Dan Hollis(goemon@anime.net)@2003.01.06 15:12:12 +0000: > On Mon, 6 Jan 2003, Jari Ruusu wrote: > > This is explained in loop-AES' README file. Look for section "Performance > > tuning for 2.4 kernels". > > As to question of how many simultaneous writes, each device backed loop can > > have preallocated number of pages in-flight simultaneously. In-flight > > meaning that a request has been sent to underlying driver (ide or scsi), but > > underlying driver has not yet acknowledged that request as completed. > > any ways to increase loop performance on 2.4.19 / AES-v1.6f ? > device-backed AES128 on duron 1200 I get 24mbyte/sec... (de)ciphering requires many work. I don't know usual performances in a such case, but from my experience with other tools (IPsec in FreeBSD, in particular) 24 MB/s seems to be a good perf. I think the processor is the bottleneck. What is your CPU usage when using cryptoloop? N. - Linux-crypto: cryptography in and on the Linux system Archive: http://mail.nl.linux.org/linux-crypto/
