Author

Topic: OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.0 - page 640. (Read 5805983 times)

Ed
member
Activity: 69
Merit: 10
New release: Version 2.2.7 - February 20, 2012
......
reject ratio higher for me, about 3% instead 0,5% at 2.1.2

my conf.
p2pool 462b252 multi merged mining
bitcoind 0.6
atiumdag 8.920.0.0 (Catalyst 11.12) / Win7 64
OpenCL 1.1 AMD-APP-SDK-v2.5 (793.1)
hero member
Activity: 616
Merit: 506
DAT here you go: http://www.fileswap.com/dl/XeABs0KKAQ/cgminer.html

Kano, I'm sorry I don't know enough about coding.  BAMT has a stop mining script that is run from gpumon on stop or restart mining.  This doesn't work after 2.1.2 so I just go to the cgminer screen session and hit q before restarting.  I'm sure it's an easy fix but the author feels 2.1.2 is the most stable and that is what he currently supports.

EDIT: should mention that's 2.2.7 w/o bitforce or cpu flags.

bamt sends: quit|\n

when cgminer should exit.

I cannot comment on how that works with 2.2.x.  There is no official build of cgminer for BAMT beyond 2.1.2.  Gave up trying to keep track of all the bugs in the 2.2.x series, figured better to wait until things settled down before pushing out an update that could potentially break a large number of working installation.  We tend to be on the conservative side with things.

Firstly, the command should be just "quit" but that isn't the actual problem
(since "quit|anything" just ignores from the "|" onwards since quit takes no notice of the parameter it's supplied)

quit without the | did not work in the version I used in creating the BAMT controller for cgminer.  I will retest this the next time we update.

However, yes it's a bug ... seems to have been around for a while now.
(I never use the quit command so I didn't notice the problem when the cgminer shutdown code changed)

But yep ... when you find a bug in something you use ... it's usually a good idea to report it Tongue

... now tracking down the cause ...

I do not use cgminer myself, therefore am not in a good place to report bugs, unless you want them secondhand.  I will encourage the folks testing cgminer and reporting trouble to me to report it via the proper channels.
legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
DAT here you go: http://www.fileswap.com/dl/XeABs0KKAQ/cgminer.html

Kano, I'm sorry I don't know enough about coding.  BAMT has a stop mining script that is run from gpumon on stop or restart mining.  This doesn't work after 2.1.2 so I just go to the cgminer screen session and hit q before restarting.  I'm sure it's an easy fix but the author feels 2.1.2 is the most stable and that is what he currently supports.

EDIT: should mention that's 2.2.7 w/o bitforce or cpu flags.

bamt sends: quit|\n

when cgminer should exit.

I cannot comment on how that works with 2.2.x.  There is no official build of cgminer for BAMT beyond 2.1.2.  Gave up trying to keep track of all the bugs in the 2.2.x series, figured better to wait until things settled down before pushing out an update that could potentially break a large number of working installation.  We tend to be on the conservative side with things.

Firstly, the command should be just "quit" but that isn't the actual problem
(since "quit|anything" just ignores from the "|" onwards since quit takes no notice of the parameter it's supplied)

However, yes it's a bug ... seems to have been around for a while now.
(I never use the quit command so I didn't notice the problem when the cgminer shutdown code changed)

But yep ... when you find a bug in something you use ... it's usually a good idea to report it Tongue

... now tracking down the cause ...
hero member
Activity: 616
Merit: 506
DAT here you go: http://www.fileswap.com/dl/XeABs0KKAQ/cgminer.html

Kano, I'm sorry I don't know enough about coding.  BAMT has a stop mining script that is run from gpumon on stop or restart mining.  This doesn't work after 2.1.2 so I just go to the cgminer screen session and hit q before restarting.  I'm sure it's an easy fix but the author feels 2.1.2 is the most stable and that is what he currently supports.

EDIT: should mention that's 2.2.7 w/o bitforce or cpu flags.

bamt sends: quit|\n

when cgminer should exit.

I cannot comment on how that works with 2.2.x.  There is no official build of cgminer for BAMT beyond 2.1.2.  Gave up trying to keep track of all the bugs in the 2.2.x series, figured better to wait until things settled down before pushing out an update that could potentially break a large number of working installation.  We tend to be on the conservative side with things.


donator
Activity: 798
Merit: 500
DAT here you go: http://www.fileswap.com/dl/XeABs0KKAQ/cgminer.html

Kano, I'm sorry I don't know enough about coding.  BAMT has a stop mining script that is run from gpumon on stop or restart mining.  This doesn't work after 2.1.2 so I just go to the cgminer screen session and hit q before restarting.  I'm sure it's an easy fix but the author feels 2.1.2 is the most stable and that is what he currently supports.

EDIT: should mention that's 2.2.7 w/o bitforce or cpu flags.
hero member
Activity: 784
Merit: 1009
firstbits:1MinerQ
Just tested, add "copytruncate" in the configuration file, so the old log file will only get truncated and keep growing
Thank You. Trying it out now.
donator
Activity: 1218
Merit: 1079
Gerald Davis
Anyone got a binary copy of linux cgminer 32bit?  BAMT is still using 32bit kernel.  Don't feel like installing all the packages to compile. 

I've compiled 2.2.6 and 2.2.7 on BAMT, how would I get them to you? FYI the API changed slightly in 2.2.x breaking the stop mining command so you have hit q in the miner screen to stop or restart mining or it just sits there and laughs at you.

Any free file hosting will be fine since cgminer isn't that big

http://www.fileswap.com/  doesn't require an account.  Just upload and get a link.
legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
Anyone got a binary copy of linux cgminer 32bit?  BAMT is still using 32bit kernel.  Don't feel like installing all the packages to compile.  

I've compiled 2.2.6 and 2.2.7 on BAMT, how would I get them to you? FYI the API changed slightly in 2.2.x breaking the stop mining command so you have hit q in the miner screen to stop or restart mining or it just sits there and laughs at you.
What was the change and how did it break BAMT?
(I have just submitted a pull request that will stop 'quit' from working unless you grant access to that command when you run cgminer - but it's not part of 2.2.7 or even accepted into the git yet)
donator
Activity: 798
Merit: 500
Anyone got a binary copy of linux cgminer 32bit?  BAMT is still using 32bit kernel.  Don't feel like installing all the packages to compile. 

I've compiled 2.2.6 and 2.2.7 on BAMT, how would I get them to you? FYI the API changed slightly in 2.2.x breaking the stop mining command so you have hit q in the miner screen to stop or restart mining or it just sits there and laughs at you.
full member
Activity: 210
Merit: 100
I know others have been able to get much higher than 266Mhas/s rates, but sadly if I increase the card's Mh/s, my system becomes unstable. I am guessing it's because I am mining on Windows, and on my main machine that's also doing some other things
You're probably guessing right. Don't worry, all in all 266 MHash/s isn't that bad for a 1120 SP card employed in a non-dedicated machine.
For a dedicated rig, however, anything lower than 300 MHash/s is inexcusable unless the card is heavily undervolted.

I really doubt it's the worksize that caused the reject rate to go up. A random packet storm seems more plausible. Care to retry?
If you happen to be using a 12.x drivers X 2.6 sdk combo --vectors 4 and --worksize 128 might boost the hash rate somewhat.
legendary
Activity: 1680
Merit: 1035
Cgminer default settings are --worksize 128 and --vectors 2 so you get those free of charge Smiley
You might want to see what --worksize 64 or --worksize 256 does for your hash rate.

The stale rate of p2pool isn't comparable to other pools as p2pool uses a much higher difficulty for its shares...
Due to high value or each high-difficulty share, I recommend you use --submit-stale when mining on p2pool.
This will make cgminer submit every share it finds, even if it thinks it is already stale. This won't make a huge difference but might net you an additional share every now and then.

Intensity is a fine tuning parameter and its optimal value is dependent on the card's hash rate.
Intensity influences the size of each uninterruptible batch of calculations the GPU performs - high intensity batches of work are larger, thus taking longer to finish.
This results in occasional situations where the GPU takes a couple of seconds too long and produces a valid but already stale result wasting the computation time.
Try lowering your intensity to 8 - if the hash rate doesn't budge you made the right call and your card didn't benefit from the higher intensity.

My general observations have been that best results are achieved (using the default two threads per gpu):
  + with <200MHash/s cards: intensity 7
  + [200..400] MHash/s : intensity 8
  + >400 MHash.s : intensity 9
Intensities beyond 9 are meant only for the new 7xxx family of cards and won't do any good when used with the old, slower cards.

Lowered intensity to I 8, and got same hash rate, so thanks for the tip. Worksize 64 for some reason made my rejection rate go up. No change between 128 and 256. So I just left that setting out and am using defalut.
I know others have been able to get much higher than 266Mhas/s rates, but sadly if I increase the card's Mh/s, my system becomes unstable. I am guessing it's because I am mining on Windows, and on my main machine that's also doing some other things
hero member
Activity: 518
Merit: 500
I get your point about it increasing U parameter but would it be best to turn it on or off Huh

Normally it is best off.

If cgminer detects a share is stale it doesn't submit it.  Those shares are marked SS.
Share submitted and designated invalid by pool are marked R.

IF cgminer stale detection is correct then SS + R should be exactly the same w/ submit stale on or off.

Turning submit stale on simply takes all the SS shares, sends them to the server who promptly rejects then.
SS = 0 and R = (SS + R)

There are a few instances where it might make sense
a) merged mining pool which gives seperate credit for NMC & BTC (not just calculate NMC reward based on BTC shares).  In that case an LP doesn't necessarily make both chains stale.  A share could be stale for one chain and not the other.

b) p2pool.  p2pool builds a share chain just like bitcoin builds a blockchain.  Since there can be re-orgs it is often useful to submit a stale share.  If a re-org occurs (and with a 10 sec block time it certainly can) then what cgminer thinks is stale could end up being valid.

TL/DR version: unless you have a specific reason you likely can keep it off.  It can't "hurt" you to turn it on you will just see more rejects at the pool level instead of SS at cgminer level.

Thank you for the detailed reply D&T ! I really appreciate your information.

So if there is no harm in submitting them then I will enable the feature. It does not seem to clogg up the GPUs but rather the network so that is fine for me.
donator
Activity: 1218
Merit: 1079
Gerald Davis
Anyone got a binary copy of linux cgminer 32bit?  BAMT is still using 32bit kernel.  Don't feel like installing all the packages to compile. 
donator
Activity: 1218
Merit: 1079
Gerald Davis
I get your point about it increasing U parameter but would it be best to turn it on or off Huh

Normally it is best off.

If cgminer detects a share is stale it doesn't submit it.  Those shares are marked SS.
Share submitted and designated invalid by pool are marked R.

IF cgminer stale detection is correct then SS + R should be exactly the same w/ submit stale on or off.

Turning submit stale on simply takes all the SS shares, sends them to the server who promptly rejects then.
SS = 0 and R = (SS + R)

There are a few instances where it might make sense
a) merged mining pool which gives seperate credit for NMC & BTC (not just calculate NMC reward based on BTC shares).  In that case an LP doesn't necessarily make both chains stale.  A share could be stale for one chain and not the other.

b) p2pool.  p2pool builds a share chain just like bitcoin builds a blockchain.  Since there can be re-orgs it is often useful to submit a stale share.  If a re-org occurs (and with a 10 sec block time it certainly can) then what cgminer thinks is stale could end up being valid.

TL/DR version: unless you have a specific reason you likely can keep it off.  It can't "hurt" you to turn it on you will just see more rejects at the pool level instead of SS at cgminer level.
full member
Activity: 210
Merit: 100
It won't make up for high network latencies and the shares which turn stale in transit.
Give it a try, just don't expect miracles.
hero member
Activity: 518
Merit: 500
So what is that submit stale flag for ?

Cheating PPS pool ops or what Huh

Thank you and forgive my ignorance but it does not really explain it in the man.
It's for the very rare occasion where cgminer might miss an opportunity to score a share when it thinks it is already stale and doesn't send it out but the pool would have accepted it.
The SS counter in cgminer windows tells you how many times cgminer didn't bother with sending a calculated share.

You can't cheat any pool because if the share is indeed stale, the pool will just classify it as such.

P2pool is specific because it uses a much higher difficulty than the rest of the pools, therefore all shares are of a much higher value.


Conman, I don't feel like delving through the code right now but wasn't the fact that --submit-stale influences Utility the reason it's off by default?
I mean the situation where a share that would otherwise be locally discarded gets sent out increasing the rating even though that share gets tossed away by the pool?

I get your point about it increasing U parameter but would it be best to turn it on or off Huh

I have really shitty internet so I think for me it might be worth turning it on but I am not sure.

Thanks !
full member
Activity: 210
Merit: 100
So what is that submit stale flag for ?

Cheating PPS pool ops or what Huh

Thank you and forgive my ignorance but it does not really explain it in the man.
It's for the very rare occasion where cgminer might miss an opportunity to score a share when it thinks it is already stale and doesn't send it out but the pool would have accepted it.
The SS counter in cgminer windows tells you how many times cgminer didn't bother with sending a calculated share.

You can't cheat any pool because if the share is indeed stale, the pool will just classify it as such.

P2pool is specific because it uses a much higher difficulty than the rest of the pools, therefore all shares are of a much higher value.


Conman, I don't feel like delving through the code right now but wasn't the fact that --submit-stale influences Utility the reason it's off by default?
I mean the situation where a share that would otherwise be locally discarded gets sent out increasing the [ U] rating even though that share gets tossed away by the pool?
full member
Activity: 210
Merit: 100
Due to high value or each high-diffigulty p2pool share, I recommend you use --submit-stale.
This will make cgminer submit every share it finds, even if it thinks it is already stale. This won't make a huge difference but might net you an additional share every now and then.
This is no longer necessary with the latest cgminer since p2pool will send the message "submitold" which cgminer now supports, so it will do it selectively for p2pool anyway.  Wink
You're a coding monster, Con. With all those updates it's hard to keep up.
Thanks.
hero member
Activity: 518
Merit: 500
Due to high value or each high-diffigulty p2pool share, I recommend you use --submit-stale.
This will make cgminer submit every share it finds, even if it thinks it is already stale. This won't make a huge difference but might net you an additional share every now and then.
This is no longer necessary with the latest cgminer since p2pool will send the message "submitold" which cgminer now supports, so it will do it selectively for p2pool anyway.  Wink

So what is that submit stale flag for ?

Cheating PPS pool ops or what Huh

Thank you and forgive my ignorance but it does not really explain it in the man.
-ck
legendary
Activity: 4088
Merit: 1631
Ruu \o/
Due to high value or each high-diffigulty p2pool share, I recommend you use --submit-stale.
This will make cgminer submit every share it finds, even if it thinks it is already stale. This won't make a huge difference but might net you an additional share every now and then.
This is no longer necessary with the latest cgminer since p2pool will send the message "submitold" which cgminer now supports, so it will do it selectively for p2pool anyway.  Wink
Jump to: