[PATCH v4 0/5] crypto: AF_ALG: add AEAD and RNG support

[PATCH v4 0/5] crypto: AF_ALG: add AEAD and RNG support

[Stephan Mueller]

Hi,

This patch set adds AEAD and RNG support to the AF_ALG interface
exported by the kernel crypto API. By extending AF_ALG with AEAD and RNG
support, all cipher types the kernel crypto API allows access to are
now accessible from userspace.

Both, AEAD and RNG implementations are stand-alone and do not depend
other AF_ALG interfaces (like hash or skcipher).

The AEAD implementation uses the same approach as provided with
skcipher by offering the following interfaces:

	* sendmsg and recvmsg interfaces allowing multiple
	  invocations supporting a threaded user space. To support
	  multi-threaded user space, kernel-side buffering
	  is implemented similarly to skcipher.

	* splice / vmsplice interfaces allowing a zero-copy
	  invocation

The RNG interface only implements the recvmsg interface as
zero-copy is not applicable.

The new AEAD and RNG interfaces are fully tested with the test application
provided at [1]. That test application exercises all newly added user space
interfaces. The testing covers:

	* use of the sendmsg/recvmsg interface

	* use of the splice / vmsplice interface

	* invocation of all AF_ALG types (aead, rng, skcipher, hash)

	* using all types of operation (encryption, decryption, keyed MD,
	  MD, random numbers, AEAD decryption with positive and negative
	  authentication verification)

	* stress testing by running all tests for 30 minutes in an
	  endless loop

	* test execution on 64 bit and 32 bit

[1] http://www.chronox.de/libkcapi.html

Changes v2:
* rebase to current cryptodev-2.6 tree
* use memzero_explicit to zeroize AEAD associated data
* use sizeof for determining length of AEAD associated data
* update algif_rng.c covering all suggestions from Daniel Borkmann
  <dborkman@redhat.com>
* addition of patch 9: add digestsize interface for hashes
* addition of patch to update documentation covering the userspace interface
* change numbers of getsockopt options: separate them from sendmsg interface
  definitions

Changes v3:
* remove getsockopt interface
* AEAD: associated data is set prepended to the plain/ciphertext
* AEAD: allowing arbitrary associated data lengths
* remove setkey patch as protection was already in the existing code

Changes v4:
* stand-alone implementation of AEAD
* testing of all interfaces offered by AEAD
* stress testing of AEAD and RNG

Stephan Mueller (5):
  crypto: AF_ALG: add user space interface for AEAD
  crypto: AF_ALG: add AEAD support
  crypto: AF_ALG: enable AEAD interface compilation
  crypto: AF_ALG: add random number generator support
  crypto: AF_ALG: enable RNG interface compilation

 crypto/Kconfig              |  18 ++
 crypto/Makefile             |   2 +
 crypto/af_alg.c             |   6 +
 crypto/algif_aead.c         | 668 ++++++++++++++++++++++++++++++++++++++++++++
 crypto/algif_rng.c          | 186 ++++++++++++
 include/crypto/if_alg.h     |   1 +
 include/uapi/linux/if_alg.h |   2 +
 7 files changed, 883 insertions(+)
 create mode 100644 crypto/algif_aead.c
 create mode 100644 crypto/algif_rng.c

-- 
2.1.0

[PATCH v4 1/5] crypto: AF_ALG: add user space interface for AEAD

[Stephan Mueller]

AEAD requires the caller to specify the following information separate
from the data stream. This information allows the AEAD interface handler
to identify the AAD, ciphertext/plaintext and the authentication tag:

        * Associated authentication data of arbitrary length and
          length

        * Length of authentication tag for encryption

Signed-off-by: Stephan Mueller <smueller@chronox.de>
---
 crypto/af_alg.c             | 6 ++++++
 include/crypto/if_alg.h     | 1 +
 include/uapi/linux/if_alg.h | 2 ++
 3 files changed, 9 insertions(+)

diff --git a/crypto/af_alg.c b/crypto/af_alg.c
index 6a3ad80..68ff113 100644
--- a/crypto/af_alg.c
+++ b/crypto/af_alg.c
@@ -421,6 +421,12 @@ int af_alg_cmsg_send(struct msghdr *msg, struct af_alg_control *con)
 			con->op = *(u32 *)CMSG_DATA(cmsg);
 			break;
 
+		case ALG_SET_AEAD_ASSOCLEN:
+			if (cmsg->cmsg_len < CMSG_LEN(sizeof(u32)))
+				return -EINVAL;
+			con->aead_assoclen = *(u32 *)CMSG_DATA(cmsg);
+			break;
+
 		default:
 			return -EINVAL;
 		}
diff --git a/include/crypto/if_alg.h b/include/crypto/if_alg.h
index d61c111..cd62bf4 100644
--- a/include/crypto/if_alg.h
+++ b/include/crypto/if_alg.h
@@ -42,6 +42,7 @@ struct af_alg_completion {
 struct af_alg_control {
 	struct af_alg_iv *iv;
 	int op;
+	unsigned int aead_assoclen;
 };
 
 struct af_alg_type {
diff --git a/include/uapi/linux/if_alg.h b/include/uapi/linux/if_alg.h
index 0f9acce..f2acd2f 100644
--- a/include/uapi/linux/if_alg.h
+++ b/include/uapi/linux/if_alg.h
@@ -32,6 +32,8 @@ struct af_alg_iv {
 #define ALG_SET_KEY			1
 #define ALG_SET_IV			2
 #define ALG_SET_OP			3
+#define ALG_SET_AEAD_ASSOCLEN		4
+#define ALG_SET_AEAD_AUTHSIZE		5
 
 /* Operations */
 #define ALG_OP_DECRYPT			0
-- 
2.1.0

Re: [PATCH v4 1/5] crypto: AF_ALG: add user space interface for AEAD

[Herbert Xu]

On Wed, Dec 03, 2014 at 08:55:42PM +0100, Stephan Mueller wrote:
> AEAD requires the caller to specify the following information separate
> from the data stream. This information allows the AEAD interface handler
> to identify the AAD, ciphertext/plaintext and the authentication tag:
> 
>         * Associated authentication data of arbitrary length and
>           length
> 
>         * Length of authentication tag for encryption
> 
> Signed-off-by: Stephan Mueller <smueller@chronox.de>

Patch applied.
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

[PATCH v4 2/5] crypto: AF_ALG: add AEAD support

[Stephan Mueller]

This patch adds the AEAD support for AF_ALG.

The implementation is based on algif_skcipher, but contains heavy
modifications to streamline the interface for AEAD uses.

