cbuchner1 / cudaminer Goto Github PK

but performance on windows xp is top notch

why is this? using a 260gtx

The help text indicates that the default for this option is 30 seconds, but the default in the code is 15.

I recently setup a system with Ubuntu 12.04, a 560ti and a 650ti.
Using CUDA 5.0 and 319.17.
Download from github, follow compile procedures.
If I launch using an F kernel, I get about 80kH/s on the 650 and 130 on the 650Ti, everything works.

If I launch using an F kernel on the 560 and a K on the 650, the cpu cannot validate the result for the 650. The 560 continues to work.

If I launch using only the 650 (filtering via the -d flag) with no specified kernel parameter, it crashes.

Thoughts?

Edit: Just tried doing this with CUDA 5.5 and i'm seeing the same behavior.

Using cudaminer from git and arch updated on 01/12/13, these errors appear when compiling:

http://pastebin.com/aRitBurE

With make install I get warning flood (literally): Cannot tell what pointer points to, assuming global memory space. This is for spinlock_kernel.cu(216) and test_kernel.cu(349) .
Should I be worried?

When I try to run cudaminer with specific configuration (doesnt matter for autotune as well) it draws horizontal color lines on top of terminal environment (tty1 for example) and gets insanely high hashrates but GPU result fails to validate on CPU.

I have NVIDIA 325.15 drivers with cuda 5.5 and gfx card is GT 610.

Hi there!

So, I'm having a great time with CudaMiner in Fedora 19 :) the issue is when I quit the aplication. I don't actually know how to terminate it - or if there is a proper way other then Ctr+c

But when I do that, my GPU becomes unaccessible:

~]$ nvidia-smi -r -i 0
Unable to determine the PCI bus id for the target device: GPU is lost

Seams like an out of bounds memory access? Have no clue...

deviceQuery Starting...

CUDA Device Query (Runtime API) version (CUDART static linking)

cudaGetDeviceCount returned 38
-> no CUDA-capable device is detected
Result = FAIL

So.. my only option to re-start cuda is by rebooting. However, my PC is remote and rebooting is unreliable unless I have someone in there to assist me manually. And I cannot recover cuda from command line either ( :/ )

Anyone has bumped into such behaviour? Thanks!

Using Mac OS 10.8.

make  all-recursive
Making all in compat
Making all in jansson
Makefile:254: .deps/dump.Po: No such file or directory
Makefile:255: .deps/hashtable.Po: No such file or directory
Makefile:256: .deps/load.Po: No such file or directory
Makefile:257: .deps/strbuffer.Po: No such file or directory
Makefile:258: .deps/utf.Po: No such file or directory
Makefile:259: .deps/value.Po: No such file or directory
make[3]: *** No rule to make target `.deps/value.Po'.  Stop.
make[2]: *** [all-recursive] Error 1
make[1]: *** [all-recursive] Error 1
make: *** [all] Error 2

I'm guessing I did something wrong and it's not actually hashing at these rates:

[ ~/opt/CudaMiner ] LD_LIBRARY_PATH=/usr/local/cuda/lib64/ ./cudaminer -H 1 -d0 -i 0 -l F56x16 -C 2 -o localhost:9332 -O user:pass
*** CudaMiner for nVidia GPUs by Christian Buchner ***
This is version 2013-11-20 (alpha)
based on pooler-cpuminer 2.3.2 (c) 2010 Jeff Garzik, 2012 pooler
Cuda additions Copyright 2013 Christian Buchner
My donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm

[2013-11-21 20:47:39] 1 miner threads started, using 'scrypt' algorithm.
[2013-11-21 20:47:39] GPU #0: GeForce GTS 450 with compute capability 2.1
[2013-11-21 20:47:39] GPU #0: interactive: 0, tex-cache: 2D, single-alloc: 1
[2013-11-21 20:47:39] GPU #0: using launch configuration F56x16
[2013-11-21 20:47:39] GPU #0: GeForce GTS 450, 28672 hashes, 152.63 khash/s
[2013-11-21 20:47:39] GPU #0: GeForce GTS 450, 774144 hashes, 1509 khash/s
[2013-11-21 20:47:44] GPU #0: GeForce GTS 450, 7569408 hashes, 1680 khash/s
[2013-11-21 20:47:49] GPU #0: GeForce GTS 450, 8429568 hashes, 1684 khash/s
[2013-11-21 20:47:54] GPU #0: GeForce GTS 450, 8429568 hashes, 1667 khash/s
[2013-11-21 20:47:59] GPU #0: GeForce GTS 450, 8343552 hashes, 1682 khash/s
[2013-11-21 20:48:04] GPU #0: GeForce GTS 450, 8429568 hashes, 1679 khash/s

Any thoughts?

I have a labtop (yes i know warning about heat) that has a Geforce 310M in it. When I run cudaminer without the -l setting it stops working. 6x6 causes the hashes not the be verified by the CPU but with 100+Kh/s. 1x1 is about 15Kh/s. What would cause this error? and do you know of a way to bump the settings up to max hashes? It has 1Gb of ram dedicated on the card.

I'm running a laptop with Quadro 3000M card (Fermi GF110GLM architecture, 450 MHz core speed and 240 shader processing units) . Installed the latest CudaMiner from 2013-12-10 (hadn't used it before) and at first was getting about 30 khash/s. I then kept tweaking and overclocking the GPU and at one point got up to 80 khash/s. I was just defaulting everything and letting autotune do its thing. Unfortunately I didn't take note of the launch configuration when I was getting this hashrate. For whatever reason I can't seem to get these rates anymore and I'm in the single digits. I ran in debug mode and it shows maximum warps = 484 (no idea what that means) and it chooses various different configurations (eg. F10x16) but always low hash rates. I tried doing an '-I F " but no dice. Even still I don't get any of the "yay!' acknowledgement messages as I was before so I'm not sure if it is even doing anything. I've put the GPU back to defaults and no difference.
What am I doing wrong?

When I leave it up to autodetection, cudaminer only finds the Tesla K40c's, leaving out the K6000. When I force it with -C 2,2,2 it crashes on startup.

Here's the output of cudaminer (2013-10-10):

[2013-10-19 13:27:01] Binding thread 0 to cpu 0
[2013-10-19 13:27:01] 3 miner threads started, using 'scrypt' algorithm.
[2013-10-19 13:27:01] Binding thread 1 to cpu 1
[2013-10-19 13:27:01] Binding thread 2 to cpu 2
[2013-10-19 13:27:01] Failed to get Stratum session id
[2013-10-19 13:27:01] Stratum difficulty set to 512
[2013-10-19 13:27:01] DEBUG: job_id='7e92' extranonce2=00000000 ntime=5262cefb
[2013-10-19 13:27:01] Stratum detected new block
[2013-10-19 13:27:02] GPU #2: x≥á♥▼─(v# with compute capability 19847128.19847128
[2013-10-19 13:27:02] GPU #2: interactive: 1, tex-cache: 2D, single-alloc: 1
[2013-10-19 13:27:02] GPU #0: Tesla K40c with compute capability 3.5
[2013-10-19 13:27:02] GPU #0: interactive: 0, tex-cache: 2D, single-alloc: 1
[2013-10-19 13:27:02] GPU #1: Tesla K40c with compute capability 3.5
[2013-10-19 13:27:02] GPU #1: interactive: 0, tex-cache: 2D, single-alloc: 1

Here's the output of NVidia-smi:

+------------------------------------------------------+
| NVIDIA-SMI 331.40     Driver Version: 331.40         |
|------------------------------+---------------------+---------------+
| GPU  Name           TCC/WDDM | Bus-Id       Disp.A | Vol Uncor.ECC |
| Fan  Temp  Perf Pwr:Usage/Cap|        Memory-Usage | GPU ComputeM. |
|==============================+=====================+===============|
|   0  Quadro K6000      WDDM  | 0000:01:00.0    Off |           Off |
| 28%   45C    P8   19W / 225W |   12232MB / 12287MB |  0%   Default |
+------------------------------+---------------------+---------------+
|   1  Tesla K40c         TCC  | 0000:02:00.0    Off |           Off |
| 33%   45C    P8   17W / 235W |      21MB / 12287MB |  0%   Default |
+------------------------------+---------------------+---------------+
|   2  Tesla K40c         TCC  | 0000:03:00.0    Off |           Off |
| 30%   37C    P8   17W / 235W |      21MB / 12287MB |  0%   Default |
+------------------------------+---------------------+---------------+

Any ideas what's going on here? Odd that cudaminer finds the Tesla's as GPU #0, and #1; when NVidia-smi reports them as GPU #1, and #2. It is because the K6000 is running WDDM? I have another rig with a K6000 running Linux, and it's working just fine. Maybe K6000 on Windows is a bit iffy? Is it perhaps a 4GB memory barrier issue under WDDM on Windows, even though the card has 12GB? The K6000 is the viz card, so I cannot set it in TCC more to test that theory, unfortunately.

Thanks for any help!

Could you please create a signed git tag?

So.. I've issued 'PATH=/usr/local/cuda-5.5/bin:$PATH make'

Everything seams cool, but then I get this output:

http://bitbin.it/XAMI3szy

For a quick resume:

test_kernel.cu(289): error: identifier "__funnelshift_l" is undefined

50 errors detected in the compilation of "/tmp/tmpxft_000018c8_00000000-6_test_kernel.cpp1.ii".
make[2]: *** [test_kernel.o] Error 2
make[2]: Leaving directory /home/<USER>/mining/CudaMiner' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory/home//mining/CudaMiner'
make: *** [all] Error 2

Anyone has an idea? Thanks

*edit: I'm on Fedora 19:

./deviceQuery
./deviceQuery Starting...

CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "GeForce GT 520"
CUDA Driver Version / Runtime Version 6.0 / 5.5
CUDA Capability Major/Minor version number: 2.1
Total amount of global memory: 1023 MBytes (1072889856 bytes)
( 1) Multiprocessors, ( 48) CUDA Cores/MP: 48 CUDA Cores
GPU Clock rate: 1620 MHz (1.62 GHz)
Memory Clock rate: 600 Mhz
Memory Bus Width: 64-bit
L2 Cache Size: 65536 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(65536), 2D=(65536, 65535), 3D=(2048, 2048, 2048)
Maximum Layered 1D Texture Size, (num) layers 1D=(16384), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(16384, 16384), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total number of registers available per block: 32768
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (65535, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 1 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device PCI Bus ID / PCI location ID: 2 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 6.0, CUDA Runtime Version = 5.5, NumDevs = 1, Device0 = GeForce GT 520
Result = PASS

I have 2 identical 560ti's in my system. No matter what options I pass telling it I have more than one card, it only creates 1 thread.
GPU-Z shows both cards in the system.
I tried running both in SLI and non SLI mode. no difference.

When trying to compile 2013-11-01 on Ubuntu 12.04. I get the following error and it will not proceed. configure: error: cannot run /bin/bash ./config.sub.

getting this error when compiling under linux.

In file included from salsa_kernel.cu:29:
miner.h:15: fatal error: curl/curl.h: No such file or directory
compilation terminated.
make[2]: *** [salsa_kernel.o] Error 1
make[2]: *** Waiting for unfinished jobs....
mv -f .deps/cudaminer-sha2.Tpo .deps/cudaminer-sha2.Po
spinlock_kernel.h(6): error: not a class or struct name

spinlock_kernel.h(21): error: identifier "cudaSharedMemConfig" is undefined

spinlock_kernel.h(21): error: identifier "cudaSharedMemBankSizeEightByte" is undefined

spinlock_kernel.cu(37): error: "KernelInterface" is not a nonstatic data member or base class of class "SpinlockKernel"

spinlock_kernel.cu(47): error: identifier "checkCudaErrors" is undefined

spinlock_kernel.cu(58): error: identifier "checkCudaErrors" is undefined

spinlock_kernel.cu(64): error: identifier "checkCudaErrors" is undefined

spinlock_kernel.cu(70): error: identifier "checkCudaErrors" is undefined

spinlock_kernel.cu(76): error: identifier "checkCudaErrors" is undefined

spinlock_kernel.cu(119): error: identifier "MyStreamSynchronize" is undefined

spinlock_kernel.cu(119): error: identifier "checkCudaErrors" is undefined

spinlock_kernel.cu(371): error: identifier "WU_PER_WARP" is undefined

spinlock_kernel.cu(378): error: identifier "WU_PER_BLOCK" is undefined

spinlock_kernel.cu(396): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(402): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(416): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(422): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(512): error: identifier "WU_PER_WARP" is undefined

spinlock_kernel.cu(519): error: identifier "WU_PER_BLOCK" is undefined

spinlock_kernel.cu(537): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(545): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(563): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(573): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(430): error: identifier "WU_PER_WARP" is undefined

spinlock_kernel.cu(437): error: identifier "WU_PER_BLOCK" is undefined

spinlock_kernel.cu(455): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(461): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(477): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(485): error: identifier "SCRATCH" is undefined

spinlock_kernel.cu(338): warning: function "operator^=(ulonglong2 &, const ulonglong2 &)" was declared but never referenced

29 errors detected in the compilation of "/tmp/tmpxft_000061de_00000000-4_spinlock_kernel.cpp1.ii".
make[2]: *** [spinlock_kernel.o] Error 2
mv -f .deps/cudaminer-scrypt.Tpo .deps/cudaminer-scrypt.Po

Hi! CudaMiner is great!

However, I am still getting a crash on exit (requiring hard reboot), once the GPUs are running.

Is there any other way to cleanly shut down the miner? Any settings which avoid crashes?

(Full disclosure: I built with CUDA 4.5... can that be the issue?)

Thanks!

Telsa M2090
RHEL 6.4 (2.6.32)
CUDA 4.5

For NVIDIA sdk 5.5 unstable ebuild the nvcc home is:
/opt/cuda/bin/

On CentOS I am getting errors on make:

Makefile:381: .deps/cudaminer-cpu-miner.Po: No such file or directory
Makefile:382: .deps/cudaminer-scrypt.Po: No such file or directory
Makefile:383: .deps/cudaminer-sha2.Po: No such file or directory
Makefile:384: .deps/cudaminer-util.Po: No such file or directory
make: *** No rule to make target `.deps/cudaminer-util.Po'. Stop.

These relate to following lines in the makefile:

include ./$(DEPDIR)/cudaminer-cpu-miner.Po
include ./$(DEPDIR)/cudaminer-scrypt.Po
include ./$(DEPDIR)/cudaminer-sha2.Po
include ./$(DEPDIR)/cudaminer-util.Po

When I comment those out make runs a bit further, resulting in a similar error message:

Makefile:254: .deps/dump.Po: No such file or directory
Makefile:255: .deps/hashtable.Po: No such file or directory
Makefile:256: .deps/load.Po: No such file or directory
Makefile:257: .deps/strbuffer.Po: No such file or directory
Makefile:258: .deps/utf.Po: No such file or directory
Makefile:259: .deps/value.Po: No such file or directory
make[3]: *** No rule to make target `.deps/value.Po'. Stop.

relating to

include ./$(DEPDIR)/dump.Po
include ./$(DEPDIR)/hashtable.Po
include ./$(DEPDIR)/load.Po
include ./$(DEPDIR)/strbuffer.Po
include ./$(DEPDIR)/utf.Po
include ./$(DEPDIR)/value.Po

Commenting those out leads to a chain of errors:

source='dump.c' object='dump.o' libtool=no
DEPDIR=.deps depmode=none /bin/sh ../../depcomp
gcc -std=gnu99 -DHAVE_CONFIG_H -I. -I../.. -O3 -c dump.c
: command not found 3:
: command not found 5:
: command not found 8:
: command not found 13:
: command not found 18:
: command not found 21:
: command not found 26:
: command not found 28:
'./../depcomp: line 29: syntax error near unexpected token in './../depcomp: line 29:case $1 in
make[3]: *** [dump.o] Error 2

And finally exiting.

