/*
* Copyright (C) 2003 Christophe Saout <christophe@saout.de>
* Copyright (C) 2004 Clemens Fruhwirth <clemens@endorphin.org>
+ * Copyright (C) 2006-2007 Red Hat, Inc. All rights reserved.
*
* This file is released under the GPL.
*/
+#include <linux/err.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/crypto.h>
#include <linux/workqueue.h>
+#include <linux/backing-dev.h>
#include <asm/atomic.h>
-#include <asm/scatterlist.h>
+#include <linux/scatterlist.h>
#include <asm/page.h>
+#include <asm/unaligned.h>
#include "dm.h"
-#define PFX "crypt: "
-
-/*
- * per bio private data
- */
-struct crypt_io {
- struct dm_target *target;
- struct bio *bio;
- struct bio *first_clone;
- struct work_struct work;
- atomic_t pending;
- int error;
-};
+#define DM_MSG_PREFIX "crypt"
+#define MESG_STR(x) x, sizeof(x)
/*
* context holding the current state of a multi-part conversion
unsigned int idx_in;
unsigned int idx_out;
sector_t sector;
- int write;
+};
+
+/*
+ * per bio private data
+ */
+struct dm_crypt_io {
+ struct dm_target *target;
+ struct bio *base_bio;
+ struct work_struct work;
+
+ struct convert_context ctx;
+
+ atomic_t pending;
+ int error;
};
struct crypt_config;
struct crypt_iv_operations {
int (*ctr)(struct crypt_config *cc, struct dm_target *ti,
- const char *opts);
+ const char *opts);
void (*dtr)(struct crypt_config *cc);
const char *(*status)(struct crypt_config *cc);
int (*generator)(struct crypt_config *cc, u8 *iv, sector_t sector);
* Crypt: maps a linear range of a block device
* and encrypts / decrypts at the same time.
*/
+enum flags { DM_CRYPT_SUSPENDED, DM_CRYPT_KEY_VALID };
struct crypt_config {
struct dm_dev *dev;
sector_t start;
*/
mempool_t *io_pool;
mempool_t *page_pool;
+ struct bio_set *bs;
+ struct workqueue_struct *io_queue;
+ struct workqueue_struct *crypt_queue;
/*
* crypto related data
*/
struct crypt_iv_operations *iv_gen_ops;
char *iv_mode;
- void *iv_gen_private;
+ union {
+ struct crypto_cipher *essiv_tfm;
+ int benbi_shift;
+ } iv_gen_private;
sector_t iv_offset;
unsigned int iv_size;
- struct crypto_tfm *tfm;
+ char cipher[CRYPTO_MAX_ALG_NAME];
+ char chainmode[CRYPTO_MAX_ALG_NAME];
+ struct crypto_blkcipher *tfm;
+ unsigned long flags;
unsigned int key_size;
u8 key[0];
};
-#define MIN_IOS 256
+#define MIN_IOS 16
#define MIN_POOL_PAGES 32
#define MIN_BIO_PAGES 8
-static kmem_cache_t *_crypt_io_pool;
-
-/*
- * Mempool alloc and free functions for the page
- */
-static void *mempool_alloc_page(gfp_t gfp_mask, void *data)
-{
- return alloc_page(gfp_mask);
-}
-
-static void mempool_free_page(void *page, void *data)
-{
- __free_page(page);
-}
+static struct kmem_cache *_crypt_io_pool;
+static void clone_init(struct dm_crypt_io *, struct bio *);
/*
* Different IV generation algorithms:
*
- * plain: the initial vector is the 32-bit low-endian version of the sector
- * number, padded with zeros if neccessary.
+ * plain: the initial vector is the 32-bit little-endian version of the sector
+ * number, padded with zeros if necessary.
+ *
+ * essiv: "encrypted sector|salt initial vector", the sector number is
+ * encrypted with the bulk cipher using a salt as key. The salt
+ * should be derived from the bulk cipher's key via hashing.