To use AEAD, the user space consumer has to use the salg_type named
"aead".

The AEAD implementation includes some overhead to calculate the size of
the ciphertext, because the AEAD implementation of the kernel crypto API
makes implied assumption on the location of the authentication tag. When
performing an encryption, the tag will be added to the created
ciphertext (note, the tag is placed adjacent to the ciphertext). For
decryption, the caller must hand in the ciphertext with the tag appended
to the ciphertext. Therefore, the selection of the used memory
needs to add/subtract the tag size from the source/destination buffers
depending on the encryption type. The code is provided with comments
explaining when and how that operation is performed.

A fully working example using all aspects of AEAD is provided at
http://www.chronox.de/libkcapi.html

Signed-off-by: Stephan Mueller <smueller@chronox.de>
---
 crypto/algif_aead.c | 668 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 668 insertions(+)
 create mode 100644 crypto/algif_aead.c

diff --git a/crypto/algif_aead.c b/crypto/algif_aead.c
new file mode 100644
index 0000000..e035dcc
--- /dev/null
+++ b/crypto/algif_aead.c
@@ -0,0 +1,668 @@
+/*
+ * algif_aeadr: User-space interface for AEAD algorithms
+ *
+ * Copyright (C) 2014, Stephan Mueller <smueller@chronox.de>
+ *
+ * This file provides the user-space API for AEAD ciphers.
+ *
+ * This file is derived from algif_skcipher.c.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ *
+ */
+
+#include <crypto/scatterwalk.h>
+#include <crypto/if_alg.h>
+#include <linux/init.h>
+#include <linux/list.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/net.h>
+#include <net/sock.h>
+
+struct aead_sg_list {
+	unsigned int cur;
+	struct scatterlist sg[ALG_MAX_PAGES];
+};
+
+struct aead_ctx {
+	struct aead_sg_list tsgl;
+	struct af_alg_sgl rsgl;
+
+	void *iv;
+
+	struct af_alg_completion completion;
+
+	unsigned long used;
+
+	unsigned int len;
+	bool more;
+	bool merge;
+	bool enc;
+
+	size_t aead_assoclen;
+	struct aead_request aead_req;
+};
+
+static inline int aead_sndbuf(struct sock *sk)
+{
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+
+	return max_t(int, max_t(int, sk->sk_sndbuf & PAGE_MASK, PAGE_SIZE) -
+			  ctx->used, 0);
+}
+
+static inline bool aead_writable(struct sock *sk)
+{
+	return PAGE_SIZE <= aead_sndbuf(sk);
+}
+
+static inline bool aead_sufficient_data(struct aead_ctx *ctx)
+{
+	unsigned as = crypto_aead_authsize(crypto_aead_reqtfm(&ctx->aead_req));
+
+	return (ctx->used >= (ctx->aead_assoclen + ctx->enc ? : as ));
+}
+static inline bool aead_readable(struct aead_ctx *ctx)
+{
+	/*
+	 * Ensure that assoc data is present, the plaintext / ciphertext
+	 * is non-zero and that the authentication tag is also present
+	 * in case of a decryption operation.
+	 *
+	 * Also, wait until all data is received before processing.
+	 */
+	return (aead_sufficient_data(ctx) && !ctx->more);
+}
+
+static void aead_put_sgl(struct sock *sk)
+{
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+	struct aead_sg_list *sgl = &ctx->tsgl;
+	struct scatterlist *sg = sgl->sg;
+	unsigned int i;
+
+	for (i = 0; i < sgl->cur; i++) {
+		if (!sg_page(sg + i))
+			continue;
+
+		put_page(sg_page(sg + i));
+		sg_assign_page(sg + i, NULL);
+	}
+	sgl->cur = 0;
+	ctx->used = 0;
+	ctx->more = 0;
+	ctx->merge = 0;
+}
+
+static int aead_wait_for_wmem(struct sock *sk, unsigned flags)
+{
+	long timeout;
+	DEFINE_WAIT(wait);
+	int err = -ERESTARTSYS;
+
+	if (flags & MSG_DONTWAIT)
+		return -EAGAIN;
+
+	set_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags);
+
+	for (;;) {
+		if (signal_pending(current))
+			break;
+		prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
+		timeout = MAX_SCHEDULE_TIMEOUT;
+		if (sk_wait_event(sk, &timeout, aead_writable(sk))) {
+			err = 0;
+			break;
+		}
+	}
+	finish_wait(sk_sleep(sk), &wait);
+
+	return err;
+}
+
+static void aead_wmem_wakeup(struct sock *sk)
+{
+	struct socket_wq *wq;
+
+	if (!aead_writable(sk))
+		return;
+
+	rcu_read_lock();
+	wq = rcu_dereference(sk->sk_wq);
+	if (wq_has_sleeper(wq))
+		wake_up_interruptible_sync_poll(&wq->wait, POLLIN |
+							   POLLRDNORM |
+							   POLLRDBAND);
+	sk_wake_async(sk, SOCK_WAKE_WAITD, POLL_IN);
+	rcu_read_unlock();
+}
+
+static int aead_wait_for_data(struct sock *sk, unsigned flags)
+{
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+	long timeout;
+	DEFINE_WAIT(wait);
+	int err = -ERESTARTSYS;
+
+	if (flags & MSG_DONTWAIT) {
+		return -EAGAIN;
+	}
+
+	set_bit(SOCK_ASYNC_WAITDATA, &sk->sk_socket->flags);
+
+	for (;;) {
+		if (signal_pending(current))
+			break;
+		prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
+		timeout = MAX_SCHEDULE_TIMEOUT;
+		if (sk_wait_event(sk, &timeout, aead_readable(ctx))) {
+			err = 0;
+			break;
+		}
+	}
+	finish_wait(sk_sleep(sk), &wait);
+
+	clear_bit(SOCK_ASYNC_WAITDATA, &sk->sk_socket->flags);
+
+	return err;
+}
+
+static void aead_data_wakeup(struct sock *sk)
+{
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+	struct socket_wq *wq;
+
+	if (!aead_readable(ctx))
+		return;
+
+	rcu_read_lock();
+	wq = rcu_dereference(sk->sk_wq);
+	if (wq_has_sleeper(wq))
+		wake_up_interruptible_sync_poll(&wq->wait, POLLOUT |
+							   POLLRDNORM |
+							   POLLRDBAND);
+	sk_wake_async(sk, SOCK_WAKE_SPACE, POLL_OUT);
+	rcu_read_unlock();
+}
+
+static int aead_sendmsg(struct kiocb *unused, struct socket *sock,
+		        struct msghdr *msg, size_t size)
+{
