OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.1

[RedGummyBear]

ckolivas,

Sometimes when i run cgminer 4.1.1.0 i get a msg saying attempted reset got err<0> libusb_success  Then it will disable one of miners saying it hasn't hashed it over 67 seconds.

Is there a way to disable this feature?  Thanks.

[1713981826]

1713981826

[1713981826]

1713981826

[1713981826]

1713981826

[-ck]

ckolivas,

Sometimes when i run cgminer 4.1.1.0 i get a msg saying attempted reset got err<0> libusb_success  Then it will disable one of miners saying it hasn't hashed it over 67 seconds.

Is there a way to disable this feature?  Thanks.

By the time it attempts to reset the usb it has already failed and is simply sending the usb device a usb reset message - i.e. it isn't the reset that's the problem. It has already failed and is trying a reset to get it going again - which works sometimes for some devices.  Disabling the reset won't change the fact the device has already failed, it will just fail one minute earlier.

[RedGummyBear]

ckolivas,

Sometimes when i run cgminer 4.1.1.0 i get a msg saying attempted reset got err<0> libusb_success  Then it will disable one of miners saying it hasn't hashed it over 67 seconds.

Is there a way to disable this feature?  Thanks.

By the time it attempts to reset the usb it has already failed and is simply sending the usb device a usb reset message - i.e. it isn't the reset that's the problem. It has already failed and is trying a reset to get it going again - which works sometimes for some devices.  Disabling the reset won't change the fact the device has already failed, it will just fail one minute earlier.

This makes sense to me.  Thank you for your time and Explanation.

[kano]

Kano before you build could you check something for the antminer s1 for me. I was recommended to try multiminer so I didn't have to change my api but the current kano version or whatever's been tweaked elsewhere leaves me with an access denied message in the program. I'm not sure exactly what it uses to communicate with the antminer but I'm hoping its a easy fix to make.

Thanks

My S1 build is standard cgminer API with the extra fields the S1 web interface wants.
If some other program can't talk to that, then that program is broken.

[kano]

Just a little point about cgminer

ckolivas 10hr remote work on the Mofarch driver:

(click for a bigger pic)

My duplo setup

(click for a bigger pic)

Edit: it's in git of course
https://github.com/ckolivas/cgminer/tree/bflsc28

[kano]

...
The problem is that I am using an already written Makefile so don't really want to mess around with it.  I have yet to find out how to pass any switches to the build, if it is actually possible, when cross compiling.

The Makefile is in the feed.
Did you not look in there when I pointed it out?
https://bitcointalk.org/index.php?topic=28402.msg7462671#msg7462671

...
The cgminer feed that I change each time:
https://github.com/kanoi/cgminer-openwrt-packages

I'll do a new S1 tonight ... and put it up for testing (and run it myself as well of course)

FYI it's broken - I've added some fixes but more work is required - will get to it soon..ish.
I'm wondering if the S1 break is related to the S2 break also ...

[PatMan]

Ain't Lego brilliant? 

[LARTAS]

Please Mr Olivas
If I buy that
http://www.amazon.com/gp/product/B000GHUIF4/ref=ox_sc_act_title_1?ie=UTF8&psc=1&smid=A1KCADVMHMIO7S
and I conect in ASIC Terraminer. Windows will get me  like USB or like COM?. I want to erase that Mixer, for me that is the problem, that mixer 2 USB In One.
Please

[-ck]

Please Mr Olivas
If I buy that
http://www.amazon.com/gp/product/B000GHUIF4/ref=ox_sc_act_title_1?ie=UTF8&psc=1&smid=A1KCADVMHMIO7S
and I conect in ASIC Terraminer. Windows will get me  like USB or like COM?. I want to erase that Mixer, for me that is the problem, that mixer 2 USB In One.
Please

If you're asking about using the cointerra Terraminer you have to open the sealed case to get to the USB slots. The outside USB slot isn't for mining.

[LARTAS]

Yes I did before but
Booth ASICS are conected to Mixer USB.
I want each ASIC to computer
Can I connect 2 Printer cables to each one ASIC to Computer?

[mdude77]

Sort of cross posted from p2pool.  I'm posting it here because Ants use some version of cgminer, so maybe the authors have an idea as to what can be done, if anything, on the Ant side.

I built a proxy that works with p2pool.  Through the proxy I pointed my S2s at p2pool to see if I could solve the "decreased hash rate" problem.  By watching the traffic I concluded that a proxy isn't going to help getting S2s working with p2pool.

I believe the problem is the constant "drop EVERYTHING you are doing and restart" messages that p2pool sends every 30 seconds.  Then it rejects all the old work coming p2pool leading to large DOA.  The Ants simply don't change to the new job ID in a timely manner.  (They also don't respond to change work requests very quickly.)

I pointed my ants to Eligius via my proxy, and observed the same problem, the Ants don't change jobs in a timely manner.  The two huge differences are:

1 - Eligius doesn't send the "clean jobs" = true every single time like p2pool does.  Yes it changes the job id, but doesn't request a clean restart.
2 - Eligius accepts work from old jobs for a period of time while the Ants switch gears.

For interest, I put code in my proxy to always set the submitted share job_id to the current job.  That stopped the ridiculous reject rate, and interestingly enough p2pool doesn't complain about it.  However I still had the "decreased hash rate" problem.

I also noticed p2pool would complain a little bit about shares being submitted over difficulty every time p2pool changed the pseudo share size.  It even did this when I had the psuedo share size forced to the highest value (1000) by appending +1000 to my address!  This could be a p2pool problem.

Any thoughts?  Kano?


M

[-ck]

Mdude, the problem is multifactorial with the S2s based on what I know about it (I don't have one and haven't looked at the code).

p2pool has to send a clean jobs message otherwise its own pseudo block chain becomes invalid as the only way it can perpetuate shares (the "found share" message) is by starting what is the equivalent of a new block in the bitcoin block chain which means it has to send the clean jobs message. If you force your miner to keep submitting work from the old share on p2pool then you may still be just as likely to find a block but your work is pretty much all going to be seen as DOA by p2pool so you won't earn anything.

Then the S2 is woeful slow at aborting work on a clean jobs (i.e. it probably doesn't stop them at all) and that's a common asic manufacturer error/limitation and it takes a while to get enough work to get itself going again.

All of these together mean the frequent restarts + s2 current code are dead in the water.

Shares over diff message after diff changes implies there is a lot of work in the queue that is still being done and it's taking ages to catch up with the diff change. If the shares over diff message doesn't go away (and some of the cheap fork Chinese miners do this) then the driver is basically borked and doesn't support frequent diff changes and probably only supports certain values.

The slush proxy won't help you with p2pool because all it does is add another layer of latency and filters out messages so you don't see the rejects upstream. It doesn't really serve any purpose using it unless you are combining miners' work to one worker upstream.

[mdude77]

Mdude, the problem is multifactorial with the S2s based on what I know about it (I don't have one and haven't looked at the code).

p2pool has to send a clean jobs message otherwise its own pseudo block chain becomes invalid as the only way it can perpetuate shares (the "found share" message) is by starting what is the equivalent of a new block in the bitcoin block chain which means it has to send the clean jobs message. If you force your miner to keep submitting work from the old share on p2pool then you may still be just as likely to find a block but your work is pretty much all going to be seen as DOA by p2pool so you won't earn anything.

Then the S2 is woeful slow at aborting work on a clean jobs (i.e. it probably doesn't stop them at all) and that's a common asic manufacturer error/limitation and it takes a while to get enough work to get itself going again.

All of these together mean the frequent restarts + s2 current code are dead in the water.

Shares over diff message after diff changes implies there is a lot of work in the queue that is still being done and it's taking ages to catch up with the diff change. If the shares over diff message doesn't go away (and some of the cheap fork Chinese miners do this) then the driver is basically borked and doesn't support frequent diff changes and probably only supports certain values.

The slush proxy won't help you with p2pool because all it does is add another layer of latency and filters out messages so you don't see the rejects upstream. It doesn't really serve any purpose using it unless you are combining miners' work to one worker upstream.

Thank you.  That confirmed most of my conclusions as well.

Slush's proxy doesn't work with p2pool.  I made my own for that purpose. 

M

[-ck]

Yes I did before but
Booth ASICS are conected to Mixer USB.
I want each ASIC to computer
Can I connect 2 Printer cables to each one ASIC to Computer?

I'm not sure I understand the question sorry. You can open your cointerra terraminer up and plug in two usb printer cables to your PC. Each card is seen as a separate device on cgminer.

[LARTAS]

Hi
Wit TERRACOIN ASICS Can I regulate power like beaglebone?.
I want to put level 7 because it breakdown the light in house each time

[-ck]

Hi
Wit TERRACOIN ASICS Can I regulate power like beaglebone?.
I want to put level 7 because it breakdown the light in house each time

--ps-load
Use a number from 1-100 probably 70 for level 7

[LARTAS]

I put that in shortcut? After put Pool and user?

[bitpop]

I can't find this anywhere but what are the difficulty units?
I think G is billions? What's T?

I know this isn't cgminer but http://imgur.com/JjfqI91

[bitpop]

I put that in shortcut? After put Pool and user?

I'd put it last

[-ck]

I can't find this anywhere but what are the difficulty units?
I think G is billions? What's T?

http://en.wikipedia.org/wiki/Metric_prefix