Bug#438792: Bug#441667: srtp: FTBFS: /bin/sh: line 1: 4402 Segmentation fault test/srtp_driver -v >/dev/null

[Philipp Kern]

On Sat, Jan 12, 2008 at 01:09:22PM +0100, Lucas Nussbaum wrote:
> srtp now builds fine on i386. So I'm closing #441667. #440665 (build
> failure on s390) and #438792 (on sparc) are still not fixed according to
> http://buildd.debian.org/pkg.cgi?pkg=srtp ...

I played a bit with the package, with the help of aurel32.  In the
version currently in the archive we have some "optimizations" which
cause unaligned accesses on many architectures.  On some data is
silently corrupted (like on arm), on some we get SIGBUS on such accesses
(like sparc).  (And there are two occurences where the result of
getopt_s gets assigned to a char instead of an int, and a wrong printf
code, but nothing not fixable.)

I am curious if the new upstream version fixes this (IMHO it should
use an aes library if possible).

Kind regards,
Philipp Kern
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://lists.alioth.debian.org/pipermail/pkg-voip-maintainers/attachments/20080113/05c970f7/attachment.pgp