getting 3 shares/minute instead of 4 shares/minute an issue with the miner or stability would be indicated.

That said, am I misinterpreting something, or is the suggestion that shares/minute is more important than MHash/s oversimplified?

The reason for taking shares/min into account rather than just MHash/s is because MHash/s is only part of the equation.  You have to take into account accuracy and efficiency.  You can compute hashes all day, but unless you're computing the right ones it's not going to do you any good.  So cgminer found a way to still maintain a decent hashing speed while also making sure that the hashes it computes are actually worth a darn.  See the full picture?

Man, you have to be minister or priest in church.
Great talk about full picture, only totally misleading.
some of your bright ideas I have to quote again:

"unless you're computing the right ones it's not going to do you any good." 
or
"making sure that the hashes it computes are actually worth a darn"   

[Internet151]

Any ETA on the windows binaries getting updated?

[kano]

The git source for this (and the latest version) is here:
https://github.com/kanoi/linux-usb-cgminer/blob/master/linux-usb-cgminer

The version below is old.

(2011-09-19: I've recently added 2 new comments about log files filling up disk space and enabling ssh login)
Check the git link to see the latest version.

Also, of course, since cgminer 2.* now does CPU/Fan control, you may not need the extra GPU/Fan control programs any more

---

How to setup a cgminer using xubuntu 11.04 live on a USB

Software
========
Short hardware comment:
Your mining computer doesn't need any HDD or CD/DVD/BD as long as it has at least 2GB of RAM, can boot USB, has some network connection to the internet and of course a reasonable mining ATI graphics card
... Or you can boot a windows PC with the USB to only do mining ... and ignore the system HDD ... wasting energy

To create the USB, you need of course a 4GB USB and temporarily need a PC with a CD (or DVD/BD) writer, a USB port and of course an internet connection to the PC

1) Download the xubuntu 11.04 desktop live CD iso for amd64
   ( look here for mirrors: http://www.xubuntu.org/getubuntu )

2) Burn it to CD then boot that temporarily on any PC with a CD/DVD/BD and a USB port (this and the next 2 step won't effect that PC)

3) Plug in your 4GB USB device and it should appear on the desktop - you can leave it's contents as long as there is at least 2.8GB free

4) Now run "Startup Disk Creator" in "Applications->System" (the system menu is the little rat in the top left corner)

(if you have no mouse you can get the menu with <ctr><esc> and navigate the menu with the arrow keys and <return> key)

From here select the boot CD as the "Source" and the USB as the "Disk to use"
lastly move the slider to 2GB for reserved extra space

The 2GB should be enough for modifications

Click: "Make Install Disk"
After about 10-15 minutes you have a base xubuntu 11.04 boot USB
(you can shut down this computer now)

5) Boot your cgminer PC with this USB stick, select "English"
   then select "Try Xubuntu without installing" and wait for the desktop to appear
   (this happens by default if you wait for the timeouts)

6) Start a terminal
   "Applications->Accessories->Terminal Emulator"

7) sudo apt-get install openssh-server screen

   if you have a problem here then it's probably coz the internet isn't available ... sort that   out by reading elsewhere about routers etc

sudo apt-get install fglrx fglrx-amdcccle fglrx-dev
   sudo sync
   sudo shutdown -r now

N.B. always do a "sudo sync" and wait for it to finish every time before shutting down the PC to ensure all data is written to the USB

9) sudo aticonfig --lsa
   this lists your ATI cards so you can see them
 sudo aticonfig --adapter=all --odgt
   this checks it can access all the cards ...

10) sudo aticonfig --adapter=all --initial
   this gets an error - no idea why but the xorg.conf is OK
 sudo sync
 sudo shutdown -r now

11) sudo aticonfig --adapter=all --odgt
   this checks it can access all the cards ...

12) get AMD-APP-SDK-v2.4-lnx64.tgz from
 http://developer.amd.com/sdks/amdappsdk/downloads/pages/default.aspx
  ( http://developer.amd.com/Downloads/AMD-APP-SDK-v2.4-lnx64.tgz )

 sudo su
 cd /opt
  (replace /home/ubuntu/ with wherever you put the file: )
 tar -xvzf /home/ubuntu/AMD-APP-SDK-v2.4-lnx64.tgz

 cd AMD-APP-SDK-v2.4-lnx64/
 cp -pv lib/x86_64/* /usr/lib/
 rsync -avl include/CL/ /usr/include/CL/
 tar -xvzf icd-registration.tgz
 rsync -avl etc/OpenCL/ /etc/OpenCL/
 ldconfig
 sync
 shutdown -r now

 You now have an OpenCL enabled xubuntu

13) cgminer:
 sudo apt-get install curl

 get the binary linux cgminer (see the bitcoin forum cgminer thread for where to get it)

 ./cgminer -n
   this shows you the GPU's it found on your PC

14) An OC option:
 sudo apt-get install libwxbase2.8-0 libwxgtk2.8-0

 http://sourceforge.net/projects/amdovdrvctrl/
  for an Over/underclocking application and get the file listed below then:
 sudo dpkg -i amdoverdrivectrl_1.2.1_amd64.deb

15) set the screen saver to ONLY blank ...

 Move the mouse to the bottom of the screen and you see a set of icons like on an Apple PC
 Click on Settings, then in the Settings window "Screensaver"
 Set "Mode:" to "Blank Screen Only"

Edit: originally left this somewhat useful addition out:
16) apt-get install ntpd
 An accurate clock is always a good idea

Initial setup complete.

========

If you want to SSH into the machine and run cgminer:
 From a terminal on the miner display each time after you boot:
  xhost +

Then after ssh into the machine:
 export DISPLAY=:0
before running cgminer

Also note, that you should force the screen to blank when mining if the ATI card is displaying the screen (using the screen saver application menu)
In my case it takes away 50Mh/s when the screen isn't blanked


This is of course just the basics ... but it should get you a computer up and running and able to run cgminer

[Viceroy]

How to setup a cgminer using xubuntu 11.04 live on a USB

This is an awesome post which, imho, deserves its own thread.

[miscreanity]

An especially usful addition would be to talk about the
radeonvolt tool to control voltage settings for stable
overclocking.

Thread is here: https://bitcointalk.org/index.php?topic=10228.msg146827#msg146827

It'd be great to pull GPU info similar to GPU-Z and edit it, but radeonvolt only supports reference 5xxx series cards currently. Until the lack of more diverse I2C code is expanded upon, its utility is virtually nil for 6xxx and non-reference cards in general.

[shaps]

[noob question]  Besides a GPU rig I also run a miner on my core2duo laptop sometimes just for fun (cpu only).  I've noticed that ufasoft does around 3x - 4x the Mhash that CGMiner does.  Any idea why or how I can get CGMiner to match the performance?

[The00Dustin]

[noob question]  Besides a GPU rig I also run a miner on my core2duo laptop sometimes just for fun (cpu only).  I've noticed that ufasoft does around 3x - 4x the Mhash that CGMiner does.  Any idea why or how I can get CGMiner to match the performance?

Install a 64-bit linux distro instead of running Windows?  (This is just a guess, the latest cgminer seems to come close to ufasoft on core 2 and older Intel processors in 64-bit Linux, but apparently [based on an earlier post in this thread] cgminer uses 64-bit optimizations to achieve that and doesn't have similar 32-bit optimizations coded).  The Windows binary is only written for 32-bit, so if I am correct in my assumption that your are running Windows, that would be the easiest solution.  The much more complex ones would be to either a) compile a 64-bit Windows version if you are running a 64-bit Windows OS or b) optimize the 32-bit code and compile a new 32-bit version.

Oh, and in either Windows compilation case, you'd also have to modify other code to get cgminer not to waste more than one core's worth of cpu on whatever it does in Windows that it doesn't do in Linux (even when only GPU mining in each).

[Meatball]

Oh, and in either Windows compilation case, you'd also have to modify other code to get cgminer not to waste more than one core's worth of cpu on whatever it does in Windows that it doesn't do in Linux (even when only GPU mining in each).

From what I gather, and I can't vouch for it's accuracy, the windows 100% CPU problem when mining has to do with the Windows OpenCL drivers.  There's a function built into them that allows the CPU to assist a GPU when it's overloaded.  So what happens is that when a GPU gets to 99/100% utilization, the CPU spins up to lend a hand, but since mining doesn't need it, the CPU just sits there running, waiting to help.

Or something to that effect...

[dostortugas]

How to setup a cgminer using xubuntu 11.04 live on a USB

This is an awesome post which, imho, deserves its own thread.

Reported to mods allready, nice post but just not useful in here.

(how to setup any ubuntu-on-a-pendrive has 100s of howtos in the internets... setting up a miner on pendrives is nothing else than on a normally installed system; although there are some limits with pendrives as you just should never hit apt-get upgrade if it is installed from a live-cd)