+	struct sock *sk = sock->sk;
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+	unsigned ivsize =
+		crypto_aead_ivsize(crypto_aead_reqtfm(&ctx->aead_req));
+	struct aead_sg_list *sgl = &ctx->tsgl;
+	struct af_alg_control con = {};
+	long copied = 0;
+	bool enc = 0;
+	bool init = 0;
+	int err = -EINVAL;
+	struct scatterlist *sg = NULL;
+	int plen = 0;
+	unsigned long len = size;
+
+	if (msg->msg_controllen) {
+		err = af_alg_cmsg_send(msg, &con);
+		if (err)
+			return err;
+
+		init = 1;
+		switch (con.op) {
+		case ALG_OP_ENCRYPT:
+			enc = 1;
+			break;
+		case ALG_OP_DECRYPT:
+			enc = 0;
+			break;
+		default:
+			return -EINVAL;
+		}
+
+		if (con.iv && con.iv->ivlen != ivsize)
+			return -EINVAL;
+
+		if (!con.aead_assoclen)
+			return -EINVAL;
+
+		/* aead_recvmsg limits the maximum AD size to one page */
+		if (con.aead_assoclen > PAGE_SIZE)
+			return -E2BIG;
+	}
+
+	lock_sock(sk);
+	if (!ctx->more && ctx->used)
+		goto unlock;
+
+	if (init) {
+		ctx->enc = enc;
+		if (con.iv)
+			memcpy(ctx->iv, con.iv->iv, ivsize);
+
+		ctx->aead_assoclen = con.aead_assoclen;
+	}
+
+	if (ctx->merge) {
+		sg = sgl->sg + sgl->cur - 1;
+		len = min_t(unsigned long, len,
+			    PAGE_SIZE - sg->offset - sg->length);
+
+		err = memcpy_fromiovec(page_address(sg_page(sg)) +
+				       sg->offset + sg->length,
+				       msg->msg_iov, len);
+		if (err)
+			goto unlock;
+
+		sg->length += len;
+		ctx->merge = (sg->offset + sg->length) & (PAGE_SIZE - 1);
+
+		ctx->used += len;
+		copied += len;
+		size -= len;
+	}
+
+	if (!aead_writable(sk)) {
+		/*
+		 * If there is more data to be expected, but we cannot write
+		 * more data, forcefully define that we do not expect more
+		 * data to invoke the AEAD operation. This prevents a deadlock
+		 * in user space.
+		 */
+		ctx->more = 0;
+		err = aead_wait_for_wmem(sk, msg->msg_flags);
+		if (err)
+			goto unlock;
+	}
+
+	len = min_t(unsigned long, size, aead_sndbuf(sk));
+	sg = sgl->sg;
+	while (len && sgl->cur < ALG_MAX_PAGES) {
+		sg = sgl->sg + sgl->cur;
+		plen = min_t(int, len, PAGE_SIZE);
+
+		if (sgl->cur >= ALG_MAX_PAGES) {
+			err = -E2BIG;
+			goto unlock;
+		}
+
+		sg_assign_page(sg, alloc_page(GFP_KERNEL));
+		err = -ENOMEM;
+		if (!sg_page(sg))
+			goto unlock;
+
+		err = memcpy_fromiovec(page_address(sg_page(sg)),
+				       msg->msg_iov, plen);
+		if (err) {
+			__free_page(sg_page(sg));
+			sg_assign_page(sg, NULL);
+			goto unlock;
+		}
+
+		sg->length = plen;
+		len -= plen;
+		ctx->used += plen;
+		copied += plen;
+		sgl->cur++;
+		size -= len;
+		ctx->merge = plen & (PAGE_SIZE - 1);
+	}
+
+	err = 0;
+
+	ctx->more = msg->msg_flags & MSG_MORE;
+
+unlock:
+	aead_data_wakeup(sk);
+	release_sock(sk);
+
+	return copied ?: err;
+}
+
+static ssize_t aead_sendpage(struct socket *sock, struct page *page,
+			     int offset, size_t size, int flags)
+{
+	struct sock *sk = sock->sk;
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+	struct aead_sg_list *sgl = &ctx->tsgl;
+	int err = -EINVAL;
+
+	if (flags & MSG_SENDPAGE_NOTLAST)
+		flags |= MSG_MORE;
+
+	if (sgl->cur >= ALG_MAX_PAGES)
+		return -E2BIG;
+
+	lock_sock(sk);
+	if (!ctx->more && ctx->used)
+		goto unlock;
+
+	if (!size)
+		goto done;
+
+	if (!aead_writable(sk)) {
+		err = aead_wait_for_wmem(sk, flags);
+		if (err)
+			goto unlock;
+	}
+
+	ctx->merge = 0;
+
+	get_page(page);
+	sg_set_page(sgl->sg + sgl->cur, page, size, offset);
+	sgl->cur++;
+	ctx->used += size;
+
+	err = 0;
+
+done:
+	ctx->more = flags & MSG_MORE;
+
+unlock:
+	aead_data_wakeup(sk);
+	release_sock(sk);
+
+	return err ?: size;
+}
+
+static int aead_recvmsg(struct kiocb *unused, struct socket *sock,
+			    struct msghdr *msg, size_t ignored, int flags)
+{
+	struct sock *sk = sock->sk;
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+	unsigned bs = crypto_aead_blocksize(crypto_aead_reqtfm(&ctx->aead_req));
+	unsigned as = crypto_aead_authsize(crypto_aead_reqtfm(&ctx->aead_req));
+	struct aead_sg_list *sgl = &ctx->tsgl;
+	struct scatterlist *sg = sgl->sg;
+	struct scatterlist assoc;
+	size_t assoclen = 0;
+	unsigned int i = 0;
+	int err = -EAGAIN;
+	unsigned long used = 0;
+	unsigned long outlen = 0;
+
+	/*
+	 * Require exactly one IOV block as the AEAD operation is a one shot
+	 * due to the authentication tag.
+	 */
+	if (msg->msg_iovlen != 1)
+		return -ENOMSG;
+
+	lock_sock(sk);
+	/*
+	* AEAD memory structure: For encryption, the tag is appended to the
+	* ciphertext which implies that the memory allocated for the ciphertext
+	* must be increased by the tag length. For decryption, the tag
+	* is expected to be concatenated to the ciphertext. The plaintext
+	* therefore has a memory size of the ciphertext minus the tag length.
+	*
+	* The memory structure for cipher operation has the following
+	* structure:
+	*	AEAD encryption input:  assoc data || plaintext
+	*	AEAD encryption output: cipherntext || auth tag
+	*	AEAD decryption input:  assoc data || ciphertext || auth tag
+	*	AEAD decryption output: plaintext
+	*/
+
+	if (!aead_readable(ctx)) {
+		err = aead_wait_for_data(sk, flags);
+		if (err)
+			goto unlock;
+	}
+
+	used = ctx->used;
+
+	err = -ENOMEM;
+	if (!aead_sufficient_data(ctx))
+		goto unlock;
+	/*
+	 * The cipher operation input data is reduced by the associated data
+	 * length as this data is processed separately later on.
+	 */
+	used -= ctx->aead_assoclen;
+
+	if (ctx->enc) {
+		/* round up output buffer to multiple of block size */
+		outlen = ((used + bs - 1) / bs * bs);
