P_Shep
Legendary
Offline
Activity: 1795
Merit: 1208
This is not OK.
|
|
July 30, 2012, 04:28:47 PM |
|
I compiled 2.6.1 for Windows and I've got it running now with 2 BFL Singles.
It looks like it tries to restart now if I pull the power plug on the Singles.
It was always supposed to... missing -> attempt to reconnect. It's a little touchy if it tries to restart while the BFLs are still booting up (the first time I tried, only one of them restarted). On a second try they both restarted, though.
Not a problem. cgminer won't interfere with the single's own boot up routine.
|
|
|
|
P_Shep
Legendary
Offline
Activity: 1795
Merit: 1208
This is not OK.
|
|
July 30, 2012, 04:37:17 PM |
|
The latest build seems to mess with something on one of my machines. After about 2 to 4 minutes, the temperature for "BFL 1" disappears, and it starts saying error: temperature celsius or something of the sort. It also starts saying send work reports: errunknown! It then goes down to 0 hashrate. It (seems) to always be the same Single. I'd blame the Single itself, but no previous build does this. It's also present in BFG's 2.6.1. Using Win7x64, and if it matters, the operating host's environment is 1x7970, 3x5850, and 7 BFL Singles in one prompt.
EDIT: Seems my other machine has fallen to this as well, it just took longer to happen. For this one, it's BFL 2, and it gives the same errors: error: get result reports: OK error: send block data reports: ERR:UNKNOWN! error: send work reports: temperature <celsius> error: send block data reports: temperature <celsius>
After the <celsius>, it will actually have a number that seems to be the correct temperature. But the temperature is not shown above, and it won't show any hash rate, and it's not being declared sick.
Maybe something's messed with the mutex there. Seems the 'get temp', which is on the watchdog thread, is interfering with the hashing thread. I'm sticking with 2.5.0... Have a solid 7 days non-stop operation at the moment. Those with FPGAs really have little use for scrypt in >2.6.0 I'll have a look at what Con/Kano did to the code at some point, but not soon. FAR too busy right now.
|
|
|
|
Epoch
Legendary
Offline
Activity: 922
Merit: 1003
|
|
July 30, 2012, 04:43:40 PM |
|
The latest build seems to mess with something on one of my machines. After about 2 to 4 minutes, the temperature for "BFL 1" disappears, and it starts saying error: temperature celsius or something of the sort. It also starts saying send work reports: errunknown! It then goes down to 0 hashrate. It (seems) to always be the same Single. I'd blame the Single itself, but no previous build does this. It's also present in BFG's 2.6.1. Using Win7x64, and if it matters, the operating host's environment is 1x7970, 3x5850, and 7 BFL Singles in one prompt.
I also see new issues with my Singles using the current cgminer and bfgminer builds (previous builds work fine). I've seen the case (more than once, across 3 machines) where one of my Singles will stop hashing, is declared SICK, and cgminer/bfgminer will try to re-initialize it but never succeeds. The only 'fix' is to restart cgminer/bfgminer. So BFL Single support seems to be broken in the current builds; I've rolled back to a previous build which works fine. YMMV.
|
|
|
|
th3.r00t
|
|
July 30, 2012, 05:30:49 PM |
|
I think there is a bug in threads running on GPU. It reports 0 Mh/s, when running --scrypt, but the card is hashing away nice.
|
|
|
|
Inaba
Legendary
Offline
Activity: 1260
Merit: 1000
|
|
July 30, 2012, 06:25:56 PM |
|
The latest build seems to mess with something on one of my machines. After about 2 to 4 minutes, the temperature for "BFL 1" disappears, and it starts saying error: temperature celsius or something of the sort. It also starts saying send work reports: errunknown! It then goes down to 0 hashrate. It (seems) to always be the same Single. I'd blame the Single itself, but no previous build does this. It's also present in BFG's 2.6.1. Using Win7x64, and if it matters, the operating host's environment is 1x7970, 3x5850, and 7 BFL Singles in one prompt.
EDIT: Seems my other machine has fallen to this as well, it just took longer to happen. For this one, it's BFL 2, and it gives the same errors: error: get result reports: OK error: send block data reports: ERR:UNKNOWN! error: send work reports: temperature <celsius> error: send block data reports: temperature <celsius>
After the <celsius>, it will actually have a number that seems to be the correct temperature. But the temperature is not shown above, and it won't show any hash rate, and it's not being declared sick.
