I've pushed a very basically working "dev_x6500" branch to GitHub.
For now, it is very unpolished and seems to be mining at around 100 Mh/s.
Topic: (OLD) BFGMiner: modular FPGA/GPU, GBT, Stratum, RPC, Avalon/Lnx/OpnWrt/PPA/W64 - page 21. (Read 260035 times)
October 24, 2012, 08:29:18 AM
October 23, 2012, 05:23:46 PM
It is working, with showing lp active and (getblocktemplate) showing up -I will report back if I have any problems with crashing
card one is showing a little slower
2,3,4 are a little faster
(probably a setting I need to adjust)
card one is showing a little slower
2,3,4 are a little faster
(probably a setting I need to adjust)
October 23, 2012, 01:42:13 PM
And I thought I was crazy solo mining at ~3 GH...
October 23, 2012, 11:46:44 AM
What I don't understand is why miner submited to pool 1. Did pool 0 went down temporaly (why there's no message about it?),
so miner temporaly switched to pool 1? I haven't set any specific multipool strategies, e.g. miner uses the default one (failover).
Without --failover-only, BFGMiner will use other pools' work (eg, from the longpoll) in some cases. Likely the share you refer to was found before updated work had been retrieved from your local bitcoind yet. so miner temporaly switched to pool 1? I haven't set any specific multipool strategies, e.g. miner uses the default one (failover).
October 23, 2012, 10:20:56 AM
BFGMiner can do the same as described in this post or not?
In theory, though I haven't tested it lately.It's ALIVEEEEEEEEEEEEEEEEEE!
What I don't understand is why miner submited to pool 1. Did pool 0 went down temporaly (why there's no message about it?),
so miner temporaly switched to pool 1? I haven't set any specific multipool strategies, e.g. miner uses the default one (failover).
One other thing, a bug maybe - occassionaly, at around 2 minutes since startup, and only once, miner shows message "Pool 0
not providing work fast enough". I have increased both queue and scan-time since I can't find anything else that could, from
my noob perspective, affect the issue but none of those two helps.
hope for a fast fix 
October 22, 2012, 06:19:26 PM
BFGMiner can do the same as described in this post or not?
In theory, though I haven't tested it lately. If it doesn't work, feel free to open a bug report on github for it.Additionally, you can use BFGMiner with Bitcoin next-test to get native longpolling from bitcoind/Bitcoin-Qt.
October 22, 2012, 03:26:18 PM
Anyone know how to get this to work with a windows utility like restartme or restart on crash? I prefer bfgminer to aoclbf/phoenix, but it crashes leaving my system gobbling power w/o mining.
I'd rather figure out what's going on that causes it to crash for you. Can you go into more detail on that?I built a functional .bat, then a .conf file, but neither works with either utility. When the restarters point to bfgminer.exe it re-launches fine, but ignores the config file and just pops up a cmd window. 
BFGMiner looks for the config file named bfgminer.conf in the "working directory" it's launched from on Windows. Hopefully those programs let you specify that?I get a BSOD/machine check exception on my main rig(i7 970) when running bfgminer usually within a couple minutes. I previously thought it was a problem with GPU memory underclocking, due to the fact that I could induce a crash in bfgminer by specifying a memclock under 625mhz. However I've since set memory clock as low as 166mhz on multiple rigs using aoclbf and then mining with bfgminer w/o issue. I'll set up a log file for bfgminer before restarting it. The program happily mines all day and night on my older s775 rig with three vid cards and speeds are reported to be only slightly higher than phoenix, but I believe phoenix is exaggerating and the gap is wider than reported.
I have bfgminer.conf setup and working when manually started. Restart on crash allows a modified path, but it doesn't seem to pick up the info in the .conf file when restarting the program and just spits out the "no url" message. Perhaps I'll try transplanting bfgminer to C:\ and see if that helps.
October 22, 2012, 03:05:57 PM
Seems to me that you would add "__" to the beginning of each line to block that line out of your config file.
ah, commenting ala JSON style, ty Eg.
"_api-port" : "4028",
"_api-listen" : true,
October 22, 2012, 03:03:10 PM
Seems to me that you would add "__" to the beginning of each line to block that line out of your config file.
ah, commenting ala JSON style, ty 
October 22, 2012, 02:50:14 PM
Seems to me that you would add "__" to the beginning of each line to block that line out of your config file.
October 22, 2012, 02:49:09 PM
add __
can you be more specific? October 22, 2012, 02:35:25 PM
Anyone know how to get this to work with a windows utility like restartme or restart on crash? I prefer bfgminer to aoclbf/phoenix, but it crashes leaving my system gobbling power w/o mining.
I'd rather figure out what's going on that causes it to crash for you. Can you go into more detail on that?I built a functional .bat, then a .conf file, but neither works with either utility. When the restarters point to bfgminer.exe it re-launches fine, but ignores the config file and just pops up a cmd window.
BFGMiner looks for the config file named bfgminer.conf in the "working directory" it's launched from on Windows. Hopefully those programs let you specify that? October 22, 2012, 02:10:44 PM
Anyone know how to get this to work with a windows utility like restartme or restart on crash? I prefer bfgminer to aoclbf/phoenix, but it crashes leaving my system gobbling power w/o mining. I built a functional .bat, then a .conf file, but neither works with either utility. When the restarters point to bfgminer.exe it re-launches fine, but ignores the config file and just pops up a cmd window.
October 22, 2012, 11:20:10 AM
add __
October 22, 2012, 03:40:50 AM
is there a way to disable a pool in the config file so i dont have to remove it?
