[OpenAFS-devel] Announce: TFS fork of OpenAFS for ipv6 and rxgk implementation testing

[Troy Benjegerdes]

Is there anything that wraps all the memory management that hcrypto
does so it can run as part of a kernel mode filesystem client driver?

On Sat, Oct 20, 2012 at 05:17:48PM -0400, Matt W. Benjamin wrote:
> Nothing else used k5ssl.  Heimdal has wide support.
> 
> Matt
> 
> ----- "Troy Benjegerdes" <hozer@hozed.org> wrote:
> 
> > After sending the email, I figured out that k5ssl had everything, so
> > now I'm fighting with trying to get it to build correctly.
> > 
> > If we have a portable k5ssl, what do we need hcrypt for?
> > 
> > On Fri, Oct 19, 2012 at 07:16:32PM -0400, Matt W. Benjamin wrote:
> > > Hi,
> > > 
> > > And of course, rxk5 did not use anything from hcrypt.  It was at the
> > cost of providing its own framework, completely portable.  We might
> > still have wanted to use builtins, of course.  I didn't think on Linux
> > we were prevented by the GPL.  Just never got around to that.
> > > 
> > > Matt
> > > 
> > > ----- "Troy Benjegerdes" <hozer@hozed.org> wrote:
> > > 
> > > > I have something now that compiles, both without rxk5, and with:
> > > > 
> > > > ./configure --enable-rxk5 --with-krb5-conf=/usr/bin/krb5-config
> > > > 
> > > > What doesn't work is it's not been tested at all, and the kernel
> > > > module for linux does not compile. Part of the problem with the 
> > > > kernel module is src/external/heimdal/hcrypt/evp.c has a whole
> > > > bunch of free/malloc/calloc stuff that is going to cause issues
> > > > when used in the kernel.
> > > > 
> > > > So one general question: Is there any code that I can pull in for
> > > > rxgk (or look at) and see how this issue is resolved.
> > > > 
> > > > Another issue: Linux has a nice set of in-kernel crypto
> > functions,
> > > > why can't we use those instead of re-inventing the wheel?
> > > > 
> > > > On Mon, Sep 10, 2012 at 06:29:08PM -0500, Troy Benjegerdes wrote:
> > > > > Please comment, clone, and fork.
> > > > > 
> > > > > https://bitbucket.org/dahozer/tfs
> > > > > 
> > > > > I will accept pretty much any change that has some tests and
> > passes
> > > > 'make check' on linux.
> > > > > _______________________________________________
> > > > > OpenAFS-devel mailing list
> > > > > OpenAFS-devel@openafs.org
> > > > > https://lists.openafs.org/mailman/listinfo/openafs-devel
> > > > _______________________________________________
> > > > OpenAFS-devel mailing list
> > > > OpenAFS-devel@openafs.org
> > > > https://lists.openafs.org/mailman/listinfo/openafs-devel
> > > 
> > > -- 
> > > Matt Benjamin
> > > The Linux Box
> > > 206 South Fifth Ave. Suite 150
> > > Ann Arbor, MI  48104
> > > 
> > > http://linuxbox.com
> > > 
> > > tel. 734-761-4689
> > > fax. 734-769-8938
> > > cel. 734-216-5309
> > _______________________________________________
> > OpenAFS-devel mailing list
> > OpenAFS-devel@openafs.org
> > https://lists.openafs.org/mailman/listinfo/openafs-devel
> 
> -- 
> Matt Benjamin
> The Linux Box
> 206 South Fifth Ave. Suite 150
> Ann Arbor, MI  48104
> 
> http://linuxbox.com
> 
> tel. 734-761-4689
> fax. 734-769-8938
> cel. 734-216-5309
> _______________________________________________
> OpenAFS-devel mailing list
> OpenAFS-devel@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-devel