What can I do to resolve this?

I got a stack smashing detected error:

*** stack smashing detected ***: cudaminer terminated

I am using ubuntu 12.10, Asus GTX 660TI, gcc4.6, cuda5

I've tried to build cudaminer on mac osx.
problems i solved:

• in scrypt.cpp

  #include <malloc.h>

  have to be

  #include <malloc/malloc.h>

  or simply deleted
• nvcc needs "-m 64" argument to build for 64bit

now i get segmentation fault :(

./cudaminer --url http://... --userpass ...:... --no-autotune
           *** CudaMiner for nVidia GPUs by Christian Buchner ***
                     This is version 2013-04-22 (alpha)
        based on pooler-cpuminer 2.2.3 (c) 2010 Jeff Garzik, 2012 pooler
               Cuda additions Copyright 2013 Christian Buchner
           My donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm
 
[2013-05-03 02:24:46] 1 miner threads started, using 'scrypt' algorithm.
[2013-05-03 02:24:47] Long-polling activated for http://notroll.in:6332/LP
[2013-05-03 02:24:47] GPU #0: GeForce 9400 with compute capability 1.1
[2013-05-03 02:24:47] GPU #0: interactive: 1, tex-cache: 0 , single-alloc: 0
[2013-05-03 02:24:50] GPU #0: using launch configuration  2x4
Segmentation fault: 11

ok i know that this is s nearly useless bugreport but i hope you can help :)

Upgraded today from a GTX560 to a GTX 780 TI. While the old card worked flawless, the new version seems to having issues with the card. Whenever running, just a handful of requests seems to be valid. Most requests are failing with the message "GeForce GTX 780 Ti result does not validate on CPU!".

This system is running on a Fedora 19 (64 bit system) with NVIDIA 331.20 and cuda 5.5.22. This issue occurs with the version from the 20th November. Unable to use a more recent version due to another issue reported.

It seems not to make a difference if setting up auto tuning or a profile and results in both cases in lots of fails. When running in debug mode, it looks like nearly all failed requests are having the same value.

Any ideas?

./cudaminer [...] -i 0 -H 1 -lT30x16 -D
*** CudaMiner for nVidia GPUs by Christian Buchner ***
This is version 2013-11-20 (alpha)
based on pooler-cpuminer 2.3.2 (c) 2010 Jeff Garzik, 2012 pooler
Cuda additions Copyright 2013 Christian Buchner
My donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm

[2013-12-05 22:16:40] 1 miner threads started, using 'scrypt' algorithm.
[...]
[2013-12-05 22:16:41] Failed to get Stratum session id
[2013-12-05 22:16:41] Stratum difficulty set to 16
[2013-12-05 22:16:41] DEBUG: job_id='7d1' extranonce2=00000000 ntime=52a0ed34
[2013-12-05 22:16:41] Stratum detected new block
[2013-12-05 22:16:42] GPU #0: GeForce GTX 780 Ti with compute capability 3.5
[2013-12-05 22:16:42] GPU #0: interactive: 0, tex-cache: 0 , single-alloc: 0
[2013-12-05 22:16:42] GPU #0: Given launch config 'T30x16' requires too much memory.
[2013-12-05 22:16:42] GPU #0: using launch configuration T30x16
[2013-12-05 22:16:42] GPU #0: GeForce GTX 780 Ti, 15360 hashes, 74.57 khash/s
[2013-12-05 22:16:42] DEBUG: hash <= target
Hash: 00000735d19f6cbda52f6bc8f9c71ea3583927ec408d8fff1226a5e2acc3ddf3
Target: 00000ffff0000000000000000000000000000000000000000000000000000000
[2013-12-05 22:16:42] GPU #0: GeForce GTX 780 Ti result does not validate on CPU!
[2013-12-05 22:16:43] DEBUG: hash <= target
Hash: 00000d3f00c065f2eeed8caea3d7518030f5e4593bdeff382f8a23dbee233f43
Target: 00000ffff0000000000000000000000000000000000000000000000000000000
[2013-12-05 22:16:43] GPU #0: GeForce GTX 780 Ti result does not validate on CPU!
[2013-12-05 22:16:43] DEBUG: hash <= target
Hash: 0000042049ea7495aca88147fef1924e1305f8d9fe40e6bf4fbac21312e4bafc
Target: 00000ffff0000000000000000000000000000000000000000000000000000000
[2013-12-05 22:16:43] GPU #0: GeForce GTX 780 Ti result does not validate on CPU!
[2013-12-05 22:16:45] DEBUG: hash <= target
Hash: 000003312bf1aba279d41c92dd969496448fc9a54ee4c87421ae64ef2d854fd2
Target: 00000ffff0000000000000000000000000000000000000000000000000000000
[2013-12-05 22:16:45] GPU #0: GeForce GTX 780 Ti result does not validate on CPU!

I have been digging for hours, but could not find anything. I have followed these instructions https://forum.litecoin.net/index.php?topic=3231.0

after the make command that's what I get:

nvcc -g -O2 -arch=compute_35 --maxrregcount=64 --ptxas-options=-v -I./compat/jansson -o titan_kernel.o -c titan_kernel.cu
g++  -g -O2 -pthread -L/usr/local/cuda/lib64  -o cudaminer cudaminer-cpu-miner.o cudaminer-util.o cudaminer-sha2.o cudaminer-scrypt.o salsa_kernel.o spinlock_kernel.o legacy_kernel.o fermi_kernel.o test_kernel.o titan_kernel.o -L/usr/lib/x86_64-linux-gnu -lcurl compat/jansson/libjansson.a -lpthread  -lcudart -fopenmp  
/usr/bin/ld: cannot find -lcudart
collect2: Fehler: ld gab 1 als Ende-Status zurück
make[2]: *** [cudaminer] Fehler 1
make[2]: Verlasse Verzeichnis '/home/rapsli/CudaMiner'
make[1]: *** [all-recursive] Fehler 1
make[1]: Verlasse Verzeichnis '/home/rapsli/CudaMiner'
make: *** [all] Fehler 2

From what I have read it doesn't find the lcudart lib? Any hints?

I'm running Ubuntu 13.04, 64 Bit

Get this during the make step:

titan_kernel.h(6): error: not a class or struct name

spinlock_kernel.h(6): error: not a class or struct name

fermi_kernel.h(6): error: not a class or struct name

legacy_kernel.h(6): error: not a class or struct name

salsa_kernel.cu(43): error: identifier "KernelInterface" is undefined

salsa_kernel.cu(85): error: identifier "KernelInterface" is undefined

salsa_kernel.cu(93): error: identifier "KernelInterface" is undefined

salsa_kernel.cu(97): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(120): error: identifier "KernelInterface" is undefined

salsa_kernel.cu(120): error: identifier "kernel" is undefined

salsa_kernel.cu(122): error: identifier "WU_PER_LAUNCH" is undefined

salsa_kernel.cu(126): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(160): error: identifier "WU_PER_LAUNCH" is undefined

salsa_kernel.cu(217): error: identifier "KernelInterface" is undefined

salsa_kernel.cu(268): error: identifier "SCRATCH" is undefined

salsa_kernel.cu(268): error: identifier "WU_PER_WARP" is undefined

salsa_kernel.cu(270): error: expression must have pointer-to-class type

salsa_kernel.cu(283): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(294): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(307): error: expression must have pointer-to-class type

salsa_kernel.cu(308): error: expression must have pointer-to-class type

salsa_kernel.cu(310): error: expression must have pointer-to-class type

salsa_kernel.cu(315): error: expression must have pointer-to-class type

salsa_kernel.cu(316): error: expression must have pointer-to-class type

salsa_kernel.cu(318): error: expression must have pointer-to-class type

salsa_kernel.cu(334): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(349): error: expression must have pointer-to-class type

salsa_kernel.cu(368): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(392): error: expression must have pointer-to-class type

salsa_kernel.cu(398): error: identifier "WU_PER_LAUNCH" is undefined

salsa_kernel.cu(398): error: identifier "WU_PER_BLOCK" is undefined

salsa_kernel.cu(410): error: expression must have pointer-to-class type

salsa_kernel.cu(438): error: expression must have pointer-to-class type

salsa_kernel.cu(446): error: expression must have pointer-to-class type

salsa_kernel.cu(449): error: expression must have pointer-to-class type

salsa_kernel.cu(451): error: expression must have pointer-to-class type

salsa_kernel.cu(467): error: expression must have pointer-to-class type

salsa_kernel.cu(560): error: expression must have pointer-to-class type

salsa_kernel.cu(568): error: expression must have pointer-to-class type

salsa_kernel.cu(570): error: expression must have pointer-to-class type

salsa_kernel.cu(571): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(581): error: expression must have pointer-to-class type

salsa_kernel.cu(582): error: expression must have pointer-to-class type

salsa_kernel.cu(584): error: expression must have pointer-to-class type

salsa_kernel.cu(589): error: expression must have pointer-to-class type

salsa_kernel.cu(590): error: expression must have pointer-to-class type

salsa_kernel.cu(592): error: expression must have pointer-to-class type

salsa_kernel.cu(596): error: expression must have pointer-to-class type

salsa_kernel.cu(605): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(629): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(642): error: identifier "WU_PER_LAUNCH" is undefined

salsa_kernel.cu(645): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(658): error: identifier "WU_PER_LAUNCH" is undefined

salsa_kernel.cu(658): error: identifier "WU_PER_BLOCK" is undefined

salsa_kernel.cu(664): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(675): error: expression must have pointer-to-class type

salsa_kernel.cu(678): error: identifier "checkCudaErrors" is undefined

salsa_kernel.cu(688): error: identifier "WU_PER_LAUNCH" is undefined

salsa_kernel.cu(691): error: identifier "checkCudaErrors" is undefined

When trying to compile the most recent version from the repository, nvcc is raising an exception and fails to compile spinlock_kernel.o.

make all-recursive
make[1]: Entering directory /home/phobeus/src/cudaminer' Making all in compat make[2]: Entering directory/home/phobeus/src/cudaminer/compat'
make[3]: Entering directory /home/phobeus/src/cudaminer/compat' make[3]: Nothing to be done forall-am'.
make[3]: Leaving directory /home/phobeus/src/cudaminer/compat' make[2]: Leaving directory/home/phobeus/src/cudaminer/compat'
make[2]: Entering directory /home/phobeus/src/cudaminer' nvcc -g -O2 -Xptxas "-abi=no -v" -arch=compute_30 --maxrregcount=63 -o spinlock_kernel.o -c spinlock_kernel.cu UNREACHABLE executed! Stack dump: 0. Running pass 'NVPTX DAG->DAG Pattern Instruction Selection' on function '@_Z28spinlock_scrypt_core_kernelBILi3EEvPjPi' nvcc error : 'cicc' died due to signal 6 nvcc error : 'cicc' core dumped make[2]: *** [spinlock_kernel.o] Error 134 make[2]: Leaving directory/home/phobeus/src/cudaminer'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/phobeus/src/cudaminer'

make: *** [all] Error 2

When using a older version with commit "88c6da6d5c2b798d1de7031e8dbcc2678f635e4b" it seems to be compiling fine.

In both cases with Kepler kernel goes faster than Fermi kernel (Last version)

In particular:

• GTX460 from 121 to 130 Khps
• GTX460M from 54 to 58 Khps

Moving to x64: Fermi performances stay the same, Kepler ones lose something (2 and 1 Khps respectively)

cudaminer.exe -d 0 -i 1 -l auto -C 0 -o stratum+tcp://global.wemineltc.com:3335 -O myname:pw

i try to start the miner with this .bat. Unfortunately the programm crashes at the detection of the right kernel launch configuration, considering F0x0 is the right one for my GPU.

My GPU is a GTX 570. Could you tell me the right launch configuration for my GPU.

Hi i have GeForce GT 430 and only have 30kh/s ?!

This error happens because the test_kernel.cu file is not compiled (there is no .o). Brute force compiling test_kernel then gives a .o, but it is not listed as part of the link list:
g++ -g -O2 -pthread -L/usr/local/cuda/lib64 -o cudaminer cudaminer-cpu-miner.o cudaminer-util.o cudaminer-sha2.o cudaminer-scrypt.o salsa_kernel.o spinlock_kernel.o legacy_kernel.o fermi_kernel.o titan_kernel.o -L/usr/lib/x86_64-linux-gnu -lcurl -Wl,-Bsymbolic-functions -Wl,-z,relro compat/jansson/libjansson.a -lpthread -lcudart -fopenmp
salsa_kernel.o: In function validate_config(char*, int&, int&, KernelInterface**, cudaDeviceProp*)': /home/drmike/Litecoin/CudaMiner-master/salsa_kernel.cu:67: undefined reference toTestKernel::TestKernel()'

Brute force adding test_kernel.o in the list allows a full compile/link and run. Is there an easy way to fix the Makefile?

Please, add some basic build instructions for dependencies and guidelines.

Can intesity settings be introduced in CudaMiner, just like cgminer works?

