Topic: OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.0 - page 107. (Read 5805527 times)

Hi all.
Just got an antminer U2. for trial but cant get it to run. I keep getting  " err-99 libusb failed to start "
Running win 7  64 bit, installed driver using zadig , WinUSB v6.1.7600.16385 .
Tried driver updates, running in and out of admin, restarting and or repluging after , even get this message with just Cgminer exe , no u2 , no command line.
Got me frizzed.
Any help would be good !

Thanks, that was fast. It works now.

And one question: In my driver (from scanwork) i return the number of hashes done (not nonces found), but in this case the hashrate does not show the correct results.
If I switch to nonces, the results for 1min, 5min etc are correct, but will loose the information about the actual hashes done ... is there a way/API to have them both or i should do it inside get_statline_before() call?
EDIT: Ended with a new line of details in get_statline() showing the expected hashrate based on the hashes done. /EDIT

Oh and one more minor bug - the details returned from get_statline_before() are destroyed and do no longer appear when zero api is called with 'all,true', but is OK with false i.e. print_summary() is causing it.
Te reason is that at line 8567 inside that function the calls are set to blank_get_statline_before and never restored

Okay, fixed it in git.

Ah yes, out of sight, out of mind because of it too. I wasn't even sure there was anywhere I could even test it to fix it.

DeepBit is still getwork only ... almost dead, because of that.

Hi all.
I've made a full release of linux 4.2.2a/bab binaries in my cgminer-binaries git.
https://github.com/kanoi/cgminer-binaries

As before, there are Fedora 18, Ubuntu 11.04, RPi Arch and RPi Raspbian
... and of course now Ant OpenWrt

The Ant OpenWRT binary cgminer-ants1-4.2.2a-7b8fb34 is (as always) described here:
https://github.com/kanoi/cgminer-binaries/tree/master/AntS1

To get the 64 bit xubuntu 11.04 binary:
wget https://github.com/kanoi/cgminer-binaries/raw/master/Ubuntu_11.04_x86_64/cgminer-4.2.2a
chmod +x cgminer-4.2.2a
mv cgminer-4.2.2a cgminer

The 5 others are:
https://github.com/kanoi/cgminer-binaries/raw/master/Fedora18_x86_64/cgminer-4.2.2a
https://github.com/kanoi/cgminer-binaries/raw/master/RPi_Arch/cgminer-4.2.2a
https://github.com/kanoi/cgminer-binaries/raw/master/RPi_Arch/cgminer-4.2.2bab
https://github.com/kanoi/cgminer-binaries/raw/master/RPi_Raspbian/cgminer-4.2.2a
https://github.com/kanoi/cgminer-binaries/raw/master/RPi_Raspbian/cgminer-4.2.2bab

All 4.2.2a are all USB only:
CFLAGS="-g -W -Wall" ./autogen.sh --enable-avalon --enable-bflsc --enable-bitforce --enable-bitfury --enable-drillbit --enable-hashfast --enable-icarus --enable-klondike --enable-modminer

All RPi 4.2.2bab are BlackArrow Bitfury GPIO and all USB:
CFLAGS="-g -W -Wall" ./autogen.sh --enable-avalon --enable-bflsc --enable-bitforce --enable-bitfury --enable-drillbit --enable-hashfast --enable-icarus --enable-klondike --enable-modminer --enable-bab
If you only have BlackArrow Bitfury GPIO devices - i.e. no USB mining devices attached also - then you can add the command line option --usb :0 when you run cgminer and it won't look for any USB devices

The BaB driver with 7 boards uses less than 10% CPU on RPI Arch - and for my 7 BA BaB boards, mines at over 270GH/s
It seems to be limited to 2 boards per SPI channel. More tends to get all sorts of errors.
Will look into resolving that ... one day.

The -g (instead of -O2) means it's a debug build so if anyone finds a problem and has core dumps enabled, it will dump a much more useful debug core.

Note I have binary folders of ckolivas official release files in my binaries git also, for if you can't get to his downloads
To get them you select the folder (e.g. 4.2.2) then click on the file you want then right-click save-as the "View Raw" link.

Where would you use getwork today?

It seems getwork is broken with this release (and 4.2.1) - all shares are rejected, but going back to 4.2 fixes the problem

When I next see Xiangfu in IRC I'll ask if he can fix the lua in their master git.

Thanks kano, I thought that might be the case.  Rather than dig through the lua code, I just reverted back to the 4.2.1 Avalon image that ckolivas had posted.  I figured it was the luci code making an incorrect assumption about the API data.  Annoying that both the Antminer and Avalon have these assumptions in them.

https://github.com/kanoi/cgminer-binaries/issues/2

To state that correctly: the luci code to display the API has been broken since it was first written - both in the Avalon and the Ant (the Ant simply copied the Avalon code and then broke it even more)
It doesn't use the field names to find the data, it numerically indexes into the data.

The API output (since I first wrote it) has always had the same key names to the same data, and that hasn't changed with 4.2.2
4.2.2 has simply added more fields to the summary between "MHS 5s" and "Found Blocks"

When you have a set of "Name"->"Data", it is foolish to not use "Name" to locate "Data"

4.2.2 breaks Cgminer Status Summary display on the Avalon Mini, reverted to 4.2.1.  4.2.2 working on Windows 7 with BE's.

New release: 4.2.2, 29th March 2014

Bugfixes and cosmetic updates.


Human readable changelog:

- Fix for memory leaks when mining solo
- Fix for a rare corruption possibility with solo mining.
- Support for a whole lot more USB hubs on windows.
- Added rules for using cointerra on PC in case anyone besides me actually does this.
- Revamped the hashrate meter to be far more reliable internally.
- Added rolling hashrates for 1 min, 5 min and 15 mins to the main display and per device to the API as well.
- Revamped the display since it was getting too wide: It now is back to the narrower size it was a few versions ago and toggles between the different data at regular intervals based on the opt log interval variable. Toggling can be disabled via the Display menu.
- Added serial number to the unique identifier on-screen when it exists instead of an arbitrary device number. This currently only tags hashfast, cointerra and BF1 devices which have serial numbers.
- Support reading configuration files with -c command from web URLs.
- Fixes for building with libcurl disabled.
- Other minor bugfixes and trivial changes.


Full changelog:

- Minor correctness fix for unnecessary free
- Clean up various curl build issues
- allow url based config files
- Frequency only needs 3 digits for cointerra statline
- Use the serial number as unique_id for cta display
- Make it possible to enable/disable the status window from switching via the
display menu
- We should not update the tv hashmeter time unless we're updating the hashrates
- Add cointerra devices to udev rules.
- Use hashfast unique id instead of number since the unique id is displayed
- Remove displayed space
- Left align the displayed unique id
- Use the hashfast opname as its unique identifier
- Display BF1 serial number as its unique identifier
- Display a unique identifier instead of a number if the device has one
- Use an alternating status display to return to a compact width of 80
characters, allowing more information to be displayed.
- No need for looking for khash hashrates in summary any more
- Fix two potential minor mem leaks
- Fix memory leaks in setup and generate work for gbt solo.
- Fix off by one malloc size error
- Fix memory leak in update_gbt_solo
- Put sanity check on decay_time to prevent updates with no time
- Add 3 rolling average hashrates to API output for summary and devs.
- Use the extra status screen real estate better, displaying rolling 1/5/15min
average hashrates as well.
- Revamp the ageing crufty hashmeter code to have proper exponential decaying
values and store rolling 1/5/15min hashrates.
- Increment total_work under control lock.
- Trivial variable reuse
- Add support for other usb3 hubs on windows

Kano . I tried to follow the instructions to get 4.2 to install on the antminer and it still shows 3.12 after a reboot. It seems to be mining but with the patch its seems its lost something.

  this seems to have gone away after another reset btw

did I miss a step and fudge something uo?

If i have a box running CGMINER as a service is there a way to monitor the stats from the CLI or linux? 

Thanks for the update.  I look forward to testing the new version when it's ready.

I will give that a try when I get home. Thanks for the help!

Found this memory leak and it's now fixed in git, but there won't be a new version out just yet as I fix other issues first.

At the moment, the code has trouble with more than 2 boards per SPI channel.
I think I know what needs to be changed, but haven't gone near it for a few weeks.
My BA setup with 7 boards gets 270GH/s no trouble since I use a BA V2 controller that has 4 SPI connectors and I put 2 boards on each (1 on one)

The --bab-options is only the numbers so for setting max,def only: --bab-options "55:54"
You can leave any blank and it will use the default for the blank ones and you only need to specify up to the ones you wish to change,