On Wed, Jul 17, 2002 at 01:55:52AM +0200, Fruhwirth Clemens wrote: > for a few days i'm trying to catch a race condition which is apparently > existing within cryptoloop. with the keen assistance of hvr and kyle i can give advice to the vmware users out there that enjoy a stable setup on a cryptoloop device: - use loop-jari instead of loop-hvr to patch loop.c. it will prevent total lockups, when vmware is involved in bulk transfers. - get the latest CVS cipher-aes.c, it's quite faster and will make your box not unresponsive that long if you're suspending your virtual machine. further i've tried loop-AES and found out the performance is much better on that issue, most likely because of the asm optimized aes-cipher. regards, clemens
Attachment:
pgp00029.pgp
Description: PGP signature
