OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.0 - page 463.

sharky112065

sr. member

Activity: 383

Merit: 250

C. Kolivas

I am having trouble compiling cgminer 2.8.4 on MinGW 32 (Windows 7 x64). The last version I compiled was 2.7.5 and that version had no problems.

Compiling from "cgminer-2.8.4.tar.bz2" on your site.

Configure complains about not having any GPU/CL support.

If I remark out the following lines in configure.ac

#if test "x$ATISTREAMSDKROOT" != x; then
#   OPENCL_FLAGS="-I$ATISTREAMSDKROOT/include $OPENCL_FLAGS"
#   OPENCL_LIBS="-L$ATISTREAMSDKROOT/lib/$ARCH_DIR $OPENCL_LIBS"
#fi

#if test "x$AMDAPPSDKROOT" != x; then
#   OPENCL_FLAGS="-I$AMDAPPSDKROOT/include $OPENCL_FLAGS"
#   OPENCL_LIBS="-L$AMDAPPSDKROOT/lib/$ARCH_DIR $OPENCL_LIBS"
#fi

and then type:
autoreconf -fvi
  and then
CFLAGS="-O2 -msse2" ./configure
  and then
make

And now, cgminer 2.8.4 compiles just fine with AMD GPU support.

$ATISTREAMSDKROOT = Empty; Does not exist on MinGW 32
$AMDAPPSDKROOT = C:\Program Files (x86)\AMD APP\ and is a valid directory.
$ARCH_DIR = Empty; Does not exist on MinGW 32

There is a $PROCESSOR_ARCHITECTURE environment variable and it is set to "x86" but using this variable could be problematic as the two AMD directories are "x86" and "x86_64" and the possible settings for the variable are "x86", "AMD64", and "IA64".

I'm sure windows users that compile on MinGW will have problems with this and it should probably be fixed.

TheHarbinger

sr. member

Activity: 378

Merit: 250

Why is it so damn hot in here?

Quote from: TheHarbinger on October 17, 2012, 09:50:10 PM

Con, I'll give the new build a whirl tomorrow since I was one of the ones that was having issues with dynamic difficulty. 2.7.6 is what I've been running, and has been rock solid for me, although it seems a few others still had problems. The only thing I did notice is that if I loose my network, the queued value (Q:) shoots through the roof and then drives the efficiency (E:) down through the floor. But it doesn't actually effect anything since those are just displayed values, and I know the queued work isn't actually increasing like a rocket ship because it can't get the work if the network is down. Roll Eyes

Okay, not seeing any issues with dynamic difficulty with 2.8.4, so hopefully that issue is dead and truly buried. I'm still seeing the weird behavior with queued work after a network loss, though this is clearly a very minor issue as it does not truly affect anything except displayed stats.

rav3n_pl

legendary

Activity: 1361

Merit: 1003

Don`t panic! Organize!

Quote from: -ck on October 19, 2012, 05:46:51 AM

I'm upgrading the status of cgminer 2.8.4 to stable status and removing 2.7.7 from the top directory. Even if some stratum bug does show up in the future, disabling stratum with --fix-protocol will still make it more stable than 2.7.7 is, and it's time to move on.

Good, but Q: is still counting in some odd way

Code:

 cgminer version 2.8.4 - Started: [2012-10-19 09:35:35]
-----------------------------------------------------------------------
 (5s):5.353M (avg):14.46Mh/s | Q:12  A:63  R:0  HW:0  E:525%  U:0.2/m
 TQ: 0  ST: 1  SS: 0  DW: 11  NB: 1  LW: 3816  GF: 0  RF: 0  WU: 0.2
 Connected to http://rav3n.dtdns.net:9332 with LP as user toy.gpu
 Block: 0000005defaf82b794f28a67e56e7482...  Started: [09:35:35]
-----------------------------------------------------------------------
 [P]ool management [G]PU management [S]ettings [D]isplay options [Q]uit
 GPU 0:  66.0C  56%    | 14.57M/14.47Mh/s | A:63 R:0 HW:0 U:0.20/m I: 3
-----------------------------------------------------------------------

 [2012-10-19 14:54:24] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:54:28] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:54:29] Accepted 89346c22 Diff 1/1 GPU 0
 [2012-10-19 14:54:29] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:54:33] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:54:34] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:54:41] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:54:50] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:54:51] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:54:56] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:55:24] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:55:33] LONGPOLL from pool 0 requested work restart
 [2012-10-19 14:55:43] LONGPOLL from pool 0 requested work restart

It is running about 6 hrs on p2pool so Q should be over 360/hour.
Where was more than 12 blocks too...

