Page 10 of 13 FirstFirst ... 89101112 ... LastLast
Results 271 to 300 of 364

Thread: bsc, new block sorting compressor

  1. #271
    Administrator Shelwien's Avatar
    Join Date
    May 2008
    Location
    Kharkov, Ukraine
    Posts
    3,205
    Thanks
    187
    Thanked 955 Times in 493 Posts

  2. #272
    Member
    Join Date
    Jun 2009
    Location
    Kraków, Poland
    Posts
    1,474
    Thanks
    26
    Thanked 121 Times in 95 Posts
    There's something wrong with this thread. I see pat357's post as the last one when viewing thread, but when viewing forum I see that Shelwien written something on the end of thread.

    Well, after I've written this post Shelwien's post showed up. Weird.

  3. #273
    Expert
    Matt Mahoney's Avatar
    Join Date
    May 2008
    Location
    Melbourne, Florida, USA
    Posts
    3,255
    Thanks
    306
    Thanked 778 Times in 485 Posts
    It was on the next page?

  4. #274
    Member
    Join Date
    Jun 2009
    Location
    Kraków, Poland
    Posts
    1,474
    Thanks
    26
    Thanked 121 Times in 95 Posts
    In the column "Last Post By" there are double-arrow images which should theoretically point to last message. In some cases (I've stumbled across a few different cases) they don't work - in case I've noticed above link sent me to the top of 9'th page and 10'th page was nonexistent (even though on forum view there was clear indication that Shelwien written the last post in the thread) till I wrote 272'th post in the thread.

  5. #275
    Tester
    Black_Fox's Avatar
    Join Date
    May 2008
    Location
    [CZE] Czechia
    Posts
    471
    Thanks
    26
    Thanked 9 Times in 8 Posts
    Quote Originally Posted by Piotr Tarsa View Post
    In the column "Last Post By" there are double-arrow images which should theoretically point to last message.
    Nope, that's not completely true. They point to last unread post of yours in that thread, which may still be on 9th page (phpBB behaves the same way BTW). This happens to me sometimes, but the "next" page is always visible. By coincidence today I read 4 topics that had one more page to display on this forum
    I am... Black_Fox... my discontinued benchmark
    "No one involved in computers would ever say that a certain amount of memory is enough for all time? I keep bumping into that silly quotation attributed to me that says 640K of memory is enough. There's never a citation; the quotation just floats like a rumor, repeated again and again." -- Bill Gates

  6. #276
    Member
    Join Date
    Jun 2009
    Location
    Kraków, Poland
    Posts
    1,474
    Thanks
    26
    Thanked 121 Times in 95 Posts
    I've clicked that double-arrow link plenty of times, refreshing the forum in the meantime and that didn't work. It only worked when I written a post after Shelwien's one. Without my post on this (10'th) page I couldn't get to this page, because there were no links to it. IIRC the problem was usually (or always) with Shelwien's posts

  7. #277
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,497
    Thanks
    733
    Thanked 659 Times in 354 Posts
    i've seen the same problem, so Piotr is right - it was the problem in forum engine

  8. #278
    Programmer Gribok's Avatar
    Join Date
    Apr 2007
    Location
    USA
    Posts
    159
    Thanks
    0
    Thanked 1 Time in 1 Post

    bsc version 3.1.0 for testing

    bsc version 3.1.0 for testing

    [+]: Compiled with CUDA Toolkit 4.2: new llvm based compiler.
    [+]: ST now uses FastSortSm20 version of b40c: should be faster for Fermi.
    [+]: Compiled with Kepler GPU support: My Kepler GPU(Quadro K1000M) actually slower than I7-3720QM.

    Please help me with testing. Note: This version is not compatible with 3.0.0

    Thanks,
    Ilya
    Attached Files Attached Files
    Enjoy coding, enjoy life!

