Date: Sun, 5 Jul 2015 09:05:19 +0300 From: Solar Designer <solar@...nwall.com> To: john-dev@...ts.openwall.com Subject: Re: PHC: Lyra2 on GPU On Sun, Jul 05, 2015 at 12:52:48AM +0200, Agnieszka Bielec wrote: > [a@...er run]$ GOMP_CPU_AFFINITY=0-31 ./john --test --format=lyra2 > Will run 32 OpenMP threads > Benchmarking: Lyra2, Generic Lyra2 [ ]... (32xOMP) DONE > Speed for cost 1 (t) of 8, cost 2 (m) of 8 > Many salts: 49792 c/s real, 1556 c/s virtual > Only one salt: 49504 c/s real, 1546 c/s virtual > > [a@...er run]$ GOMP_CPU_AFFINITY=0-31 ./john --test --format=lyra2-lm > Will run 32 OpenMP threads > Benchmarking: Lyra2-lm, Generic Lyra2 [ ]... (32xOMP) DONE > Speed for cost 1 (t) of 8, cost 2 (m) of 8 > Many salts: 38848 c/s real, 1263 c/s virtual > Only one salt: 42272 c/s real, 1376 c/s virtual What's the difference between Lyra2 and Lyra2-lm? Can you set reasonable comments for them, please? In place of the "Generic Lyra2 [ ]". Is the memory (de)allocation overhead kept out of the loop already? I previously suggested that you reuse yescrypt's yescrypt-platform.c: *_region() functions for this, so that we use the same memory allocation mechanism for all PHC finalists. Alexander
Powered by blists - more mailing lists
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.