just curious: why is eligius including 1 satoshi payouts to 18d3HV2bm94UyY4a9DrPfoZ17sXuiDQq2B in many of its blocks, like here? isn't it "bad" under new bloat rules? It helps identify/track genuine Eligius blocks. I clean them up regularly so they don't bloat the UTXO set. |
|
|
|
OK, any comments about the protocol proposal?
It's missing several things: - A way to know when a job has been started (and on which boards/chips) and completed (and how much of the nonce range was in fact scanned).
- A way to know which specific slave board and chip found the nonce in question.
- Access to temperature sensors.
- Control of any LEDs.
- Commands to get and/or control clock frequencies, voltages, etc.
- Some way to get error checking information/feedback, to make dynamic clocking more accurate.
Additionally, "cancel" is misspelled. IMO there is no real value for asking the device for its current hashrate. It might be nice if the protocol were prepared to handle non-Bitcoin POW algorithms too, but this could probably be done as an extension. Finally, if any vendors are actually willing to implement this protocol, KNK should get a BIP number assigned by gmaxwell and post the draft on the Bitcoin wiki for further review. |
|
|
|
Any idea if/when you'll be adding the getwork server to the Windows binaries? No, I'm not prepared to compile myself (my head would explode trying). The main problem here lies with libmicrohttpd, which doesn't really support Windows (some specific versions will build with a POSIX simulation layer, which itself only works on 32-bit). So, someone would need to do one of these: - Port libmicrohttpd to Windows proper.
- Fix PlibC to work on not only just 32-bit Windows.
- Port BFGMiner to support another HTTP server library (I'm not aware of any decent alternatives, unfortunately).
wizkid057 thought he might have time to look into these options. |
|
|
|
"if you run cgminer, it will corrupt the real driver (until re-plug or reboot), which is likely why it isn't working."
Well that's the kind of info ppl need to know. That's why I added it to the README FAQ in 3.2.0  as for the /dev/ttyUSB never seen it show up as that Did you build your kernel without the driver? Device Drivers --->
[*] USB support --->
<M> USB Serial Converter support --->
<M> USB CP210x family of UART Bridge Controllers and what about lsusb
Bus 001 Device 007: ID 10c4:ea60 Cygnal Integrated Products, Inc. CP210x UART Bridge / myAVR mySmartUSB light This only means it appears on the USB bus, not that you have the driver for it. |
|
|
|
I have a usb Block Erupter that I run on my Gentoo-3.7.9 Linux box but I cant get it to run on bfgminer  it runs on cgminer fine so its not a driver problem, I just don't like cgminer cgminer doesn't use the driver, so that doesn't follow at all. In fact, if you run cgminer, it will corrupt the real driver (until re-plug or reboot), which is likely why it isn't working. I did a debug with -T -S /dev/bus/usb/001/007
[2013-08-26 09:37:33] Icarus Detect: Attempting to open /dev/bus/usb/001/007
[2013-08-26 09:37:33] Icarus Detect: Test failed at /dev/bus/usb/001/007: get 12011001, should: 000187a2 The device will be one of /dev/ttyUSB*, not /dev/bus/usb/... |
|
|
|
Anyone have this in use on Avalon Asic? I seen theirs beta or test version has this been updated at all?
There's an avalonhost-raminst script you can copy to the Avalon host system and use, but it's not really documented yet. Or you can just plug the Avalon boards into a normal PC. |
|
|
|
I don't see why that involves stopping. If anything, it would make stopping harder, as you'd have to tell the other cores to stop their part.
I am not quite sure i understand this. BF chips are not scanning all possible nonces in the range and they also can't cancel the calculation of a job, because of a hardware limitation. Oh, I hadn't noticed the former. What cscape meant is that it is hard to scan the missing nonces and they may not even be in some specific range which can be passed to another hardware, not that we should stop scanning. I'm confused as to what the original argument was then.. No reason to go out of your way to scan missing nonce ranges either, unless you're trying to produce work on a severely underpowered machine. Considering how much is being invested in ASICs, it's silly to try to go super-cheap on the controller. Especially when mining rigs really should have their own full bitcoin node running. So this just reinforces that it's easier to generate more work, as cscape said. |
|
|
|
Why stop scanning at all? There's no benefit to ignore the rest of the nonce space.
It's a limitation of the hardware. The bitfury chips have 756 separate hashing cores, all working in parallel, with each core hardwired to search a fixed 1/1024th of the nonce space. I don't see why that involves stopping. If anything, it would make stopping harder, as you'd have to tell the other cores to stop their part. |
|
|
|
Why would you care about scanning the rest of the nonces ? It's much easier to ignore that part of the nonce space, and create some new work instead. Why stop scanning at all? There's no benefit to ignore the rest of the nonce space. |
|
|
|
|
Anyone care to provide me with a debug log (or backtrace) of these crashes? It's been rock solid for me :/
|
|
|
|
I`ve been using BFG for a while and love it, and just found out about the request diff function so I thought I`d enter it into my bat file.
Thing is, BFG doesn`t seem to recognise this and carries on as before? If it didn't recognise it, it'd give you an error and exit. But it's inevitably just a request. The pool can choose to ignore it or honour it. Can you even have 3 as a difficulty? I thought it was 1,2,4,8,16.... It's possible, sure. It doesn't even need the be a whole number. Powers of two are just easier to deal with. |
|
|
|
NEW VERSION 3.2.0, AUGUST 30 2013
This release is the first of a new stable-focused 3.2.x branch. Many new improvements have been added, plus a merge from cgminer 3.4.0. BFGMiner 3.1.x is now discontinued, but bugfixes have been backported to 3.0.7 and 2.10.13.
...
Does not compile  My system: x86/Ubuntu/Linaro ./configure: all disabled except bitforce and ztex. I'm using the source from the tgz in the first post. First error: CC bfgminer-httpsrv.o
In file included from httpsrv.c:16:0:
/usr/include/microhttpd.h:893:55: error: unknown type name ‘va_list’
make[2]: *** [bfgminer-httpsrv.o] Error 1 An #include <stdarg.h> is missing. Fixing this, I get the next error: CC bfgminer-driver-getwork.o
driver-getwork.c: In function ‘getwork_gen_error’:
driver-getwork.c:109:9: warning: implicit declaration of function ‘MHD_create_response_from_buffer’ [-Wimplicit-function-declaration]
driver-getwork.c:109:85: error: ‘MHD_RESPMEM_MUST_FREE’ undeclared (first use in this function)
driver-getwork.c:109:85: note: each undeclared identifier is reported only once for each function it appears in
driver-getwork.c: In function ‘handle_getwork’:
driver-getwork.c:169:2: warning: implicit declaration of function ‘MHD_basic_auth_get_username_password’ [-Wimplicit-function-declaration]
driver-getwork.c:169:7: warning: assignment makes pointer from integer without a cast [enabled by default]
driver-getwork.c:173:3: warning: implicit declaration of function ‘MHD_queue_basic_auth_fail_response’ [-Wimplicit-function-declaration]
driver-getwork.c:252:58: error: ‘MHD_RESPMEM_MUST_FREE’ undeclared (first use in this function)
make[2]: *** [bfgminer-driver-getwork.o] Error 1 Looks like a very old (pre-2011) libmicrohttpd. Can you confirm the version? grep Version $(pkg-config --variable=pcfiledir libmicrohttpd)/libmicrohttpd.pc helps, but I want the builtin getwork server  I believe the oldest supported version is 0.9.5, released January 2011. This should be fixed. Is there a way to get the current share difficulty from the API? Would be useful for some monitoring tools.
The "pools" method has this info per-pool. I guess someone should add a way to determine which pool(s) are actively in use? Cool. "Last Share Difficulty" would give me that info, right? Sortof. It won't work until a share is found at the difficulty. Also, keep in mind that it's possible for devices to be hashing different difficulty works, around changes. To find out which pool is connected, I'd have to walk the list from priority 0-N and check which one is alive.