EDIT:
Some time later O_o

Code:

 cgminer version 2.8.4 - Started: [2012-10-19 09:35:35]
--------------------------------------------------------------------------------
 (5s):14.57M (avg):13.18Mh/s | Q:66382  A:81  R:5  HW:0  E:0%  U:0.2/m
 TQ: 0  ST: 1  SS: 0  DW: 34  NB: 1  LW: 4852  GF: 18  RF: 0  WU: 0.2
 Connected to http://rav3n.dtdns.net:9332 with LP as user toy.gpu
 Block: 0000005defaf82b794f28a67e56e7482...  Started: [09:35:35]
--------------------------------------------------------------------------------
 [P]ool management [G]PU management [S]ettings [D]isplay options [Q]uit
 GPU 0:  65.5C  54%    | 14.56M/13.18Mh/s | A:81 R:5 HW:0 U:0.17/m I: 3
--------------------------------------------------------------------------------

 [2012-10-19 17:40:57] LONGPOLL from pool 0 requested work restart

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

I'm upgrading the status of cgminer 2.8.4 to stable status and removing 2.7.7 from the top directory. Even if some stratum bug does show up in the future, disabling stratum with --fix-protocol will still make it more stable than 2.7.7 is, and it's time to move on.

dlasher

sr. member

Activity: 467

Merit: 250

Quote from: -ck on October 18, 2012, 03:28:32 PM

Quote from: dlasher on October 18, 2012, 03:05:28 PM

Interesting failure state today.

2 minutes. It's there already.

Hmmm... I wonder what went wrong then.. I'll have to go back and look at logs again, see if I can dig anything up. In the mean time, I updated all the 2.8.3's to 2.8.4, hope there's magic in the upgrade.

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: dlasher on October 18, 2012, 03:05:28 PM

Interesting failure state today.

I have some (2.8.3) miners behind stratum 1.1.1 proxy.. primary pool stratum was pointed at failed. (stratum crashed). cgminer stayed connected to proxy, which stayed up. proxy couldn't get work from primary pool, stayed idle, all miners went idle, cgminer never failed down to next pools in the chain.

It would be nice to be able to set some sort of threshold.. (And shoot me conman if you already do this and I just missed it)..in cgminer. If I don't get any work from a pool in {threshold} minutes, in spite of being TCP/UDP/etc up, I should mark it dead. (I swear cgminer already does this for getwork.. I see messages about "pool not providing work fast enough")

(FWIW, I'm going to post something over in Slush's proxy thread, because I think his proxy should have marked itself dead, but somehow cgminer should know a pool is "starving", shouldn't it?)

2 minutes. It's there already.

EDIT: The problem is the proxy may have been lying and continuing to send what looked like fresh work to cgminer. It's time to stop using the proxy Wink

dlasher

sr. member

Activity: 467

Merit: 250

Interesting failure state today.

I have some (2.8.3) miners behind stratum 1.1.1 proxy.. primary pool stratum was pointed at failed. (stratum crashed). cgminer stayed connected to proxy, which stayed up. proxy couldn't get work from primary pool, stayed idle, all miners went idle, cgminer never failed down to next pools in the chain.

It would be nice to be able to set some sort of threshold.. (And shoot me conman if you already do this and I just missed it)..in cgminer. If I don't get any work from a pool in {threshold} minutes, in spite of being TCP/UDP/etc up, I should mark it dead. (I swear cgminer already does this for getwork.. I see messages about "pool not providing work fast enough")

(FWIW, I'm going to post something over in Slush's proxy thread, because I think his proxy should have marked itself dead, but somehow cgminer should know a pool is "starving", shouldn't it?)

jjiimm_64

legendary

Activity: 1876

Merit: 1000

Quote from: Karasur on October 18, 2012, 09:07:50 AM

it seems that 2.8.4 uses gpu more aggressive...
Some of my radeon 7970 (1170/1050/65C) work well with 2.7.7 but with 2.8.4 are sick after 2-3 min btc mining.

you must have free electricity or something... 1170! The other guy was right, lower your clock for stability. if you run at 950 you can lower the volt to 0.949 and cut your electricity usage in half and only go down about 70Mh

Vbs

hero member

Activity: 504

Merit: 500

Quote from: Nite69 on October 18, 2012, 11:21:01 AM

Quote from: -ck on October 17, 2012, 03:44:21 PM

Quote from: Nite69 on October 17, 2012, 11:38:14 AM

Quote from: Vbs on October 16, 2012, 02:28:54 PM

Nope. Any time you change 1 bit on the input, you change all bits on the output hash (SHA-256). If it were like you said above, it would be a pretty crappy cryptographic function in the first place! Grin

Not all. If it would be so, changing 2 or any even number of bits would result in the same hash. Optimally one bit changes about half of the result hash.
Bt the real problem is that changing a bit affets on that what other bits affect. Anyway, I also think it would not work. Just a stupid idea.

No, a good hash function, and sha256 is that, will make the results have absolutely nothing in common even if only 1 bit is changed. This is called the avalanche effect and is a most desirable property.

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

That's what I practically told. But if one input bit allways changes *all* the output bits, then the function is not a hash, but it is binary "NOT". And that means they (inputs with 1bit difference) would practically have almost all in common with the results. So optimally (as your link also noted), about half of the output bits are changed when one input bit changes.
But this conversation is quite stupid. We all agree, but are still arguing. And quite much off-topic.

Of course, if you use a binary system for representing something, at the bit-level there can only be 0's or 1's. So, even if the bits don't seem to change much at a low level between the hashes of 2 different data, the hashes themselves are quite different because they can only be interpreted at the word level, and they are built to minimize the correlation with original data.

Think of a hash like a "random" function. Although not all bits change between runs of random trials, at a higher level the numbers are still random, and follow no predictable pattern.

jjiimm_64

legendary

Activity: 1876

Merit: 1000

Quote from: amigaman on October 18, 2012, 01:37:45 PM

Quote from: Karasur on October 18, 2012, 09:07:50 AM

it seems that 2.8.4 uses gpu more aggressive...
Some of my radeon 7970 (1170/1050/65C) work well with 2.7.7 but with 2.8.4 are sick after 2-3 min btc mining.

I have the same probs, ckolivas told me it's from stratum use.
You will have to use non-stratum pools (or disable stratum with the option, see help) or reduce engine clock.

isn't aggressiveness tied to intensity? lower the intensity

amigaman

sr. member

Activity: 406

Merit: 250

Quote from: Karasur on October 18, 2012, 09:07:50 AM

it seems that 2.8.4 uses gpu more aggressive...
Some of my radeon 7970 (1170/1050/65C) work well with 2.7.7 but with 2.8.4 are sick after 2-3 min btc mining.

I have the same probs, ckolivas told me it's from stratum use.
You will have to use non-stratum pools (or disable stratum with the option, see help) or reduce engine clock.

Nite69

sr. member

Activity: 477

Merit: 500

Quote from: -ck on October 17, 2012, 03:44:21 PM

Quote from: Nite69 on October 17, 2012, 11:38:14 AM

Quote from: Vbs on October 16, 2012, 02:28:54 PM

Nope. Any time you change 1 bit on the input, you change all bits on the output hash (SHA-256). If it were like you said above, it would be a pretty crappy cryptographic function in the first place! Grin

Not all. If it would be so, changing 2 or any even number of bits would result in the same hash. Optimally one bit changes about half of the result hash.
Bt the real problem is that changing a bit affets on that what other bits affect. Anyway, I also think it would not work. Just a stupid idea.

No, a good hash function, and sha256 is that, will make the results have absolutely nothing in common even if only 1 bit is changed. This is called the avalanche effect and is a most desirable property.

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

That's what I practically told. But if one input bit allways changes *all* the output bits, then the function is not a hash, but it is binary "NOT". And that means they (inputs with 1bit difference) would practically have almost all in common with the results. So optimally (as your link also noted), about half of the output bits are changed when one input bit changes.
But this conversation is quite stupid. We all agree, but are still arguing. And quite much off-topic.

BlackPrapor

hero member

Activity: 628

Merit: 504

Quote from: Karasur on October 18, 2012, 09:07:50 AM

it seems that 2.8.4 uses gpu more aggressive...
Some of my radeon 7970 (1170/1050/65C) work well with 2.7.7 but with 2.8.4 are sick after 2-3 min btc mining.

dunno about getting sick, but my 6950 now able to run with I=12, whereas in 2.7.x and all previous versions anything above 9 was causing cgminer to use 100% of CPU. So, now I'm able to squeeze 400Mhs from 6950 with AMD APP 2.4 and catalyst 12.8 drivers, running at 965Mhz Grin

and almost 440Mhs from 6970.

Karasur

sr. member

Activity: 272

Merit: 250

it seems that 2.8.4 uses gpu more aggressive...
Some of my radeon 7970 (1170/1050/65C) work well with 2.7.7 but with 2.8.4 are sick after 2-3 min btc mining.

wind

member

Activity: 125

Merit: 10

again 2.8.4 doesn't compile on Windows with opencl enabled

Commenting this code in configure.ac solves the problem

Code:

if test "x$AMDAPPSDKROOT" != x; then
	OPENCL_FLAGS="-I$AMDAPPSDKROOT/include $OPENCL_FLAGS"
	OPENCL_LIBS="-L$AMDAPPSDKROOT/lib/$ARCH_DIR $OPENCL_LIBS"
fi

Krak

hero member

Activity: 591

Merit: 500

Quote from: BitMinerN8 on October 17, 2012, 11:13:34 PM

Very cool, now if you can get it to support a USB hub and 5 or 6 ASICs when they start shipping. Grin

Are you mining over 3G/4G or WiFi?

The phone is on wifi in the video. Wink

dietwice

newbie

Activity: 58

Merit: 0

2.8.4 on Win7 x64 still crashes on network failures.

Being on Wi-Fi it survived switching to another network just by switching to backup pool and then reconnecting to the main one (Slush).
But on switching network back it crashed.

Problem signature:
  Problem Event Name:   APPCRASH
  Application Name:   cgminer.exe
  Application Version:   0.0.0.0
  Application Timestamp:   507f309d
  Fault Module Name:   StackHash_0a9e
  Fault Module Version:   0.0.0.0
  Fault Module Timestamp:   00000000
  Exception Code:   c0000005
  Exception Offset:   6f43203a
  OS Version:   6.1.7601.2.1.0.256.1
  Locale ID:   1058
  Additional Information 1:   0a9e
  Additional Information 2:   0a9e372d3b4ad19135b953a78882e789
  Additional Information 3:   0a9e
  Additional Information 4:   0a9e372d3b4ad19135b953a78882e789

BitMinerN8

hero member

Activity: 626

Merit: 500

Mining since May 2011.

Quote from: kaerf on October 17, 2012, 08:20:50 PM

check it out. cgminer on android Grin

http://www.youtube.com/watch?v=c9hBRljvZPI

Very cool, now if you can get it to support a USB hub and 5 or 6 ASICs when they start shipping. Grin

Are you mining over 3G/4G or WiFi?

TheHarbinger

sr. member

Activity: 378

Merit: 250

Why is it so damn hot in here?

Con, I'll give the new build a whirl tomorrow since I was one of the ones that was having issues with dynamic difficulty. 2.7.6 is what I've been running, and has been rock solid for me, although it seems a few others still had problems. The only thing I did notice is that if I loose my network, the queued value (Q:) shoots through the roof and then drives the efficiency (E:) down through the floor. But it doesn't actually effect anything since those are just displayed values, and I know the queued work isn't actually increasing like a rocket ship because it can't get the work if the network is down. Roll Eyes

Mobius

hero member

Activity: 988

Merit: 1000

Quote from: kaerf on October 17, 2012, 08:20:50 PM

check it out. cgminer on android Grin

http://www.youtube.com/watch?v=c9hBRljvZPI

Coolness

Topic: OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.0 - page 463. (Read 5806015 times)