[strongSwan] AES-GCM for ESP

Randy Wyatt rwwyatt01 at gmail.com
Wed Jul 8 17:03:42 CEST 2015


tcrypt is simply to measure the performance of the encryption algorithm.
It takes several arguments.  The first argument is the algorithm.  The
second is the type.  The third is mode. The fourth is mask, and the fifth
is time.

The error message you are seeing is actually a little deceiving. There is
not actually a problem with the kernel symbols. It solely means one of the
arguments are wrong.  You can inspect tcrypt.c in the kernel sources to see
the encryption tests.  From what I can tell,  your argument to alg is wrong.

I think the wiki page needs to be updated.  I will wait for someone more
experienced than I am to validate my findings.

The tests start at line 952 in tcrypt.c.  I would specifically look at line
1229.  It still gives the resources temporarily unavailable as mentioned in
the wiki.

Regards,
Randy

On Tue, Jul 7, 2015 at 2:17 PM, Joshua Liebow-Feeser <joshlf13 at gmail.com>
wrote:

> Hi,
>
> I'm trying to set up strongSwan to use AES-GCM as the encryption algorithm
> for ESP. I'd like this to be parallelized, so I'm trying to use pcrypt, but
> whenever I try to instantiate the proper kernel support, I get errors:
>
> $ sudo modprobe tcrypt alg="pcrypt(rfc4106(gcm(aes-aesni)))" type=3
> FATAL: Error inserting tcrypt
> (/lib/modules/3.13.0-55-generic/kernel/crypto/tcrypt.ko): Unknown symbol in
> module, or unknown parameter (see dmesg)
> $ dmesg | tail -n 1
> [...] tcrypt: one or more tests failed!
>
> Any idea what's going on here or how I could fix it? Thanks!
> Cheers,
> Josh
>
> _______________________________________________
> Users mailing list
> Users at lists.strongswan.org
> https://lists.strongswan.org/mailman/listinfo/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.strongswan.org/pipermail/users/attachments/20150708/9d5481f7/attachment.html>


More information about the Users mailing list