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

[ZombieDeity]

I just wanted to say that I <3 CGMiner.  Thank you for all your work in making it so great.  I was hesitant to switch from Phoenix, since I was getting good hash rates and decent stability, but I couldn't pass this up when I read the list of features.

I love the failover policy and all the measures you've implemented to prevent idling.  I'm getting consistently higher hash rates now and I can breathe easier knowing that if my primary pools get overloaded or DDOSed, I'm still making money somewhere.

Also, with Phoenix, I was running a separate process for each of my six GPUs (3x6990) in a bunch of screen sessions.  This made it really tricky to determine my overall performance and to quickly catch the GPUs that crapped out no and then.  I basically had to run "watch -n3 aticonfig --adapter=all --odgt" to get an indicator of performance from temps.  I'm really pleased that I can now see everything that's happening with all my GPUs as well as a nice summary of current and average hashrate.

Thanks again!

[1714097921]

1714097921

[1714097921]

1714097921

[JinTu]

Looks like the 1.5.3 configure script inadvertently dropped support for alternate cURL library path support via --with-libcurl:

Code:

$ ./configure --with-libcurl=/usr/local/lib/

configure: WARNING: unrecognized options: --with-libcurl
.
.

[zaytsev]

Looks like the 1.5.3 configure script inadvertently dropped support for alternate cURL library path support via --with-libcurl:

Yes, because it's the crappy way to do it. The right way is to export a variable to specify the PKG_CONFIG_PATH to the correct *.pc file. See my RPMs for reference...

[ancow]

Looks like the 1.5.3 configure script inadvertently dropped support for alternate cURL library path support via --with-libcurl:

Yes, because it's the crappy way to do it. The right way is to export a variable to specify the PKG_CONFIG_PATH to the correct *.pc file. See my RPMs for reference...

At least it worked on OSX. Using pkg-config exclusively breaks the whole OSX build...

[tigereye]

I'm getting this error on one of my rigs and am unsure how to resolve it:

Code:

Error: Getting program info CL_PROGRAM_BINARY_SIZES. (clGetPlatformInfo)

Has anyone else run into this?
Does anyone have a suggestion I could try?

Thanks in advance!

[Ali]

Is it possible to use a proxy-server with cgminer?

I can't find any option about it in the first post.

[Caesium]

I have a couple of improvement requests if I may

1) I've been purposefully feeding cgminer some getworks that will never be accepted for my pool testing purposes and noticed a pattern - even if the first share it sends back in a getwork is rejected, it still carries on processing that getwork anyway, sometimes leading to a second (and maybe even third or more?) reject out of the same work unit. I think cgminer should abandon a work unit if a share in it is rejected, as subsequent submissions are very likely to be rejected as well.

2) it would be nice to be able to PageUp and PageDown the log in the lower part of the display

3) optional millisecond timestamp display would be useful - maybe a --developer mode or something for people who want exact precision?

[zaytsev]

Is it possible to use a proxy-server with cgminer?
I can't find any option about it in the first post.

I think I told you that you don't need any options for that, just export the http_proxy variable before starting cgminer.

[Ali]

Is it possible to use a proxy-server with cgminer?
I can't find any option about it in the first post.

I think I told you that you don't need any options for that, just export the http_proxy variable before starting cgminer.

Sorry, I forgot about this.

I'm trying to run it ONLY for CPU-mining. Had to install AMD's Stream SDK.
When I start with this options:

C:\cgminer>cgminer.exe --cpu-threads 2 --gpu-threads 0 -o http://uswest.btcguild
.com:8332/

I get this error:

[2011-08-02 10:57:54] Error: Getting Device IDs (num)
clDevicesNum returned error

Can I skip that step somehow so it just mines with the CPU?

[sirky]

I tried cgminer on a couple of my boxes last night and woke up to find 3/4 GPU marked as DEAD on one and 2/4 DEAD on another. Trying to restart the DEAD GPU doesnt seem to do anything on either machine (restarting the ones still running did seem to restart them OK). Both boxes have been running for weeks using poclbm without errors. Running the 1.5.3 binary on Ubuntu (3xHD5870 and HD6310).

Anyone else seeing problems with DEAD GPU that won't recover unless cgminer is restarted?

BB.

I do too, but only on my linux boxes.

Actually, I have switched to persistent mode of Linuxcoin, and this does not seem to be happening anymore. Are you just running off of a USB stick? Perhaps persistent mode will fix you as well.

[dostortugas]

Can I skip that step somehow so it just mines with the CPU?

nothing personal Ali, it was just your post to let me become mad!

This thread is filled with CPU miners complaining about stuff, I do not get why you people still care that much about CPU mining?! My CPUs get 3MH/s each even my crappy 2 year old nvidia graphics does 28MH/s using less power... want it? 5 btc + shipping! If it is a game, pay for it before you play with it, you know ck's BTC address! You won't ever return any amount of investment with CPU mining. Let me do the math for you, ~2 valid shares/hour per core, 2 BTC sent to ck for fixing up some cpu related stuff in cgminer == 4.6 CPU-Years without increase in difficulty to pay the 2 BTC at current difficulty... think of the environment, you waste +95W for 2 BTC a year (reference is Athlon II X4 in this case and difficulty is not raised)...

Buy a $200 gpu and have it payed in 3-5 month including your powerbills... or buy the 2 BTC for $20

[twobits]

Looks like the 1.5.3 configure script inadvertently dropped support for alternate cURL library path support via --with-libcurl:

Yes, because it's the crappy way to do it. The right way is to export a variable to specify the PKG_CONFIG_PATH to the correct *.pc file. See my RPMs for reference...

