[freenet-dev] propose we get rid of jbigi/jcpuid and move to GNU classpath BigInteger

2011-08-06 Thread Matthew Toseland 
On Wednesday 27 Jul 2011 22:10:44 Ximin Luo wrote:
> benefits of GNU BigInteger (from GNU classpath):
> 
> - we don't need to maintain jbigi
>   - we don't need to maintain jcpuid
>   - jcpuid's per-cpu speed improvements are probably negligible anyway; the
> bigger speed increase is java->native, not native->native-special-for-your-cpu
> - GNU BigInteger is much faster than our jbigi implementation
> 
> (we don't need to depend on the entirety of GNU classpath, we can just include
> the bits we need: https://bitbucket.org/dfdeshom/gmp-java)
> 
> # on latest fred-staging:
> # assumes you are on debian, with gcj-4.6-jre-lib installed:
> 
> $ git checkout bigint-gnu-gmp
> $ ant unit -Dtest.class=net.i2p.util.NativeBigIntegerTest 
> -Dtest.benchmark=true
> [...]
> unit:
> [junit] Running net.i2p.util.NativeBigIntegerTest
> [junit] Testsuite: net.i2p.util.NativeBigIntegerTest
> [junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 26.878 sec
> [junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 26.878 sec
> [junit] - Standard Output ---
> [junit] DEBUG: Warming up the random number generator...
> [junit] DEBUG: Random number generator warmed up
> [junit] INFO: 200 runs complete without any errors
> [junit] native run time:  4121ms (20ms each)
> [junit] java run time:8465ms (42ms each)
> [junit] gnu run time: 1776ms (8ms each)
> [junit] native = 48.682811577082106% of pure java time
> [junit] gnu = 20.98050797401063% of pure java time
> [junit] DEBUG: Warming up the random number generator...
> [junit] DEBUG: Random number generator warmed up
> [junit] INFO: 200 runs complete without any errors
> [junit] native run time:  6254ms (31ms each)
> [junit] java run time:3987ms (19ms each)
> [junit] gnu run time: 2190ms (10ms each)
> [junit] native = 156.85979433157763% of pure java time
> [junit] gnu = 54.92851768246802% of pure java time
> [junit] -  ---
> [junit] - Standard Error -
> [junit] INFO: Optimized native BigInteger library
> 'net/i2p/util/libjbigi-linux-x86_64.so' loaded from resource
> [junit] -  ---
> [junit]
> [junit] Testcase: testModPow took 14.436 sec
> [junit] Testcase: testDoubleValue took 12.438 sec
> [...]
> 
> 
> -- 
> GPG: 4096R/5FBBDBCE
> https://github.com/infinity0
> https://bitbucket.org/infinity0
> https://launchpad.net/~infinity0
> 
Looks good to me. I assume it won't make the jar dramatically bigger?
-- next part --
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL:

Re: [freenet-dev] propose we get rid of jbigi/jcpuid and move to GNU classpath BigInteger

2011-08-06 Thread Matthew Toseland 
On Wednesday 27 Jul 2011 22:10:44 Ximin Luo wrote:
> benefits of GNU BigInteger (from GNU classpath):
> 
> - we don't need to maintain jbigi
>   - we don't need to maintain jcpuid
>   - jcpuid's per-cpu speed improvements are probably negligible anyway; the
> bigger speed increase is java->native, not native->native-special-for-your-cpu
> - GNU BigInteger is much faster than our jbigi implementation
> 
> (we don't need to depend on the entirety of GNU classpath, we can just include
> the bits we need: https://bitbucket.org/dfdeshom/gmp-java)
> 
> # on latest fred-staging:
> # assumes you are on debian, with gcj-4.6-jre-lib installed:
> 
> $ git checkout bigint-gnu-gmp
> $ ant unit -Dtest.class=net.i2p.util.NativeBigIntegerTest 
> -Dtest.benchmark=true
> [...]
> unit:
> [junit] Running net.i2p.util.NativeBigIntegerTest
> [junit] Testsuite: net.i2p.util.NativeBigIntegerTest
> [junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 26.878 sec
> [junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 26.878 sec
> [junit] - Standard Output ---
> [junit] DEBUG: Warming up the random number generator...
> [junit] DEBUG: Random number generator warmed up
> [junit] INFO: 200 runs complete without any errors
> [junit] native run time:  4121ms (20ms each)
> [junit] java run time:8465ms (42ms each)
> [junit] gnu run time: 1776ms (8ms each)
> [junit] native = 48.682811577082106% of pure java time
> [junit] gnu = 20.98050797401063% of pure java time
> [junit] DEBUG: Warming up the random number generator...
> [junit] DEBUG: Random number generator warmed up
> [junit] INFO: 200 runs complete without any errors
> [junit] native run time:  6254ms (31ms each)
> [junit] java run time:3987ms (19ms each)
> [junit] gnu run time: 2190ms (10ms each)
> [junit] native = 156.85979433157763% of pure java time
> [junit] gnu = 54.92851768246802% of pure java time
> [junit] -  ---
> [junit] - Standard Error -
> [junit] INFO: Optimized native BigInteger library
> 'net/i2p/util/libjbigi-linux-x86_64.so' loaded from resource
> [junit] -  ---
> [junit]
> [junit] Testcase: testModPow took 14.436 sec
> [junit] Testcase: testDoubleValue took 12.438 sec
> [...]
> 
> 
> -- 
> GPG: 4096R/5FBBDBCE
> https://github.com/infinity0
> https://bitbucket.org/infinity0
> https://launchpad.net/~infinity0
> 
Looks good to me. I assume it won't make the jar dramatically bigger?


signature.asc
Description: This is a digitally signed message part.
___
Devl mailing list
Devl@freenetproject.org
http://freenetproject.org/cgi-bin/mailman/listinfo/devl

[freenet-dev] propose we get rid of jbigi/jcpuid and move to GNU classpath BigInteger

2011-07-27 Thread Ximin Luo 
benefits of GNU BigInteger (from GNU classpath):

- we don't need to maintain jbigi
  - we don't need to maintain jcpuid
  - jcpuid's per-cpu speed improvements are probably negligible anyway; the
bigger speed increase is java->native, not native->native-special-for-your-cpu
- GNU BigInteger is much faster than our jbigi implementation

(we don't need to depend on the entirety of GNU classpath, we can just include
the bits we need: https://bitbucket.org/dfdeshom/gmp-java)

# on latest fred-staging:
# assumes you are on debian, with gcj-4.6-jre-lib installed:

$ git checkout bigint-gnu-gmp
$ ant unit -Dtest.class=net.i2p.util.NativeBigIntegerTest -Dtest.benchmark=true
[...]
unit:
[junit] Running net.i2p.util.NativeBigIntegerTest
[junit] Testsuite: net.i2p.util.NativeBigIntegerTest
[junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 26.878 sec
[junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 26.878 sec
[junit] - Standard Output ---
[junit] DEBUG: Warming up the random number generator...
[junit] DEBUG: Random number generator warmed up
[junit] INFO: 200 runs complete without any errors
[junit] native run time:4121ms (20ms each)
[junit] java run time:  8465ms (42ms each)
[junit] gnu run time:   1776ms (8ms each)
[junit] native = 48.682811577082106% of pure java time
[junit] gnu = 20.98050797401063% of pure java time
[junit] DEBUG: Warming up the random number generator...
[junit] DEBUG: Random number generator warmed up
[junit] INFO: 200 runs complete without any errors
[junit] native run time:6254ms (31ms each)
[junit] java run time:  3987ms (19ms each)
[junit] gnu run time:   2190ms (10ms each)
[junit] native = 156.85979433157763% of pure java time
[junit] gnu = 54.92851768246802% of pure java time
[junit] -  ---
[junit] - Standard Error -
[junit] INFO: Optimized native BigInteger library
'net/i2p/util/libjbigi-linux-x86_64.so' loaded from resource
[junit] -  ---
[junit]
[junit] Testcase: testModPow took 14.436 sec
[junit] Testcase: testDoubleValue took 12.438 sec
[...]


-- 
GPG: 4096R/5FBBDBCE
https://github.com/infinity0
https://bitbucket.org/infinity0
https://launchpad.net/~infinity0

-- next part --
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 900 bytes
Desc: OpenPGP digital signature
URL:

[freenet-dev] propose we get rid of jbigi/jcpuid and move to GNU classpath BigInteger

2011-07-27 Thread Ximin Luo 
benefits of GNU BigInteger (from GNU classpath):

- we don't need to maintain jbigi
  - we don't need to maintain jcpuid
  - jcpuid's per-cpu speed improvements are probably negligible anyway; the
bigger speed increase is java->native, not native->native-special-for-your-cpu
- GNU BigInteger is much faster than our jbigi implementation

(we don't need to depend on the entirety of GNU classpath, we can just include
the bits we need: https://bitbucket.org/dfdeshom/gmp-java)

# on latest fred-staging:
# assumes you are on debian, with gcj-4.6-jre-lib installed:

$ git checkout bigint-gnu-gmp
$ ant unit -Dtest.class=net.i2p.util.NativeBigIntegerTest -Dtest.benchmark=true
[...]
unit:
[junit] Running net.i2p.util.NativeBigIntegerTest
[junit] Testsuite: net.i2p.util.NativeBigIntegerTest
[junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 26.878 sec
[junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 26.878 sec
[junit] - Standard Output ---
[junit] DEBUG: Warming up the random number generator...
[junit] DEBUG: Random number generator warmed up
[junit] INFO: 200 runs complete without any errors
[junit] native run time:4121ms (20ms each)
[junit] java run time:  8465ms (42ms each)
[junit] gnu run time:   1776ms (8ms each)
[junit] native = 48.682811577082106% of pure java time
[junit] gnu = 20.98050797401063% of pure java time
[junit] DEBUG: Warming up the random number generator...
[junit] DEBUG: Random number generator warmed up
[junit] INFO: 200 runs complete without any errors
[junit] native run time:6254ms (31ms each)
[junit] java run time:  3987ms (19ms each)
[junit] gnu run time:   2190ms (10ms each)
[junit] native = 156.85979433157763% of pure java time
[junit] gnu = 54.92851768246802% of pure java time
[junit] -  ---
[junit] - Standard Error -
[junit] INFO: Optimized native BigInteger library
'net/i2p/util/libjbigi-linux-x86_64.so' loaded from resource
[junit] -  ---
[junit]
[junit] Testcase: testModPow took 14.436 sec
[junit] Testcase: testDoubleValue took 12.438 sec
[...]


-- 
GPG: 4096R/5FBBDBCE
https://github.com/infinity0
https://bitbucket.org/infinity0
https://launchpad.net/~infinity0



signature.asc
Description: OpenPGP digital signature
___
Devl mailing list
Devl@freenetproject.org
http://freenetproject.org/cgi-bin/mailman/listinfo/devl