Pages:
Author

Topic: OLD: BFGMiner 4.10.0: GBT+Stratum, RPC, Mac/Linux/Win64, Spondoolies SP30 - page 29. (Read 308349 times)

newbie
Activity: 37
Merit: 0
Very cool.
MultiMiner is great for windows and Mac but this gives a Pi setup that should help lots of people.
Thanks.

GRID CONTROLA CE (Community Edition) Raspberry Pi image now available for download and testing... I hope I did this right.

https://mega.co.nz/#!5QJkwYyC!-7AJ_qTmcWBmNlS93A1ybQf-oW2CZUZlwYkIUak-yFw

4GB SD card image based on HASHRA BLADE CONTROLA 1.4.6.1 and Luke-Jr BFGMiner 4.2. I have only tested it with 10 GridSeed Minis and 1 Blade. This image has xrdp (Remote Desktop) pre-installed for your convenience so you may want to change credentials if your Raspi is exposed to the public side. HASHRA set credentials below:

Default Username: hashra
Default Password: hashra

Root access:
User: pi
Password: 3K4Hb8FMeZjQZJEX6scYzZa

Will try to create separate support thread if there is any interest (don't want to spam Luke's).

Vadim 

Wow I hope they keep this supported. I don't see it on their website. Where did you find it?
newbie
Activity: 3
Merit: 0
I've got one antminerU2 running on the latest build on a BTCGuild pool and it's doing fine. Having got one working I decided to add my 2nd unit. It showed up just like the first one and I went and added another pool with the same URL as my first pool and a 2nd worker user name. Despite all of this the 2nd unit isn't getting any work. CPU is only running at 13 of 2000 Mhz it is capable of. BTCGuild shows now work for my 2nd worker. Is there some way to assign a worker to a device? Does bfgminer spit out a log somewhere that details access errors and such to a device, or do you have to setup a redirect?

Well apparently the AC adapter on my hub can handle the one amp it says it can. Plugged 2nd ant directly into the computer and it's working now.
hero member
Activity: 840
Merit: 1002
Great... and good luck!

BTW - using MultiMiner for the first time in a Windows VM just to get my R-boxes going... any plans for hashrate/temp charting in MultiMiner?

If you enable MobileMiner integration and pay for the minimal subscription at http://web.mobileminerapp.com there are hashrate and temp charts both farm wide, per-rig and per-device.

I've considered putting something into MultiMiner and just never landed on an implementation I like. I'll probably do it for 4.0 though. 4.0 includes a rewrite of the UI (just the front-end) in WebKit for better cross-platform experience and would be a good place to start introducing charting I think.
hero member
Activity: 798
Merit: 1000
Nwoolls - saw on github that you were going to be receiving an R-Box soon to look at the issue with running it on Macs with 4.3.0.

Did you get it yet?

Running a VM of Windows or Ubuntu just to run the R-Boxes on a Mac is a bit of a pain. Smiley
Worst case, I guess I could break out one of the RaspPi's or BeagleBones I have, but would rather not. Both platforms are pretty finnicky.

It's due today. Hopefully I'll be able to work out the kinks for a 4.3.1 update.

Great... and good luck!

BTW - using MultiMiner for the first time in a Windows VM just to get my R-boxes going... any plans for hashrate/temp charting in MultiMiner?
hero member
Activity: 840
Merit: 1002
Nwoolls - saw on github that you were going to be receiving an R-Box soon to look at the issue with running it on Macs with 4.3.0.

Did you get it yet?

Running a VM of Windows or Ubuntu just to run the R-Boxes on a Mac is a bit of a pain. Smiley
Worst case, I guess I could break out one of the RaspPi's or BeagleBones I have, but would rather not. Both platforms are pretty finnicky.

It's due today. Hopefully I'll be able to work out the kinks for a 4.3.1 update.
hero member
Activity: 798
Merit: 1000
Nwoolls - saw on github that you were going to be receiving an R-Box soon to look at the issue with running it on Macs with 4.3.0.

Did you get it yet?