Having the same problem on Linux... I'm not entirely convinced it's not the hardware, but it seems unlikely I have 3 units failing like that.
|
If you're searching these lines for a point, you've probably missed it. There was never anything there in the first place.
|
|
|
farfie
Newbie
Offline
Activity: 63
Merit: 0
|
|
July 30, 2012, 06:48:39 PM |
|
Having the same problem on Linux... I'm not entirely convinced it's not the hardware, but it seems unlikely I have 3 units failing like that.
Highly doubt it's your hardware Inaba. I went back 2.5.0 and everything's smooth as butter again.
|
|
|
|
ThiagoCMC
Legendary
Offline
Activity: 1204
Merit: 1000
฿itcoin: Currency of Resistance!
|
|
July 30, 2012, 07:27:51 PM |
|
Hi! For my two 5870 @ 930 / 1300, I'm getting: GPU_USE_SYNC_OBJECTS=1 DISPLAY=:0 ./cgminer/cgminer --scrypt -o http://MY_IP:9327 -u ltc5870.2 -p X --shaders 1600 -I 17 ~300kH from each NiceIf I run with: GPU_MAX_ALLOC_PERCENT=100, I see: [2012-07-30 16:21:05] Error -5: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel) [2012-07-30 16:21:05] GPU 1 failure, disabling!
GPU0 seems to work with GPU_MAX_ALLOC_PERCENT... Any better setup for 5870s?! On top on Ubuntu 12.04 64 bits, Catalyst 12.6 and SDK v2.6. Tks! Thiago
|
|
|
|
btckeeper
Newbie
Offline
Activity: 13
Merit: 0
|
|
July 30, 2012, 07:40:56 PM |
|
Hello all. How can cgminer autostart in Linux Ubuntu 12.04 ? Script like #!/bin/bash cd ~ cd /cgminer-2.6.1-x86_64-built/ ./cgminer -o http://pool:port -u usr_name -p pass -I 9
is not working Thanks for help
|
|
|
|
ThiagoCMC
Legendary
Offline
Activity: 1204
Merit: 1000
฿itcoin: Currency of Resistance!
|
|
July 30, 2012, 07:50:24 PM |
|
Hello all. How can cgminer autostart in Linux Ubuntu 12.04 ? Script like #!/bin/bash cd ~ cd /cgminer-2.6.1-x86_64-built/ ./cgminer -o http://pool:port -u usr_name -p pass -I 9
is not working Thanks for help Hi! I'm using this: cat /etc/init/btc-miner.conf
description "Start BTC Mining" start on runlevel [2345] stop on runlevel [!2345] kill timeout 30 script exec /usr/bin/screen -dmS CGMiner su -c '/home/miner/miner-default' end script
file /home/miner/miner-default /home/miner/miner-default: symbolic link to `miner-user1'
miner-user1 Bash scrypt: #! /bin/sh DISPLAY=:0 /home/miner/cgminer/cgminer -c /home/miner/cgminer-user1.conf
file /home/miner/cgminer /home/miner/cgminer: symbolic link to `cgminer-2.6.1-x86_64-built'
Also the X server start-up: cat /etc/init/xserver.conf
description "Start X Server only for mining" start on runlevel [2345] stop on runlevel [!2345] kill timeout 30 script exec /usr/bin/X 2>&1 end script
Only SSH access to my rigs... Cheers! Thiago
|
|
|
|
SAC
|
|
July 30, 2012, 09:00:37 PM |
|
Hi! For my two 5870 @ 930 / 1300, I'm getting: GPU_USE_SYNC_OBJECTS=1 DISPLAY=:0 ./cgminer/cgminer --scrypt -o http://MY_IP:9327 -u ltc5870.2 -p X --shaders 1600 -I 17 ~300kH from each NiceIf I run with: GPU_MAX_ALLOC_PERCENT=100, I see: [2012-07-30 16:21:05] Error -5: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel) [2012-07-30 16:21:05] GPU 1 failure, disabling!