October 18, 2012, 08:13:02 AM
NEW VERSION 2.8.3, OCTOBER 18 2012
2.8.1 made an old, rare dereference-after-free error a lot more common in some cases (due to locking the output mutex for new debug info) so I finally got a proper chance to track it down. That crash, along with a variety of other bugs (including some scrypt fixes from Jake and Con) have been fixed for 2.8.3.
2.9.0 is still in the works to bring the long-anticipated X6500 support (which is now coming along nicely). Depending on testing and reviews before then, it may also include solo mining and/or Stratum support - so if either of those features are important to you, please test the latest code (in git branches) and let me know!
Human readable changelog:
Full changelog
2.8.1 made an old, rare dereference-after-free error a lot more common in some cases (due to locking the output mutex for new debug info) so I finally got a proper chance to track it down. That crash, along with a variety of other bugs (including some scrypt fixes from Jake and Con) have been fixed for 2.8.3.
2.9.0 is still in the works to bring the long-anticipated X6500 support (which is now coming along nicely). Depending on testing and reviews before then, it may also include solo mining and/or Stratum support - so if either of those features are important to you, please test the latest code (in git branches) and let me know!
Human readable changelog:
- Various bugs fixed, no new features.
Full changelog
- Update to libblkmaker 0.1.3
- Use explicit host to BE functions in scrypt code instead of hard coding byteswap everywhere.
- Ease the checking on allocation of padbuffer8 in the hope it works partially anyway on an apparently failed call.
- Round target difficulties down to be in keeping with the rounding of detected share difficulties.
- String alignment to 4 byte boundaries and optimisations for bin<->hex conversions.
- Fix GPU memory allocation size for scrypt
- Fix access violation with scrypt mining
- Bugfix: Only free rpc_req after using it, not before
- Bugfix: Increment work->pool->staged inside of mutex to avoid work being freed (and staged decremented) before we dereference it
- Revert "No need for extra variable in hash_push.": The extra variable is needed to avoid a rare dereference-after-free error.
- In opencl_free_work, make sure to still flush results in dynamic mode.
- Workaround: Debug log only after dec_queued, to make a free/use race more rare
- Bugfix: Remove redundant \n in debug messages
- Bugfix: Free rpc_req in pool_active and longpolls
- README: Explicitly provide Ubuntu package name for libjansson-dev
- Bugfix: Include flash_led bool in cgpu_info for Icarus-but-not-BitForce builds, since Cairnsmore uses it
- Only check work block id against pool's if the pool has a known block id
- Avoid clearing pool->block_id unless we really are changing pools
October 16, 2012, 02:16:45 PM
Does BFGMiner include Stratum now?
No. If the beta Stratum branch gets enough testing it might make 2.9.0. 
October 16, 2012, 01:37:28 PM
Does BFGMiner include Stratum now?
October 16, 2012, 01:47:47 AM
...
Also, cablepair is presumably using ModMiner devices, which cgminer doesn't work with at all (because they stopped merging updates to the driver before the production devices existed).
You really should actually check some of the things you say ...Also, cablepair is presumably using ModMiner devices, which cgminer doesn't work with at all (because they stopped merging updates to the driver before the production devices existed).
instead of making up stuff ...
You didn't send pull requests for fixing MMQ ...
First Modminer commit merged Tue Jun 12 03:02:39 2012 -0700 from pull request 219
https://github.com/ckolivas/cgminer/pull/219
https://github.com/ckolivas/cgminer/commit/76f96f4717a113129b9b3482acae8b40ff45871d
Merge Jun 12 19:55:37 2012 -0700 of pull request 221
https://github.com/ckolivas/cgminer/pull/221
https://github.com/ckolivas/cgminer/commit/67611949dcef3a36986c88d894a2c4dd6726ab8c
Merge Jun 13 17:16:27 2012 -0700 of pull request 223
https://github.com/ckolivas/cgminer/pull/223
https://github.com/ckolivas/cgminer/commit/8f76d15fae3c3107747691838deb20d4c2ff0696
Merge Jun 19 04:42:32 2012 -0700 of pull request 227
https://github.com/ckolivas/cgminer/pull/227
https://github.com/ckolivas/cgminer/commit/f70577b09731b665e46779fe3449edb1d9c20afa
They are the pull requests you sent for MMQ - where are the ones that were rejected?
I checked up to the 23rd of July Pull Request
https://github.com/ckolivas/cgminer/pull/271
Before I got bored looking for them - were they hidden somewhere after that?
October 14, 2012, 01:06:12 PM
possible bug to report:
I am seeing 2.8.2 crash on multiple linux machines about once every 24 hours
by crash I mean it just closes itself one min you see shares being accepted the next min you see $ command prompt - no error msg given
dont know if its a segmentation fault or what but it happens once a day
This doesn't help identify or fix the problem at all. Please get some information and open an Issue. I suggest building with CFLAGS="-ggdb -O0" (add this onto the ./autogen.sh line) and running with "ulimit -c unlimited; valgrind bfgminer I am seeing 2.8.2 crash on multiple linux machines about once every 24 hours
by crash I mean it just closes itself one min you see shares being accepted the next min you see $ command prompt - no error msg given
dont know if its a segmentation fault or what but it happens once a day
You might want to switch to cgminer and avoid these issues https://bitcointalksearch.org/topic/official-cgminer-mining-software-thread-for-linuxwinosxmipsarmr-pi-4110-28402
No, most of such issues tend to come from cgminer-originated changes. Also, cablepair is presumably using ModMiner devices, which cgminer doesn't work with at all (because they stopped merging updates to the driver before the production devices existed). Jump to: