[viff-devel] Yet another SFDL parser
Hi everybody, I've just sent 22 patches to the viff-patches list: http://thread.gmane.org/gmane.comp.cryptography.viff.patches/104 They represent yet another attempt at making a SFDL parser. I say yet another since Mikkel and Sigurd has been working on one too, but unfortunately that seems to have lost some momentum. Or at least I haven't heard anything about it recently :-) At the WP4 meeting in Amsterdam we talked about comparing VIFF to FairPlay on the same inputs, so we should move forward with building the converter. I don't care which parser is used -- I just played around with it to learn more about Pyparsing. However, regardless of the choice, we should discuss who will commit to take the lead on it. -- Martin Geisler VIFF (Virtual Ideal Functionality Framework) brings easy and efficient SMPC (Secure Multiparty Computation) to Python. See: http://viff.dk/. ___ viff-devel mailing list (http://viff.dk/) viff-devel@viff.dk http://lists.viff.dk/listinfo.cgi/viff-devel-viff.dk
Re: [viff-devel] Yet another SFDL parser
Mikkel Krøigård [EMAIL PROTECTED] writes: Quoting Martin Geisler [EMAIL PROTECTED]: I've just sent 22 patches to the viff-patches list: http://thread.gmane.org/gmane.comp.cryptography.viff.patches/104 They represent yet another attempt at making a SFDL parser. I say yet another since Mikkel and Sigurd has been working on one too, but unfortunately that seems to have lost some momentum. Or at least I haven't heard anything about it recently :-) We haven't lost any momentum on the parser. In fact it has been done for quite a while. We have, however, lost momentum on the rest of the compiler. Right, that was sort of what I meant :-) I don't care which parser is used -- I just played around with it to learn more about Pyparsing. However, regardless of the choice, we should discuss who will commit to take the lead on it. We should have probably talked about that before we made 3 of them. I think we did agree on me, but somehow Sigurd and I ended up doing more or less the same thing concurrently. It would be great if you could go forward with the compiler! In Amsterdam we were asked about how long it would take, and I didn't have a good answer. Can you try to give a time estimate that we can tell the FairPlay guys? -- Martin Geisler VIFF (Virtual Ideal Functionality Framework) brings easy and efficient SMPC (Secure Multiparty Computation) to Python. See: http://viff.dk/. ___ viff-devel mailing list (http://viff.dk/) viff-devel@viff.dk http://lists.viff.dk/listinfo.cgi/viff-devel-viff.dk
[viff-devel] Benchmarks on a mac
Here is a transcript of my profiling run. I disabled ssl, because I had some trouble with the keys, hope that does not matter too much. sigurdmeldgaard$ python benchmark.py --profile --no-ssl -c 1 player-1.ini Seeding random generator with random seed 9520 Using field elements (65 bit modulus) I am player 1, will mul 1 numbers Using the base runtime: viff.passive.PassiveRuntime. Not using SSL Listening on port 9000 Will connect to Player 2: camel13:9001 Will connect to Player 3: camel14:9002 Starting reactor ### Starting reactor with profiling Need no preprocessing Runtime ready, generating shares Synchronizing test start. Starting test in 3 Starting test in 2 Starting test in 1 Starting test now Started parallel test Total time used: 21.186 sec Time per parallel test operation: 2 ms ** Synchronizing shutdown... done. Closing connections... done. Stopping reactor... done. Loading profiling statistics... done. 2176272 function calls (1852904 primitive calls) in 33.202 CPU seconds Ordered by: internal time, call count List reduced from 207 to 40 due to restriction 40 ncalls tottime percall cumtime percall filename:lineno(function) 366 20.9230.057 26.6800.073 selectreactor.py:93(doSelect) 173360/866802.5240.0009.8540.000 defer.py:314(_runCallbacks) 90013 1.0700.0001.3150.000 field.py:371(__mul__) 307 0.9710.0035.6750.018 basic.py:321(dataReceived) 6 0.7720.0000.9040.000 runtime.py:579(_expect_data) 16668 0.5550.0002.1890.000 shamir.py:30(share) 203364/933640.5180.0006.1410.000 defer.py:172(addCallbacks) 260039 0.4780.0000.4780.000 field.py:339(__init__) 33340 0.3600.0000.8330.000 runtime.py:311(sendData) 20004 0.3480.0001.8110.000 runtime.py:189(__init__) 8 0.3370.0004.6550.000 runtime.py:272(stringReceived) 113350/433500.2970.0005.4900.000 defer.py:189(addCallback) 80004 0.2970.0000.4080.000 field.py:342(__add__) 6 0.2700.0001.3980.000 runtime.py:613(_expect_share) 70021/46678 0.2530.0004.3540.000 defer.py:291(_startRunCallbacks) 1 0.2260.0000.6050.000 shamir.py:96(recombine) 33342 0.2040.0000.2900.000 abstract.py:161(write) 1 0.1870.0004.7900.000 passive.py:163(share_recombine) 70021/46678 0.1800.0004.4770.000 defer.py:233(callback) 70010 0.1740.0000.3630.000 runtime.py:66(__init__) 2 0.1670.0001.6700.000 passive.py:389(shamir_share) 80006/70004 0.1670.0001.3370.000 runtime.py:213(_callback_fired) 30002 0.1530.0007.9910.000 runtime.py:358(inc_pc_wrapper) 23336 0.1440.0000.1440.000 random.py:217(_randbelow) 23336 0.1420.0000.2860.000 random.py:147(randrange) 30006 0.1390.0001.8010.000 runtime.py:597(_exchange_shares) 33342 0.1330.0000.4730.000 basic.py:338(sendString) 1 0.0880.0886.5056.505 benchmark.py:247(run_test) 1 0.0800.0006.1670.001 passive.py:138(mul) 1 0.0730.0732.2342.234 benchmark.py:197(begin) 20004 0.0660.0002.0600.000 runtime.py:221(gather_shares) 70020 0.0590.0000.0590.000 defer.py:166(__init__) 33346 0.0590.0000.0860.000 abstract.py:264(startWriting) 1 0.0580.0004.8490.000 runtime.py:559(callback_wrapper) 6 0.0580.0000.1110.000 runtime.py:615(lambda) 1 0.0520.0005.1660.001 runtime.py:539(schedule_callback) 33342 0.0510.0000.0510.000 struct.py:54(pack) 33628 0.0490.0000.0490.000 struct.py:77(unpack) 33340 0.0470.0000.8810.000 runtime.py:315(sendShare) 23336 0.0390.0000.3250.000 random.py:211(randint) ___ viff-devel mailing list (http://viff.dk/) viff-devel@viff.dk http://lists.viff.dk/listinfo.cgi/viff-devel-viff.dk
[viff-devel] Repositories on hg.viff.dk
Good evening everybody, If you would like space for a Mercurial repository on hg.viff.dk, then it's easy -- all I need is your SSH public key. I've already added a repository for Marcel -- I got a first patch from him today! If anybody else would like a similar repository for outgoing changes, then let me know. Mikkel and Thomas: you already have access to viff.dk. You can easily create your own repositories by going to ~/repos on viff.dk and cloning an existing repository. Then edit the .hg/hgrc file to describe it, look in ~/repos/viff/.hg/hgrc for a starting point. -- Martin Geisler VIFF (Virtual Ideal Functionality Framework) brings easy and efficient SMPC (Secure Multiparty Computation) to Python. See: http://viff.dk/. pgpA9V8KPoj59.pgp Description: PGP signature ___ viff-devel mailing list (http://viff.dk/) viff-devel@viff.dk http://lists.viff.dk/listinfo.cgi/viff-devel-viff.dk
