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

[kano]

go through each device on zadig ensuring its on winusb

did that. they are. also followed the steps here:
https://bitcointalk.org/index.php?topic=220450.0

everything has a comport. after following those steps (including a reboot) bfgminer's working again, but cgminer remains with the same error.

That page is for the old serial-USB driver for 3.1.1 and before.

For 3.2.0 (and later) you need to use the WinUSB driver as I mentioned above.

[arklan]

go through each device on zadig ensuring its on winusb

did that. they are. also followed the steps here:
https://bitcointalk.org/index.php?topic=220450.0

everything has a comport. after following those steps (including a reboot) bfgminer's working again, but cgminer remains with the same error.

That page is for the old serial-USB driver for 3.1.1 and before.

For 3.2.0 (and later) you need to use the WinUSB driver as I mentioned above.

i now see that zadig did infact... not do it's job. it's running now, as admin. waiting on completion.

thanks for the help. i feel like an idiot.

[arklan]

did i correctly see that cgminer doesn't like usb 3.0 hubs?

if so, that is likely the explination why, when i just sucessfully installed the winusb driver, it gave no errors, but ALSO said no devices found.

wait, no... cause if that was the case it wouldn't find them ever, regardless of driver...

argh.

[kano]

crap got another after restarting
http://screencast.com/t/8K88aK1xwNky
I guess i have to only watch the ghs. But i still dont trust it was hotplugged
and no i didnt unplug them.

Yes windows seems to sometimes get an error that on linux means to drop the device.
On windows it may just mean I need to reinitialise it.
It's at the top of the todo list now.

Anyway once it does drop the device (ZOMBIE), hotplug will pick it up again within 5 seconds.
You can see in the API ... the number of devices will grow by 1 each time this happens (and also each time, a new ZOMBIE will appear with nodev = true)
The ZOMBIE name comes from the fact that I will (soon?) come up with a way to map the hotplug devices back onto the old matching ZOMBIE device ... and revive it ... it's not completely dead yet

Anyway, this is all display related, the total hashing power should still be the same with a small loss per device (of between 0 and 5 seconds) each time one goes ZOMBIE and then hotplugs.
The 'going ZOMBIE' process is immediate - no delay at all, as soon as a short list of errors occurs, it a slam dunk on the head, ZOMBIE straight away.
If I work out a tidy way to revive it before giving up on it, the only gain will be the time from ZOMBIE to hotplug that is currently lost.
It will simply be the underlying usbutils code that makes the ZOMBIE flag, intercepting the ZOMBIE happening and trying to revive it before giving up on it.

[kano]

did i correctly see that cgminer doesn't like usb 3.0 hubs?

if so, that is likely the explination why, when i just sucessfully installed the winusb driver, it gave no errors, but ALSO said no devices found.

wait, no... cause if that was the case it wouldn't find them ever, regardless of driver...

argh.

No, someone did report a problem with USB 3.0 once before
It may indeed be, that this is causing your problem also.
OK USB 3.0 also now added to the top of the todo list also.

[kano]

... and yes this is all Windows ... just in case anyone else was wondering.

Linux should hopefully be rock solid.

[bitpop]

I reported USB 3 issues, but it was mostly when using more than a few devices. 1-2 should work fine. But usb 3 is unreliable at this time. Especially with things like mobo drivers trying to add stupid shit like UASP and USB SCSI.

[Schland]

settings for mining BTC? ... thread concurrency, shaders, ...

Those aren't settings for BTC mining.

Thanks for trying to help, anyone else?

[bitpop]

I dont get it. What the hell could they be for then? LTC?
What you're looking for is going to be ages old when gpu mining first started. Search for those keywords and settings vs "guide"

settings for mining BTC? ... thread concurrency, shaders, ...

Those aren't settings for BTC mining.

Thanks for trying to help, anyone else?

[os2sam]

Does anyone have the link to the CGmimer fine tuning settings for mining BTC? I used the search but could not find it. It's the one around thread concurrency, shaders, worksizes and their factors.

I used this, two years ago.

https://en.bitcoin.it/wiki/Mining_hardware_comparison

[steganos]

I wrote a bash script that will check to see if a newer version of cgminer is available.  If there is a newer version the script will download the pre-built version and extract it. 

Here is the script if anybody is interrested:

Code:

#!/bin/bash
server=http://ck.kolivas.org
path=/apps/cgminer/

#; Get the file name for the latest version
cgminer=`wget -q -O- $server$path | egrep -o cgminer-[0-9\.]+-x86_64-built.tar.bz2 | sort -V | tail -1`

#; Get the folder name that the downloaded file should be extracted to
folder="${cgminer%.tar.bz2}"
version="${folder%-x86_64-built}"
version="cgminer ${version##*-}"

echo "Latest version is $version"

#; Check to see if this version is already installed.
if [ ! -d $folder ]
then
        echo "Downloading $version..."
        #; Download the file
        wget -q $server$path$cgminer

        echo "Installing $version..."
        #; Extract cgminer
        tar -xjf $cgminer
        #; Create a symbolic link to the new folder so we can keep older versions and the latest version can always be found in "cgminer"
        if [ -L cgminer ]
        then
                rm cgminer
        fi
        ln -s $folder cgminer
        #; Remove the downloaded file
        rm $cgminer
else
        echo "Already have $version"
fi

[Dunkelheit667]

Strange things happened here on my rpi (Raspbian wheezy). 3.2.0 detected my Lancelot (BLT), plug and pray is working fine. But no luck with any of my Block Erupter USBs, not detected, no plug and play at all.

All devices connected using the same powered usb 2.0 hub. These are working fine with 3.1.1 using -S options.

Whenever I quit 3.2.0 and try running 3.1.1 again, it can't access the USB devices anymore, have to reboot the rpi...

Any hints are welcome.

[NR3000]

I'm getting errors in os x 10.8.3. Any ideas how to fix it?

./configure --enable-scrypt --enable-icarus PKG_CONFIG_PATH=/usr/local/opt/curl/lib/pkgconfig CFLAGS="-g -O2 -Wall"
sudo make install

"
<..>
usbutils.c:2576: error: redefinition of ‘union semun’
usbutils.c: In function ‘resource_unlock’:
usbutils.c:2878: warning: implicit declaration of function ‘semtimedop’
make[1]: *** [cgminer-usbutils.o] Error 1
make: *** [install-recursive] Error 1
"
Thank you!

[fydel]

Strange things happened here on my rpi (Raspbian wheezy). 3.2.0 detected my Lancelot (BLT), plug and pray is working fine. But no luck with any of my Block Erupter USBs, not detected, no plug and play at all.

All devices connected using the same powered usb 2.0 hub. These are working fine with 3.1.1 using -S options.

Whenever I quit 3.2.0 and try running 3.1.1 again, it can't access the USB devices anymore, have to reboot the rpi...

Any hints are welcome.

I am having exactly the same problem on the pi. My ZTEXs Quads are recognised without problems but the Eruptor Gizmodos won't show up.
libusb-1.0.0 is installed.

Perhaps it is working with Kano because he is using the Fedora dist on the Raspberry instead of Wheezy?

[ewitte]

First time messing with Linux version on Raspberry Pi does it just cancel out if not mining devices are found?  Its 2-3 months before my Avalon devices come in.

[kano]

Strange things happened here on my rpi (Raspbian wheezy). 3.2.0 detected my Lancelot (BLT), plug and pray is working fine. But no luck with any of my Block Erupter USBs, not detected, no plug and play at all.

All devices connected using the same powered usb 2.0 hub. These are working fine with 3.1.1 using -S options.

Whenever I quit 3.2.0 and try running 3.1.1 again, it can't access the USB devices anymore, have to reboot the rpi...

Any hints are welcome.

pastebin (don't paste here) the output of the end of 'dmesg' when you plug the AMU in
also 2 cgminer outputs:
1) cgminer -n
2) cgminer -D -T -verbose ... your options ... 2> debug.log
and let that run for about 30 seconds - and pastebin it or upload the log somewhere

I don't have a powered hub, I guess I better get one to see what's happening myself
(tried to get one last week and oddly none of the shops I went to had one ... except one that was crap)

One other thing to try is to leave it unplugged for about a minute then see if that fixes it
I had mine go into a weird state not being recognised after using different OSs and being left unplugged for a while fixed it for me
The error I got when this happened in the debug log was an error 4 setting the baud
I might try see if that is reproduceable - the error and the unplug fix, and if so, when it happens print a message to do that.

[kano]

First time messing with Linux version on Raspberry Pi does it just cancel out if not mining devices are found?  Its 2-3 months before my Avalon devices come in.

Not in the latest version.

[ewitte]

Not in the latest version.

Odd.  Get this on the benchmark (immediately )

 [2013-05-31 15:00:09] Started at
 [2013-05-31 15:00:09] Pool: Benchmark
 [2013-05-31 15:00:09] Runtime: 380564 hrs : 0 mins : 9 secs
 [2013-05-31 15:00:09] Average hashrate: 0.0 Kilohash/s
 [2013-05-31 15:00:09] Solved blocks: 0
 [2013-05-31 15:00:09] Best share difficulty: 0
 [2013-05-31 15:00:09] Share submissions: 0
 [2013-05-31 15:00:09] Accepted shares: 0
 [2013-05-31 15:00:09] Rejected shares: 0
 [2013-05-31 15:00:09] Accepted difficulty shares: 0
 [2013-05-31 15:00:09] Rejected difficulty shares: 0
 [2013-05-31 15:00:09] Hardware errors: 0
 [2013-05-31 15:00:09] Utility (accepted shares / min): 0.00/min               
 [2013-05-31 15:00:09] Work Utility (diff1 shares solved / min): 0.00/min

 [2013-05-31 15:00:09] Stale submissions discarded due to new blocks: 0         
 [2013-05-31 15:00:09] Unable to get work from server occasions: 0             
 [2013-05-31 15:00:09] Work items generated locally: 0
 [2013-05-31 15:00:09] Submitting work remotely delay occasions: 0             
 [2013-05-31 15:00:09] New blocks detected on network: 0

 [2013-05-31 15:00:09] Summary of per device statistics:

[pekv2]

Something I noticed that last version didn't do.

This version is frequently changing difficulty. Is this suppose to happen? Scrypt mining for litecoin.

Code:

Pool 0 difficulty changed to

 Accepted f3c31b0a Diff 208/128 GPU 0
 Pool 0 difficulty changed to 64.000000
 Accepted 81bdd518 Diff 157/128 GPU 1
 Accepted f732d5b3 Diff 67/64 GPU 1

[ewitte]

Never mind

# ./cgminer -D -T --verbose
 [2013-05-31 15:02:15] Started cgminer 3.2.0
 [2013-05-31 15:02:15] All devices disabled, cannot mine!
#