users-mac:cudaminer-master user$ ./configure
checking build system type... i386-apple-darwin12.5.0
checking host system type... i386-apple-darwin12.5.0
checking target system type... i386-apple-darwin12.5.0
checking for a BSD-compatible install... /usr/local/bin/ginstall -c
checking whether build environment is sane... yes
checking for a thread-safe mkdir -p... /usr/local/bin/gmkdir -p
checking for gawk... no
checking for mawk... no
checking for nawk... no
checking for awk... awk
checking whether make sets $(MAKE)... yes
checking whether to enable maintainer-specific portions of Makefiles... no
checking for style of include used by make... GNU
checking for gcc... gcc
checking whether the C compiler works... yes
checking for C compiler default output file name... a.out
checking for suffix of executables...
checking whether we are cross compiling... no
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether gcc accepts -g... yes
checking for gcc option to accept ISO C89... none needed
checking dependency style of gcc... gcc3
checking for gcc option to accept ISO C99... -std=gnu99
checking how to run the C preprocessor... gcc -std=gnu99 -E
checking for grep that handles long lines and -e... /usr/bin/grep
checking for egrep... /usr/bin/grep -E
checking whether gcc -std=gnu99 needs -traditional... no
checking whether gcc -std=gnu99 and cc understand -c and -o together... yes
checking dependency style of gcc -std=gnu99... gcc3
checking for ranlib... ranlib
checking for g++... g++
checking whether we are using the GNU C++ compiler... yes
checking whether g++ accepts -g... yes
checking dependency style of g++... gcc3
checking for gcc -std=gnu99 option to support OpenMP... -fopenmp
checking for ANSI C header files... yes
checking for sys/types.h... yes
checking for sys/stat.h... yes
checking for stdlib.h... yes
checking for string.h... yes
checking for memory.h... yes
checking for strings.h... yes
checking for inttypes.h... yes
checking for stdint.h... yes
checking for unistd.h... yes
checking sys/endian.h usability... no
checking sys/endian.h presence... no
checking for sys/endian.h... no
checking sys/param.h usability... yes
checking sys/param.h presence... yes
checking for sys/param.h... yes
checking syslog.h usability... yes
checking syslog.h presence... yes
checking for syslog.h... yes
checking for sys/sysctl.h... yes
checking whether be32dec is declared... no
checking whether le32dec is declared... no
checking whether be32enc is declared... no
checking whether le32enc is declared... no
checking for size_t... yes
checking for working alloca.h... yes
checking for alloca... yes
checking for getopt_long... yes
checking whether we can compile AVX code... no
configure: WARNING: The assembler does not support the AVX instruction set.
checking for json_loads in -ljansson... yes
checking for pthread_create in -lpthread... yes
checking for gawk... (cached) awk
checking for curl-config... /usr/bin/curl-config
checking for the version of libcurl... 7.24.0
checking for libcurl >= version 7.15.2... yes
checking whether libcurl is usable... yes
checking for curl_free... yes
configure: creating ./config.status
config.status: creating Makefile
config.status: creating compat/Makefile
config.status: creating compat/jansson/Makefile
config.status: creating cpuminer-config.h
config.status: cpuminer-config.h is unchanged
config.status: executing depfiles commands
users-mac:cudaminer-master user$ make
make all-recursive
Making all in compat
make[3]: Nothing to be done for all-am'. gcc -std=gnu99 -DHAVE_CONFIG_H -I. -msse2 -fopenmp -pthread -fno-strict-aliasing -g -O2 -MT cudaminer-cpu-miner.o -MD -MP -MF .deps/cudaminer-cpu-miner.Tpo -c -o cudaminer-cpu-miner.otest -f 'cpu-miner.c' || echo './'cpu-miner.c mv -f .deps/cudaminer-cpu-miner.Tpo .deps/cudaminer-cpu-miner.Po gcc -std=gnu99 -DHAVE_CONFIG_H -I. -msse2 -fopenmp -pthread -fno-strict-aliasing -g -O2 -MT cudaminer-util.o -MD -MP -MF .deps/cudaminer-util.Tpo -c -o cudaminer-util.otest -f 'util.c' || echo './'util.c mv -f .deps/cudaminer-util.Tpo .deps/cudaminer-util.Po gcc -std=gnu99 -DHAVE_CONFIG_H -I. -msse2 -fopenmp -pthread -fno-strict-aliasing -g -O2 -MT cudaminer-sha2.o -MD -MP -MF .deps/cudaminer-sha2.Tpo -c -o cudaminer-sha2.otest -f 'sha2.c' || echo './'sha2.c mv -f .deps/cudaminer-sha2.Tpo .deps/cudaminer-sha2.Po g++ -DHAVE_CONFIG_H -I. -msse2 -fopenmp -pthread -fno-strict-aliasing -g -O2 -MT cudaminer-scrypt.o -MD -MP -MF .deps/cudaminer-scrypt.Tpo -c -o cudaminer-scrypt.otest -f 'scrypt.cpp' || echo './'`scrypt.cpp
mv -f .deps/cudaminer-scrypt.Tpo .deps/cudaminer-scrypt.Po
nvcc -g -O2 -arch=compute_10 --maxrregcount=64 --ptxas-options=-v -o salsa_kernel.o -c salsa_kernel.cu
/usr/local/include/c++/4.8.1/cstdlib(178): error: identifier "__int128" is undefined

/usr/local/include/c++/4.8.1/cstdlib(179): error: identifier "__int128" is undefined

2 errors detected in the compilation of "/tmp/tmpxft_00000c9c_00000000-6_salsa_kernel.cpp1.ii".
make[2]: *** [salsa_kernel.o] Error 2
make[1]: *** [all-recursive] Error 1
make: *** [all] Error 2

Error 39 error : C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v5.0\bin/../open64/lib//be.exe returned non-zero status 1 D:\CUDA\LTC\CudaMiner-master\CUDACOMPILE

In file included from scrypt.cpp:46:
/usr/include/malloc.h:3:2: error: #error "<malloc.h> has been replaced by <stdlib.h>"
scrypt.cpp: In member function 'const uint32x4_t uint32x4_t::operator>>(int) const':
scrypt.cpp:78: error: shift must be an immediate
scrypt.cpp: In member function 'const uint32x4_t uint32x4_t::operator<<(int) const':
scrypt.cpp:77: error: shift must be an immediate
scrypt.cpp:772: confused by earlier errors, bailing out
*** [cudaminer-scrypt.o] Error code 1

I have the following compile error with the latest pull, I was able to compile the previous releases from git without issues
Ubuntu 12.04.3 LTS x86_64
root@server:# dpkg --list | grep -i cuda
ii cuda 5.5-22 CUDA meta-package
ii cuda-5-5 5.5-22 CUDA 5.5 meta-package
ii cuda-command-line-tools-5-5 5.5-22 CUDA command-line tools
ii cuda-core-5-5 5.5-22 CUDA core tools
ii cuda-core-libs-5-5 5.5-22 CUDA native runtime library
ii cuda-documentation-5-5 5.5-22 CUDA documentation
ii cuda-extra-libs-5-5 5.5-22 CUDA native libraries
ii cuda-headers-5-5 5.5-22 CUDA native headers
ii cuda-license-5-5 5.5-22 CUDA licenses
ii cuda-misc-5-5 5.5-22 CUDA miscellaneous tools
ii cuda-repo-ubuntu1204 5.5-0 CUDA repo configuration files.
ii cuda-samples-5-5 5.5-22 CUDA example applications
ii cuda-visual-tools-5-5 5.5-22 CUDA visual tools
root@server:#

nvcc -g -O2 -arch=compute_10 --maxrregcount=64 --ptxas-options=-v -o test_kernel.o -c test_kernel.cu
test_kernel.cu(90): error: no suitable constructor exists to convert from "int" to "uint4"

test_kernel.cu(90): error: no suitable conversion function from "uint4" to "unsigned int" exists

test_kernel.cu(91): error: no suitable constructor exists to convert from "int" to "uint4"

test_kernel.cu(91): error: no suitable conversion function from "uint4" to "unsigned int" exists

test_kernel.cu(92): error: no suitable constructor exists to convert from "int" to "uint4"

test_kernel.cu(92): error: no suitable conversion function from "uint4" to "unsigned int" exists

test_kernel.cu(93): error: no suitable constructor exists to convert from "int" to "uint4"

test_kernel.cu(93): error: no suitable conversion function from "uint4" to "unsigned int" exists

test_kernel.cu(155): error: identifier "__ldg" is undefined

test_kernel.cu(191): error: no suitable constructor exists to convert from "int" to "uint4"

test_kernel.cu(191): error: no operator "*" matches these operands
operand types are: int * uint4

test_kernel.cu(191): error: identifier "__ldg" is undefined

test_kernel.cu(192): error: no suitable constructor exists to convert from "int" to "uint4"

test_kernel.cu(192): error: no operator "*" matches these operands
operand types are: int * uint4

test_kernel.cu(193): error: no suitable constructor exists to convert from "int" to "uint4"

test_kernel.cu(193): error: no operator "*" matches these operands
operand types are: int * uint4

test_kernel.cu(194): error: no suitable constructor exists to convert from "int" to "uint4"

test_kernel.cu(194): error: no operator "*" matches these operands
operand types are: int * uint4

test_kernel.cu(244): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(245): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(246): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(247): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(250): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(251): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(252): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(253): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(256): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(257): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(258): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(259): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(262): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(263): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(264): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(265): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(268): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(269): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(270): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(271): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(274): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(275): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(276): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(277): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(280): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(281): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(282): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(283): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(286): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(287): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(288): error: identifier "__funnelshift_l" is undefined

test_kernel.cu(289): error: identifier "__funnelshift_l" is undefined

50 errors detected in the compilation of "/tmp/tmpxft_000025c3_00000000-6_test_kernel.cpp1.ii".
make[2]: *** [test_kernel.o] Error 2
make[2]: Leaving directory /root/CudaMiner' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory/root/CudaMiner'
make: *** [all] Error 2

The original path defined in the Makefile for nvcc, after running configure.sh /usr/local/bin/Cuda/nvcc is not where the Ubuntu nvidia-cuda-toolkit installs it. I renamed the pathname to /usr/bin/nvcc and compilation succeeded.

Hello,

Maybe I am not understanding correctly but with 2D cache in interactive mode I am getting 265KH/s using K192x3 @ 70% Use. With T192x1 I am getting only 200KH/s, with T192x2 245KH/s, T192x3 then drops down to

Asus OCII GTX780 OCed to 1306/6610

auto tuning results in T99x1 at 105KH/s

Cheers,
Al

A couple of days ago I had a problem with cudaminer. After this line it stopped working completely:

"stratum_recv_line failed to parse a newline-terminated string"

Looks like some sort of protocol error of the pool communication to me. Anyway, it would be nice if on this sort of error cudaminer automatically reconnects to the pool instead of ceasing to mine.

archlinux,3.11.6kernel,gtx260,325.15driver,cuda 5.5.22.

[2013-11-29 08:27:16] 4 miner threads started, using 'scrypt' algorithm.
[2013-11-29 08:27:16] Binding thread 0 to cpu 0
[2013-11-29 08:27:16] Binding thread 3 to cpu 3
[2013-11-29 08:27:16] Binding thread 2 to cpu 2
[2013-11-29 08:27:16] Binding thread 1 to cpu 1
[2013-11-29 08:27:17] Starting Stratum on stratum+tcp://127.0.0.1:3333
[2013-11-29 08:27:17] Stratum session id: ae6812eb4cd7735a302a8a9dd95cf71f
[2013-11-29 08:27:17] Stratum difficulty set to 256
[2013-11-29 08:27:17] DEBUG: job_id='35' extranonce2=000000 ntime=5297df74
[2013-11-29 08:27:17] Stratum detected new block
[2013-11-29 08:28:15] DEBUG: job_id='36' extranonce2=000000 ntime=5297dfb0
[2013-11-29 08:28:20] GPU #0: GeForce GTX 260 with compute capability 1.3
[2013-11-29 08:28:20] GPU #0: interactive: 1, tex-cache: 0 , single-alloc: 0
[2013-11-29 08:28:20] GPU #2: with compute capability -1261933287.32553
segment fault

On Arch Linux with the latest nvidia drivers (319.17-2, https://www.archlinux.org/packages/extra/x86_64/nvidia/) and nvcc 5.0, VO2.1221,
after running the configuration script, a makefile is produced for my NVIDIA GTS 250 that specifies "-arch=compute_10".

Compiling with this option causes an arithmetic error when auto-tune tries to run. Auto-tune output shows all 0s for x1, then crashes with either a floating point exception or an arithmetic error; the output looks like this:

[2013-05-17 19:07:24] GPU #0: Performing auto-tuning (Patience...)
[2013-05-17 19:07:24]        x1    x2    x3    x4    x5    x6    x7    x8
[2013-05-17 19:07:24]   1:  0.0                                           kh/s
... (many similar lines)...
[2013-05-17 19:07:24] GPU #0:    0.00 khash/s with configuration  0x0
[2013-05-17 19:07:24] GPU #0: using launch configuration  0x0
Floating point exception (core dumped)

Using gdb, I found that on line 1156 of salsa_kernel.cu,

dim3  grid(WU_PER_LAUNCH/WU_PER_BLOCK, 1, 1);

WU_PER_BLOCK was evaluating to zero, resulting in a division by zero.

I went through the code, trying to understand how these things were being computed, and ended up seeing that, at line 842, immediately after the line

checkCudaErrors(cudaMemcpy(d_idata, h_idata, mem_size, cudaMemcpyHostToDevice)),

checking the return value of "cudaPeekAtLastError" in gdb gave me error 13, or "cudaErrorInvalidSymbol." I continued debugging, and saw that after the call of run_normal_kernel on line 878 of salsa_kernel.cu, peeking at the last cuda error returned 8. Looking at http://www.cs.cmu.edu/afs/cs/academic/class/15668-s11/www/cuda-doc/html/group__CUDART__TYPES_g3f51e3575c2178246db0a94a430e0038.html, I saw that 8 was cudaErrorInvalidDeviceFunction; checking in gdb confirmed this.

I googled this error, and found this udacity link: http://forums.udacity.com/questions/100032430/cuda-error-invalid-device-function-cudagetlasterror, where the commenter reports removing the '-arch' argument from the nvcc call solved his problem.

Checking the makefile, I saw nvcc was being called with the option -arch=compute_10; I removed this argument, compiled, and autotune finished successfully. The pool I was hooked up to confirmed the hashes I was sending were correct.

I've created a pull request that removes the arch specification. After this is pulled, I'll likely make this a package on arch linux's user repository.

Thanks for your work!

Hello,
Thanks for the very nice program! I run Archlinux fully updated and i am not sure if it's a cuda 5 thing or something else, but i can only build CudaMiner once i add the following flag to the actual compilation lines at Makefile.am:
-ccbin=/usr/bin/g++
Here is the diff:
diff --git a/Makefile.am b/Makefile.am

index 6265742..0dcdce3 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -32,10 +32,10 @@ cudaminer_LDADD             = @LIBCURL@ @JANSSON_LIBS@ @PTHREAD_LIBS@ @WS2_LIBS@ @CUDA_LIBS
 cudaminer_CPPFLAGS     = -msse2 @LIBCURL_CPPFLAGS@ @OPENMP_CFLAGS@

 .cu.o:
-       $(NVCC) @CFLAGS@ -arch=compute_10 --maxrregcount=124 --ptxas-options=-v $(JANSSON_INCLUDES) -o $@ -c $<
+       $(NVCC) @CFLAGS@ -ccbin=/usr/bin/g++ -arch=compute_10 --maxrregcount=124 --ptxas-options=-v $(JANSSON_INCLUDES) -o $@ -c $<

 spinlock_kernel.o: spinlock_kernel.cu
-       $(NVCC) @CFLAGS@ -arch=compute_11 --maxrregcount=124 --ptxas-options=-v $(JANSSON_INCLUDES) -o $@ -c $<
+       $(NVCC) @CFLAGS@ -ccbin=/usr/bin/g++ -arch=compute_11 --maxrregcount=124 --ptxas-options=-v $(JANSSON_INCLUDES) -o $@ -c $<

 titan_kernel.o: titan_kernel.cu
-       $(NVCC) @CFLAGS@ -arch=compute_35 --maxrregcount=64 --ptxas-options=-v $(JANSSON_INCLUDES) -o $@ -c $<
+       $(NVCC) @CFLAGS@ -ccbin=/usr/bin/g++ -arch=compute_35 --maxrregcount=64 --ptxas-options=-v $(JANSSON_INCLUDES) -o $@ -c $<

On the gentoo system I just installed nvidia sdk 5.5, the system is 64bit and after install I have both 32bit and 64bit versions of libraries:
/opt/cuda/lib/libcudart.so
/opt/cuda/lib/libcudart.so.5.5
/opt/cuda/lib/libcudart.so.5.5.22
/opt/cuda/lib/libcudart_static.a
/opt/cuda/lib64/libcudart.so
/opt/cuda/lib64/libcudart.so.5.5
/opt/cuda/lib64/libcudart.so.5.5.22
/opt/cuda/lib64/libcudart_static.a

I am really not sure in this multilib world if it is necessary to have installed following but I did it:

Emerging (1 of 2) app-emulation/emul-linux-x86-opengl-20131008
Installing (1 of 2) app-emulation/emul-linux-x86-opengl-20131008
Emerging (2 of 2) app-emulation/emul-linux-x86-gtklibs-20131008
Installing (2 of 2) app-emulation/emul-linux-x86-gtklibs-20131008

When trying to make cudaminer I receive( with "strace make" command):
fcntl(3, F_SETFD, FD_CLOEXEC) = 0
rt_sigprocmask(SIG_BLOCK, [HUP INT QUIT TERM XCPU XFSZ], NULL, 8) = 0
vfork() = 9221
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
wait4(-1, [{WIFEXITED(s) && WEXITSTATUS(s) == 0}], 0, NULL) = 9221
--- SIGCHLD {si_signo=SIGCHLD, si_code=CLD_EXITED, si_pid=9221, si_status=0, si_utime=0, si_stime=0} ---
rt_sigreturn() = 9221
rt_sigprocmask(SIG_BLOCK, [HUP INT QUIT TERM XCPU XFSZ], NULL, 8) = 0
vfork() = 9222
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
wait4(-1, [{WIFEXITED(s) && WEXITSTATUS(s) == 0}], 0, NULL) = 9222
--- SIGCHLD {si_signo=SIGCHLD, si_code=CLD_EXITED, si_pid=9222, si_status=0, si_utime=0, si_stime=0} ---
rt_sigreturn() = 9222
rt_sigprocmask(SIG_BLOCK, [HUP INT QUIT TERM XCPU XFSZ], NULL, 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
stat("cpuminer-config.h", {st_mode=S_IFREG|0644, st_size=5469, ...}) = 0
fstat(1, {st_mode=S_IFCHR|0620, st_rdev=makedev(136, 0), ...}) = 0
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7ffbe5a56000
write(1, "make all-recursive\n", 20make all-recursive
) = 20
rt_sigprocmask(SIG_BLOCK, [HUP INT QUIT TERM XCPU XFSZ], NULL, 8) = 0
vfork() = 9223
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
wait4(-1, make[1]: Entering directory /root/cudaminer' Making all in compat make[2]: Entering directory/root/cudaminer/compat'
make[3]: Entering directory /root/cudaminer/compat' make[3]: Nothing to be done forall-am'.
make[3]: Leaving directory /root/cudaminer/compat' make[2]: Leaving directory/root/cudaminer/compat'
make[2]: Entering directory /root/cudaminer' g++ -O3 -pthread -L/usr/local/cuda/lib64 -o cudaminer cudaminer-cpu-miner.o cudaminer-util.o cudaminer-sha2.o cudaminer-scrypt.o salsa_kernel.o spinlock_kernel.o legacy_kernel.o fermi_kernel.o test_kernel.o titan_kernel.o -lcurl -lssl -lcrypto -lssl -lcrypto -lz -lrt -ljansson -lpthread -lcudart -fopenmp /usr/lib/gcc/x86_64-pc-linux-gnu/4.6.3/../../../../x86_64-pc-linux-gnu/bin/ld: cannot find -lcudart collect2: ld returned 1 exit status make[2]: *** [cudaminer] Error 1 make[2]: Leaving directory/root/cudaminer'
make[1]: * [all-recursive] Error 1
make[1]: Leaving directory `/root/cudaminer'
[{WIFEXITED(s) && WEXITSTATUS(s) == 2}], 0, NULL) = 9223
--- SIGCHLD {si_signo=SIGCHLD, si_code=CLD_EXITED, si_pid=9223, si_status=2, si_utime=0, si_stime=0} ---
rt_sigreturn() = 9223
write(2, "make: ", 6make: ) = 6
write(2, "* [all] Error 2", 17*** [all] Error 2) = 17
write(2, "\n", 1
) = 1
rt_sigprocmask(SIG_BLOCK, [HUP INT QUIT TERM XCPU XFSZ], NULL, 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
chdir("/root/cudaminer") = 0
close(1) = 0
munmap(0x7ffbe5a56000, 4096) = 0
exit_group(2) = ?
+++ exited with 2 +++

Any hint?

I'm using cudaminer on Linux Mint 14 Nadia, trying to connect to burnside's Litecoin pool. With these settings:

./cudaminer -o stratum+tcp://ltc-stratum.kattare.com:3333 -O [uname]:[pword]

I get the following output on the first run:

Starting Stratum on stratum+tcp://ltc-stratum.kattare.com:3333
thread 23719 create failed

And this output on the second run:

Starting Stratum on stratum+tcp://ltc-stratum.kattare.com:3333
Stratum authentication failed
...retry after 15 seconds
thread 23724 create failed

The latest version of cudaminer says it has stratum support, so I don't need to use a proxy, right?

I'm running cudaminer on an opensuse 12.2 with a GTX TITAN with 6GB onboard and 16GB Mainboard Memory. the nvidia driver version is 331.20.
I get the "result does not validate on CPU" error permanently.
Is there a way or a feature to disable this verification? And what can be the reasons that this test fails?

Hi,

I'm using the 1th of december release of cudaminer, with CUDA 5.5 on an opensuse system with 1 GTX 460. Autoconfig selects F56x2 or F14x8 randomly. When I run cudaminer, I switch to a real terminal and drop the GUI to save GPU memory for mining. I set the device to interactive.

However, three days ago I noticed weird green and yellow pixels on my screen when I start cudaminer. I assumed this to be caused by the screen buffer overlapping with the memory used for computation. This was not a problem, but I noticed that after a while my results do not longer validate. On my screen this occurs when console text overwrites these green areas. I redirected all output of the command into a file to avoid these issues but it only works for a while because sometimes the screen buffers do get touched and cause failing to occur. Having the computation data in the screen buffers doesn't seem to be a good idea anyway.

How should I resolve these issues? I'm not 100% convinced this is a bug, rather a combination of circumstances. I could try to use CUDA 5.0 perhaps? or run in it from a graphical session so perhaps cudaminer avoids these memory regions? (although it may cost some khash, right now faillure means some hours of mining downtime which is worse).

00:02.0 VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controller (rev 02) (prog-if 00 [VGA controller])

./cudaminer -u xxx-p xxxx -o stratum+tcp://mining.eu.hypernova.pw:3333 -d 0 -i 0 -l auto
*** CudaMiner for nVidia GPUs by Christian Buchner ***
This is version 2013-07-13 (alpha)
based on pooler-cpuminer 2.3.2 (c) 2010 Jeff Garzik, 2012 pooler
Cuda additions Copyright 2013 Christian Buchner
My donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm

[2013-07-17 16:59:47] Starting Stratum on stratum+tcp://mining.eu.hypernova.pw:3333
[2013-07-17 16:59:47] 1 miner threads started, using 'scrypt' algorithm.
[2013-07-17 16:59:47] Stratum detected new block
[2013-07-17 16:59:48] GPU #0: starting up...

[2013-07-17 16:59:48] GPU #0: with compute capability 402655552.32735
[2013-07-17 16:59:48] GPU #0: interactive: 0, tex-cache: 0 , single-alloc: 0
[2013-07-17 16:59:48] GPU #0: Performing auto-tuning (Patience...)
[2013-07-17 16:59:48] GPU #0: 0.00 khash/s with configuration 0x0
[2013-07-17 16:59:48] GPU #0: using launch configuration 0x0
Floating point exception (core dumped)

I have been playing for the last couple of hours trying to sort the probelm out and to be honest I am not really sure what the problem actually is.

I am using an Amazon EC2 CG1.4Xlarge with Amazon Linux AMI (HVM) 2013.09.1 - ami-d1bfe4b8 onboard which is loosely based on Fedora/ Red Hat I believe and before the below steps I performed the following to ready the unit. I previously tried Cuda 5.5 which was working well but given some research today I reset the unit and installed 5.0 from a fresh image with no luck.

I have put a dump of the directory contents at the bottom as well.

sudo yum -y groupinstall "Development Tools"
sudo yum -y install git libcurl-devel python-devel screen rsync yasm numpy openssl-devel
sudo yum -y update kernel* selinux-policy*

sudo reboot

wget http://developer.download.nvidia.com/compute/cuda/5_0/rel-update-1/installers/cuda_5.0.35_linux_64_fedora16-1.run
sudo sh cuda_5.0.35_linux_64_fedora16-1.run

git clone https://github.com/cbuchner1/CudaMiner

[ec2-user@ip-172-31-28-158 ~]$ cd CudaMiner
[ec2-user@ip-172-31-28-158 CudaMiner]$ git reset -–hard 88c6da6
fatal: ambiguous argument '–hard': unknown revision or path not in the working tree.
Use '--' to separate paths from revisions, like this:
'git [...] -- [...]'
[ec2-user@ip-172-31-28-158 CudaMiner]$ ./configure
checking build system type... x86_64-unknown-linux-gnu
checking host system type... x86_64-unknown-linux-gnu
checking target system type... x86_64-unknown-linux-gnu
checking for a BSD-compatible install... /usr/bin/install -c
checking whether build environment is sane... yes
checking for a thread-safe mkdir -p... /bin/mkdir -p
checking for gawk... gawk
checking whether make sets $(MAKE)... yes
checking whether to enable maintainer-specific portions of Makefiles... no
checking for style of include used by make... GNU
checking for gcc... gcc
checking whether the C compiler works... yes
checking for C compiler default output file name... a.out
checking for suffix of executables...
checking whether we are cross compiling... no
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether gcc accepts -g... yes
checking for gcc option to accept ISO C89... none needed
checking dependency style of gcc... gcc3
checking for gcc option to accept ISO C99... -std=gnu99
checking how to run the C preprocessor... gcc -std=gnu99 -E
checking for grep that handles long lines and -e... /bin/grep
checking for egrep... /bin/grep -E
checking whether gcc -std=gnu99 needs -traditional... no
checking whether gcc -std=gnu99 and cc understand -c and -o together... yes
checking dependency style of gcc -std=gnu99... gcc3
checking for ranlib... ranlib
checking for g++... g++
checking whether we are using the GNU C++ compiler... yes
checking whether g++ accepts -g... yes
checking dependency style of g++... gcc3
checking for gcc -std=gnu99 option to support OpenMP... -fopenmp
checking for ANSI C header files... yes
checking for sys/types.h... yes
checking for sys/stat.h... yes
checking for stdlib.h... yes
checking for string.h... yes
checking for memory.h... yes
checking for strings.h... yes
checking for inttypes.h... yes
checking for stdint.h... yes
checking for unistd.h... yes
checking sys/endian.h usability... no
checking sys/endian.h presence... no
checking for sys/endian.h... no
checking sys/param.h usability... yes
checking sys/param.h presence... yes
checking for sys/param.h... yes
checking syslog.h usability... yes
checking syslog.h presence... yes
checking for syslog.h... yes
checking for sys/sysctl.h... yes
checking whether be32dec is declared... no
checking whether le32dec is declared... no
checking whether be32enc is declared... no
checking whether le32enc is declared... no
checking for size_t... yes
checking for working alloca.h... yes
checking for alloca... yes
checking for getopt_long... yes
checking whether we can compile AVX code... yes
checking whether we can compile XOP code... yes
checking whether we can compile AVX2 code... yes
checking for json_loads in -ljansson... no
checking for pthread_create in -lpthread... yes
checking for gawk... (cached) gawk
checking for curl-config... /usr/bin/curl-config
checking for the version of libcurl... 7.33.0
checking for libcurl >= version 7.15.2... yes
checking whether libcurl is usable... yes
checking for curl_free... yes
configure: creating ./config.status
config.status: creating Makefile
config.status: creating compat/Makefile
config.status: creating compat/jansson/Makefile
config.status: creating cpuminer-config.h
config.status: cpuminer-config.h is unchanged
config.status: executing depfiles commands
[ec2-user@ip-172-31-28-158 CudaMiner]$ PATH=/usr/local/cuda-5.0/bin:$PATH make
make all-recursive
make[1]: Entering directory /home/ec2-user/CudaMiner' Making all in compat make[2]: Entering directory/home/ec2-user/CudaMiner/compat'
Making all in jansson
make[3]: Entering directory /home/ec2-user/CudaMiner/compat/jansson' make[3]: Nothing to be done forall'.
make[3]: Leaving directory /home/ec2-user/CudaMiner/compat/jansson' make[3]: Entering directory/home/ec2-user/CudaMiner/compat'
make[3]: Nothing to be done for all-am'. make[3]: Leaving directory/home/ec2-user/CudaMiner/compat'
make[2]: Leaving directory /home/ec2-user/CudaMiner/compat' make[2]: Entering directory/home/ec2-user/CudaMiner'
g++ -g -O2 -pthread -L/usr/local/cuda/lib64 -o cudaminer cudaminer-cpu-miner.o cudaminer-util.o cudaminer-sha2.o cudaminer-scrypt.o salsa_kernel.o spinlock_kernel.o legacy_kernel.o fermi_kernel.o test_kernel.o titan_kernel.o -lcurl compat/jansson/libjansson.a -lpthread -lcudart -fopenmp
/usr/local/cuda/lib64/libcudart.so: undefined reference to memcpy@GLIBC_2.14' collect2: ld returned 1 exit status make[2]: *** [cudaminer] Error 1 make[2]: Leaving directory/home/ec2-user/CudaMiner'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/ec2-user/CudaMiner'
make: *** [all] Error 2
[ec2-user@ip-172-31-28-158 CudaMiner]$

[ec2-user@ip-172-31-28-158 CudaMiner]$ ls
= cpu-miner.c INSTALL salsa_kernel.h
aclocal.m4 cpuminer-config.h install-sh salsa_kernel.o
AUTHORS cpuminer-config.h.in legacy_kernel.cu scrypt.cpp
autogen.sh cudaminer-cpu-miner.o legacy_kernel.h sha2.c
ChangeLog cudaminer-scrypt.o legacy_kernel.o spinlock_kernel.cu
compat cudaminer-sha2.o LICENSE spinlock_kernel.h
compat.h cudaminer.sln LICENSE.txt spinlock_kernel.o
compile cudaminer-util.o Makefile stamp-h1
config.guess cudaminer.vcxproj Makefile.am test_kernel.cu
config.log cudaminer.vcxproj.filters Makefile.in test_kernel.h
config.status cudaminer.vcxproj.user miner.h test_kernel.o
config.sub depcomp missing titan_kernel.cu
configure elist.h NEWS titan_kernel.h
configure.ac fermi_kernel.cu README titan_kernel.o
configure.sh fermi_kernel.h README.txt util.c
COPYING fermi_kernel.o salsa_kernel.cu
[ec2-user@ip-172-31-28-158 CudaMiner]$