*
- * ess_iv: "encrypted sector|salt initial vector", the sector number is
- * encrypted with the bulk cipher using a salt as key. The salt
- * should be derived from the bulk cipher's key via hashing.
+ * benbi: the 64-bit "big-endian 'narrow block'-count", starting at 1
+ * (needed for LRW-32-AES and possible other narrow block modes)
+ *
+ * null: the initial vector is always zero. Provides compatibility with
+ * obsolete loop_fish2 devices. Do not use for new devices.
*
* plumb: unimplemented, see:
* http://article.gmane.org/gmane.linux.kernel.device-mapper.dm-crypt/454
}
static int crypt_iv_essiv_ctr(struct crypt_config *cc, struct dm_target *ti,
- const char *opts)
+ const char *opts)
{
- struct crypto_tfm *essiv_tfm;
- struct crypto_tfm *hash_tfm;
+ struct crypto_cipher *essiv_tfm;
+ struct crypto_hash *hash_tfm;
+ struct hash_desc desc;
struct scatterlist sg;
unsigned int saltsize;
u8 *salt;
+ int err;
if (opts == NULL) {
- ti->error = PFX "Digest algorithm missing for ESSIV mode";
+ ti->error = "Digest algorithm missing for ESSIV mode";
return -EINVAL;
}
/* Hash the cipher key with the given hash algorithm */
- hash_tfm = crypto_alloc_tfm(opts, CRYPTO_TFM_REQ_MAY_SLEEP);
- if (hash_tfm == NULL) {
- ti->error = PFX "Error initializing ESSIV hash";
- return -EINVAL;
- }
-
- if (crypto_tfm_alg_type(hash_tfm) != CRYPTO_ALG_TYPE_DIGEST) {
- ti->error = PFX "Expected digest algorithm for ESSIV hash";
- crypto_free_tfm(hash_tfm);
- return -EINVAL;
+ hash_tfm = crypto_alloc_hash(opts, 0, CRYPTO_ALG_ASYNC);
+ if (IS_ERR(hash_tfm)) {
+ ti->error = "Error initializing ESSIV hash";
+ return PTR_ERR(hash_tfm);
}
- saltsize = crypto_tfm_alg_digestsize(hash_tfm);
+ saltsize = crypto_hash_digestsize(hash_tfm);
salt = kmalloc(saltsize, GFP_KERNEL);
if (salt == NULL) {
- ti->error = PFX "Error kmallocing salt storage in ESSIV";
- crypto_free_tfm(hash_tfm);
+ ti->error = "Error kmallocing salt storage in ESSIV";
+ crypto_free_hash(hash_tfm);
return -ENOMEM;
}
- sg.page = virt_to_page(cc->key);
- sg.offset = offset_in_page(cc->key);
- sg.length = cc->key_size;
- crypto_digest_digest(hash_tfm, &sg, 1, salt);
- crypto_free_tfm(hash_tfm);
+ sg_init_one(&sg, cc->key, cc->key_size);
+ desc.tfm = hash_tfm;
+ desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP;
+ err = crypto_hash_digest(&desc, &sg, cc->key_size, salt);
+ crypto_free_hash(hash_tfm);
+
+ if (err) {
+ ti->error = "Error calculating hash in ESSIV";
+ kfree(salt);
+ return err;
+ }
/* Setup the essiv_tfm with the given salt */
- essiv_tfm = crypto_alloc_tfm(crypto_tfm_alg_name(cc->tfm),
- CRYPTO_TFM_MODE_ECB |
- CRYPTO_TFM_REQ_MAY_SLEEP);
- if (essiv_tfm == NULL) {
- ti->error = PFX "Error allocating crypto tfm for ESSIV";