GPU0 seems to work with GPU_MAX_ALLOC_PERCENT... Any better setup for 5870s?! On top on Ubuntu 12.04 64 bits, Catalyst 12.6 and SDK v2.6. Tks! Thiago Clocks are too high I had similar until I started playing around after noticing a 5850 of mine that will only go to a core of 765 was getting the same Kh/s as my 5870s high clocked so now I use 770,1050 and get 356Kh/s on Ubuntu 11.04 64 bits, Catalyst 12.6 and SDK v2.7. Oh and forget the --shaders it does not give you the correct .bin file used it is always lower than if using --thread-concurrency 8000,8000 in your case which will give you a 8000.bin file used as opposed to on my install somewhere in the 7000s.bin it was using with the --shaders. The startup file I use below the 5850s seem to like 8192 as opposed to the 7200 it should be for 5x shaders that seems to be the sweet spot for all my other cards this on the new 2.6.1 code I have the pools set in the .cgminer.conf. The -I 18 gives me the same amount of stales as the -I 17 does but get a few more kh/s out of the cards 19 or 20 give a little higher speeds but results in double or triple the stales respectively so in the end leave you with a lesser overall effective speed. cat ltc.sh export GPU_USE_SYNC_OBJECTS=1 export GPU_MAX_ALLOC_PERCENT=100 export DISPLAY=:0
~/cgminer-ltc --scrypt --worksize 256,256,256 --thread-concurrency 8000,8000,8192 --vectors 4,4,4 --gpu-threads 2 -I 18,18,18 -g 1 --auto-fan --auto-gpu --temp-target 81
My method for getting the most from the cards. 1. Start with --thread-concurrency at 4x and 5x shaders with an 8192 thrown in to see what gets you the fastest speed. 2. Move onto the core clock raising it up/down until you get the fastest kh/s the core is usually always lower than what you think it should be. 3. Now on to the memory clock raising it up/down until you find the sweet spot that gives you your total highest speed you will get, this again is usually lower than you think it should be on an over clock. 4. Play around some more after you have what you think is the highest speeds as sometimes that extra/lesser core/memory 5-10mhz will find a new sweet spot for you resulting in an extra 5-10Kh/s. Always go with an ending number of 0 or 5 as the others seem to screw things up.
|
|
|
|
ddd1
|
|
July 30, 2012, 09:09:11 PM |
|
I have auto-fan activated and have gpu-fan 0-25%, so I want to have at most 25% fan speed.
So I want the cgminer start from 25% then decrease but it starts from 50% fan speed then goes down to 0% 1100rpm?
Cooler is wery good and only needs wery low % fan speed.
"intensity" : "7,9,9", "vectors" : "1,1,1", "worksize" : "64,64,64", "kernel" : "poclbm,poclbm,poclbm", "gpu-engine" : "0-0,0-0", "gpu-fan" : "0-25,0-25,0-25",
"auto-fan" : true,
"gpu-memclock" : "0,0", "gpu-memdiff" : "0,0", "gpu-powertune" : "0,0", "gpu-vddc" : "0.000,0.000", "temp-cutoff" : "90,90,90", "temp-overheat" : "89,89,89", "temp-target" : "88,88,88", "api-port" : "4028", "expiry" : "120", "gpu-dyninterval" : "7", "gpu-platform" : "0", "gpu-threads" : "2", "log" : "5", "queue" : "1", "retry-pause" : "5", "scan-time" : "60", "temp-hysteresis" : "3", "shares" : "0", "kernel-path" : "/usr/local/bin"
|
|
|
|
BlackPrapor
|
|
July 30, 2012, 09:14:22 PM |
|
The latest build seems to mess with something on one of my machines. After about 2 to 4 minutes, the temperature for "BFL 1" disappears, and it starts saying error: temperature celsius or something of the sort. It also starts saying send work reports: errunknown! It then goes down to 0 hashrate. It (seems) to always be the same Single. I'd blame the Single itself, but no previous build does this. It's also present in BFG's 2.6.1. Using Win7x64, and if it matters, the operating host's environment is 1x7970, 3x5850, and 7 BFL Singles in one prompt.
EDIT: Seems my other machine has fallen to this as well, it just took longer to happen. For this one, it's BFL 2, and it gives the same errors: error: get result reports: OK error: send block data reports: ERR:UNKNOWN! error: send work reports: temperature <celsius> error: send block data reports: temperature <celsius>
After the <celsius>, it will actually have a number that seems to be the correct temperature. But the temperature is not shown above, and it won't show any hash rate, and it's not being declared sick.
Having the same problem on Linux... I'm not entirely convinced it's not the hardware, but it seems unlikely I have 3 units failing like that. Inaba, I just had a miner crash, but couldn't see what was the reason. Previously I used BFG 2.5.1 and you know that I had same problems with units you have now. I'll keep my eye on the miners, and hope that its not the same error again.
|
There is no place like 127.0.0.1 In blockchain we trust
|
|
|
kano
Legendary
Offline
Activity: 4592
Merit: 1851
Linux since 1997 RedHat 4
|
|
July 30, 2012, 09:50:52 PM |
|
Having the same problem on Linux... I'm not entirely convinced it's not the hardware, but it seems unlikely I have 3 units failing like that.
Highly doubt it's your hardware Inaba. I went back 2.5.0 and everything's smooth as butter again. Does the problem occur during throttling? Do you always specify "bitforce:" on the front of the serial device specification? One change that I added (that's in 2.6.0 and 2.6.1) is there is now a timeout in BFL if you specify "bitforce:" (there was always a timeout for BFL if you didn't specify "bitforce:" so I changed it work the same way in both situations) Thus in the log there should be a message saying there was a timeout at the time the problem occurred if that is the cause of it. (BFL needs to be able to timeout rather than hang if something goes wrong on linux - it used to hang sometimes when there were comm errors) However, I guess it could be LP related when it tries to abort work that could be during a throttle? If you can get it to fail either see if there is a timeout message (and post a pastebin of the log for up to 1 minute before it happens) or if there is no timeout message, run it in "-D" debug mode and post a pastebin of the last minute or so of that log up to when it fails Or course, visiting FreeNode IRC #cgminer will be easier ... My testing of this is on my BFL that has hardware issues lately so I put the fastest bitstream on it and thus was expecting to see any possible problems But if it is during an LP then that's not easy to make happen ... Edit: oh there was also another change that luke-jr did at the same time (I didn't notice) - so I guess it could be either of our changes. https://github.com/ckolivas/cgminer/commit/cf36331d815e7b87131d547b92b9ceaa218d114d
|
|
|
|
Luke-Jr
Legendary
Offline
Activity: 2576
Merit: 1186
|
|
July 30, 2012, 10:03:30 PM |
|
FWIW, any BFGMiner issues (including bugs in the FPGA drivers, that CGMiner copied from BFGMiner) should be reported in detail here. I'm also available in #Eligius for real-time troubleshooting.
|
|
|
|
kano
Legendary
Offline
Activity: 4592
Merit: 1851
Linux since 1997 RedHat 4
|
|
July 30, 2012, 10:18:17 PM |
|
FWIW, any BFGMiner issues (including bugs in the FPGA drivers, that CGMiner copied from BFGMiner) should be reported in detail here. I'm also available in #Eligius for real-time troubleshooting. Hmm - well the actual correct term is a pull request you sent to cgminer ... but anyway ...
|
|
|
|
farfie
Newbie
Offline
Activity: 63
Merit: 0
|
|
July 30, 2012, 10:36:00 PM |
|
Does the problem occur during throttling? Do you always specify "bitforce:" on the front of the serial device specification? One change that I added (that's in 2.6.0 and 2.6.1) is there is now a timeout in BFL if you specify "bitforce:" (there was always a timeout for BFL if you didn't specify "bitforce:" so I changed it work the same way in both situations) Thus in the log there should be a message saying there was a timeout at the time the problem occurred if that is the cause of it. (BFL needs to be able to timeout rather than hang if something goes wrong on linux - it used to hang sometimes when there were comm errors) However, I guess it could be LP related when it tries to abort work that could be during a throttle? If you can get it to fail either see if there is a timeout message (and post a pastebin of the log for up to 1 minute before it happens) or if there is no timeout message, run it in "-D" debug mode and post a pastebin of the last minute or so of that log up to when it fails Or course, visiting FreeNode IRC #cgminer will be easier ... My testing of this is on my BFL that has hardware issues lately so I put the fastest bitstream on it and thus was expecting to see any possible problems But if it is during an LP then that's not easy to make happen ... Edit: oh there was also another change that luke-jr did at the same time (I didn't notice) - so I guess it could be either of our changes. https://github.com/ckolivas/cgminer/commit/cf36331d815e7b87131d547b92b9ceaa218d114dMy devices aren't throttling, and yes I always specify "bitforce:". I also don't believe the problem would be related to a long poll, since both machines would have received long polls near the same time, but one started having singles stop hashing much sooner than the other. I think that makes sense anyway, I'm no expert.
|
|
|
|
-ck (OP)
Legendary
Offline
Activity: 4256
Merit: 1645
Ruu \o/
|
|
July 30, 2012, 11:04:48 PM |
|
FWIW, any BFGMiner issues (including bugs in the FPGA drivers, that CGMiner copied from BFGMiner) should be reported in detail here. I'm also available in #Eligius for real-time troubleshooting. Hmm - well the actual correct term is a pull request you sent to cgminer ... but anyway ... That's fine, I can't see what luke-jr says anyway since ignoring him unless someone else quotes him and only allow him to speak to me in c code. Since he still seems to have the same attitude and spams my thread, I can just ignore his code as well. Rather than showing humility it seems he is getting worse, so perhaps it's time to refuse to take any more code.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
Luke-Jr
Legendary
Offline
Activity: 2576
Merit: 1186
|
|
July 30, 2012, 11:21:28 PM |
|
FWIW, any BFGMiner issues (including bugs in the FPGA drivers, that CGMiner copied from BFGMiner) should be reported in detail here. I'm also available in #Eligius for real-time troubleshooting. Hmm - well the actual correct term is a pull request you sent to cgminer ... but anyway ... That's fine, I can't see what luke-jr says anyway since ignoring him unless someone else quotes him and only allow him to speak to me in c code. Since he still seems to have the same attitude and spams my thread, I can just ignore his code as well. Rather than showing humility it seems he is getting worse, so perhaps it's time to refuse to take any more code. And so Con's fork deviates yet even further... guess that means he'll be rejecting the bitforce bugfix.
|
|
|
|
kano
Legendary
Offline
Activity: 4592
Merit: 1851
Linux since 1997 RedHat 4
|
|
July 30, 2012, 11:44:16 PM |
|
Does the problem occur during throttling? Do you always specify "bitforce:" on the front of the serial device specification? One change that I added (that's in 2.6.0 and 2.6.1) is there is now a timeout in BFL if you specify "bitforce:" (there was always a timeout for BFL if you didn't specify "bitforce:" so I changed it work the same way in both situations) Thus in the log there should be a message saying there was a timeout at the time the problem occurred if that is the cause of it. (BFL needs to be able to timeout rather than hang if something goes wrong on linux - it used to hang sometimes when there were comm errors) However, I guess it could be LP related when it tries to abort work that could be during a throttle? If you can get it to fail either see if there is a timeout message (and post a pastebin of the log for up to 1 minute before it happens) or if there is no timeout message, run it in "-D" debug mode and post a pastebin of the last minute or so of that log up to when it fails Or course, visiting FreeNode IRC #cgminer will be easier ... My testing of this is on my BFL that has hardware issues lately so I put the fastest bitstream on it and thus was expecting to see any possible problems But if it is during an LP then that's not easy to make happen ... Edit: oh there was also another change that luke-jr did at the same time (I didn't notice) - so I guess it could be either of our changes. https://github.com/ckolivas/cgminer/commit/cf36331d815e7b87131d547b92b9ceaa218d114dMy devices aren't throttling, and yes I always specify "bitforce:". I also don't believe the problem would be related to a long poll, since both machines would have received long polls near the same time, but one started having singles stop hashing much sooner than the other. I think that makes sense anyway, I'm no expert. What are you running on? There seems to be an ongoing bug fix on bug fix on bug fix for this at the moment (nothing to do with me) The first bugfix pshep did was mostly to remove the 2 lines of luke-jr's commit (which has gone in the chain of bug fixes and not returned) I might be able to make a version that just removes that commit and compile it for you until the mess is sorted out by the others. If it's windows - then yep that's OK also I can make one of them. Let me know.
|
|
|
|
ShadesOfMarble
Donator
Hero Member
Offline
Activity: 543
Merit: 500
|
|
July 30, 2012, 11:46:33 PM |
|
I had to restart cgminer 2.6.1 on two different rigs because the API did no longer respond to any request (it did not even time out!?) Mining was still working and this has never happened with earlier versions.
Anyone experienced the same problem?
|
|
|
|
|