Has anyone been able to run bfgminer on a Windows XP 32bit OS? Will it run? when I run bfgminer with no switches I get:
"All devices disabled. cannot mine!"
Would this mean that the ASIC USB driver is not installed?
Thanks,
Probably, does Device Manager show it installed?
|
|
|
Eligius is now the 3rd pool to pass the 2-years-old marker line, and the only pool over 2 years old to remain feeless. Happy birthday, Eligius!
Side question: Would anyone be interested in donating toward getting Eligius an industrial membership in the Bitcoin Foundation and/or sending wizkid057 and/or myself to the San Jose Bitcoin Conference in a few weeks?
|
|
|
NEW VERSION 3.0.2, APRIL 28 2013Human readable changelog:Full changelog- Receive failures in recv_line should unconditionally fail.
- Use sock_blocks function for stratum send and receive.
- Avoid applog under stratum_lock in __stratum_send.
- Create an OS specific sock_blocks function.
- There should be no error response code with return value 0 in recv_line.
- Check for errors on stratum recv for any recv return value less than 1 and only parse the response if it's positive.
- Avoid applog under stratum_lock in recv_line.
|
|
|
I'm curious, why have you decided not to include the FPGA bitstreams in the windows binaries anymore? Makes the download much larger, mainly, and most miners don't need them. Seemed reasonable to have the few who do continue using the bitstreams they already have, or download them as the separate ZIP.
|
|
|
I compiled bfgminer 3.0.1 for Ubuntu and when I start it I get the command line version not the display version. Any help would be great.
See dependencies at the top of README. Especially the part about ncurses.
|
|
|
Note BFGMiner (with fewer bugs and better support) has official OpenWrt packages.
|
|
|
Is there a particular OS/Python version you have been successful with? I run Gentoo + Python 3.1 or 3.2. Another possibility (more likely perhaps) is that you miscompiled the midstate module for your system... be sure to check CFLAGS in the Makefile.
|
|
|
This is all I get when I try running ./eloipool.py: root@pool:~/eloipool# ./eloipool.py Illegal instruction
Any ideas? I checked the log but it's empty. Almost certainly something wrong with your OS or Python.
|
|
|
So they do have mining software that supports it.
Onboard the blade. These are smart/ethernet ASICs, not USB ASICs.
|
|
|
Could everyone who was experiencing the crash-on-pool-switch please answer: are you GPU mining?
|
|
|
The samples are just the USB device, not the blade. And do you really want to wait "many months from now" before any mining software supports your hardware? I'm sorry, explain to me how ASICMINER is mining without "any mining software". The blades have their custom software. The USB device (which isn't done yet AFAIK) requires a host system.
|
|
|
What about those who were promised free sample boards? TBH, if they do send out sample boards, they should only be sent out many months from now when the price/board is not 70btc. You don't just give someone something worth $10,000, shareholder(or board member) or not. The samples are just the USB device, not the blade. And do you really want to wait "many months from now" before any mining software supports your hardware?
|
|
|
NEW VERSION 3.0.1, APRIL 24 2013This release removes bitstreams from the Windows binary ZIPs. If you are mining with ModMiner, X6500, or ZTEX FPGAs you will need to download the source code ZIP as well and copy over the "bitstreams" directory. Human readable changelog:- IMPORTANT: Bitstreams for ModMiner, X6500, and ZTEX FPGA boards are no longer included with the Windows builds and must be copied from the source distribution.
- Bitforce SC: Switched to using bulk queue mode for all ASIC-based devices.
- Bitforce SC: Fix reinitialization-on-failure, so devices can autorecover from some problems.
- OpenCL: Support for mining with completely free (non-proprietary) Mesa/LLVM OpenCL on Linux (requires git Mesa/LLVM).
- OpenCL: Include platform in kernel binary filenames to avoid using stale binaries from other SDK installs.
- Added network hashrate (based on block difficulty) to block status line.
Full changelog- Bugfix: configure: Move actual roundl macro back to miner.h after math.h should be included
- Bugfix: configure: Use dummy pointer to correctly detect roundl when conftest main is missing argc parameter
- Bugfix: configure: Use variable argument to roundl to prevent compilers from optimizing it out entirely
- Remove bitstreams from Windows binary distributions (README directs users to download source and copy them)
- make-release: Remove autom4te.cache from distributed source
- Bugfix: Omit --no-opencl-binaries option from build if OpenCL is not being compiled
- Bugfix: Check that all pools have URIs set before starting
- Bugfix: bitforce: Make noncebuf large enough for max qresults +1 (for OK line)
- opencl: Ability to avoid using binary kernels with new --no-opencl-binaries option
- README: Include jansson PKG_CONFIG_PATH in example for Mac
- Include trailing \0 with coinbase sigs if there's room
- Differentiate socket closed from socket error in recv_line.
- Add new best share info to verbose logging.
- Add notice for when network diff is changed.
- convert sleep(const) to nmsleep()
- Rename longpoll threads according to what pool they're associated with
- miner.php report 'Last Valid Work' as time before request
- API V1.25 - add 'Last Valid Work' time for each device
- add 'count' to customsummarypage 'calc'
- Bugfix: ztex: Initialize fw_buf pointer to NULL so a free before allocation is safe
- Cleanup when stratum curl fails to initialise.
- LTC text typo
- Recreate cURL for new stratum connections, and clear stratum_notify on suspending them
- clear_stratum_shares: Rename diff_stale variable to diff_cleared
- MMQ it's a bitstream
- Update a pool's last work time when the work is popped as well as staged.
- Extend stratum connections another minute (total 2 minutes) after the last work item was staged, and maintain last_work_time for non-stratum pools.
- Fix --benchmark generating valid work for cgminer.
- Bugfix: Correct pdbuilder result directory
- Omit add_serial_all code when serial support is not wanted
- Use configure to detect presence of roundl to avoid redefining an actual function (possibly inline)
- Bugfix: roundl: Add needed parenthesis to perform ?: before +
- Bugfix: ft232r: Defer allocating structure until after USB endpoint is successfully opened, so it won't leak in case of failure
- Bugfix: ztex: Free bitstream in memory when done with it
- Bugfix: Safely handle all-space cURL debug messages, should they ever happen
- Silence warnings about poor format usage for quit()
- Apply noreturn and printf-format attributes to quit() function definition
- Bugfix: set_serial_rts get flags to manipulate them correctly
- Bugfix: Missing return for /dev globbing
- Bugfix: Free unused work when retrying failed lp request
- Display processor name with thread disabled/re-enabled messages
- Move best share to top summary line, and add network difficulty to block line
- opencl: Default to phatk kernel for Mesa platform
- opencl: Default to single thread with Mesa OpenCL
- opencl: Check for Mesa OpenCL and avoid using binary kernels with it
- bitforce: Never increase bulkqueue poll wait time during queue underruns
- bitforce: Start off polling bulk queue every 100ms
- bitforce: Log device queue size after getting bulk results
- bitforce: Ensure bulkqueue polling occurs at least once a second
- opencl: Include OpenCL platform in kernel binary filenames
- bitforce: Use bulk queue mode for all SC devices
- Bugfix: bitforce: When reinitializing, free all known works to avoid decrementing reset queued counter
- bitforce: Handle timeout during ZOX as cleanly as possible
- Bugfix: bitforce: Make reinitialization more complete and safe
- Bugfix: bitforce: Close opened fd if reinit fails
- Bugfix: bitforce: Retry ZGX until device is NOT busy
- bitforce: Log when zero queued results are received
|
|
|
Why build without OpenCL support anyway? I have absolutely no fecking idea! When it came up it made me smile, the Raspberry PI is not known for its outstanding GPU performance. In my defence though, I did use the '--disable-opencl' flag for configure. Yes, that's what found this bug (I don't usually build without OpenCL support)
|
|
|
Hi Luke, small compile error to report;- Fixed. Why build without OpenCL support anyway?
|
|
|
|