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

[-ck]

sorry for double post,

No it does not work with the bitcoin-qt in solo mode.

The rpcallow is now on the raspi ip, the win version was not started and the message appear
direct after starting cgminer. It probe for alive pool and display the error, after that it switch over and work fine.

Maybe it is an issue on the raspi Network?

Did you build the raspi version with libcurl ? If you don't then it can't do solo mining.

[o_solo_miner]

it is quite a bit ago, I will check and recompile.
Is there a way how to check that quick?

EDIT: Libcurl was installed so I normaly build with it, but to be shure I am recompiling now.
The raspi2 is fast compared to the old version....

[o_solo_miner]

Recompile finished, same result, sorry.

Libcurl4-dev was not switched off.

Anything else I might have ruined?

[-ck]

Recompile finished, same result, sorry.

Libcurl4-dev was not switched off.

Anything else I might have ruined?

Not in cgminer. What does your rpcallowip have in it on your bitcoin.conf ?

[o_solo_miner]

  Sorry for wasting your time, found the problem, just checking...

[-ck]

Please be aware that solo mining to your own bitcoind/bitcoin-qt does NOT support M of N addresses (i.e. bitcoin addresses starting with 3). At some stage in the future I will fix this in cgminer but in the interim please only use 1x addresses when mining to your own bitcoin node.

[o_solo_miner]

Yes, it was my Fault!

I got a copy of bitcoin.conf for not screwing it up, and guess what I did?

EDIT:
To clear that for all with a similar problem:
The bitcoin.conf got not the correct rpcallowip entry for the client.

I wish I could delete the last entrys in that thread to not beeing identifyed as a "Honk".

Thank you for your help!

I did not see it until I checked the vlan on the cisco and I've got a copy of the conf in the same folder.
The date of the files showed me what I edited.

EDIT: I wonder about the number of connection to bitcoin-qt, both cgminer run "visible" good.
What kind of problem should I expect then on cgminer or bitcoin-qt?

[wariner]

Hi -ck,

thank you for your job!

I have 3 Ant S5, if i install this version http://ck.kolivas.org/apps/cgminer/antminer/s5/4.9.0-150105/ the fan go to constantly 100%.
I'm doing something wrong?

[pinhead666]

Is it possible to modify cgminer is S3 to look for more than two hashboards? I have a test-board with 4-connectors (Tx and Rx lines + pwm). And I can also test with more devices connected to usb-hub. At least kernel log shows that software installs drivers for cp2102 usb/uart adapter (even with pic in circuit).

[kano]

Is it possible to modify cgminer is S3 to look for more than two hashboards? I have a test-board with 4-connectors (Tx and Rx lines + pwm). And I can also test with more devices connected to usb-hub. At least kernel log shows that software installs drivers for cp2102 usb/uart adapter (even with pic in circuit).

Prolly change one of the S3 settings I put in driver-bitmain.h ...

Look for #ifdef USE_ANT_S3

No I've no idea ... it would probably be the combination of the settings re: the chains etc. to deal with more work.

However, the number of boards/chips is returned by the fpga/hardware and that doesn't come from cgminer settings
(if it's not seeing the extra boards/chips)

[pinhead666]

Is it possible to modify cgminer is S3 to look for more than two hashboards? I have a test-board with 4-connectors (Tx and Rx lines + pwm). And I can also test with more devices connected to usb-hub. At least kernel log shows that software installs drivers for cp2102 usb/uart adapter (even with pic in circuit).

Prolly change one of the S3 settings I put in driver-bitmain.h ...

Look for #ifdef USE_ANT_S3

No I've no idea ... it would probably be the combination of the settings re: the chains etc. to deal with more work.

However, the number of boards/chips is returned by the fpga/hardware and that doesn't come from cgminer settings
(if it's not seeing the extra boards/chips)

Thanks for an answer. I suspected that pic's code is written for only two boards (what I can understand from hex-dump of it). I then connected second pic to hub's port 2 and that way kernel log shows both pic's are installed with bitmainbl driver. Second pic sends test-work to hashboards but it won't get any work after that. Pic nr 1 works though with 2 boards (and second pic installed). That made me wonder if I can modify drivers to look for more than 2 boards. It is also possible that only first downstream usb-port works with these drivers. I have to do more testing.

[kano]

S3 is serial (yeah that sux) so you'd prolly have to modify the detection to detect and configure both devices if they show up as 2 serial devices.

[pinhead666]

Thanks again. These answers clarified many things and helped to narrow down my options.

[d57heinz]

Kano,
edit:  meant to direct this to ck,

got a quick question.. is there a way you can incorporate the voting process into cgminer so that each mining machine can vote individually vs the pool wallet submitting the vote? What some people would like to do is stay on the pool of there liking and still vote with there mining hash regardless of the pool operators views;).  By allowing only the pool operators control, that centralizes this rather drastically!  Sorry if this was asked or is a stupid question:)

edit,,  i guess problem would still be that there is two asic manufacture backing two of the largest pools,  so really i think the issue still lies in getting mining back to decentralized means..   Ive always thought one way to get it back decentralized would be to limit the amount of hash each node can have.  One can only run so many instances of bitcoind before it becomes a problem... i dont know what would be a good number to limit it too.. say 100 ths per node.  im sure that could be worked out later maybe with the size of the network thats a bit too small. It would then become too costly to setup a massive farm or at least more prohibitive anyway.. not like there wouldnt be ways to circumvent it... Just thinking about all the issue we have with centralization.. Block propagation, network protocols being held up by one entity, and im sure it goes on and on..

Best Regards
d57heinz

[-ck]

got a quick question.. is there a way you can incorporate the voting process into cgminer so that each mining machine can vote individually vs the pool wallet submitting the vote? What some people would like to do is stay on the pool of there liking and still vote with there mining hash regardless of the pool operators views;).  By allowing only the pool operators control, that centralizes this rather drastically!  Sorry if this was asked or is a stupid question:)

I assume by "vote" you mean what transactions are included and what goes into the coinbase, in which case the answer is realistically no. By design pooled mining gives that control to the pool. While it is possible to redesign the protocol so the miners design their own work and send it to a pool, the ability to make the pools and mining majority change to such a system is quite beyond the level of improbability. The bulk of the hashrate you see on the network is not personal miners as they only account for maybe 5% of the network now and swaying those will have zero net effect and only make mining more difficult for them. If you're talking about political debates of BIP101 versus mainline bitcoin, that's even more unlikely as the choice of protocol is done and verified at the pool mining node, and pools consciously choose what software to run on their bitcoin node. When you are pooled mining, you are all mining to that pool's bitcoin node so you can't change its protocol from the mining end in any meaningful way.

[Lorker]

ty for the good software

[matt4054]

ty for the good software

...and thank you for the long term, continuous, constant and consistant support throughout the Bitcoin ages

[durandal]

I'm not sure where the best place to put this is, so I'll leave it here to save someone else a couple of hours of tinkering.

I'm running an Antminer U3, and was struggling to get decent speeds on a linux host, and the bitmain branch kept going "errr, no". The windows binary was spot on, but the windows machine sounds like a jet engine so for the sake of marital harmony and my leccy bill, it was going to have to be the linux netbook. I'll fiddle with MinePeon at some point to make it really small, but this is a good way to start.

The steps I followed are:

Clone the 4.9.2 git tree into a directory.
git clone [latest tree]
./autogen.sh
./configure --enable-icarus
make (and when you're happy it works, sudo make install)

To start, I've made a little script. So nano (or vim, or emacs, or edlin if you're hardcore) cgm

Code:

#/bin/bash
sudo cgminer --au3-freq 250 --au3-volt 810

chmod +x cgm

And away you go. The frequency/voltage settings are at the higher end of the chart, but heatsinks are only warm to the touch and CGMiner is reporting 63Gh/s, which is well within expectations. No excessive rejections either, so it seems like the right trade-off between speed and longevity.


Edited to correct a typo

[altcointalk14]

Hello,

I am not sure I have really found 3 blocks in 8 days.   Looking for advice on how to really check if this is a lucky machine or NOT.
(side note- I am really tempted to solo on CKSolo, but the ckproxy setup has troubled me)

This machine cgminer is still running, before it reboots and I loose the data, is there anyway to grab the logs/shares info from this miner?  Bitmain S4.

The 10.70.130.10:8885 is my stratum proxy server I gather all my mining power then point to antpool.

I did switch all my hashrate over to nicehash for about 8-10 hours about 5 of 6 days ago and maybe that is where the found blocks came from, but I have 50+ machines and none show any blocks except this one. 

Here is where I am confused:
79102380900   btc diff
273901451   S4 page below best share
1449843295860  stratum proxy server gui best share

[altcointalk14]

OK, I found this in the cgminer readme, so probably no history on the machine?? 

Anyway to add this option to the existing bitmain system cgminer start command?  (S4 says 3.2GB free, but S7 says only 115MB free)

from cgminer readme:

If you start cgminer with the --sharelog option, you can get detailed
information for each share found. The argument to the option may be "-" for
standard output (not advisable with the ncurses UI), any valid positive number
for that file descriptor, or a filename.

To log share data to a file named "share.log", you can use either:
./cgminer --sharelog 50 -o xxx -u yyy -p zzz 50>share.log
./cgminer --sharelog share.log -o xxx -u yyy -p zzz

For every share found, data will be logged in a CSV (Comma Separated Value)
format