Date:	Wed, 12 Nov 2014 05:24:53 +0100
From:	Stephan Mueller <smueller@...onox.de>
To:	Herbert Xu <herbert@...dor.apana.org.au>
Cc:	Marek Vasut <marex@...x.de>,
	"David S. Miller" <davem@...emloft.net>,
	Jason Cooper <cryptography@...edaemon.net>,
	Grant Likely <grant.likely@...retlab.ca>,
	'Geert Uytterhoeven' <geert@...ux-m68k.org>,
	"Joy M. Latten" <jmlatten@...ux.vnet.ibm.com>,
	Jonathan Corbet <corbet@....net>,
	LKML <linux-kernel@...r.kernel.org>, linux-crypto@...r.kernel.org
Subject: [PATCH v3 03/13] crypto: Documentation - userspace interface spec

The userspace interface of the kernel crypto API is documented with
 * a general explanation
 * a discussion of the memory in-place operation
 * the description of the message digest API
 * the description of the symmetric cipher API

The documentation refers to libkcapi as a working example on how to use
the kernel crypto API from user space.

Signed-off-by: Stephan Mueller <smueller@...onox.de>
---
 Documentation/crypto/crypto-API-userspace.txt | 205 ++++++++++++++++++++++++++
 1 file changed, 205 insertions(+)
 create mode 100644 Documentation/crypto/crypto-API-userspace.txt

diff --git a/Documentation/crypto/crypto-API-userspace.txt b/Documentation/crypto/crypto-API-userspace.txt
new file mode 100644
index 0000000..ac619cd
--- /dev/null
+++ b/Documentation/crypto/crypto-API-userspace.txt
@@ -0,0 +1,205 @@
+Introduction
+============
+
+The concepts of the kernel crypto API visible to kernel space is fully
+applicable to the user space interface as well. Therefore, the kernel crypto API
+high level discussion for the in-kernel use cases applies here as well.
+
+The major difference, however, is that user space can only act as a consumer
+and never as a provider of a transformation or cipher algorithm.
+
+The following covers the user space interface exported by the kernel crypto
+API. A working example of this description is libkcapi that can be obtained from
+[1]. That library can be used by user space applications that require
+cryptographic services from the kernel.
+
+Some details of the in-kernel kernel crypto API aspects do not
+apply to user space, however. This includes the difference between synchronous
+and asynchronous invocations. The user space API call is fully synchronous.
+In addition, only a subset of all cipher types are available as documented
+below.
+
+
+User space API general remarks
+==============================
+
+The kernel crypto API is accessible from user space. Currently, the following
+ciphers are accessible:
+
+	* Message digest including keyed message digest (HMAC, CMAC)
+
+	* Symmetric ciphers
+
+Note, AEAD ciphers are currently not supported via the symmetric cipher
+interface.
+
+The interface is provided via Netlink using the type AF_ALG. In addition, the
+setsockopt option type is SOL_ALG. In case the user space header files do not
+export these flags yet, use the following macros:
+
+#ifndef AF_ALG
+#define AF_ALG 38
+#endif
+#ifndef SOL_ALG
+#define SOL_ALG 279
+#endif
+
+A cipher is accessed with the same name as done for the in-kernel API calls.
+This includes the generic vs. unique naming schema for ciphers as well as the
+enforcement of priorities for generic names.
+
+To interact with the kernel crypto API, a Netlink socket must be created by
+the user space application. User space invokes the cipher operation with the
+send/write system call family. The result of the cipher operation is obtained
+with the read/recv system call family.
+
+The following API calls assume that the Netlink socket descriptor is already
+opened by the user space application and discusses only the kernel crypto API
+specific invocations.
+
+To initialize a Netlink interface, the following sequence has to be performed
+by the consumer:
+
+	1. Create a socket of type AF_ALG with the struct sockaddr_alg parameter
+	   specified below for the different cipher types.
+
+	2. Invoke bind with the socket descriptor
+
+	3. Invoke accept with the socket descriptor. The accept system call
+	   returns a new file descriptor that is to be used to interact with
+	   the particular cipher instance. When invoking send/write or recv/read
+	   system calls to send data to the kernel or obtain data from the
+	   kernel, the file descriptor returned by accept must be used.
+
+In-place cipher operation
+=========================
+
+Just like the in-kernel operation of the kernel crypto API, the user space
+interface allows the cipher operation in-place. That means that the input buffer
+used for the send/write system call and the output buffer used by the read/recv
+system call may be one and the same. This is of particular interest for
+symmetric cipher operations where a copying of the output data to its final
+destination can be avoided.
+
+If a consumer on the other hand wants to maintain the plaintext and the
+ciphertext in different memory locations, all a consumer needs to do is to
+provide different memory pointers for the encryption and decryption operation.
+
+Message digest API
+==================
+
+The message digest type to be used for the cipher operation is selected when
+invoking the bind syscall. bind requires the caller to provide a filled
+struct sockaddr data structure. This data structure must be filled as follows:
+
+struct sockaddr_alg sa = {
+	.salg_family = AF_ALG,
+	.salg_type = "hash", /* this selects the hash logic in the kernel */
+	.salg_name = "sha1" /* this is the cipher name */
+};
+
+The salg_type value "hash" applies to message digests and keyed message digests.
+Though, a keyed message digest is referenced by the appropriate salg_name.
+Please see below for the setsockopt interface that explains how the key can be
+set for a keyed message digest.
+
+Using the send() system call, the application provides the data that should be
+processed with the message digest. The send system call allows the following
+flags to be specified:
+
+	* MSG_MORE: If this flag is set, the send system call acts like a
+		    message digest update function where the final hash is not
+		    yet calculated. If the flag is not set, the send system call
+		    calculates the final message digest immediately.
+
+With the recv() system call, the application can read the message digest from
+the kernel crypto API. If the buffer is too small for the message digest, the
+flag MSG_TRUNC is set by the kernel.
+
+In order to set a message digest key, the calling application must use the
+setsockopt() option of ALG_SET_KEY. If the key is not set the HMAC operation is
+performed without the initial HMAC state change caused by the key.
+
+
+Symmetric cipher API
+====================
+
+The operation is very similar to the message digest discussion. During
+initialization, the struct sockaddr data structure must be filled as follows:
+
+struct sockaddr_alg sa = {
+	.salg_family = AF_ALG,
+	.salg_type = "skcipher", /* this selects the symmetric cipher */
+	.salg_name = "cbc(aes)" /* this is the cipher name */
+};
+
+Before data can be sent to the kernel using the write/send system call family,
+the consumer must set the key. The key setting is described with the setsockopt
+invocation below.
+
+Using the sendmsg() system call, the application provides the data that should
+be processed for encryption or decryption. In addition, the IV is specified
+with the data structure provided by the sendmsg() system call.
+
+The sendmsg system call parameter of struct msghdr is embedded into the
+struct cmsghdr data structure. See recv(2) and cmsg(3) for more information
+on how the cmsghdr data structure is used together with the send/recv system
+call family. That cmsghdr data structure holds the following information
+specified with a separate header instances:
+
+	* specification of the cipher operation type with one of these flags:
+		ALG_OP_ENCRYPT - encryption of data
+		ALG_OP_DECRYPT - decryption of data
+
+	* specification of the IV information marked with the flag ALG_SET_IV
+
+The send system call family allows the following flag to be specified:
+
+	* MSG_MORE: If this flag is set, the send system call acts like a
+		    cipher update function where more input data is expected
+		    with a subsequent invocation of the send system call.
+
+Note: The kernel reports -EINVAL for any unexpected data. The caller must
+make sure that all data matches the constraints given in /proc/crypto for the
+selected cipher.
+
+With the recv() system call, the application can read the result of the
+cipher operation from the kernel crypto API. The output buffer must be at least
+as large as to hold all blocks of the encrypted or decrypted data. If the output
+data size is smaller, only as many blocks are returned that fit into that
+output buffer size.
+
+Setsockopt interface
+====================
+
+In addition to the read/recv and send/write system call handling to send and
+retrieve data subject to the cipher operation, a consumer also needs to set
+the additional information for the cipher operation. This additional information
+is set using the setsockopt system call that must be invoked with the file
+descriptor of the open cipher (i.e. the file descriptor returned by the
+accept system call).
+
+Each setsockopt invocation must use the level SOL_ALG.
+
+The setsockopt interface allows setting the following data using the mentioned
+optname:
+
+	* ALG_SET_KEY -- Setting the key. Key setting is applicable to:
+
+		- the skcipher cipher type (symmetric ciphers)
+
+		- the hash cipher type (keyed message digests)
+
+User space API example
+======================
+
+Please see [1] for libkcapi which provides an easy-to-use wrapper around the
+aforementioned Netlink kernel interface. [1] also contains a test application
+that invokes all libkcapi API calls.
+
+[1] http://www.chronox.de/libkcapi.html
+
+Author
+======
+
+Stephan Mueller <smueller@...onox.de>
-- 
2.1.0


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists