On Tue, 2006-05-16 at 16:58 +0200, Michael Buesch wrote: > I think most (all?) of the machines, OpenWRT runs on, are running > a bcm43xx wireless chip. This chip has a hardware random number > generator. patches to utilize it recently went into -mm. > But I must admit, we don't know how it generates random numbers. > But someone did some RNG tests on it in the past (I think it was > Johannes). I did, but no predictability tests, only FIPS 140-2 or so. It checked out pretty well but someone claimed that the RNG was predictable because it was probably only used for some of the coding for wireless. I don't know how to test this hypothesis as I don't know enough about QAM and whatever codings are used there (nor do I know where you'd need (pseudo-)random numbers) johannes
Attachment:
signature.asc
Description: This is a digitally signed message part
- References:
- /dev/random on Linux
- From: Jonathan Day <[email protected]>
- Re: /dev/random on Linux
- From: Muli Ben-Yehuda <[email protected]>
- Re: /dev/random on Linux
- From: Kyle Moffett <[email protected]>
- Re: /dev/random on Linux
- From: Michael Buesch <[email protected]>
- /dev/random on Linux
- Prev by Date: Re: jffs2 build fixes
- Next by Date: Re: [PATCH 2/3] reliable stack trace support (x86-64)
- Previous by thread: Re: /dev/random on Linux
- Next by thread: [PATCH 2/3] moxa: remove pointless check of 'tty' argument vs NULL
- Index(es):