+		/* add the size needed for the auth tag to be created */
+		outlen += as;
+	} else {
+		/* output data size is input without the authentication tag */
+		outlen = used - as;
+		/* round up output buffer to multiple of block size */
+		outlen = ((outlen + bs - 1) / bs * bs);
+	}
+
+	/* ensure output buffer is sufficiently large */
+	if (msg->msg_iov->iov_len < outlen)
+		goto unlock;
+
+	outlen = af_alg_make_sg(&ctx->rsgl, msg->msg_iov->iov_base, outlen, 1);
+	err = outlen;
+	if (err < 0)
+		goto unlock;
+
+	err = -EINVAL;
+	/*
+	 * first chunk of input is AD -- one scatterlist entry is one page,
+	 * and we process only one scatterlist, the maximum size of AD is
+	 * one page
+	 */
+	sg_init_table(&assoc, 1);
+	sg_set_page(&assoc, sg_page(sg), ctx->aead_assoclen, sg->offset);
+	aead_request_set_assoc(&ctx->aead_req, &assoc, ctx->aead_assoclen);
+
+	/* point sg to cipher/plaintext start */
+	assoclen = ctx->aead_assoclen;
+	for(i = 0; i < ctx->tsgl.cur; i++) {
+		sg = sgl->sg + i;
+		if (sg->length <= assoclen) {
+			assoclen -= sg->length;
+			if (i >= ctx->tsgl.cur)
+				goto unlock;
+		} else {
+			sg->length -= assoclen;
+			sg->offset += assoclen;
+			break;
+		}
+	}
+
+	aead_request_set_crypt(&ctx->aead_req, sg, ctx->rsgl.sg, used, ctx->iv);
+
+	err = af_alg_wait_for_completion(ctx->enc ?
+					 crypto_aead_encrypt(&ctx->aead_req) :
+					 crypto_aead_decrypt(&ctx->aead_req),
+					 &ctx->completion);
+
+	af_alg_free_sg(&ctx->rsgl);
+
+	if (err)
+		goto unlock;
+
+	aead_put_sgl(sk);
+
+	err = 0;
+
+unlock:
+	aead_wmem_wakeup(sk);
+	release_sock(sk);
+
+	return err ? err : outlen;
+}
+
+static int aead_setsockopt(struct socket *sock, int level, int optname,
+			   char __user *optval, unsigned int optlen)
+{
+	struct sock *sk = sock->sk;
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+	const struct af_alg_type *type;
+	int err = -ENOPROTOOPT;
+
+	lock_sock(sk);
+	type = ask->type;
+
+	if (level != SOL_ALG || !type)
+		goto unlock;
+
+	switch (optname) {
+	case ALG_SET_AEAD_AUTHSIZE:
+		err = crypto_aead_setauthsize(
+				crypto_aead_reqtfm(&ctx->aead_req), optlen);
+		if (err)
+			goto unlock;
+	}
+
+unlock:
+	release_sock(sk);
+
+	return err;
+}
+
+static unsigned int aead_poll(struct file *file, struct socket *sock,
+				  poll_table *wait)
+{
+	struct sock *sk = sock->sk;
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+	unsigned int mask;
+
+	sock_poll_wait(file, sk_sleep(sk), wait);
+	mask = 0;
+
+	if (aead_readable(ctx))
+		mask |= POLLIN | POLLRDNORM;
+
+	if (aead_writable(sk))
+		mask |= POLLOUT | POLLWRNORM | POLLWRBAND;
+
+	return mask;
+}
+
+static struct proto_ops algif_aead_ops = {
+	.family		=	PF_ALG,
+
+	.connect	=	sock_no_connect,
+	.socketpair	=	sock_no_socketpair,
+	.getname	=	sock_no_getname,
+	.ioctl		=	sock_no_ioctl,
+	.listen		=	sock_no_listen,
+	.shutdown	=	sock_no_shutdown,
+	.getsockopt	=	sock_no_getsockopt,
+	.mmap		=	sock_no_mmap,
+	.bind		=	sock_no_bind,
+	.accept		=	sock_no_accept,
+
+	.release	=	af_alg_release,
+	.sendmsg	=	aead_sendmsg,
+	.sendpage	=	aead_sendpage,
+	.recvmsg	=	aead_recvmsg,
+	.poll		=	aead_poll,
+	.setsockopt	=	aead_setsockopt,
+};
+
+static void *aead_bind(const char *name, u32 type, u32 mask)
+{
+	return crypto_alloc_aead(name, type, mask);
+}
+
+static void aead_release(void *private)
+{
+	crypto_free_aead(private);
+}
+
+static int aead_setkey(void *private, const u8 *key, unsigned int keylen)
+{
+	return crypto_aead_setkey(private, key, keylen);
+}
+
+static void aead_sock_destruct(struct sock *sk)
+{
+	struct alg_sock *ask = alg_sk(sk);
+	struct aead_ctx *ctx = ask->private;
+	unsigned int ivlen = crypto_aead_ivsize(
+				crypto_aead_reqtfm(&ctx->aead_req));
+
+	aead_put_sgl(sk);
+	sock_kzfree_s(sk, ctx->iv, ivlen);
+	sock_kfree_s(sk, ctx, ctx->len);
+	af_alg_release_parent(sk);
+}
+
+static int aead_accept_parent(void *private, struct sock *sk)
+{
+	struct aead_ctx *ctx;
+	struct alg_sock *ask = alg_sk(sk);
+	unsigned int len = sizeof(*ctx) + crypto_aead_reqsize(private);
+	unsigned int ivlen = crypto_aead_ivsize(private);
+
+	ctx = sock_kmalloc(sk, len, GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+	memset(ctx, 0, len);
+
+	ctx->iv = sock_kmalloc(sk, ivlen, GFP_KERNEL);
+	if (!ctx->iv) {
+		sock_kfree_s(sk, ctx, len);
+		return -ENOMEM;
+	}
+	memset(ctx->iv, 0, ivlen);
+
+	ctx->len = len;
+	ctx->used = 0;
+	ctx->more = 0;
+	ctx->merge = 0;
+	ctx->enc = 0;
+	ctx->tsgl.cur = 0;
+	ctx->aead_assoclen = 0;
+	af_alg_init_completion(&ctx->completion);
+	sg_init_table(ctx->tsgl.sg, ALG_MAX_PAGES);
+
+	ask->private = ctx;
+
+	aead_request_set_tfm(&ctx->aead_req, private);
+	aead_request_set_callback(&ctx->aead_req, CRYPTO_TFM_REQ_MAY_BACKLOG,
+				  af_alg_complete, &ctx->completion);
+
+	sk->sk_destruct = aead_sock_destruct;
+
+	return 0;
+}
+
+static const struct af_alg_type algif_type_aead = {
+	.bind		=	aead_bind,
+	.release	=	aead_release,
+	.setkey		=	aead_setkey,
+	.accept		=	aead_accept_parent,
+	.ops		=	&algif_aead_ops,
+	.name		=	"aead",
+	.owner		=	THIS_MODULE
+};
+
+static int __init algif_aead_init(void)
+{
+	return af_alg_register_type(&algif_type_aead);
+}
+
+static void __exit algif_aead_exit(void)
+{
+	int err = af_alg_unregister_type(&algif_type_aead);
+	BUG_ON(err);
+}
+
+module_init(algif_aead_init);
+module_exit(algif_aead_exit);
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Stephan Mueller <smueller@chronox.de>");
+MODULE_DESCRIPTION("AEAD kernel crypto API user space interface");
-- 
2.1.0

