|
Message-ID: <0adac16dd599d77fde51e7b5da561701@smtp.hushmail.com> Date: Tue, 24 Jul 2012 13:57:09 +0200 From: magnum <john.magnum@...hmail.com> To: john-users@...ts.openwall.com Subject: Re: Confusing --mem-file-size=SIZE description On 2012-07-24 13:42, magnum wrote: > On 2012-07-24 12:06, Frank Dittrich wrote: >> The parameter --mem-file-size is also mentioned in doc/README.mpi. >> Here, the special SIZE values 0, 1, 2 are mentioned. >> Apparently, for MPI this parameter is mis-used for something completely >> different. Especially the --mem-file-size=0 meaning in MPI mode is >> unfortunate, since without MPI, this means: load the complete file into >> memory. If you have enough memory, this is usually what you want. > > --mem=0 is not different - it's actually where the non-MPI behavior > stems from. MPI or not, it is a short-cut for saying "buffer the file, > no matter its size". Actually --mem=1 is also the same, MPI or not. And it's neither a mis-use nor a special case: You just declare that you will buffer at most one byte - effectively turning off preload. magnum
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.