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

[m3ta]

(lesson: roll back to 11.6)

11.8->11.6 yields me -2 Mh/s, but with the CPUs at roughly 0% now, that very small loss is perfectly bearable.

[1713874566]

1713874566

[1713874566]

1713874566

[-ck]

(lesson: roll back to 11.6)

11.8->11.6 yields me -2 Mh/s, but with the CPUs at roughly 0% now, that very small loss is perfectly bearable.

Indeed in a MH/watt equation this makes a lot of sense since it will rise significantly. This was always what cgminer was supposed to be capable of. The code in cgminer itself should be incredibly low CPU overhead. It's the drivers and windows' pthread library which use up a lot of CPU.

[m3ta]

(lesson: roll back to 11.6)

11.8->11.6 yields me -2 Mh/s, but with the CPUs at roughly 0% now, that very small loss is perfectly bearable.

Indeed in a MH/watt equation this makes a lot of sense since it will rise significantly. This was always what cgminer was supposed to be capable of[/i]. The code in cgminer itself should be incredibly low CPU overhead. It's the drivers and windows' pthread library which use up a lot of CPU.

"should be", "was supposed"?
You're too modest. From the eyes of this 'htop' fan, cgminer does that, and so much more. :]

[d3m0n1q_733rz]

Look, do you want to to know why we shouldn't remove CPU functionality from cgminer?  It's simple.  CPU mining is where a lot of us got our start.  We couldn't afford to get decent graphics cards to mine with so we CPU mined and worked our way up to small quantities like .1 BTC where we invest it in something like Ponzie to build it up faster to get a graphics card.
In short, those who need some way to begin are the ones who will need this the most.  The children; don't forget the children!

[-ck]

Look, do you want to to know why we shouldn't remove CPU functionality from cgminer?  It's simple.  CPU mining is where a lot of us got our start.  We couldn't afford to get decent graphics cards to mine with so we CPU mined and worked our way up to small quantities like .1 BTC where we invest it in something like Ponzie to build it up faster to get a graphics card.
In short, those who need some way to begin are the ones who will need this the most.  The children; don't forget the children!

That keeps occurring to me to. However, it doesn't change the fact that code needs maintenance and effort and working on one thing means it sacrifices the time I have to work on another. Not to mention that people donate BTC for GPU mining code, which is hard to ignore as an incentive.

[twmz]

Look, do you want to to know why we shouldn't remove CPU functionality from cgminer?  It's simple.  CPU mining is where a lot of us got our start.  We couldn't afford to get decent graphics cards to mine with so we CPU mined and worked our way up to small quantities like .1 BTC where we invest it in something like Ponzie to build it up faster to get a graphics card.
In short, those who need some way to begin are the ones who will need this the most.  The children; don't forget the children!

That keeps occurring to me to. However, it doesn't change the fact that code needs maintenance and effort and working on one thing means it sacrifices the time I have to work on another. Not to mention that people donate BTC for GPU mining code, which is hard to ignore as an incentive.

Frankly, this is all that matters.  The pros and cons of CPU mining aside, if maintaining CPU mining isn't something that ckolivas is interested it, then it is reasonable for it to be cut.  If you want CPU mining bad enough, invest the time and maintain the code yourself.  It's open source and in GIT.  Fork it and go crazy.

[-ck]

I'm also most open to pulling changes into my tree, and kind people like znort and others have already contributed code too.

[geek-trader]

Look, do you want to to know why we shouldn't remove CPU functionality from cgminer?  It's simple.  CPU mining is where a lot of us got our start.  We couldn't afford to get decent graphics cards to mine with so we CPU mined and worked our way up to small quantities like .1 BTC where we invest it in something like Ponzie to build it up faster to get a graphics card.
In short, those who need some way to begin are the ones who will need this the most.  The children; don't forget the children!

That keeps occurring to me to. However, it doesn't change the fact that code needs maintenance and effort and working on one thing means it sacrifices the time I have to work on another. Not to mention that people donate BTC for GPU mining code, which is hard to ignore as an incentive.

So what you are saying is, you are writing code for FREE, and people donate to you for GPU code, and no one donates to you for CPU code.

Drop the CPU code.  I don't see the problem here.  

[d3m0n1q_733rz]

I'm also most open to pulling changes into my tree, and kind people like znort and others have already contributed code too.

You know what MIGHT get people more active?  I was just thinking of using drop-in miner algorithms.  It would keep them nice and neat while making them easier to work on, change and play around with.  Granted, it would make cgminer more of a front-end for the actual algorithms, but if all of the algorithms are tossed into a folder (as SSE2/SSE4 algorithms are) marked algos or something of the sort, the miner can search the folder for the main algorithms startup files allowing for each algorithm to use as many or as few files as they need in subfolders or just flat-out have the startup file contain everything needed.  Just include a small readme of memory mapped locations to send the hashed results to and the miner/front-end can do the rest.  This will help modularize your program, make it easy to work on and promote others to jump in with their own works.
I know it's a bit of an overhaul, but it could serve fruitful.  What does everyone else think?  Viable alternative?  Let the drop-ins determine what device to mine with and just pass messages between the drop-ins and the cgminer front-end?  Granted, there would need to be a method of passing settings to the drop-in and specifying the name based upon either a compiled list or a typed one.

[-ck]

It's a good idea, but it needs even more work on my part for the cpu mining . If someone wants to write such an interface, be my guest, and I'll pull it into the tree.

[d3m0n1q_733rz]

It's a good idea, but it needs even more work on my part for the cpu mining . If someone wants to write such an interface, be my guest, and I'll pull it into the tree.

Why?  I was saying let the mining algorithms have their own settings, send their own commands to whatever devices and have the cgminer as a sort of universal interface for them all.  Yeah, there would be one last modification to each of them to make them all fit into the interface, but this would allow for dynamic starting, stopping, setting...ect.  The biggest problem I can think of is compiling the algorithms since each one wouldn't necessarily be an official part of cgminer.  But, with this setup, you wouldn't even necessarily have to close out of the miner or stop mining to refresh the algorithm list and drop-in a new algorithm to use on-the-fly.  (Silently wishes I could program worth a darn)  You see, all your miner would do is report back what device is being used and what the hashing rate is while sending work to the algorithms and sending results to the pools.
It lessens what you need to keep maintained and lets other people take the algorithm work off your hands (if they will) and have fun writing new algorithms for the miner.  Theoretically, this would mean that your miner could be used for more than just bitcoin mining too; but let's not get too far ahead of the thought process.   

[-ck]

Why? Because someone has to make the code modular the way you describe.

[dishwara]

http://securityresponse.symantec.com/security_response/writeup.jsp?docid=2011-030906-0727-99&vid=42292

cgminer-1.5.6-win32.zip is clean.

[-ck]

Excellent. Welcome to the moronic world that is windows virus management. Since you clearly DO know what cgminer is for, have fun working around that.

[shaps]

Since we are on the topic of CPU mining, I do have a question:

On my Core2Duo laptop, win7
CGMiner (best algo): 2.1 MHash
Ufasoft: 6.0 MHash

On my Core i7 laptop, win7
CGMiner (best algo): 9.3 MHash
Ufasofr: 15.7 MHash

I obviously realize that these minuscule compared to what a GPU does but I'm curious what there is such a large performance difference.

ckolivas - any idea why?

[-ck]

No windows 64 bit care factor included.

[The00Dustin]

<big false positive image removed>

http://securityresponse.symantec.com/security_response/writeup.jsp?docid=2011-030906-0727-99&vid=42292

cgminer-1.5.6-win32.zip is clean.

I'm still for CPU mining support being maintained (in spite of the fact that I don't use cgminer for cpu mining because ufasoft works better in Windows), but this might be the single best argument for removing CPU mining from cgminer.  If someone is distributing this with a botnet, they probably wouldn't bother if it couldn't CPU mine since non-user processes can't GPU mine in WinVista/7 and plenty of XP machines don't have relevant hardware.  That said, dishwara, please find out how to report this as a false positive to whoever controls "File Insight" before the other AV companies start copying the definition (not that they would do that, surely all of the AV companies find and research all of the different malwares on their own [just like they run and thoroughly test the executables which would obviousy show them that ufasoft and cgminer aren't malware]  )

[dishwara]

Is posting what Norton said to only a particular version is that much wrong doing to go & explain every AV companies in the world?

or any one has the ability to prove that every single file given in this forum consisting of 1000's of members & 10000' of post was NEVER a scam or virus or trojan contained file?

[The00Dustin]

There is no guarantee, but if you got a malware notice that you believe is false, you should report it.  Technically they should then do their due diligence.  If your are incorrect and they determine it is maltare, they should let you know that and not remove it from their definitions.

[Meatball]

The problem with the virus thing is that CGMiner in these cases probably is a virus/malware.  These asshat script kiddies that are using botnets and distributing the CPU miners for mining is going to end up making all the AV companies start flagging the mining programs as viruses.  They're being installed without the users knowledge and using up a ton of system resources, of course they'll get flagged by the AV companies.

We can keep reporting this, but unless these guys stop using the botnets, it's going to be a constant battle to keep telling the AV companies these are good programs.