Seems to be a way that works on more systems  and with local built binaries is the right way to do it.  Dependance on a packaging system  is the crappy way to do it.

[plantucha]

Can I skip that step somehow so it just mines with the CPU?

nothing personal Ali, it was just your post to let me become mad!

This thread is filled with CPU miners complaining about stuff, I do not get why you people still care that much about CPU mining?! My CPUs get 3MH/s each even my crappy 2 year old nvidia graphics does 28MH/s using less power... want it? 5 btc + shipping! If it is a game, pay for it before you play with it, you know ck's BTC address! You won't ever return any amount of investment with CPU mining. Let me do the math for you, ~2 valid shares/hour per core, 2 BTC sent to ck for fixing up some cpu related stuff in cgminer == 4.6 CPU-Years without increase in difficulty to pay the 2 BTC at current difficulty... think of the environment, you waste +95W for 2 BTC a year (reference is Athlon II X4 in this case and difficulty is not raised)...

Buy a $200 gpu and have it payed in 3-5 month including your powerbills... or buy the 2 BTC for $20

well fo number of reasons

because my

1. cpu make 20Mh/s
2. cpu is already in rig and doing nothing (load 1%)
3. in 5 years will cost 1BTC many thousands $ so even making almost nothing or loosing - will be profit in future if you keep it
4. because this is only decent cpu miner
5. On my rig is  GPU mining 20% more effective with combination Phoenix + Smartcoin
    it has more Mhash/s and failover with more options. Cgminer is running for CPU only under Smartcoin, so I don't care about Cgminer GPU hashing at all.

[Ali]

I don't want to talk about the reasons why I cpu-mine. It leads to nothing but arguments.

All I want to know is how to get cgminer to work without an openCl-device.

[spiccioli]

Hi ckolivas,

today I decided to give your miner a try, it has so many features that I could not resist anymore

Sadly, after a couple of minutes both GPUs in my rig ( a 5850 and a 5870 on a linuxcoin v0.2b system) are declared SICK and then DEAD.

This is without any fancy parameters, just -o -u -p and -I 8 (it is a dedicated rig) and -Q 2 to have a little more buffer (so I thought).

Is there some other parameter I can try?

spiccioli

ps. my rig is running since june with phoenix and poclmb without problems, so I don't think this is due to hardware/OS issues.

[sirky]

Hi ckolivas,

today I decided to give your miner a try, it has so many features that I could not resist anymore

Sadly, after a couple of minutes both GPUs in my rig ( a 5850 and a 5870 on a linuxcoin v0.2b system) are declared SICK and then DEAD.

This is without any fancy parameters, just -o -u -p and -I 8 (it is a dedicated rig) and -Q 2 to have a little more buffer (so I thought).

Is there some other parameter I can try?

spiccioli

ps. my rig is running since june with phoenix and poclmb without problems, so I don't think this is due to hardware/OS issues.

ckolivas is on a two week vacation.

Are you running persistent? I had the same problem before I starting running in persistent mode with a r/w file.

[dostortugas]

I don't want to talk about the reasons why I cpu-mine. It leads to nothing but arguments.

All I want to know is how to get cgminer to work without an openCl-device.

I'm being a bit of an eco freak there , you can do what ever you want... but ck wont do anything to cpu support if not paid for so I guess the calculation is important even if your cpu does 25MH/s!

Without opencl it should just say that there is no opencl and still compile without oCL support, have you tried with opencl (sdk, etc) installed to your system? (even if you do not use it...)

[Ali]

Yeah, I installed it (and therefore got rid of the missing openCl.dll-error).

Now I get this:

C:\cgminer>cgminer
[2011-08-02 14:51:15] Error: Getting Device IDs (num)
clDevicesNum returned error

[spiccioli]

Hi ckolivas,

today I decided to give your miner a try, it has so many features that I could not resist anymore

Sadly, after a couple of minutes both GPUs in my rig ( a 5850 and a 5870 on a linuxcoin v0.2b system) are declared SICK and then DEAD.

This is without any fancy parameters, just -o -u -p and -I 8 (it is a dedicated rig) and -Q 2 to have a little more buffer (so I thought).

Is there some other parameter I can try?

spiccioli

ps. my rig is running since june with phoenix and poclmb without problems, so I don't think this is due to hardware/OS issues.

ckolivas is on a two week vacation.

Are you running persistent? I had the same problem before I starting running in persistent mode with a r/w file.

Yes, I'm persistent.

I've found that with a -g 1 it does not happen anymore (it's been running for half a hour now), but I loose a couple Mh/s.

spiccioli.

ps. I had to install curl since it's not installed in linuxcoin by default.

[sirky]

Hi ckolivas,

today I decided to give your miner a try, it has so many features that I could not resist anymore

Sadly, after a couple of minutes both GPUs in my rig ( a 5850 and a 5870 on a linuxcoin v0.2b system) are declared SICK and then DEAD.

This is without any fancy parameters, just -o -u -p and -I 8 (it is a dedicated rig) and -Q 2 to have a little more buffer (so I thought).

Is there some other parameter I can try?

spiccioli

ps. my rig is running since june with phoenix and poclmb without problems, so I don't think this is due to hardware/OS issues.

ckolivas is on a two week vacation.

Are you running persistent? I had the same problem before I starting running in persistent mode with a r/w file.

Yes, I'm persistent.

I've found that with a -g 1 it does not happen anymore (it's been running for half a hour now), but I loose a couple Mh/s.

spiccioli.

ps. I had to install curl since it's not installed in linuxcoin by default.

Do you have linuxcoin final? Everything should just work with it.