[PATCH] Use Intel AES instruction set if it exists.
jra at samba.org
Fri Sep 1 15:48:26 UTC 2017
On Fri, Sep 01, 2017 at 10:06:46PM +1200, Andrew Bartlett wrote:
> Sure. I don't know how others feel about a mandatory dependency, but I
> think we should work to get out of the crypto business, even if it
Great ! What is the timeframe for your Catalyst Team to
create the libnettle code we need ?
We've been sitting on a severe performance deficiency
for many years now (the Intel crypto code was added
to the Linux kernel in *2008*) and we have done
*NOTHING* to fix this in this time.
Netgear has offered us a solution (needed a little
work, but I was able to fix this up in a couple of
days) that fixes this issue right now.
I don't think it's sensible to ignore working code
in this way.
I'm happy to move us to libnettle, but how much
work is this going to take ? Who is committing
to do this work ? What timeframe ?
Continuing to ignore performance issues for the sake of
"jam tomorrow" is not wise IMHO.
More information about the samba-technical