Re: [PATCH v4 2/5] crypto: AF_ALG: add AEAD support

[Herbert Xu]

On Wed, Dec 03, 2014 at 08:57:24PM +0100, Stephan Mueller wrote:
>
> +	if (ctx->merge) {
> +		sg = sgl->sg + sgl->cur - 1;
> +		len = min_t(unsigned long, len,
> +			    PAGE_SIZE - sg->offset - sg->length);
> +
> +		err = memcpy_fromiovec(page_address(sg_page(sg)) +
> +				       sg->offset + sg->length,
> +				       msg->msg_iov, len);
> +		if (err)
> +			goto unlock;
> +
> +		sg->length += len;
> +		ctx->merge = (sg->offset + sg->length) & (PAGE_SIZE - 1);
> +
> +		ctx->used += len;
> +		copied += len;
> +		size -= len;
> +	}

Any reason why you got rid of the outer loop here? This will cause
short writes I think.

> +static struct proto_ops algif_aead_ops = {
> +	.family		=	PF_ALG,
> +
> +	.connect	=	sock_no_connect,
> +	.socketpair	=	sock_no_socketpair,
> +	.getname	=	sock_no_getname,
> +	.ioctl		=	sock_no_ioctl,
> +	.listen		=	sock_no_listen,
> +	.shutdown	=	sock_no_shutdown,
> +	.getsockopt	=	sock_no_getsockopt,
> +	.mmap		=	sock_no_mmap,
> +	.bind		=	sock_no_bind,
> +	.accept		=	sock_no_accept,
> +
> +	.release	=	af_alg_release,
> +	.sendmsg	=	aead_sendmsg,
> +	.sendpage	=	aead_sendpage,
> +	.recvmsg	=	aead_recvmsg,
> +	.poll		=	aead_poll,
> +	.setsockopt	=	aead_setsockopt,

No it should go into the parent setsockopt.  Perhaps add a setsockopt
to af_alg_type in order to keep this out of the generic code.

Thanks,
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: [PATCH v4 2/5] crypto: AF_ALG: add AEAD support

[Stephan Mueller]

Am Freitag, 5. Dezember 2014, 23:46:06 schrieb Herbert Xu:

Hi Herbert,

> On Wed, Dec 03, 2014 at 08:57:24PM +0100, Stephan Mueller wrote:
> > +	if (ctx->merge) {
> > +		sg = sgl->sg + sgl->cur - 1;
> > +		len = min_t(unsigned long, len,
> > +			    PAGE_SIZE - sg->offset - sg->length);
> > +
> > +		err = memcpy_fromiovec(page_address(sg_page(sg)) +
> > +				       sg->offset + sg->length,
> > +				       msg->msg_iov, len);
> > +		if (err)
> > +			goto unlock;
> > +
> > +		sg->length += len;
> > +		ctx->merge = (sg->offset + sg->length) & (PAGE_SIZE - 1);
> > +
> > +		ctx->used += len;
> > +		copied += len;
> > +		size -= len;
> > +	}
> 
> Any reason why you got rid of the outer loop here? This will cause
> short writes I think.

You are absolutely right. I removed it as I do not have the multiple sgl 
entries. But now as you mentioned it, I still need it if size > 
aead_sndbuf(sk).

This will be fixed in the next installment.
> 
> > +static struct proto_ops algif_aead_ops = {
> > +	.family		=	PF_ALG,
> > +
> > +	.connect	=	sock_no_connect,
> > +	.socketpair	=	sock_no_socketpair,
> > +	.getname	=	sock_no_getname,
> > +	.ioctl		=	sock_no_ioctl,
> > +	.listen		=	sock_no_listen,
> > +	.shutdown	=	sock_no_shutdown,
> > +	.getsockopt	=	sock_no_getsockopt,
> > +	.mmap		=	sock_no_mmap,
> > +	.bind		=	sock_no_bind,
> > +	.accept		=	sock_no_accept,
> > +
> > +	.release	=	af_alg_release,
> > +	.sendmsg	=	aead_sendmsg,
> > +	.sendpage	=	aead_sendpage,
> > +	.recvmsg	=	aead_recvmsg,
> > +	.poll		=	aead_poll,
> > +	.setsockopt	=	aead_setsockopt,
> 
> No it should go into the parent setsockopt.  Perhaps add a setsockopt
> to af_alg_type in order to keep this out of the generic code.

I was thinking about that for quite a while. My thought for the current 
approach was that the actual cipher operation happens in the child FD (i.e. 
after accept). AAD is delivered to that FD. Therefore, I thought that the size 
of the AAD can be specific to that operational FD.

If we move it to the parent setsockopt, all child FDs have the same AAD size. 
If you think that this is the right course of action, I can surely implement 
that.

Would you please be so kind and help me understand when some operations are 
intended for the parent FD and when for the child FD?

Thanks
-- 
Ciao
Stephan

Re: [PATCH v4 2/5] crypto: AF_ALG: add AEAD support

[Stephan Mueller]

Am Freitag, 5. Dezember 2014, 22:51:51 schrieb Stephan Mueller:

Hi Herbert,

> > > +static struct proto_ops algif_aead_ops = {
> > > +	.family		=	PF_ALG,
> > > +
> > > +	.connect	=	sock_no_connect,
> > > +	.socketpair	=	sock_no_socketpair,
> > > +	.getname	=	sock_no_getname,
> > > +	.ioctl		=	sock_no_ioctl,
> > > +	.listen		=	sock_no_listen,
> > > +	.shutdown	=	sock_no_shutdown,
> > > +	.getsockopt	=	sock_no_getsockopt,
> > > +	.mmap		=	sock_no_mmap,
> > > +	.bind		=	sock_no_bind,
> > > +	.accept		=	sock_no_accept,
> > > +
> > > +	.release	=	af_alg_release,
> > > +	.sendmsg	=	aead_sendmsg,
> > > +	.sendpage	=	aead_sendpage,
> > > +	.recvmsg	=	aead_recvmsg,
> > > +	.poll		=	aead_poll,
> > > +	.setsockopt	=	aead_setsockopt,
> > 
> > No it should go into the parent setsockopt.  Perhaps add a setsockopt
> > to af_alg_type in order to keep this out of the generic code.
> 
> I was thinking about that for quite a while. My thought for the current
> approach was that the actual cipher operation happens in the child FD (i.e.
> after accept). AAD is delivered to that FD. Therefore, I thought that the
> size of the AAD can be specific to that operational FD.
> 
> If we move it to the parent setsockopt, all child FDs have the same AAD
> size. If you think that this is the right course of action, I can surely
> implement that.
> 
> Would you please be so kind and help me understand when some operations are
> intended for the parent FD and when for the child FD?

While implementing that request, I thought about setting the auth size as part 
of the msg control in sendmsg instead of setsockopt. This would save us a 
system call and thus CPU cycles.

Thus may I propose that instead of using setsockopt,  ALG_SET_AEAD_AUTHSIZE 
should be added to af_alg_cmsg_send the same way as ALG_SET_AEAD_ASSOCLEN.

Along that line, wouldn't it make sense to get rid of the setsockopt in 
general and move even setting the key into the msg control and thus into 
af_alg_cmsg_send? I understand that this is a change to the user space API.

Thanks

-- 
Ciao
Stephan

Re: [PATCH v4 2/5] crypto: AF_ALG: add AEAD support

[Herbert Xu]

On Sat, Dec 06, 2014 at 09:08:54PM +0100, Stephan Mueller wrote:
>
> While implementing that request, I thought about setting the auth size as part 
> of the msg control in sendmsg instead of setsockopt. This would save us a 
> system call and thus CPU cycles.

No, this is a tfm property and therefore you can't set them on the
child sockets sicne two child sockets would share the same tfm.

Cheers,
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: [PATCH v4 2/5] crypto: AF_ALG: add AEAD support

[Herbert Xu]

On Fri, Dec 05, 2014 at 10:51:51PM +0100, Stephan Mueller wrote:
>
> Would you please be so kind and help me understand when some operations are 
> intended for the parent FD and when for the child FD?

If it changes the tfm it goes in the parent FD otherwise it's the
child FD.

Cheers,
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: [PATCH v4 2/5] crypto: AF_ALG: add AEAD support

[Stephan Mueller]

Am Freitag, 5. Dezember 2014, 23:46:06 schrieb Herbert Xu:

Hi Herbert,

> > +static struct proto_ops algif_aead_ops = {
> > +	.family		=	PF_ALG,
> > +
> > +	.connect	=	sock_no_connect,
> > +	.socketpair	=	sock_no_socketpair,
> > +	.getname	=	sock_no_getname,
> > +	.ioctl		=	sock_no_ioctl,
> > +	.listen		=	sock_no_listen,
> > +	.shutdown	=	sock_no_shutdown,
> > +	.getsockopt	=	sock_no_getsockopt,
> > +	.mmap		=	sock_no_mmap,
> > +	.bind		=	sock_no_bind,
> > +	.accept		=	sock_no_accept,
> > +
> > +	.release	=	af_alg_release,
> > +	.sendmsg	=	aead_sendmsg,
> > +	.sendpage	=	aead_sendpage,
> > +	.recvmsg	=	aead_recvmsg,
> > +	.poll		=	aead_poll,
> > +	.setsockopt	=	aead_setsockopt,
> 
> No it should go into the parent setsockopt.  Perhaps add a setsockopt
> to af_alg_type in order to keep this out of the generic code.

What about adding a setauthsize to af_alg_type that is called from the parent 
setsockopt? I would like to keep all user space interfaces as close together 
as possible. With a new setsockopt callback, I fear that user space interface 
handling may be scattered around.

Currently I am simply thinking about the following:

partent setsockopt:

        case ALG_SET_AEAD_AUTHSIZE:
                if (sock->state == SS_CONNECTED)
                        goto unlock;
                if (!type->setauthsize)
                        goto unlock;
                err = type->setauthsize(ask->private, optlen);

af_alg_type:

int (*setauthsize)(void *private, unsigned int authsize);

-- 
Ciao
Stephan

[PATCH v4 3/5] crypto: AF_ALG: enable AEAD interface compilation

[Stephan Mueller]

Enable compilation of the AEAD AF_ALG support and provide a Kconfig
option to compile the AEAD AF_ALG support.

Signed-off-by: Stephan Mueller <smueller@chronox.de>
---
 crypto/Kconfig  | 9 +++++++++
 crypto/Makefile | 1 +
 2 files changed, 10 insertions(+)

diff --git a/crypto/Kconfig b/crypto/Kconfig
index 87bbc9c..0b2affc 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -1505,6 +1505,15 @@ config CRYPTO_USER_API_SKCIPHER
 	  This option enables the user-spaces interface for symmetric
 	  key cipher algorithms.
 
+config CRYPTO_USER_API_AEAD
+	tristate "User-space interface for AEAD cipher algorithms"
+	depends on NET
+	select CRYPTO_AEAD
+	select CRYPTO_USER_API
+	help
+	  This option enables the user-spaces interface for AEAD
+	  cipher algorithms.
+
 config CRYPTO_HASH_INFO
 	bool
 
diff --git a/crypto/Makefile b/crypto/Makefile
index 1445b91..593fd3c 100644
--- a/crypto/Makefile
+++ b/crypto/Makefile
@@ -99,6 +99,7 @@ obj-$(CONFIG_CRYPTO_GHASH) += ghash-generic.o
 obj-$(CONFIG_CRYPTO_USER_API) += af_alg.o
 obj-$(CONFIG_CRYPTO_USER_API_HASH) += algif_hash.o
 obj-$(CONFIG_CRYPTO_USER_API_SKCIPHER) += algif_skcipher.o
+obj-$(CONFIG_CRYPTO_USER_API_AEAD) += algif_aead.o
 
 #
 # generic algorithms and the async_tx api
-- 
2.1.0

[PATCH v4 4/5] crypto: AF_ALG: add random number generator support

[Stephan Mueller]

This patch adds the random number generator support for AF_ALG.

A random number generator's purpose is to generate data without
requiring the caller to provide any data. Therefore, the AF_ALG
interface handler for RNGs only implements a callback handler for
recvmsg.

The following parameters provided with a recvmsg are processed by the
RNG callback handler:

	* sock - to resolve the RNG context data structure accessing the
	  RNG instance private to the socket

	* len - this parameter allows userspace callers to specify how
	  many random bytes the RNG shall produce and return. As the
	  kernel context for the RNG allocates a buffer of 128 bytes to
	  store random numbers before copying them to userspace, the len
	  parameter is checked that it is not larger than 128. If a
	  caller wants more random numbers, a new request for recvmsg
	  shall be made.

The size of 128 bytes is chose because of the following considerations:

	* to increase the memory footprint of the kernel too much (note,
	  that would be 128 bytes per open socket)

	* 128 is divisible by any typical cryptographic block size an
	  RNG may have

	* A request for random numbers typically only shall supply small
	  amount of data like for keys or IVs that should only require
	  one invocation of the recvmsg function.

Note, during instantiation of the RNG, the code checks whether the RNG
implementation requires seeding. If so, the RNG is seeded with output
from get_random_bytes.

A fully working example using all aspects of the RNG interface is
provided at http://www.chronox.de/libkcapi.html

Signed-off-by: Stephan Mueller <smueller@chronox.de>
---
 crypto/algif_rng.c | 186 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 186 insertions(+)
 create mode 100644 crypto/algif_rng.c

diff --git a/crypto/algif_rng.c b/crypto/algif_rng.c
new file mode 100644
index 0000000..fc25869
--- /dev/null
+++ b/crypto/algif_rng.c
@@ -0,0 +1,186 @@
+/*
+ * algif_rng: User-space interface for random number generators
+ *
+ * This file provides the user-space API for random number generators.
+ *
+ * Copyright (C) 2014, Stephan Mueller <smueller@chronox.de>
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, and the entire permission notice in its entirety,
+ *    including the disclaimer of warranties.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. The name of the author may not be used to endorse or promote
+ *    products derived from this software without specific prior
+ *    written permission.
+ *
+ * ALTERNATIVELY, this product may be distributed under the terms of
+ * the GNU General Public License, in which case the provisions of the GPL2
+ * are required INSTEAD OF the above restrictions.  (This clause is
+ * necessary due to a potential bad interaction between the GPL and
+ * the restrictions contained in a BSD-style copyright.)
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
+ * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, ALL OF
+ * WHICH ARE HEREBY DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT
+ * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
+ * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
+ * USE OF THIS SOFTWARE, EVEN IF NOT ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ */
+
+#include <linux/module.h>
+#include <crypto/rng.h>
+#include <linux/random.h>
+#include <crypto/if_alg.h>
+#include <linux/net.h>
+#include <net/sock.h>
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Stephan Mueller <smueller@chronox.de>");
+MODULE_DESCRIPTION("User-space interface for random number generators");
+
+struct rng_ctx {
+#define MAXSIZE 128
+	u8 result[MAXSIZE];
+	unsigned int len;
+	struct crypto_rng *drng;
+};
+
+static int rng_recvmsg(struct kiocb *unused, struct socket *sock,
+		       struct msghdr *msg, size_t len, int flags)
+{
+	struct sock *sk = sock->sk;
+	struct alg_sock *ask = alg_sk(sk);
+	struct rng_ctx *ctx = ask->private;
+	int err = -EFAULT;
+
+	if (len == 0)
+		return 0;
+	if (len > MAXSIZE)
+		len = MAXSIZE;
+
+	lock_sock(sk);
+	len = crypto_rng_get_bytes(ctx->drng, ctx->result, len);
+	if (len < 0)
+		goto unlock;
+
+	err = memcpy_toiovec(msg->msg_iov, ctx->result, len);
+	memzero_explicit(ctx->result, len);
+
+unlock:
+	release_sock(sk);
+
+	return err ? err : len;
+}
+
+static struct proto_ops algif_rng_ops = {
+	.family		=	PF_ALG,
+
+	.connect	=	sock_no_connect,
+	.socketpair	=	sock_no_socketpair,
+	.getname	=	sock_no_getname,
+	.ioctl		=	sock_no_ioctl,
+	.listen		=	sock_no_listen,
+	.shutdown	=	sock_no_shutdown,
+	.getsockopt	=	sock_no_getsockopt,
+	.mmap		=	sock_no_mmap,
+	.bind		=	sock_no_bind,
+	.accept		=	sock_no_accept,
+	.setsockopt	=	sock_no_setsockopt,
+	.poll		=	sock_no_poll,
+	.sendmsg	=	sock_no_sendmsg,
+	.sendpage	=	sock_no_sendpage,
+
+	.release	=	af_alg_release,
+	.recvmsg	=	rng_recvmsg,
+};
+
+static void *rng_bind(const char *name, u32 type, u32 mask)
+{
+	return crypto_alloc_rng(name, type, mask);
+}
+
+static void rng_release(void *private)
+{
+	crypto_free_rng(private);
+}
+
+static void rng_sock_destruct(struct sock *sk)
+{
+	struct alg_sock *ask = alg_sk(sk);
+	struct rng_ctx *ctx = ask->private;
+
+	memzero_explicit(ctx->result, sizeof(ctx->result));
+	sock_kfree_s(sk, ctx, ctx->len);
+	af_alg_release_parent(sk);
+}
+
+static int rng_accept_parent(void *private, struct sock *sk)
+{
+	struct rng_ctx *ctx;
+	struct alg_sock *ask = alg_sk(sk);
+	unsigned int len = sizeof(*ctx);
+	int seedsize = crypto_rng_seedsize(private);
+	int ret = -ENOMEM;
+
+	ctx = sock_kmalloc(sk, len, GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+	memset(ctx->result, 0, sizeof(ctx->result));
+
+	ctx->len = len;
+
+	if (seedsize) {
+		u8 *buf = kmalloc(seedsize, GFP_KERNEL);
+		if (!buf)
+			goto err;
+		get_random_bytes(buf, seedsize);
+		ret = crypto_rng_reset(private, buf, len);
+		kzfree(buf);
+		if (ret)
+			goto err;
+	}
+
+	ctx->drng = private;
+	ask->private = ctx;
+	sk->sk_destruct = rng_sock_destruct;
+
+	return 0;
+
+err:
+	sock_kfree_s(sk, ctx, len);
+	return ret;
+}
+
+static const struct af_alg_type algif_type_rng = {
+	.bind		=	rng_bind,
+	.release	=	rng_release,
+	.accept		=	rng_accept_parent,
+	.ops		=	&algif_rng_ops,
+	.name		=	"rng",
+	.owner		=	THIS_MODULE
+};
+
+static int __init rng_init(void)
+{
+	return af_alg_register_type(&algif_type_rng);
+}
+
+void __exit rng_exit(void)
+{
+	int err = af_alg_unregister_type(&algif_type_rng);
+	BUG_ON(err);
+}
+
+module_init(rng_init);
+module_exit(rng_exit);
-- 
2.1.0

Re: [PATCH v4 4/5] crypto: AF_ALG: add random number generator support

[Herbert Xu]

On Wed, Dec 03, 2014 at 08:59:01PM +0100, Stephan Mueller wrote:
>
> +static int rng_recvmsg(struct kiocb *unused, struct socket *sock,
> +		       struct msghdr *msg, size_t len, int flags)
> +{
> +	struct sock *sk = sock->sk;
> +	struct alg_sock *ask = alg_sk(sk);
> +	struct rng_ctx *ctx = ask->private;
> +	int err = -EFAULT;
> +
> +	if (len == 0)
> +		return 0;
> +	if (len > MAXSIZE)
> +		len = MAXSIZE;
> +
> +	lock_sock(sk);

This lock simply protects ctx->result.  Since you're using a
tiny buffer why not just put it on the stack?

> +		u8 *buf = kmalloc(seedsize, GFP_KERNEL);
> +		if (!buf)
> +			goto err;
> +		get_random_bytes(buf, seedsize);
> +		ret = crypto_rng_reset(private, buf, len);

I think you should leave the seeding and the seed to the user.
Perhaps do it through setsockopt (on the parent socket).

Cheers,
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: [PATCH v4 4/5] crypto: AF_ALG: add random number generator support

[Stephan Mueller]

Am Freitag, 5. Dezember 2014, 23:53:59 schrieb Herbert Xu:

Hi Herbert,

> On Wed, Dec 03, 2014 at 08:59:01PM +0100, Stephan Mueller wrote:
> > +static int rng_recvmsg(struct kiocb *unused, struct socket *sock,
> > +		       struct msghdr *msg, size_t len, int flags)
> > +{
> > +	struct sock *sk = sock->sk;
> > +	struct alg_sock *ask = alg_sk(sk);
> > +	struct rng_ctx *ctx = ask->private;
> > +	int err = -EFAULT;
> > +
> > +	if (len == 0)
> > +		return 0;
> > +	if (len > MAXSIZE)
> > +		len = MAXSIZE;
> > +
> > +	lock_sock(sk);
> 
> This lock simply protects ctx->result.  Since you're using a
> tiny buffer why not just put it on the stack?

When I developed the DRBG code, I got comments that 128 byte variables shall 
not be on the stack in kernel code.

But if you agree that I can put a 128 byte variable on the stack, I will see 
it done.

> 
> > +		u8 *buf = kmalloc(seedsize, GFP_KERNEL);
> > +		if (!buf)
> > +			goto err;
> > +		get_random_bytes(buf, seedsize);
> > +		ret = crypto_rng_reset(private, buf, len);
> 
> I think you should leave the seeding and the seed to the user.
> Perhaps do it through setsockopt (on the parent socket).

Sure. But please note that the seeding happens only when seedsize > 0. Such 
seeding therefore is not performed for krng, and the DRBG because both seed 
automatically.

Therefore, may I propose the following: We offer a setsockopt for (re)seeding. 
For all RNGs with seedsize > 0, we return EAGAIN for recvmsg until a 
setsockopt for at least seedsize is provided. That would imply that krng and 
DRBG would be usable without seeding from user space.

-- 
Ciao
Stephan

[PATCH v4 5/5] crypto: AF_ALG: enable RNG interface compilation

[Stephan Mueller]

Enable compilation of the RNG AF_ALG support and provide a Kconfig
option to compile the RNG AF_ALG support.

Signed-off-by: Stephan Mueller <smueller@chronox.de>
---
 crypto/Kconfig  | 9 +++++++++
 crypto/Makefile | 1 +
 2 files changed, 10 insertions(+)

diff --git a/crypto/Kconfig b/crypto/Kconfig
index 0b2affc..8e968dc 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -1514,6 +1514,15 @@ config CRYPTO_USER_API_AEAD
 	  This option enables the user-spaces interface for AEAD
 	  cipher algorithms.
 
+config CRYPTO_USER_API_RNG
+	tristate "User-space interface for random number generator algorithms"
+	depends on NET
+	select CRYPTO_RNG
+	select CRYPTO_USER_API
+	help
+	  This option enables the user-spaces interface for random
+	  number generator algorithms.
+
 config CRYPTO_HASH_INFO
 	bool
 
diff --git a/crypto/Makefile b/crypto/Makefile
index 593fd3c..c109df5 100644
--- a/crypto/Makefile
+++ b/crypto/Makefile
@@ -100,6 +100,7 @@ obj-$(CONFIG_CRYPTO_USER_API) += af_alg.o
 obj-$(CONFIG_CRYPTO_USER_API_HASH) += algif_hash.o
 obj-$(CONFIG_CRYPTO_USER_API_SKCIPHER) += algif_skcipher.o
 obj-$(CONFIG_CRYPTO_USER_API_AEAD) += algif_aead.o
+obj-$(CONFIG_CRYPTO_USER_API_RNG) += algif_rng.o
 
 #
 # generic algorithms and the async_tx api
-- 
2.1.0