Running a VM of Windows or Ubuntu just to run the R-Boxes on a Mac is a bit of a pain. Smiley
Worst case, I guess I could break out one of the RaspPi's or BeagleBones I have, but would rather not. Both platforms are pretty finnicky.
newbie
Activity: 3
Merit: 0
I've got one antminerU2 running on the latest build on a BTCGuild pool and it's doing fine. Having got one working I decided to add my 2nd unit. It showed up just like the first one and I went and added another pool with the same URL as my first pool and a 2nd worker user name. Despite all of this the 2nd unit isn't getting any work. CPU is only running at 13 of 2000 Mhz it is capable of. BTCGuild shows now work for my 2nd worker. Is there some way to assign a worker to a device? Does bfgminer spit out a log somewhere that details access errors and such to a device, or do you have to setup a redirect?
hero member
Activity: 520
Merit: 500
4.3.0 Works wonders on windows xp!!!  Super stable, super lower HW (using CGminer with same settings had 10x as many HW).  Going to use the RPi and see if I get any lib_usb_errors on it.  If not, I can switch to that and monitor from anywhere...Is there some way to configure BFGMiner so that I can use the internet on my phone to make changes, restart the miners, that kind of thing, rather than using putty and a computer?  I do have LemonSSH for my android phone, but being able to go to my IP address and punch something in would be handy too...
hero member
Activity: 840
Merit: 1002
For those interested, here are updated 703n FW bins for BFGMiner 4.3:

https://github.com/nwoolls/BFGMiner-OpenWrt-Tools/releases/tag/v4.3.0-pr3

Tested with DualMiner, ZeusMiner, GridSeed 5-chip, G-Blade, LittleFury, AntMiner USB, Bi*Fury and Hex*Fury.
legendary
Activity: 2576
Merit: 1186
NEW VERSION 4.3.0, JUNE 30 2014

I plan to move 3.10.x to stable soon, so if you're still using 3.5.x, please upgrade and report any regressions (note them as such).

Human readable changelog:
  • URI parameter #getcbaddr to request coinbase address from bitcoind or compatible servers for solo mining.
  • When running on systems with a configured Bitcoin Core server, it is automatically configured as a final failover and used for local block submission (GBT pools only) to reduce the risk of a stale block.
  • bitforce: Implement some last-minute Monarch interface changes.
  • zeusminer: New driver for these scrypt ASIC devices.

