On 01/16/2014 01:07 PM, Castor Fu wrote: > > > > On Fri, Jan 10, 2014 at 9:00 PM, Jens Axboe <jaxboe@xxxxxxxxxxxx > <mailto:jaxboe@xxxxxxxxxxxx>> wrote: > > > > Jens Axboe (1): > Add option to manually seed the random generators > > > Hi Jens: Have you seen the work on counter-based random number > generators? They might be applicable here. > > The basic idea is that if you encrypt the value of a a counter, you can > get usefully random > bits. You can do something like use some of the bits for thread number, > file number, etc. > and be able to generate the expected state without cycling through the RNG. I know about these, you end up using some of the same techniques on data reliability on flash as well. But I don't think it applies very well here, as fio isn't really interested in fast-forwarding to a specific sequence. The above change is just meant to be able to specifically seed the random generator so you can control the sequence. You don't care about the specific sequence at all, just whether or not it's different with seed X vs seed Y, and that seed X will always generate the same sequence. -- Jens Axboe -- To unsubscribe from this list: send the line "unsubscribe fio" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
