On Thu, Jul 05, 2012 at 11:39:02PM +0200, Pawel Jakub Dawidek wrote: > > sysctls are not all Giant locked. KERN_ARND is marked MPSAFE, so it does > > not > > use Giant: > > It doesn't really matter. Our in-kernel random generator has its own > giant lock, so is basically single threaded. Compare results of those > two on some SMP machine: > > # dd if=/dev/random of=/dev/null bs=1m count=1024 > > # sh -c 'for i in `jot $(sysctl -n hw.ncpu)`; do dd if=/dev/random > of=/dev/null bs=1m count=128 & done'
It doesn't really matter for KERN_ARND because it is rarely reseeded and consume most of its time on internal arc4 permutations. That is why strong KERN_ARND seeding after boot is required (currently it is very weak and easily guessed as I already mention) - started right after boot applications all suffers from weak seeding up to the next rare KERN_ARND and arc4random() both reseeds.. -- http://ache.vniz.net/
pgpSJkSo7cdy5.pgp
Description: PGP signature
