> Date: Mon, 23 Aug 2010 14:09:03 +0200
quoted text
> From: Mike Belopuhov <mkb@crypt.org.ru>
> 
> On Sun, Aug 22, 2010 at 19:40 +0000, Christian Weisgerber wrote:
> > Mike Belopuhov <mkb@crypt.org.ru> wrote:
> > 
> > > This is a first diff of a series that brings support for the
> > > Galois/Counter Mode for AES in OpenBSD.
> > 
> > \o/
> > 
> > Comments below.
> > 
> 
> thanks for the feedback!
> 
> > > --- crypto/cryptosoft.c	2 Jul 2010 02:40:15 -0000	1.54
> > > +++ crypto/cryptosoft.c	20 Aug 2010 15:38:50 -0000
> > > @@ -492,6 +492,147 @@ swcr_authcompute(struct cryptop *crp, st
> > >  }
> > >  
> > >  /*
> > > + * Apply a combined encryption-authentication transformation
> > > + */
> > > +int
> > > +swcr_combined(struct cryptop *crp)
> > > +{
> > > +	unsigned char iv[EALG_MAX_BLOCK_LEN], blk[EALG_MAX_BLOCK_LEN];
> > ...
> > > +	/* Do any required special finalization */
> > > +	switch (crda->crd_alg) {
> > > +		case CRYPTO_AES_128_GMAC:
> > > +		case CRYPTO_AES_192_GMAC:
> > > +		case CRYPTO_AES_256_GMAC:
> > > +			/* length block */
> > > +			bzero(blk, blksz);
> > > +			blkp = (uint32_t *)blk + 1;
> > > +			*blkp = htobe32(crda->crd_len * 8);
> > > +			blkp = (uint32_t *)blk + 3;
> > > +			*blkp = htobe32(crde->crd_len * 8);
> > > +			axf->Update(&ctx, blk, blksz);
> > > +			break;
> > > +	}
> > 
> > panic: unaligned access
> > This is exactly the rdate/ntp.c problem again.  You can't assume
> > that a char array is int32-aligned.
> > 
> 
> true!  well, almost.  let me tell you something about GCC:
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33594

Ah, that's a good find.  I had been looking at LOCAL_ALIGNMENT before,
but wasn't sure that was the way to go.

I'm testing the diff from that bug report now on sparc64 and an
equivalent diff on hppa.