From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:50429) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aMCyo-0002nB-33 for qemu-devel@nongnu.org; Thu, 21 Jan 2016 06:00:48 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1aMCym-0002S8-SV for qemu-devel@nongnu.org; Thu, 21 Jan 2016 06:00:42 -0500 Date: Thu, 21 Jan 2016 11:00:30 +0000 From: "Daniel P. Berrange" Message-ID: <20160121110030.GG19835@redhat.com> References: <1453311539-1193-1-git-send-email-berrange@redhat.com> <1453311539-1193-5-git-send-email-berrange@redhat.com> <20160121075137.GF31960@ad.usersys.redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <20160121075137.GF31960@ad.usersys.redhat.com> Subject: Re: [Qemu-devel] [PATCH v2 04/17] crypto: add support for generating initialization vectors Reply-To: "Daniel P. Berrange" List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Fam Zheng Cc: Kevin Wolf , qemu-devel@nongnu.org, qemu-block@nongnu.org On Thu, Jan 21, 2016 at 03:51:37PM +0800, Fam Zheng wrote: > On Wed, 01/20 17:38, Daniel P. Berrange wrote: > > +static int qcrypto_ivgen_essiv_init(QCryptoIVGen *ivgen, > > + const uint8_t *key, size_t nkey, > > + Error **errp) > > +{ > > + uint8_t *salt; > > + size_t nhash; > > + QCryptoIVGenESSIV *essiv = g_new0(QCryptoIVGenESSIV, 1); > > + > > + nhash = qcrypto_hash_digest_len(ivgen->hash); > > + /* Salt must be larger of hash size or key size */ > > + salt = g_new0(uint8_t, nhash > nkey ? nhash : nkey); > > + > > + if (qcrypto_hash_bytes(ivgen->hash, (const gchar *)key, nkey, > > + &salt, &nhash, > > + errp) < 0) { > > + g_free(essiv); > > + return -1; > > + } > > + > > + essiv->cipher = qcrypto_cipher_new(ivgen->cipher, > > + QCRYPTO_CIPHER_MODE_ECB, > > + salt, nkey, > > + errp); > > + if (!essiv->cipher) { > > + g_free(essiv); > > + g_free(salt); > > + return -1; > > + } > > + > > + g_free(salt); > > + ivgen->private = essiv; > > + > > + return 0; > > +} > > + > > +static int qcrypto_ivgen_essiv_calculate(QCryptoIVGen *ivgen, > > + uint64_t sector, > > + uint8_t *iv, size_t niv, > > + Error **errp) > > +{ > > + QCryptoIVGenESSIV *essiv = ivgen->private; > > + size_t ndata = qcrypto_cipher_get_block_len(ivgen->cipher); > > + uint8_t *data = g_new(uint8_t, ndata); > > + > > + sector = cpu_to_le64((uint32_t)sector); > > Why casting to 32 bit? Bugtastic. > > + memcpy(data, (uint8_t *)§or, ndata); > > + if (sizeof(sector) < ndata) { > > + memset(data + sizeof(sector), 0, ndata - sizeof(sector)); > > + } > > + > > + if (qcrypto_cipher_encrypt(essiv->cipher, > > + data, > > + data, > > + ndata, > > + errp) < 0) { > > + g_free(data); > > + return -1; > > + } > > + > > + if (ndata > niv) { > > + ndata = niv; > > + } > > + memcpy(iv, data, ndata); > > + if (ndata < niv) { > > + memset(iv + ndata, 0, niv - ndata); > > + } > > + g_free(data); > > + return 0; > > +} > > + > > +static void qcrypto_ivgen_essiv_cleanup(QCryptoIVGen *ivgen) > > +{ > > + QCryptoIVGenESSIV *essiv = ivgen->private; > > + > > + qcrypto_cipher_free(essiv->cipher); > > + g_free(essiv); > > +} > > + > > + > > +struct QCryptoIVGenDriver qcrypto_ivgen_essiv = { > > + qcrypto_ivgen_essiv_init, > > It'd be more readable to explicitly write > > .init = qcrypto_ivgen_essiv_init, > .calculate = ..., > .cleanup = ... > > but it is fine since there are only three operations now. Yeah, I normally do it that way anyway, so will change it. Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|