The one with the lowest priority and alive status is the one that is being connected to? Only in Failover mode. |
|
|
|
Is there a way to get the current share difficulty from the API? Would be useful for some monitoring tools.
The "pools" method has this info per-pool. I guess someone should add a way to determine which pool(s) are actively in use? |
|
|
|
|
I'm in touch with the author of this fan site and asked her to look into it.
Apparently it's related to some "Linkbucks" thing that's supposed to simply be a click-through with ads.
Should be removed soon, if it's actually malware. :/
Edit: She says she'll be updating the links to directly download 3.2.0 from my site in a few hours.
Edit: I've confirmed the site now links to the official downloads directly.
|
|
|
|
NEW VERSION 3.2.0, AUGUST 30 2013This release is the first of a new stable-focused 3.2.x branch. Many new improvements have been added, plus a merge from cgminer 3.4.0. BFGMiner 3.1.x is now discontinued, but bugfixes have been backported to 3.0.7 and 2.10.13. Human readable changelog:- modminer, x6500 & ztex: Completely removed bitstreams from BFGMiner distribution, and replaced with instructions on where to find them officially. OpenWrt users should note that the repositories still include packages, but under new names.
- New embedded getwork server to drive Block Erupter Blades and similar devices off BFGMiner. Instructions in README.ASIC. Currently not included in OpenWrt or Windows binaries.
- Fixed stratum pool-switch crash regressed in BFGMiner 3.1.0.
- Added stratum support for --request-diff by attempting to send a mining.suggest_target notification to pools when connecting.
- erupter & icarus: Various optimizations to catch racing nonces, and otherwise reduce hardware errors.
- Enable simple notifications for idle/sick/dead devices via --cmd-idle, --cmd-sick and --cmd-dead options that execute commands when the events occur.
- Support for probing/hotplugging new devices from the TUI ([M]anage, plus key).
- Simple multicast RPC detection. cg
- Imported Aaron D. Gifford's fast SHA256 implementation (used for share validation on CPU). cg
- Due to random font issues, Unicode is disabled by default. Use --unicode to enable it.
Full changelog:- cpu: sse2_64: Rename sha256_init to sha256_init_sse2 to avoid conflict with new sha2.c
- httpsrv: Some older versions of libmicrohttpd need stdint.h included first
- make-release: Include libmicrohttpd-10.dll if it exists
- Fixes column alignment in decimal fields. Workaround for printf rounding up when formatting decimals into limited width.
- New hidden --unittest option. No longer runs unit tests at startup by default, for faster startup. Added unit test for width printing of decimal numbers.
- Bugfix: configure: More fixing BSD sed syntax for curses header search
- Bugfix: configure: More fixing BSD sed syntax for libusb header search
- README: Update solo mining docs
- erupter: Continue searching a job until the end, even if an earlier result is found
- icarus: Do hwerror-triggered reopen before sending the next job the first time, to avoid having to resend it later
- icarus: Avoid sending a new job if the nonce found was for one before the current job anyway
- icarus: Double-buffer work to solve nonces found before work change takes effect
- icarus: Abstract nonce processing
- Bugfix: icarus: Avoid trying to submit 0 on comms error
- icarus: Skip unnecessary nonce memcpy
- icarus: Make state->last_work a pointer and store it on the heap
- README.RPC: Remove usbstats mention
- README: Add FAQ regarding cgminer messing up drivers
- README.RPC: Fix miner name
- README.RPC: Correct to mention other supported devices for pgaidentify
- api.c fix mcast debug message bug
- README.RPC: break all lines at most 80 characters
- Update the API Multicast documentation
- miner.php implement API Multicast handling to automatically find your local net miners
- Bugfix: RPC: Fix log format types in io_flush
- Set RT_LIBS correctly from autoconf detection
- Explicitly check for clock_nanosleep and only use it when available
- Further integrate cgsleep API into BFGMiner's timer system, so clock_nanosleep is only used on platforms with CLOCK_MONOTONIC
- Convert cgtimer_t to struct timeval
- Bugfix: Fix BSD sed syntax for curses header search
- Bugfix: configure: Really fix BSD sed syntax for libusb header search
- README.RPC: Mention multicast detection
- README: Include --mcast-* options in usage
- Use ccan's standard char* set/show functions for --api-mcast-{addr,code}
- Bugfix: RPC: Use the same mcast code in reply, and log it correctly
- API mcast only reply to remote IP's that are allowed access
- Initial API Multicast response v0.1 to find cgminer APIs
- Check for cnx_needed on each loop through wait_lp_current.
- Return positive for cnx_needed when no_work is true.
- Add no_work bool to set when we are in an underrun situation
- Reorder support names alphabetically.
- We don't want to continue into the hash_pop function if the getq is frozen.
- Carve out the unused portions of sha2 implementation.
- Import Aaron D. Gifford's fast sha256 implementation.
- Use cloned work when finding avalon results since another thread can discard the work item while it's in use.
- Provide a variant of find_work_bymidstate that returns a clone of the found work.
- Use timespecs on windows as cgtimer_t to capitalise on the higher resolution clock changes.
- Abstract out the conversion of system time to an lldiv_t in decimicroseconds.
- Use our own gettimeofday implementation on windows for it to be consistent across ming builds and higher resolution.
- Provide cgtimer_sub helper functions.
- Provide cgtimer_to_ms helper functions.
- Rename cgsleep_prepare_r as cgtimer_time to get time in cgtimer_t format and call cgsleep_prepare_r as a macro for cgtimer_time
- TimeBeginPeriod and TimeEndPeriod do not add significant overhead when run the entire time for cgminer so avoid trying to maintain balanced numbers of them for specific time calls to simplify code.
- Replace all references to the old n*sleep functions with the equivalent cgsleep_*s replacements.
- timeGetTime uses huge resources on windows so revert to using timevals for its implementation of cgtimer_t
- Quotient/remainder error in ms division.
- Provide cgtimer_to_timeval helper functions.
- Provide a timeval_to_cgtime helper function to reuse values.
- Simplify cgsleep code for windows by using a typedef for cgtimer_t that resolves to clock resolution, using that internally.
- On windows use the higher accuracy timegettime function to really get 1ms clock and timer accuracy.
- Fix missed endtimeperiod in overrun timer on windows.
- Make cgsleep_us_r take an int64_t for us.
- Make the cgsleep functions build on windows.
- Set high resolution timing on windows within the cgsleep functions.
- Provide reentrant versions of cgsleep functions to allow start time to be set separately from the beginning of the actual sleep, allowing scheduling delays to be counted in the sleep.
- Make the nmsleep and nusleep functions use the new cgsleep functions internally till functions are migrated to the new cgsleep API.
- Add a ms_to_timespec helper function, and create a cgsleep_ms function that uses absolute timers with clock_nanosleep to avoid overruns.
- Add rt lib linkage to enable use of clock_nanosleep functions with older glibc.
- Add a timeraddspec helper function.
- Provide a us_to_timespec helper function.
- Provide a us_to_timeval helper function.
- Add helper functions to convert timespec to timeval and vice versa.
- Bugfix: SGW: Discard work from log only by expiry, so post-startup hardware errors are truly only hardware errors
- bitforce: Make voltages available to RPC
- bitforce: Save voltages as array of longs internally
- Bugfix: Clear device bad nonces when zeroing stats
- cpu & opencl: Defer RUNONCE to actual autodetection, so they can be added once after startup
- Bugfix: Avoid crash activating [M]anage devices with no currently defined devices
- TUI: Support for adding new devices using the plus key from [M]anage devices
- Bugfix: Use add_serial function for scan_serial, so that "all" keyword works correctly
- SGW: Add support for new X-Hashes-Done header to allow devices to more accurately report their work
- SGW: Add X-Mining-Identifier header to inform devices what they are represented by in BFGMiner
- SGW: Use JSON for 401 response; add Server HTTP header
- AUTHORS: Add contributor Josh Lehan
- Changed comparison constants to allow for floating-point rounding
- Install README.ASIC for Block Erupter related drivers as well as Avalon
- make-release: Windows users don't need README.Debian
- README.OpenWrt: Import from BFGMiner downloads
- Silence Windows warning about send data signedness
- Fix block info
- API/miner.php add some % fields
- Don't yield on grabbing the read lock variant of cglocks.