Full changelog:
  • Bugfix: zeusminer: Include config.h
  • openwrt: Disable libusb via --without-libusb
  • Avoid waiting for local bitcoind test if we have other pools defined already
  • lowl-pci: Silence warning on big endian
  • bitfury: Portability fixes, including big endian compatibility
  • Bugfix: bifury: strtoll already implies native endian, no need to flip it again
  • bitforce: Additional sanity checks, including ensuring our final queued count matches that seen by the device
  • bitforce: Go back to ignoring race-triggered sanity check warnings for ZqX
  • bitforce: Refactor FLB and queue flush sanity checks
  • bitforce: Use binary FLB protocol for Monarch
  • bitforce: Add bitforce_read function to deal with binary responses
  • bitforce: Support for queue id based jobs
  • Make work->device_id unsigned
  • bitforce: Support device explicit "Scan Interval" to override hardcoded sleep times
  • bitforce: Allow queued_max to exceed BITFORCE_MAX_QUEUED_MAX iff the firmware provides an explicit Queue Depth
  • bitforce: Fix bitforce_chips_to_plan_for to work beyond 32 chips
  • configure: Clean up zeusminer logic
  • icarus: Use estimated hashes for hw errors (more accurate hashrate for devices where hw error values vary significantly from hashes performed)
  • zeusminer: Bugfix: calculate work_division based on chips and cores
  • Replace swabn with bswap_32mult (matching swap32yes parameters and performance)
  • Bugfix: zeusminer: Correctly encode job header regardless of native endian
  • zeusminer: support added for ZeusMiner scrypt ASICs
  • icarus: refactoring: commenting ICARUS_INFO members
  • icarus: Accept any power of two as a valid work_division
  • icarus: option added for Icarus-based drivers to ignore the golden nonce during probe
  • util: is_power_of_two function
  • Bugfix: Display proper statline R and HW error percents for values < 1
  • openwrt: Provide scrypt option
  • Benchmark: Debuglog headers generated
  • Bugfix: cpu: Fix scrypt CPU miner
  • Bugfix: scrypt: Fix scrypt hash postprocessing for big endian
  • Bugfix: scrypt: Correct endian handling in PBKDF2_SHA256_80_128
  • scrypt: Add unit test for scrypt_1024_1_1_256_sp
  • scrypt: Add unit tests for PBKDF2_SHA256_80_128 and salsa20_8
  • Bugfix: Copy ALL the data with swap32to
  • Automatically add pool configured from bitcoin.conf for failover only
  • Add failover_only pool flag, to avoid using it regardless of strategy
  • Accept --default-config option to search and load default configuration files, even if user is specifying their own
  • Use appdata_file_call to find BFGMiner config file(s)
  • util: appdata_file_call and appdata_file_find_first functions
  • ccan: Update to latest version (which includes floatval option support)
  • Bugfix: Adapt quit function to new logging design
  • Workaround GCC 4.6 initialisation bugs with BYTES_INIT
  • Bugfix: Use int rather than ssize_t to type-match %d expectation
  • Minor refactor of text-only mode: avoid wasting bottom line of console, and save status lines as they are replaced by log items
  • Bugfix: Wake up wait_lpcurrent if its conditions for returning may have changed
  • Defer determining --quit-summary default until it is needed, so hotplugged devices affect it
  • Bugfix: In benchmark mode, only remove the new pools after putting them in the array, since the removal will try to remove it from the array
  • Bugfix: Initialise struct pool in add_pool rather than halfway through startup
  • URI parameter #getcbaddr to request coinbase address from bitcoind or compatible servers
  • pool_set_uri function
  • util: bytes_eq and bytes_assimilate
  • Replace struct _cbscript_t with bytes_t for simplicity
  • Bugfix: Parse userpass option with strchr rather than strtok to avoid usernameless passwords getting into the username field and being displayed
  • Bugfix: Free old coinbase script before setting a new one
  • util: upper_power_of_two_u32 function
  • Avoid unnecessarily finding drivers in a loop
  • lowl-pci: Never probe via -S *:all
  • configure: Succeed even if the libblkmaker submodule is missing (only --with-system-libblkmaker)
  • rockminer: Reset task timeout when queuing work
  • rockminer: When we detect a task failure, check if we ought to grab a new work rather than resend
  • rockminer: Add a time limit before resending work even outside of midtask state
  • Bugfix: rockminer: Keep count of requested jobs properly
  • make-release: Include ccan LICENSE symlinks as symlinks even in ZIP
  • make-release: Ensure ccan licenses get included in source
  • Bugfix: make-release: Correctly avoid unused parts of ccan
  • make-release: Do builds with the official source release, to test that it works
  • Bugfix: DevAPI: Make for_each_logical_proc work correctly
  • Avoid double hashing of shares
  • Create work_hash function to abstractly produce work->hash from work->data regardless of algorithm used
hero member
Activity: 520
Merit: 500
I've got 3 Gridseed 5 chips.  I've got each of them running separate instances of BFGMiner so I can monitor them easily.  2 of them seem to run fine but one of them hashes for a while, then the speed starts to drop, then it goes to nothing.  I'm using windows XP Pro.  They all ran fine on Linux together on this current computer which I reformatted back to windows.  Don't ask me why, I just did.  I'm not sure if there are driver issues, or what the deal is.  On that machine on linux using CGminer 3.7.2, no problems.  Using an RPi and CGMiner I get lib_usb_error after 1/2 an hour and BFGMiner seems to only run 2 of them.  I may have had an incorrect line in the command (I was using --set-device \\.\=COMX:freq=xxx instead of --set-device gridseed@\\.\comx:freq=xxx) so it seemed like everything was running much slower.  I'll give that a whirl when I get home.  Any suggestions though to help the Windows version?  I've tried powered USB hubs, switching to different USB ports, rebooting, power cycling the gridseeds...nothing helps.