  9. #279
    Expert
    Matt Mahoney's Avatar
    Join Date
    May 2008
    Location
    Melbourne, Florida, USA
    Posts
    3,255
    Thanks
    306
    Thanked 778 Times in 485 Posts
    I only tested for x86 max compression (-b328 -T). It is faster than 3.00 but less compression. http://mattmahoney.net/dc/text.html#1640

  10. #280
    Programmer Gribok's Avatar
    Join Date
    Apr 2007
    Location
    USA
    Posts
    159
    Thanks
    0
    Thanked 1 Time in 1 Post
    Source code now available online @ libbsc.com
    Enjoy coding, enjoy life!

  11. #281
    Member
    Join Date
    May 2008
    Location
    Germany
    Posts
    410
    Thanks
    37
    Thanked 60 Times in 37 Posts
    @gribok: Wish you all the best in the new year!

    i have seen - your website https://github.com/IlyaGrebnov/libbsc is alive!

    you have changed the License for libbsc from GPL to Apache Version 2.0

    Will there be a adaption of bsc to the new CUDA 5.0 API in the near future?

    especially i am interested in the mode st order 5 and st order 6, which you are implemented for CPU and for GPU too!

    best regards

  12. #282
    Programmer Gribok's Avatar
    Join Date
    Apr 2007
    Location
    USA
    Posts
    159
    Thanks
    0
    Thanked 1 Time in 1 Post
    Quote Originally Posted by joerg View Post
    @gribok: Wish you all the best in the new year!

    i have seen - your website https://github.com/IlyaGrebnov/libbsc is alive!

    you have changed the License for libbsc from GPL to Apache Version 2.0

    Will there be a adaption of bsc to the new CUDA 5.0 API in the near future?

    especially i am interested in the mode st order 5 and st order 6, which you are implemented for CPU and for GPU too!

    best regards
    bsc is distributed now by Apache License, so now it can be used in commercial applications. Sorry, but I don't have any time right now to work on bsc.
    Enjoy coding, enjoy life!

  13. #283
    Member
    Join Date
    Jan 2013
    Location
    USA
    Posts
    28
    Thanks
    0
    Thanked 2 Times in 2 Posts
    Is it at all possible with a block sorting algorithm to make the data `even more sorted` by running the algorithm on the encoded data?

  14. #284
    Expert
    Matt Mahoney's Avatar
    Join Date
    May 2008
    Location
    Melbourne, Florida, USA
    Posts
    3,255
    Thanks
    306
    Thanked 778 Times in 485 Posts
    BWT is a transform that sorts a block of bytes by context, which brings together bytes that appear in similar contexts. It does not compress by itself, but it allows the data to be compressed by a fast adapting order 0 model, which is generally faster and uses less memory than a high order context model that would otherwise be needed. At http://mattmahoney.net/dc/dce.html#Section_55 you can see the effect of a BWT on text.

    Encoded data (compressed or encrypted) is not predictable by context. Therefore, sorting by context will not do any good.

  15. #285
    Administrator Shelwien's Avatar
    Join Date
    May 2008
    Location
    Kharkov, Ukraine
    Posts
    3,205
    Thanks
    187
    Thanked 955 Times in 493 Posts
    > Encoded data (compressed or encrypted) is not predictable by context. Therefore, sorting by context will not do any good.

    actually it is (compressed at least) and block sorting compression can gain a few bytes, same as CM,
    but it'd have to be a bitwise sort+postcoder, not bsc.

  16. #286
    Tester
    Nania Francesco's Avatar
    Join Date
    May 2008
    Location
    Italy
    Posts
    1,565
    Thanks
    220
    Thanked 146 Times in 83 Posts
    I tried several options this program in WCC 2013 there are mistakes, please the author to test and release a new version if possible!

  17. #287
    Member
    Join Date
    Dec 2011
    Location
    Cambridge, UK
    Posts
    452
    Thanks
    141
    Thanked 155 Times in 103 Posts
    I'm trying to build from source under Linux with Cuda support (Cuda-5.5) and get errors:

    Code:
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -DLIBBSC_CUDA_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -DNDEBUG -c libbsc/st/st.cpp
    In file included from libbsc/st/b40c/radix_sort/../radix_sort/../util/spine.cuh:28:0,
                     from libbsc/st/b40c/radix_sort/../radix_sort/problem_instance.cuh:26,
                     from libbsc/st/b40c/radix_sort/enactor.cuh:26,
                     from libbsc/st/st.cu:65,
                     from libbsc/st/st.cpp:58:
    libbsc/st/b40c/radix_sort/../radix_sort/../util/../util/cuda_properties.cuh: In constructor 'b40c::util::CudaProperties::CudaProperties()':
    libbsc/st/b40c/radix_sort/../radix_sort/../util/../util/cuda_properties.cuh:168:63: error: cannot resolve overloaded function 'FlushKernel' based on conversion to type 'const void*'
    libbsc/st/b40c/radix_sort/../radix_sort/../util/../util/cuda_properties.cuh: In constructor 'b40c::util::CudaProperties::CudaProperties(int)':
    libbsc/st/b40c/radix_sort/../radix_sort/../util/../util/cuda_properties.cuh:183:63: error: cannot resolve overloaded function 'FlushKernel' based on conversion to type 'const void*'
    (Amongst many others, but I'm hoping the cause of this error is the same as all the rest).

    I don't know enough about C++ templates to understand why it's trying to convert to const void *, but given the number of other problems I'm suspecting an incompatibility somewhere else.

    Do you have any hints for building under Linux Ubuntu 12.04?

    Thanks.

  18. #288
    The Founder encode's Avatar
    Join Date
    May 2006
    Location
    Moscow, Russia
    Posts
    3,964
    Thanks
    367
    Thanked 341 Times in 134 Posts
    Quick test with overclocked GTX 780 (ASUS GTX 780 DirectCU II OC)
    Code:
    Z:\>bsc e enwik9 enwik9.bsc -m6
    This is bsc, Block Sorting Compressor. Version 3.1.0. 8 July 2012.
    Copyright (c) 2009-2012 Ilya Grebnov <Ilya.Grebnov@gmail.com>.
    
    enwik9 compressed 1000000000 into 202545396 in 10.296 seconds.
    
    Z:\>bsc e enwik9 enwik9.bsc -m6 -G
    This is bsc, Block Sorting Compressor. Version 3.1.0. 8 July 2012.
    Copyright (c) 2009-2012 Ilya Grebnov <Ilya.Grebnov@gmail.com>.
    
    enwik9 compressed 1000000000 into 202545396 in 5.928 seconds.
    
    Z:\>bsc e enwik9 enwik9.bsc -b160 -m6
    This is bsc, Block Sorting Compressor. Version 3.1.0. 8 July 2012.
    Copyright (c) 2009-2012 Ilya Grebnov <Ilya.Grebnov@gmail.com>.
    
    enwik9 compressed 1000000000 into 191678466 in 11.996 seconds.
    
    Z:\>bsc e enwik9 enwik9.bsc -b160 -m6 -G
    This is bsc, Block Sorting Compressor. Version 3.1.0. 8 July 2012.
    Copyright (c) 2009-2012 Ilya Grebnov <Ilya.Grebnov@gmail.com>.
    
    enwik9 compressed 1000000000 into 191678466 in 7.722 seconds.

  19. #289
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    873
    Thanks
    460
    Thanked 175 Times in 85 Posts
    Is it true that BSC still does not contain preprocessor for 32 bit tables and executables?
    Would it be difficult to add?

    For suffix sorting, BSC uses libdivsufsort by Yuta Mori. Is there a chance to get a compile with Yuta mori's SAIS?
    its results look promising: https://code.google.com/p/libdivsufs...ACA_Benchmarks
    SAIS can be found here: https://sites.google.com/site/yuta256/sais
    Last edited by Stephan Busch; 31st October 2013 at 00:48.

  20. #290
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,497
    Thanks
    733
    Thanked 659 Times in 354 Posts
    Quote Originally Posted by Gribok View Post
    This is not normal if size is different. I will try to dig into this issue. But it maybe because of overclocked GPU/CPU.
    i've finally recheked compression on my GPU, performed 1000 runs of ST7 & ST8 compression+decompression overnight. there are no errors and all compressed files had the same size

    sometimes in 2012 i've updated my videocard BIOS, afair due to stability problems in (gpu-accelerated) Chrome. This update increased GPU Voltage in order to ensure its correct work and was also required for correct gaming. So the entire problem was due to incorrect fabric GPU overclocking and it manifests itself in gaming as well as gpgpu computations

    although i still fear that GPU/GDDR5 are less reliable than CPU and DDR3/4, just because their "mission" is less critical

  21. #291
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,611
    Thanks
    30
    Thanked 65 Times in 47 Posts
    I've done some QA for BSC and found a number of bugs, crashes included.
    I tried to contact gribok about it but got no reply.
    Is anyone interested in fixing this stuff?

  22. #292
    Member Skymmer's Avatar
    Join Date
    Mar 2009
    Location
    Russia
    Posts
    681
    Thanks
    37
    Thanked 168 Times in 84 Posts
    I'm interested. It would be nice if you'll also provide the data you have tested on, so we can check on other systems. Anyway, with or without your samples, I'm interested

  23. #293
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,611
    Thanks
    30
    Thanked 65 Times in 47 Posts
    OK, I have 1 sample on my PC, I'll post that now. Others will follow. This one happens on unmodified github master, just compiled with clang's address sanitizer.
    Code:
    katmacadapc% ./bsc e
    result1/crashes/id:000000,sig:06,src:000000,op:havoc,rep:128 /tmp/3 -M4 -ca
    -e2 
    This is bsc, Block Sorting Compressor. Version 3.1.0. 8 July 2012.
    Copyright (c) 2009-2012 Ilya Grebnov <Ilya.Grebnov@gmail.com>.
    
    Compressing
    result1/crashes/id:000000,sig:06,src:000000,op:havoc,re(00%)=================================================================
    ==4685==ERROR: AddressSanitizer: heap-buffer-overflow on address
    0x60300000efdf at pc 0x0000004d2a91 bp 0x7fffebc6edf0 sp 0x7fffebc6ede8
    READ of size 1 at 0x60300000efdf thread T0
        #0 0x4d2a90 (/home/madamczyk/afl-1.94b/bsc/bsc+0x4d2a90)
        #1 0x4c91c0 (/home/madamczyk/afl-1.94b/bsc/bsc+0x4c91c0)
        #2 0x4ce1c7 (/home/madamczyk/afl-1.94b/bsc/bsc+0x4ce1c7)
        #3 0x7f900e88aeac (/lib/x86_64-linux-gnu/libc.so.6+0x1eeac)
        #4 0x4c807c (/home/madamczyk/afl-1.94b/bsc/bsc+0x4c807c)
    
    0x60300000efdf is located 1 bytes to the left of 29-byte region
    [0x60300000efe0,0x60300000effd)
    allocated by thread T0 here:
        #0 0x4aacab (/home/madamczyk/afl-1.94b/bsc/bsc+0x4aacab)
        #1 0x4e60ea (/home/madamczyk/afl-1.94b/bsc/bsc+0x4e60ea)
        #2 0x4ce1c7 (/home/madamczyk/afl-1.94b/bsc/bsc+0x4ce1c7)
    
    SUMMARY: AddressSanitizer: heap-buffer-overflow ??:0 ??
    Shadow bytes around the buggy address:
      0x0c067fff9da0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
      0x0c067fff9db0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
      0x0c067fff9dc0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
      0x0c067fff9dd0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
      0x0c067fff9de0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
    =>0x0c067fff9df0: fa fa fa fa fa fa fa fa fa fa fa[fa]00 00 00 05
      0x0c067fff9e00: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
      0x0c067fff9e10: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
      0x0c067fff9e20: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
      0x0c067fff9e30: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
      0x0c067fff9e40: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
    Shadow byte legend (one shadow byte represents 8 application bytes):
      Addressable:           00
      Partially addressable: 01 02 03 04 05 06 07 
      Heap left redzone:       fa
      Heap right redzone:      fb
      Freed heap region:       fd
      Stack left redzone:      f1
      Stack mid redzone:       f2
      Stack right redzone:     f3
      Stack partial redzone:   f4
      Stack after return:      f5
      Stack use after scope:   f8
      Global redzone:          f9
      Global init order:       f6
      Poisoned by user:        f7
      Container overflow:      fc
      ASan internal:           fe
    ==4685==ABORTING
    Attached Files Attached Files
    • File Type: 7z 1.7z (191 Bytes, 50 views)

  24. #294
    Member Skymmer's Avatar
    Join Date
    Mar 2009
    Location
    Russia
    Posts
    681
    Thanks
    37
    Thanked 168 Times in 84 Posts
    Native 3.1.0 x64 build crashes on my Win7 x64 too. Native x86 build works.
    Tried random byte values without luck but 0x7F value definitely triggers the crash.

  25. #295
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,611
    Thanks
    30
    Thanked 65 Times in 47 Posts
    OK, the next part. It's all I have so far, but I may find more. I expect there's a lot more, but I don't know how much CPU time I'll spend on it. I will be more dedicated if I see fixes.
    The following bugs are in the decoder. BSC checksums it's data. It's quite robust, I executed it over million times with randomised data and saw no problems. However, if there's a risk of forged input, checksums are of no help.
    I could generate random data with correct checksums, but it was easier to disable checking them. There's a patch for libbsc.cpp attached. To reproduce issues compile it with clang -fsanitize=undefined -fno-sanitize=alignment -fsanitize=memory.
    The results were terrible. BSC would crash on over 90% of files thrown at it (though they were derivatives of just a few, so there may be selection bias). There's a set of 25 files attached. Each represents a different stack trace, so it has a potential of being a different bug, though I expect some to be duplicates.
    To sum up, aside from the 1-byte file there's nothing to worry as long as all your inputs can be trusted to come from legitimate source (even if they may be corrupted).
    Otherwise, bsc is extremely unreliable.
    I have to say I'm very disappointed, I expected it to be much better.
    Attached Files Attached Files
    Last edited by m^2; 27th October 2015 at 23:35.

  26. The Following User Says Thank You to m^2 For This Useful Post:

    Bulat Ziganshin (29th October 2015)

  27. #296
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,611
    Thanks
    30
    Thanked 65 Times in 47 Posts
    Another drop of decompressor bugs, until I see fixes it's the last one.
    There are at least 7 new paths (maybe more as I can't easily compare outputs from different sessions).
    Attached Files Attached Files

  28. #297
    Member
    Join Date
    Aug 2016
    Location
    India
    Posts
    36
    Thanks
    2
    Thanked 0 Times in 0 Posts

    Do you know how to enable CUDA support for libbsc in Linux?

    In MakeFile, the compilation is for openmp. We need to set -DLIBBSC_CUDA_SUPPORT. But still
    no effect over GPU. How the make file could be modified, so that libbsc could run on GPU?

    If I enable the flags -DLIBBSC_SORT_TRANSFORM_SUPPORT -DLIBBSC_CUDA_SUPPORT in make file, I am getting the following error

    Code:
    g++ -g -Wall -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3 -fomit-frame-pointer -fstrict-aliasing -ffast-math -DLIBBSC_SORT_TRANSFORM_SUPPORT -DLIBBSC_CUDA_SUPPORT -DNDEBUG bsc.cpp -o bsc -L. -lbsc
    ./libbsc.a(st.o): In function `bsc_st_encode':
    /home/bsc-3.1.0-src/libbsc/st/st.cpp:1015: undefined reference to `bsc_st_encode_cuda'
    ./libbsc.a(st.o): In function `bsc_st_init':
    /home/bsc-3.1.0-src/libbsc/st/st.cpp:65: undefined reference to `bsc_st_cuda_init'
    collect2: error: ld returned 1 exit status
    make: *** [bsc] Error 1
    I am using linux (Ubuntu 14.04 and CentOS 7.2)

  29. #298
    Member
    Join Date
    Aug 2016
    Location
    India
    Posts
    36
    Thanks
    2
    Thanked 0 Times in 0 Posts

    Do you know how to enable CUDA support for libbsc in Linux?

    Quote Originally Posted by Bulat Ziganshin View Post
    i use nvidia drivers 275.33, you can get it here: http://www.nvidia.ru/Download/index.aspx?lang=ru
    I am not able to enable CUDA support for libbsc in linux via makefile. How could it be resolved?

  30. #299
    Administrator Shelwien's Avatar
    Join Date
    May 2008
    Location
    Kharkov, Ukraine
    Posts
    3,205
    Thanks
    187
    Thanked 955 Times in 493 Posts
    Well, there's st.cu file in the source with bsc_st_encode_cuda() - I guess compiling that st.cu with cuda would help.

  31. #300
    Member
    Join Date
    Aug 2016
    Location
    India
    Posts
    36
    Thanks
    2
    Thanked 0 Times in 0 Posts

    libbsc not running in GPU

    Quote Originally Posted by Shelwien View Post
    Well, there's st.cu file in the source with bsc_st_encode_cuda() - I guess compiling that st.cu with cuda would help.
    Eventhough the compilation has been done with cuda for st.cu, no effect. Meaning libbsc is not at all running in GPU. If libbsc linux source package for cuda is available, it will be very helpful

    Code:
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/adler32/adler32.cpp
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/bwt/divsufsort/divsufsort.c
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/bwt/bwt.cpp
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/coder/coder.cpp
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/coder/qlfc/qlfc.cpp
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/coder/qlfc/qlfc_model.cpp
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/filters/detectors.cpp
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/filters/preprocessing.cpp
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/libbsc/libbsc.cu
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/lzp/lzp.cpp
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/platform/platform.cpp
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ -c libbsc/st/st.cu
    rm -f libbsc.a
    ar cq libbsc.a adler32.o divsufsort.o bwt.o coder.o qlfc.o qlfc_model.o detectors.o preprocessing.o libbsc.o lzp.o platform.o st.o 
    ranlib libbsc.a
    nvcc -DLIBBSC_SORT_TRANSFORM_SUPPORT -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O3  -lcuda -lcudart -DLIBBSC_CUDA_SUPPORT -DNDEBUG -I/usr/local/cuda/include/ -L/usr/local/cuda/lib64/ bsc.cu -o bsc -L. -lbsc
    This is how my makefile works. make ran successfully

    bsc execution

    Code:
    time ./bsc e /home/1.tar /home/11.tar.bsc
    still it is running in CPU. while doing nvidia-smi, no process related to bsc added there. Is there any GPU architecture specification for libbsc to run?
    Last edited by Vanns; 24th August 2016 at 09:53.

Page 10 of 13 FirstFirst ... 89101112 ... LastLast

Similar Threads

  1. Brute forcing Delta block size
    By SvenBent in forum Data Compression
    Replies: 2
    Last Post: 2nd May 2009, 12:44
  2. Block sorting for LZ compression
    By Bulat Ziganshin in forum Forum Archive
    Replies: 15
    Last Post: 14th April 2007, 15:37

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •