Topic: BitCrack - A tool for brute-forcing private keys - page 67. (Read 76850 times)

Oh my goodness, you just reminded me something. I keep forgetting that CUDA toolkit 11 dropped compute cap 30 so that you need to make it at least 35.


See this message. Some function is feeding bad pointers to batchBeginAdd() and batchBeginAddWithDouble(), and these are passed to MulModP() --> copyBigInt() to the point where array access of either of copyBigInt() parameters generates athe Misaligned Access exception everyone here is getting.

It might be something in host code that's giving batchBeginAdd some pointer that's incremented by 1 and not aligned or something. I didn't get a chance to check.

I've done a bit of CUDA development, I might be able to help out.  I have a 3070 card arriving by courier in a few days.

Can you tell me any specific error messages you are getting?   

yes that is an edited vanitsearch used there, used with wildcards. * stronger cards and more means more chances,
or to say lesser wildcards. good luck there,

Noname, COMPUTE_CAP=35 this in the makefile instead of COMPUTE_CAP=30 makes it work for me same with
your files as with yoyodapro's files, thanks for the help there mate
,

With what tool have you generated keys ? VanitySearch ? Thanks

https://github.com/brichard19/BitCrack/pull/148/commits

I will be grateful for any thoughts.

I compiled for you.
try it. if it starts, I will give the value MAKE ()

https://drive.google.com/drive/folders/1LJNcU_CGvWJONH0fJf6KEWi0qmxxW5Ro?usp=sharing

NVCCFLAGS=-std=c++11 -gencode=arch=compute_50,code=sm_50 -gencode=arch=compute_52,code=sm_52 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_61,code=compute_61 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -Xptxas="-v" -Xcompiler "${CXXFLAGS}"

I compiled for you.
try it. if it starts, I will give the value MAKE
or try changing MAKE and compile it yourself.
The CUDA version is very good.
https://github.com/djarumlights/BitCrack

Thanks for the help a ton, and only changed the c_c to 35 in my case. made the solution file work as a charm,
otherwise no issues yet.
is there a verbose mode on this bitcrack as of no others have that, like show where it is now instead of continue file.
changed to 1024 now.
thanks guys

I sometimes got that error when changing the compute cap and arch to different values from whatever the GPU's using e.g to compute_70 and sm_70 when working with a compute cap 86 GPU.

Try building the program again but with COMPUTE_CAP variable changed to 61 , that is the corresponding value for Pascal GPUs.


I doubt that aligning the thread size to 1024 will fix an unsupported PTX binary problem since according to the error, the CUDA program cannot even be read by the GPU.

The worst a misaligned thread size (not aligned to 32 threads i.e the size of a warp) can do is degrade performance.

try -t 128 -b 256 -t 1024

i am using a 1070, 128 1000 256

What is your hardware and -t -b -p settings?

 Error: misaligned address
?

cubitcrack does - the provided PTX was compiled with an unsupported toolchain

clbitcrack works fine,

why are you using OPENCL?
CUDA 2080TI = 1350mh/s

Yes that's right.
I am currently using version (Pica's) https://github.com/neutron220/BitCrack

but based on the above, when finding (example) 30m points, I would like to have a random search within the range between the points. I think this will speed up the search.
and it also gives a chance for people with slow and old cards to play this game. as fast cards go through consistently much faster and we have no chance.

Awesome man, I'll send you a PM and see if I can get this thing working. Hopefully there is some kind of debugger bundled with the CUDA toolkit for Windows  I think it will make the problem more obvious than cuda-gdb.

I will update this post or just make a new one if I can find the optimal settings. Maybe someone can make bitcrack keep a list of GPUs known in advance and load optimal settings for them (preferably by means of some benchmarking program).

https://github.com/djarumlights/BitCrack (rndom only CUDA)
https://github.com/neutron220/BitCrack (random all) low speed RX480 8gb 100m/s

Ive successfully compiled a release for Windows that utilizes Cuda 11.2.

https://github.com/yoyodapro/BitCrack/releases/tag/v11.2-beta

Ive compiled cuBitCrack.exe and clBitCrack.exe using 11.2 targets. clBitCrack confirmed working using -t 128 -b 256 -p 1024, if anyone could provide the recommended settings for these cards it would be greatly appreciated! cuBitCrack still not working for 20** and 30** series cards, hoping someone can help me figure it out 

[2021-01-13.23:39:32] [Info] Compression: compressed
[2021-01-13.23:39:32] [Info] Starting at: 0000000000000000000000000000000000000000000000008000000000000001
[2021-01-13.23:39:32] [Info] Ending at:   000000000000000000000000000000000000000000000000FFFFFFFFFFFFFFFF
[2021-01-13.23:39:32] [Info] Counting by: 0000000000000000000000000000000000000000000000000000000000000002
[2021-01-13.23:39:32] [Info] Compiling OpenCL kernels...
[2021-01-13.23:39:45] [Info] Initializing GeForce RTX 3070
[2021-01-13.23:39:49] [Info] Generating 33,554,432 starting points (1280.0MB)
[2021-01-13.23:39:51] [Info] 10.0%
[2021-01-13.23:39:52] [Info] 20.0%
[2021-01-13.23:39:52] [Info] 30.0%
[2021-01-13.23:39:52] [Info] 40.0%
[2021-01-13.23:39:52] [Info] 50.0%
[2021-01-13.23:39:52] [Info] 60.0%
[2021-01-13.23:39:52] [Info] 70.0%
[2021-01-13.23:39:52] [Info] 80.0%
[2021-01-13.23:39:53] [Info] 90.0%
[2021-01-13.23:39:53] [Info] 100.0%
[2021-01-13.23:39:53] [Info] Done
GeForce RTX 3070 3072 / 8192MB | 1 target 936.15 MKey/s (40,802,189,312 total) [00:00:38][2021-01-13.23:40:33]

[2021-01-13.23:39:32] [Info] Compression: compressed
[2021-01-13.23:39:32] [Info] Starting at: 0000000000000000000000000000000000000000000000008000000000000000
[2021-01-13.23:39:32] [Info] Ending at:   000000000000000000000000000000000000000000000000FFFFFFFFFFFFFFFF
[2021-01-13.23:39:32] [Info] Counting by: 0000000000000000000000000000000000000000000000000000000000000002
[2021-01-13.23:39:32] [Info] Compiling OpenCL kernels...
[2021-01-13.23:39:32] [Info] Initializing GeForce RTX 2080
[2021-01-13.23:39:35] [Info] Generating 33,554,432 starting points (1280.0MB)
[2021-01-13.23:39:38] [Info] 10.0%
[2021-01-13.23:39:38] [Info] 20.0%
[2021-01-13.23:39:39] [Info] 30.0%
[2021-01-13.23:39:39] [Info] 40.0%
[2021-01-13.23:39:39] [Info] 50.0%
[2021-01-13.23:39:39] [Info] 60.0%
[2021-01-13.23:39:39] [Info] 70.0%
[2021-01-13.23:39:39] [Info] 80.0%
[2021-01-13.23:39:39] [Info] 90.0%
[2021-01-13.23:39:40] [Info] 100.0%
[2021-01-13.23:39:40] [Info] Done
GeForce RTX 2080 3072 / 8192MB | 1 target 710.05 MKey/s (38,050,725,888 total) [00:00:51][2021-01-13.23:40:33]

Very interesting, was looking for a tool like this.
Definitely going to give it a try.