Thanks!
hero member
Activity: 840
Merit: 1002
Some performance numbers running 10x ZeusMiner ASICs (6 chips) on a Raspberry Pi:



Code:
./bfgminer --scrypt -S noauto -S zus:all --set zus@/dev/ttyUSB6:ignore_golden_nonce=1 -o usa4.wemineltc.com:3333 -u x -p x
 
 [2014-06-30 02:41:37]
Summary of runtime statistics:
 
 [2014-06-30 02:41:37] Started at [2014-06-30 01:41:17]
 [2014-06-30 02:41:37] Pool: http://usa4.wemineltc.com:3333
 [2014-06-30 02:41:37] Runtime: 1 hrs : 0 mins : 20 secs
 [2014-06-30 02:41:37] Average hashrate: 13.6 Megahash/s
 [2014-06-30 02:41:37] Solved blocks: 0
 [2014-06-30 02:41:37] Best share difficulty: 15
 [2014-06-30 02:41:37] Share submissions: 498
 [2014-06-30 02:41:37] Accepted shares: 498
 [2014-06-30 02:41:37] Rejected shares: 0 + 0 stale (0.00%)
 [2014-06-30 02:41:37] Accepted difficulty shares: 10
 [2014-06-30 02:41:37] Rejected difficulty shares: 0
 [2014-06-30 02:41:37] Hardware errors: 45
 [2014-06-30 02:41:37] Network transfer:  77.8 /  64.1 kB  ( 21.5 /  17.7  B/s)
 [2014-06-30 02:41:37] Efficiency (accepted shares * difficulty / 2 KB): 0.15
 [2014-06-30 02:41:37] Utility (accepted shares / min): 8.26/min
 
 [2014-06-30 02:41:37] Unable to get work from server occasions: 0
 [2014-06-30 02:41:37] Work items generated locally: 2788
 [2014-06-30 02:41:37] Submitting work remotely delay occasions: 0
 [2014-06-30 02:41:37] New blocks detected on network: 16
 
 [2014-06-30 02:41:37] Summary of per device statistics:
 
 [2014-06-30 02:41:37] ZUS0       | 20s: 1.76 avg: 1.36 u: 1.11 Mh/s | A:46 R:0+0(none) HW:2/4.7%
 [2014-06-30 02:41:37] ZUS1       | 20s: 1.84 avg: 1.36 u: 0.91 Mh/s | A:35 R:0+0(none) HW:16/ 27%
 [2014-06-30 02:41:37] ZUS2       | 20s: 1.66 avg: 1.35 u: 1.13 Mh/s | A:44 R:0+0(none) HW:3/5.5%
 [2014-06-30 02:41:37] ZUS3       | 20s: 1.30 avg: 1.36 u: 1.39 Mh/s | A:56 R:0+0(none) HW:2/3.8%
 [2014-06-30 02:41:37] ZUS4       | 20s: 1.55 avg: 1.36 u: 0.95 Mh/s | A:38 R:0+0(none) HW:16/ 30%
 [2014-06-30 02:41:37] ZUS5       | 20s: 1.44 avg: 1.36 u: 1.74 Mh/s | A:72 R:0+0(none) HW:1/1.5%
 [2014-06-30 02:41:38] ZUS6       | 20s: 2.20 avg: 1.35 u: 1.07 Mh/s | A:47 R:0+0(none) HW:1/2.5%
 [2014-06-30 02:41:38] ZUS7       | 20s: 1.85 avg: 1.35 u: 1.13 Mh/s | A:44 R:0+0(none) HW:4/7.6%
 [2014-06-30 02:41:38] ZUS8       | 20s: 1.79 avg: 1.34 u: 1.36 Mh/s | A:54 R:0+0(none) HW:0/none
 [2014-06-30 02:41:38] ZUS9       | 20s: 1.76 avg: 1.36 u: 1.60 Mh/s | A:62 R:0+0(none) HW:0/none
 [2014-06-30 02:41:38]
Shutdown signal received.
hero member
Activity: 840
Merit: 1002
I'm excited to say that version 4.3 (and the current source in Git) now supports mining Scrypt using ASICs on Big Endian hardware, such as the 703n router running OpenWrt:

Code:
GSD0       | 20s:360.1 avg:361.6 u:171.3 kh/s | A:2 R:0+0(none) HW:0/none
ZUS0       | 20s: 1.45 avg: 1.42 u: 1.26 Mh/s | A:14 R:0+0(none) HW:1/6.3%
 [2014-06-29 23:35:34] Accepted 00ab553a ZUS 0  Diff 5m/3m
ALL 20s: 3.43 avg: 3.14 u: 2.94 Mh/s | A:35 R:0+0(none) HW:3/7.9%
ZUS1       | 20s: 1.40 avg: 1.43 u: 1.50 Mh/s | A:18 R:0+0(none) HW:2/9.5%
 [2014-06-29 23:35:46] Accepted 00622f22 ZUS 1  Diff 10m/3m
GSD0       | 20s:361.0 avg:361.6 u:152.6 kh/s | A:2 R:0+0(none) HW:0/none
ALL 20s: 2.88 avg: 3.03 u: 2.74 Mh/s | A:36 R:0+0(none) HW:3/7.7%
 [2014-06-29 23:35:55] Accepted 00d3ca5f ZUS 1  Diff 4m/3m
ZUS0       | 20s: 1.46 avg: 1.42 u: 1.22 Mh/s | A:15 R:0+0(none) HW:1/5.9%
 [2014-06-29 23:35:55] Accepted 00e8d459 ZUS 0  Diff 4m/3m
 [2014-06-29 23:35:59] Accepted 00d622b8 ZUS 0  Diff 4m/3m
 [2014-06-29 23:36:00] Accepted 0058b7d8 ZUS 0  Diff 11m/3m
 [2014-06-29 23:36:06] Accepted 00281489 ZUS 0  Diff 24m/3m
 [2014-06-29 23:36:10] Accepted 005315cc ZUS 1  Diff 12m/3m
ALL 20s: 3.02 avg: 3.05 u: 2.94 Mh/s | A:42 R:0+0(none) HW:3/6.7%
GSD0       | 20s:361.4 avg:361.6 u:137.5 kh/s | A:2 R:0+0(none) HW:0/none
ZUS0       | 20s: 1.46 avg: 1.43 u: 1.37 Mh/s | A:19 R:0+0(none) HW:1/4.8%
 [2014-06-29 23:36:19] Accepted 009af945 ZUS 0  Diff 6m/3m
 [2014-06-29 23:36:20] Accepted 000cfb89 ZUS 0  Diff 77m/3m
 [2014-06-29 23:36:22] Accepted 00228d7f GSD 0  Diff 28m/3m
 [2014-06-29 23:36:24] Accepted 00e108f1 ZUS 0  Diff 4m/3m
 [2014-06-29 23:36:25] Accepted 00ef9dea ZUS 0  Diff 4m/3m
 [2014-06-29 23:36:32] Accepted 00631445 ZUS 0  Diff 10m/3m
ALL 20s: 2.70 avg: 2.99 u: 3.10 Mh/s | A:48 R:0+0(none) HW:3/5.9%

Hardware:

http://wiki.openwrt.org/toh/tp-link/tl-wr703n

Bins:

https://www.dropbox.com/s/0ffntt6uto4n5fw/openwrt-ar71xx-generic-tl-wr703n-v1-squashfs-factory.bin
https://www.dropbox.com/s/f1a2ng2ox4800j1/openwrt-ar71xx-generic-tl-wr703n-v1-squashfs-sysupgrade.bin

Note: I give the FW bins nothing more than an "It Didn't Brick Mine" seal of approval.  Grin

Edit: Here's an updated release, with notes and everything. This also includes the scripts I used to cross-compile BFGMiner and to create the OpenWrt image itself.
newbie
Activity: 58
Merit: 0
Is there ever going to be a method for hotswapping kernels between SHA and Scrypt?  Kinda want to get a dualminer, but don't want to completely shut off the more-lucrative SHA mining completely to mine Doge

I'm not sure what you mean by "hotswapping kernels". Kernels are for GPU mining, not ASIC mining.

BFGMiner already supports Dual Mode mining with DualMiner ASICs, just not with the 5-chip GridSeed Orbs. I doubt that will ever be supported, as there is very little demand. Most controller software has already disabled or deprecated mining SHA with the Scrypt ASICs and newer Scrypt ASICs don't support the feature at all.

Okay, let me get to the deep problem, then we can come back and correct my crappy translation to shallow problem Smiley  I have a Radeon 4350 (what can I say, AGP board and not a lot of round tuit to gut it), and mined scrypt coins and bleeding-edge alts with it fine until Ubuntu trusty changed things so badly that the fglrx-legacy blew up and I couldn't mine anymore (they SAY that the rv710 will eventually get opencl in the radeon drivers, but I ain't holding my breath). I found I liked mining, so grabbed a cheap antminer to tide me over, knowing that I'd have to go SHA. Of course, when I fired up bfgminer right afterward, I had to move the old config file that had the --scrypt option within (which I now know was only to get the radeon to work with scrypt, and became mostly irrelevant when trusty ruined my rig's drivers) and all my scrypt pools because I was (understandably) getting all hwerrors.  This is fun, but I wanna get a dualminer and back to scrypt asap, and rotate through pools (and altcoins) like I used to, except use the antminer when I'm on a SHA pool and not have to do extra stuff when I rotate through a scrypt pool (ultimately, I'd like the radeon to follow the bouncy pools, but that can wait until the radeon actually plays opencl in the open-source driver).  TBH, I had thought that the dualminer wasn't even an option until I looked at the changelog the other day and realized that dualminers work natively under bfgminer (dualminer itself says you need a specially patched cgminer, which they haven't gotten the round tuit to get working in linux).  So how can I get bfgminer to happily realize the proper hardware to use when I point it at a given pool, or am I just spinning my wheels because of a mistaken belief (see radeon mess) that bfgminer needs assistance in doing the right thing?  
hero member
Activity: 840
Merit: 1002
Is there ever going to be a method for hotswapping kernels between SHA and Scrypt?  Kinda want to get a dualminer, but don't want to completely shut off the more-lucrative SHA mining completely to mine Doge

I'm not sure what you mean by "hotswapping kernels". Kernels are for GPU mining, not ASIC mining.

BFGMiner already supports Dual Mode mining with DualMiner ASICs, just not with the 5-chip GridSeed Orbs. I doubt that will ever be supported, as there is very little demand. Most controller software has already disabled or deprecated mining SHA with the Scrypt ASICs and newer Scrypt ASICs don't support the feature at all.
newbie
Activity: 58
Merit: 0
Is there ever going to be a method for hotswapping kernels between SHA and Scrypt?  Kinda want to get a dualminer, but don't want to completely shut off the more-lucrative SHA mining completely to mine Doge
full member
Activity: 361
Merit: 100
Noticed when moving from 3.10 to 4.20 Block Erupter BES units are now recognized as BEE's when using -S erupter:all. Doesn't seem to be much difference in hash rate over time. Anything to be concerned about? TIA.

Just a visual change (bugfix).

Thanks for replying. Also receiving this message on most startups and at random times therafter.

[2014-06-28 20:01:37] Failed to open Device Parameters registry key in lowl-vcom.c _vcom_devinfo_scan_windows__hubport():440: The system cannot find the file specified

hero member
Activity: 840
Merit: 1002
Noticed when moving from 3.10 to 4.20 Block Erupter BES units are now recognized as BEE's when using -S erupter:all. Doesn't seem to be much difference in hash rate over time. Anything to be concerned about? TIA.

Just a visual change (bugfix).
full member
Activity: 361
Merit: 100
Noticed when moving from 3.10 to 4.20 Block Erupter BES units are now recognized as BEE's when using -S erupter:all. Doesn't seem to be much difference in hash rate over time. Anything to be concerned about? TIA.
newbie
Activity: 2
Merit: 0
I tried using git to get the latest version of BFGMiner, but when I compiled it, BFGMiner wouldn't recognize my Rockminer? I'm not sure what I did wrong.

http://i.imgur.com/pP4v04H.jpg
Pages:
Jump to: