Re: [FFmpeg-devel] libavutil: optimize camellia cipher

2015-02-14 Thread supraja reddy
Hello, Here are the results. lavu CAMELLIA size: 1048576 runs: 1024 time: 21.476 +- 0.034 crypto CAMELLIA size: 1048576 runs: 1024 time: 20.184 +- 0.043 gcrypt CAMELLIA size: 1048576 runs: 1024 time: 64.362 +- 1.031 tomcrypt CAMELLIA size: 1048576

Re: [FFmpeg-devel] libavutil: optimize camellia cipher

2015-02-14 Thread Giorgio Vazzana
Hi Supraja, could you also post the result of crypto_bench now to see how fast our implementation is compared to the others? Thanks. Giorgio Vazzana ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] libavutil: optimize camellia cipher

2015-02-10 Thread Michael Niedermayer
On Tue, Feb 10, 2015 at 09:31:24PM +0530, supraja reddy wrote: > Hello, > > > +AV_WB64(y,KE); > > > + > > KE=SP[0][y[0]]^SP[1][y[1]]^SP[2][y[2]]^SP[3][y[3]]^SP[4][y[4]]^SP[5][y[5]]^SP[6][y[6]]^SP[7][y[7]]; > > > +return KE; > > > > have you tried it without the y[] array ? > > storing and

Re: [FFmpeg-devel] libavutil: optimize camellia cipher

2015-02-10 Thread supraja reddy
Hello, > +AV_WB64(y,KE); > > + > KE=SP[0][y[0]]^SP[1][y[1]]^SP[2][y[2]]^SP[3][y[3]]^SP[4][y[4]]^SP[5][y[5]]^SP[6][y[6]]^SP[7][y[7]]; > > +return KE; > > have you tried it without the y[] array ? > storing and loading from memory may (or may not) be slower I have tried without y[] array.

Re: [FFmpeg-devel] libavutil: optimize camellia cipher

2015-02-09 Thread Michael Niedermayer
On Tue, Feb 10, 2015 at 12:44:23AM +0530, supraja reddy wrote: > Hello, > > I have attached a patch to optimize the camellia block cipher. > I ran crypto_bench on the optimized version. Here are the results of > runtime using crypto_bench. > > Before the changes : > lavu CAMELLIA size:

[FFmpeg-devel] libavutil: optimize camellia cipher

2015-02-09 Thread supraja reddy
Hello, I have attached a patch to optimize the camellia block cipher. I ran crypto_bench on the optimized version. Here are the results of runtime using crypto_bench. Before the changes : lavu CAMELLIA size: 1048576 runs: 1024 time: 32.541 +- 0.044 After the changes: lavu C