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

[RvdE]

[..]

Compiling 1.6.1 is a different story though. I notice the configure script checks for curl_easy_init and after that does the pkg-config check for libcurl. Since it doesn't know where to look for curl yet, the former check obviously fails. Also, you use LIBCURL_CPPFLAGS, but on FreeBSD (and possibly other systems), pkg-config only returns LIBCURL_CFLAGS. Since they should be the same, could you change it to make use of CFLAGS instead of CPPFLAGS ?

libcurl.m4 defines LIBCURL_CPPFLAGS and CFLAGS would be something completely different as far as I understand it, so you're missing the m4 file from some package.

The configure script from the tarball (1.6.2) contains the following:

Code:

pkg_cv_LIBCURL_CFLAGS=`$PKG_CONFIG --cflags "libcurl >= 7.15.6" 2>/dev/null`

It doesn't set LIBCURL_CPPFLAGS anywhere (which makes pkg-config detection obsolete IMO). Also, in the Makefile.am it should be changed to LIBCURL_CFLAGS instead of LIBCURL_CPPFLAGS.

[zaytsev]

Yay! It compiles with curl from RHEL5 again! CPU-only RPMs are ready: http://rpm.zaytsev.net/test/cgminer/ . I will edit the RPM post later.

[sharky112065]

If anyone is interested, thought this might be helpful to some.

If you run Cgminer on Windows 7 (should work on other versions) this startup .cmd will launch Cgminer using only one CPU core and outputs log info to a log file that has the date and time in the filename.

Open notepad.exe and put the following in (changing username and password and your servers of course):

Code:

rem del /F /Q *.bin

start /B /affinity 8 cgminer -o http://yourprimaryserver:port -u yourworkername -p your password -o http://yourbackupserver:port -u yourworkername -p yourpassword -I 9 2>"%date:~4,2%-%date:~7,2%-%date:~12,2%%time:~0,2%-%time:~3,2%-%time:~6,2%.log"

pause

This will use core number 4.

1st core = 1
2nd core = 2
3rd core = 4
4th core = 8
5th core = 10 ( hexadecimal 16 )
6th core = 20 ( hexadecimal 32 )

now if you want something to run on just one core, you pick the right number like this, just multiply by 2 until you get to the core you want.

if you want something to run on two cores, you ADD their values:

1st and 2nd core = 1 + 2 = 3
1st and 3rd core = 1 + 4 = 5
2nd and 3rd core = 2 + 4 = 6
2nd and 4th core = 2 + 8 = A

Same thing if you want more than two cores, just add each core numerical value up and use the sum as the /affinity parameter

[-ck]

Yay! It compiles with curl from RHEL5 again! CPU-only RPMs are ready: http://rpm.zaytsev.net/test/cgminer/ . I will edit the RPM post later.

Yes I figured out what I had broken from the old curl at last. Please note that the old curl does not have support for sockoptions that make longpoll reliable. So while it is compiling with older libcurls, it may not be as reliable with longpoll on some servers.

[abracadabra]

keep up the good work love the app

[skEwb]

I'm having some issues with the Windows release of 1.6.2. It will not activate 2/3 of my GPU's it fails to lock the thread on the GPU's. If I go back to 1.6.0 everything is fine. What sort of output do I need to provide for this issue to be troubleshooted? Screenshot or some sort of log?

[phase]

I wrote a dead-simple shell script to start cgminer, and it works, but the peculiar thing is that it starts up fine from an existing terminal, but it won't start if I just double-click the script and select run in terminal, or when I add it to Ubuntu start-up applications (or try to call it from another script). It didn't work before on Linuxcoin and now I'm still having this problem with Ubuntu. I've looked around everywhere but I can't find a working fix, any hints?

[The LT]

I wrote a dead-simple shell script to start cgminer, and it works, but the peculiar thing is that it starts up fine from an existing terminal, but it won't start if I just double-click the script and select run in terminal, or when I add it to Ubuntu start-up applications (or try to call it from another script). It didn't work before on Linuxcoin and now I'm still having this problem with Ubuntu. I've looked around everywhere but I can't find a working fix, any hints?

Either use an option to point it to the kernels or add a cd /path/to/cgminer before the line starting the binary.

Use a wrapper utility such as start-stop-daemon, suppress all output with 'really-quiet' flag.

[phase]

Either use an option to point it to the kernels or add a cd /path/to/cgminer before the line starting the binary.

Use a wrapper utility such as start-stop-daemon, suppress all output with 'really-quiet' flag.

Thanks, however the script already does that, it cd's to the cgminer dir and then ./ starts it with a config file argument, also located in the same directory. I suppose I'm asking in the wrong place, since I also have the same problem with a script that starts 4 instances of AMDOverdriveCtrl. The scripts both work fine when launched from an existing terminal (./script). It just doesn't work when I call the scripts from another script (or start-up script) or when I try to double-click it and select run in terminal. *sigh* Why is linux always such a b*tch, there ought to be a Steve Jobs for the linux community, to simplify these overly complicated, very basic tasks.

I don't know if I could use start-stop daemon, since I have to launch two instances of cgminer each with different parameters. Thanks for the heads-up though.

[Ghostofkobra]

Linux and Unix os's are build up around a few very easy concepts which makes them quite predictable.
I almost only work with Linux in commercial environments so i do not have any experience with Ubuntu...

Nevertheless its almost impossible to answer the question w/o knowing the output of the doubleclicked .sh file
seeing your script and maybe an "ls -l" of the directory where cgminer is located.

One reason might be that the user you are logged into X with does not have the correct permissions to access cgminer.
At least that is where i could but my 2 bitcents w/o knowing more.


Best
/GoK

[phase]

Thanks, I figured it out, it was missing the LD_LIBRARY_PATH which the script seems to need if you want to launch it without starting it from an existing terminal window. As for the script starting 4 instances of AMDOverdriveCtrl, I had to run that script directly instead of via a new terminal, and then apparently it works fine without terminal windows and without exiting. Some fine hours wasted nitpicking, but finally it works, time to move her to the basement and get a good (silent) night sleep.

If it can be of help to anyone, here are my scripts (modify, launch via start.sh and run 'directly', not in a new terminal):

start.sh

Code:

#!/bin/bash
./oca.sh &
sleep 5
gnome-terminal -x bash -c "./miner1.sh" &
gnome-terminal -x bash -c "./miner2.sh" &

oca.sh

Code:

#!/bin/bash
AMDOverdriveCtrl -i 0 gpu1.ovdr &
AMDOverdriveCtrl -i 10 gpu2.ovdr &
AMDOverdriveCtrl -i 3 gpu3.ovdr &
AMDOverdriveCtrl -i 4 gpu4.ovdr &

miner1.sh

Code:

#!/bin/bash
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH":/path/to/AMD-APP-SDK-v2.4-lnx64/lib/x86_64/"
cgminer -d 0 -c cfg.json --submit-stale

miner2.sh

Code:

#!/bin/bash
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH":/path/to/AMD-APP-SDK-v2.4-lnx64/lib/x86_64/"
cgminer -d 1 -d 2 -d 3 -c cfg.json -I 8 --submit-stale

[tenzor]

Segfault then trying to  switch from pool without LP to second pool with LP. Second pool was added manually after programm starts.
ubuntu 11.04 x32, cgminer 1.6.2, git version

[wknight]

I just wanted to say.. I moved to this mining software yesterday and noticed a huge increase.. thanks so much! Donation on its way!

[cablepair]

ive been using cgminer for awhile. I must say I absolutely love it. Favorite mining software by far. So whats new in 1.6.2?

thanks!

(mining at 3.5 g/hash exclusively with cgminer)

[-ck]

Thanks for positive feedback.

Major feature upgrade in the works. Somewhat below the radar and invisible at the moment, but available in the new git branch adl_support till it's complete.

[zaytsev]

FYI as I migrated to 1.6.2 from 1.6.1 and 1.6.0, my U went up from 4.0 to 4.2 and 4.4 respectively. Great job ck, I am really impressed highly appreciate your efforts  (in spite of the persistent curses bug lol).

There is something I find weird in 1.6.1/2 though. Does this E indicator actually make sense? The total E indicator seems to work properly.

Code:

 [P]ool management [G]PU management [S]ettings [D]isplay options [Q]uit
 GPU 0: [313.7 / 313.8 Mh/s] [Q:2  A:858  R:8  HW:0  E:42900%  U:4.39/m]

[cablepair]

I upgraded all my rigs to 1.6.2, one thing I am noticing is the m/hash is fluctuating a great deal more than it was. Before it only fluctuated maybe 5-10 mhash at most

now I am seeing fluctuations on some of my cards in the 50+ range

for instance my 5830 is going down to 283 than up to 360
my 5870's dont seem to be fluctuating as much as that but are fluctuating more than usual
same with the 6870's, and 5850's

any ideas?

[cablepair]

never mind. That was with default settings. When I started using -I 9 I am getting steady solid speeds.

Thanks again for such an awesome piece of mining software, one donation coming your way!

[PLaci1982]

Yet another newb section topic about a problem with the miner...

https://bitcointalk.org/index.php?topic=41152.0

[teflone]

crashes after I run cgminer

and answer the first 4 qquestions..

After password..  hit enter..  crashes the window..

any ideas?

Im win 7 64.