Topic: OLD: BFGMiner 3.10.0: modular ASIC+FPGA, GBT+Strtm, RPC, Mac/Lnx/W64, AntU1, DRB - page 150. (Read 1193223 times)

tssk i wanted to change to mining onto litecoin or similar coins
however cgminer and bfgminer simply crash again

i used as command bfgminer --scrypt -I13 and this error shows when start the 32 bit version
 
Failed to init bfd from (d:\Miner\bfgminer-3.1.2-win32\bfgminer.exe)
0x417b73 : d:\Miner\bfgminer-3.1.2-win32\bfgminer.exe : [unknown file]
Failed to init bfd from (d:\Miner\bfgminer-3.1.2-win32\bfgminer.exe)
0x41d8f3 : d:\Miner\bfgminer-3.1.2-win32\bfgminer.exe : [unknown file]
Failed to init bfd from (d:\Miner\bfgminer-3.1.2-win32\pthreadGC2.dll)
0x6248671b : d:\Miner\bfgminer-3.1.2-win32\pthreadGC2.dll : pthread_create
Failed to init bfd from (C:\Windows\SYSTEM32\msvcrt.dll)
0x7732906a : C:\Windows\SYSTEM32\msvcrt.dll : ui64tow
Failed to init bfd from (C:\Windows\SYSTEM32\msvcrt.dll)
0x77329147 : C:\Windows\SYSTEM32\msvcrt.dll : iswalnum
Failed to init bfd from (C:\Windows\SYSTEM32\KERNEL32.DLL)
0x77b1850d : C:\Windows\SYSTEM32\KERNEL32.DLL : BaseThreadInitThunk
Failed to init bfd from (C:\Windows\SYSTEM32\ntdll.dll)
0x77d3bf39 : C:\Windows\SYSTEM32\ntdll.dll : RtlInitializeExceptionChain
Failed to init bfd from (C:\Windows\SYSTEM32\ntdll.dll)
0x77d3bf0c : C:\Windows\SYSTEM32\ntdll.dll : RtlInitializeExceptionChain

when using the 64 bit one nothing happens just see apllication stopped working in windows

hi i would appreciate your help. i'm running bfgminer on a Pi. i have 3 icarus boards, 1 cm1 and 1 bfl single fpga
i did the following (made it a script so as to build it every so often):

no errors on build

also on boot:


i run it with either
bfgminer -c ../miner.txt -S all
or
bfgminer -c ../miner.txt -S /dev/ttyUSB0 -S /dev/ttyUSB1 -S /dev/ttyUSB2 -S /dev/ttyUSB3 -S /dev/ttyUSB4 -S /dev/ttyUSB5 -S /dev/ttyUSB6 -S /dev/ttyUSB7

and the bfl is not detected or bfgminer stays in a blank screen.

even though all 0 to 7 ttyUSB exist

any ideas?

edit:
if i use bfgminer -c ../miner.txt  it starts ok but no ICA start.

by the way i've mined with them before but not all together.

Sounds like GPU.

You just broke your driver. WinUSB is the wrong one for Erupters.
It won't work until you figure out how to undo that and get the right driver.
http://www.silabs.com/products/mcu/pages/usbtouartbridgevcpdrivers.aspx

hey I just got one of the blade erupter little asic usb miners.
it is my first miner. I got bfgminer to work and got into the deepbit pool.  

But my miner is going at most 64 Mh per second and fell to 27 Mh. I have it on "d" for intensity.

And I have no idea if I am using my GPU in my computer or it is my usb miner.

The program didn't see it until I changed to WinUSB driver.

Any thoughts appreciated

I cant figure out how to upload a pic, otherwise I would put p the screen shot of my miner

Thanks

***UPDATE****

First of all, thanks so much for the advice and link to the UART page. It is the help and support on this forum that really makes this tech and currency exciting and able to thrive.  But I started with the UART driver and read that the cgminer couldnt use the driver to mine, that only the winusb driver could mine. But, conversely, once I installed it, Everytime I would log into deepbit, cgminer immediately fatal errors. So I got it working with BFMiner with the crappy WinUSB driver.  So now, I downloaded the newer UART driver from the linked page, and the mining programs dont see the device.



http://pit.deepbit.net:8332
Username:

Password:

 [2013-07-16 16:58:15] Probing for an alive pool
 [2013-07-16 16:58:16] Long-polling activated for http://pit.deepbit.net:8332/li
stenChannel (getwork)
 [2013-07-16 16:58:16] Error -6: Creating Command Queue. (clCreateCommandQueue)
 [2013-07-16 16:58:16] Failed to init GPU thread 0, disabling device 0
 [2013-07-16 16:58:16] Restarting the GPU from the menu will not fix this.
 [2013-07-16 16:58:16] Try restarting BFGMiner.
Press enter to continue:


 bfgminer version 2.10.11 - Started: [2013-07-16 16:58:16] - [  0 days 00:00:58]
--------------------------------------------------------------------------------
 5s:  0.0 avg:  0.0 u:  0.0 kh/s | A:0 R:0 S:0 HW:0 U:0.0/m
 ST: 4  DW: 0  GW: 4  LW: 0  GF: 0  NB: 1  AS: 0  RF: 0  E: 0.00
 Connected to pit.deepbit.net diff 1 with LP as user
 Block: ...c36eb323bf433f12  Diff:26.2M  Started: [16:58:16]  Best share: 0
--------------------------------------------------------------------------------
 [P]ool management  S ettings Display options Qit
 OCL 0:  59.0C         | OFF  /  0.0/  0.0kh/s | A:0 R:0 HW:0 U:0.00/m
--------------------------------------------------------------------------------

 [2013-07-16 16:58:15] Probing for an alive pool
 [2013-07-16 16:58:16] Long-polling activated for http://pit.deepbit.net:8332/li
stenChannel (getwork)
 [2013-07-16 16:58:16] Error -6: Creating Command Queue. (clCreateCommandQueue)
 [2013-07-16 16:58:16] Failed to init GPU thread 0, disabling device 0
 [2013-07-16 16:58:16] Restarting the GPU from the menu will not fix this.
 [2013-07-16 16:58:16] Try restarting BFGMiner.
Press enter to continue:

 [2013-07-16 16:58:57] Error -6: Creating Command Queue. (clCreateCommandQueue)
 [2013-07-16 16:58:57] Failed to init GPU thread 1, disabling device 0
 [2013-07-16 16:58:57] Disabling extra threads due to dynamic mode.
 [2013-07-16 16:58:57] Tune dynamic intensity with --gpu-dyninterval


Any help from you guys with an Asic on usb on windows xp would be much appreciated.  I'll say it right now: I already intend to buy a raspberry pi on friday for a dedicated linux asic mining rig. But I would like to see proof of concept on my xp machine because I bought the hardware from ebay and want to make sure it actually will go 333 Mh.

To make sure the thread is seen, I will post again at the end of the queue. Thanks.

@Luke-Jr: finally got around to posting a repo on github of the Bitfury source. and am finally on way to a working bitstream. >72 hrs left to P&R le sigh tis what i get for using a dual core Pentium(newer)

https://github.com/Taugeran/Bitfury

-Taugeran

Edit: building straight from source as is following instructions. not sure if changes need made    anyone more knowledgeable with Spartan6 fpgas who care to chip in their 2 satoshi. please pm

More likely a driver bug.

hmmm cant seem to gpu mine on my fossilized Radeon 5450 but i have in the past

Output of bfgminer -n


are the call fails to adl a possible windows bug?

using latest 3.1.3 x32 build

Hey, I'm running 13.4 drivers on Ubuntu 13.04 amd64, I had the problem with the BFI_INT as LukeJr described, with CGMiner, it was fixed with BFGMiner with SHA-256, but I still have errors mining scrypt hashes. What could I do?. Thanks. Right now I have to decrease intensity a lot to make it work, it gives me half the windows hashrate.

edit: it was my own dumbness, I forgot to export gpu alloc 100.

What is your hashrate with 2 x hd 7970?

Totally forgot the fact that APUs have Northern Light GPU cores. Used to working with a standard CPU, I feel pretty stupid now haha, but thanks!

Poclbm was using the CPU cores and Bfgminer is using OpenCL and the GPU stream processors on the chip. Bfgminer doesn't have native CPU support on windows anymore.

NEW VERSION 3.1.3, JULY 11 2013

Special thanks to iongchun for bisecting, debugging, and fixing the 100% CPU issue.

Human readable changelog:

• Fix 100% CPU usage hang with GBT/getwork pools.
• Make staged work underrun detection less overly aggressive.
• Generate baud rate list from OS on *nix (fixes Mac/BSD build).

Full changelog

• Bugfix: Reset staged_full flag when discarding (stale) popped work, or increasing the queue minimum
• Bugfix: Only trigger staged work underrun if a mining thread was actually waiting for it (and do so sooner, before it has the work made)
• bytes_cpy: avoid malloc and memcpy when size is zero
• fix infinite loop in bytes_cpy when size is zero
• Bugfix: Generate iospeeds_local.h based on termios.h defines, and only try to use POSIX standard if that fails

Okay, this just seems totally weird to me. I'm screwing around with different miners on my AMD A10-4600m APU. I started using Poclbm and BFGMiner. Poclbm got me about 3 MH/s utilizing 100% APU power, which seemed normal to me. Now with BFGMiner, I get approx. 100 MH/s and only using 5% of my APU. This is totally mindfucking me, please explain? Haha, and I can tell this is true by looking at my mining pool account status, which I'm acutally giving shares too a lot faster than Poclbm. Is my APU really supposed to be this fast at 100%? My CPU monitor might just be screwing with me about the percentage.

**Edit: Well, the hashrate is slowly lowering,  but is still a lot faster than GUIMiner and looking at Task Manager, it's almost like BFGMiner isn't running.

http://gyazo.com/18ac57cb1322a4e05b97e2e9ed1a41d5

adding a queue size of 254 ultimately ends up the same in time bfg starts giving the messages as mentioned in my previous post
i got the jala running as well on bfgminer 3.1.2 and guess what lol after a day running fine it also started giving the same message even with a queue size of 254
however still not sure why bfg crashes sometimes without any message whats wrong

adding the following commands to the startup batch seems to helped me get it to run more stable but does not solve the underrun messages

setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1

I see this hang.  It occurs in bytes_cpy.  A size of 0 results in an infinite loop.

Here's the results:


Description of the problem in my previous post:

https://bitcointalksearch.org/topic/m.2685683

I am seeing a similar speed increase if i fiddle with the queue as well.  I expect that the auto adjustment of the queue size is capped at some function of the hashpower - but yes - i find moving it upwards it helps - also when i increase the difficulty some and twiddle with it.

I have found it to stop trying to auto adjust enventually when i get the number big enough and it does seem to be based on the amount of hashing and possibly number of devices as well.

Whatever you did Luke - it is making a difference for me - thanks!  Donation coming your way shortly. 

Would like to be able to use it on my rigs with X6500's - (win7 x64) but 3.1.1 is working great for them as well..

well after adding the queue to 64 in the config i see it still fail [ staged work underrun error ] still given and constant

i wonder if its normal to need a buffer so big with just 2 x 7970 which do run on version 3.1.1 with a queue size of 0

anyway upped the queue to massive 254 to see how that runs

i get the feeling the program is slightly faster on these gpu's even while they are heavy underclocked

Works great for me, under Linux.  I haven't noticed any problems with bfgminer 3.1.2 crashing or taking 100% CPU.  Maybe the problem could be isolated to Windows?

Compiled my entire Bitcoin stack from source, on Gentoo:

bitcoin 0.8.3 (bitcoin-qt target)
p2pool 13.1 (no compilation needed, it's a Python program which is interpreted)
bfgminer 3.1.2

I rather like bfgminer, it follows the "configure" standard, and drops right in!  Rather impressive.

Even the Satoshi bitcoin client wasn't this cleanly written, as it still required a little tweaking before it would compile: http://forums.gentoo.org/viewtopic-p-7344252.html#7344252

Command lines I've been using for the stack, which mines on my little Erupter:


If scripting, be sure to change "cd" commands to the directories you are using on your system (it is critically important to start from the correct directory for p2pool).

And, you probably want to change your payout address accordingly

Josh

Luke, any chance of integrating bitfury support into bfgminer?

https://github.com/legkodymov/cgminer