- util.c expand quit to show file/func/line
- We should be using a cg_wlock initially in generating stratum work before downgrading the lock.
- Add the ability to downgrade a write variant of the cglocks.
- Yield after releasing a lock in case we are on a device with limited CPU resources.
- Fix --scrypt being required before scrypt intensities on command line or not working at all via config files.
- Limit intensity range according to whether scrypt is in use or not.
- Do not allow benchmark mode to be used with scrypt.
- miner.php format Best Share
- README.ASIC block erupter USB brief
- Check for negative wait time in socket_full.
- Adjust socket wait timeout in recv_line according to how long we've already waited to avoid a 60 second wait dropping to 1 second due to a blocked socket.
- force type checking on curses
- logging - size check sprintf
- Only use length-counted variants of format_unit and percentf
- Implement snprintf-like versions of format_unit and percentf
- miner - size check all sprintf
- size check get_datestamp/get_timestamp and remove unused cgpu->init
- make all statline overflow safe
- Convert the decay_time function into one that truly creates an exponentially decaying average over opt_log_interval.
- GPU fan rpm display 9999 when it overflows
- Change mode on python file.
- Only update hashmeter if we have done hashes or haven't updated longer than the log interval, fixing a us/ms error.
- README.ASIC: Document usage with Block Erupter Blades
- README: Add mention of libmicrohttpd to dependencies
- SGW: Include hash1 in work
- SGW: Include application/json Content-Type header
- Fix build without libmicrohttpd
- Allow startup to proceed with no devices, as long as RPC or HTTP are listening
- SGW: Refuse to issue new work to disabled devices
- SGW: Respond with stale rejection if share is known to be immediately stale
- Expose HTTP getwork username in Manage TUI and devdetails RPC
- Add --cmd-idle notification command for REST/WAIT conditions
- bfgminer-rpc: Accept unlimited size replies
- api-example.py: Accept unlimited size replies
- RPC: Rewrite io_data to cleanly handle unlimited sized responses
- Configure options to build --without-libmicrohttpd support
- RPC: setconfig can now change http-port (enabling or disabling the http service included)
- Embedded HTTP server to handle getwork-based mining devices
- inc_hw_errors2 function can handle a bad nonce without a known work
- Split up scan_serial function to enable internally adding ad-hoc cgpu
- Helper functions bfg_json_obj_string and share __json_array_string
- bytes_t: Add bytes_shift and bytes_nullterminate
- Fix configure help for --without-sensors
- README: Document --with-system-libblkmaker
- Bugfix: Use BSD-friendly sed syntax for libusb header search
- Bugfix: Handle bitstreams properly
- Bugfix: Skip search for addwstr since it breaks --with-curses=preference
- Bugfix: configure: Find correct curses include path without *-config
- README.FPGA: Further clarify ZTEX setup
- Display a friendly error directing to README.FPGA when bitstream cannot be loaded
- Remove bitstreams from BFGMiner distribution entirely, and include pointers where to find them in README.FPGA
- ztex: Use load_bitstream_bytes for .bin files
- fpgautils: load_bitstream_bytes support for Intel HEX format (.ihx) files
- Add missing 16-bit byteswap macros
- fpgautils: load_bitstream_bytes function to load a bitstream into a bytes_t
- bytes_t functions: bytes_init, bytes_append, bytes_reset
- Look for bitstreams in /usr/share/bitstreams
- Disable Unicode support by default (use --unicode to enable)
- make-release: Skip stripping debug info from Windows EXEs, for now
- Bugfix: Ensure work variable is assigned before checking its thr_id
- Bugfix: notifier_init (Windows): setsockopt needs an int for SO_REUSEADDR
- Bugfix: Avoid turning totals red just because a processor is idle
- Enable notifications for sick/dead with --cmd-sick and --cmd-dead options that execute commands when the event occurs
- RPC: Add cpu enable/disable/restart
- windows-build: Update to use libpdcursesw.dll
- Ensure socket error messages are used for socket errors on Windows
- fpgautils: Attempt to use Linux advisory locks on serial devices
- Bugfix: README: --temp-cutoff sets the maximum temperature before cutoff, not temperature that triggers cutoff
- Bugfix: Avoid incrementing dev_thermal_cutoff_count when just updating timestamp on status
- Detect curses support for wide characters during configure
- configure: Check for more variants of curses library names
- Bugfix: ztex: Avoid trying to format non-libusb error with libusb error name
- Bugfix: Avoid trying to assign const use_unicode variable with --no-unicode option
- Remove long-unused opt_time variable
- RPC: Avoid exposing Coinbase-Sig when it isn't supported
- Bugfix: Fix build with libblkmaker < 0.2
- Attempt stratum mining.suggest_target before mining.subscribe, if --request-diff is used
- Retry stratum if initiation fails for any reason after we have sent something (assuming there is more older variants we can try)
- Bugfix: Restore delay for authorization on stratum mining.get_transactions
- Bugfix: Fix TUI-only build
- bitforce: Extend pgaset _cmd to variable-length commands
- Bugfix: Use red for total processor count
- bitforce: Undocumented _cmd1 pgaset for experimenting with firmware
- Bugfix: curses: use_default_colors() when possible
- Bugfix: Calculate scrolling range with new cursor lines
- Bugfix: Ensure use_unicode and have_unicode_degrees constants are defined for curses-less builds
- Bugfix: Replace block_timeval with (time_t) block_time
- README: Update for --no-unicode option
|
|
|
|
I just don't know how to ensure that the bfgminer process won't be terminated when I logout my ssh session. I tried backgrounding the process which does background it, but still gets killed upon logout.
What does everyone else do? I always run BFGMiner in a GNU Screen session. BASH also lets you disown background programs. waht is the json format for using the --scan-serial flag in a config file
would like to do -S OpenCL:noauto -S erupter:/dev/cu.SLAB_USBtoUART<#> JSON array. "scan-serial":["opencl:noauto", "erupter:/dev/cu.SLAB_USBtoUART<#>"] |
|
|
|
|
TheSeven has a draft protocol for something like this that he might perhaps contribute.
But realistically, I don't think vendors are going to use the same protocol.
There are also practical problems due to how these devices work internally - any protocol that superset the behaviour they all need would be overly complicated and inevitably fail to address the needs of a future device.
Perhaps a binary protocol similar to stratum could work for future devices.
This requires much more memory, processing power, and testing (in theoretical worst-case scenarios, NOT simply mining bitcoin!) than current MCUs have.
|
|
|
|
I've updated the download links with Eligius's mirror. Hopefully my server will be online soon :/ --request-diff only works going up, not back down; and is used as a means to lower TCP/IP network chatter and bandwidth by Rasing the difficulty that your miner accepts as shares It asks the pool for a specific difficulty. The pool can decide to oblige, ignore, or do the opposite. What is "BW" in bfgminer? Bandwidth. |
|
|
|
It looks like a bad USB cable or something. :/ |
|
|
|
Luke,
Is there a way for bfgminer to set the minimum difficulty? I always thought pools decide what the minimum difficulty suppose to be and they tell miners what is the share difficulty. But just wondering if there is a way to tell the pool, "hey, I'm fast, don't bother sending me shares lower than XXX".
Thanks,
af_newbie
See --request-diff (GBT only until 3.2.0, released any hour now...) |
|
|
|
|
Show Posts