+ essiv_tfm = crypto_alloc_cipher(cc->cipher, 0, CRYPTO_ALG_ASYNC);
+ if (IS_ERR(essiv_tfm)) {
+ ti->error = "Error allocating crypto tfm for ESSIV";
kfree(salt);
- return -EINVAL;
+ return PTR_ERR(essiv_tfm);
}
- if (crypto_tfm_alg_blocksize(essiv_tfm)
- != crypto_tfm_alg_ivsize(cc->tfm)) {
- ti->error = PFX "Block size of ESSIV cipher does "
- "not match IV size of block cipher";
- crypto_free_tfm(essiv_tfm);
+ if (crypto_cipher_blocksize(essiv_tfm) !=
+ crypto_blkcipher_ivsize(cc->tfm)) {
+ ti->error = "Block size of ESSIV cipher does "
+ "not match IV size of block cipher";
+ crypto_free_cipher(essiv_tfm);
kfree(salt);
return -EINVAL;
}
- if (crypto_cipher_setkey(essiv_tfm, salt, saltsize) < 0) {
- ti->error = PFX "Failed to set key for ESSIV cipher";
- crypto_free_tfm(essiv_tfm);
+ err = crypto_cipher_setkey(essiv_tfm, salt, saltsize);
+ if (err) {
+ ti->error = "Failed to set key for ESSIV cipher";
+ crypto_free_cipher(essiv_tfm);
kfree(salt);
- return -EINVAL;
+ return err;
}
kfree(salt);
- cc->iv_gen_private = (void *)essiv_tfm;
+ cc->iv_gen_private.essiv_tfm = essiv_tfm;
return 0;
}
static void crypt_iv_essiv_dtr(struct crypt_config *cc)
{
- crypto_free_tfm((struct crypto_tfm *)cc->iv_gen_private);
- cc->iv_gen_private = NULL;
+ crypto_free_cipher(cc->iv_gen_private.essiv_tfm);
+ cc->iv_gen_private.essiv_tfm = NULL;
}
static int crypt_iv_essiv_gen(struct crypt_config *cc, u8 *iv, sector_t sector)
{
- struct scatterlist sg = { NULL, };
-
memset(iv, 0, cc->iv_size);
*(u64 *)iv = cpu_to_le64(sector);
+ crypto_cipher_encrypt_one(cc->iv_gen_private.essiv_tfm, iv, iv);
+ return 0;
+}
+
+static int crypt_iv_benbi_ctr(struct crypt_config *cc, struct dm_target *ti,
+ const char *opts)
+{
+ unsigned int bs = crypto_blkcipher_blocksize(cc->tfm);
+ int log = ilog2(bs);
+
+ /* we need to calculate how far we must shift the sector count
+ * to get the cipher block count, we use this shift in _gen */
+
+ if (1 << log != bs) {
+ ti->error = "cypher blocksize is not a power of 2";
+ return -EINVAL;
+ }
+
+ if (log > 9) {
+ ti->error = "cypher blocksize is > 512";
+ return -EINVAL;
+ }
+
+ cc->iv_gen_private.benbi_shift = 9 - log;
+
+ return 0;
+}
+
+static void crypt_iv_benbi_dtr(struct crypt_config *cc)
+{
+}
+
+static int crypt_iv_benbi_gen(struct crypt_config *cc, u8 *iv, sector_t sector)
+{
+ __be64 val;
+
+ memset(iv, 0, cc->iv_size - sizeof(u64)); /* rest is cleared below */
+
+ val = cpu_to_be64(((u64)sector << cc->iv_gen_private.benbi_shift) + 1);
+ put_unaligned(val, (__be64 *)(iv + cc->iv_size - sizeof(u64)));
+
+ return 0;
+}
- sg.page = virt_to_page(iv);
- sg.offset = offset_in_page(iv);
- sg.length = cc->iv_size;
- crypto_cipher_encrypt((struct crypto_tfm *)cc->iv_gen_private,
- &sg, &sg, cc->iv_size);
+static int crypt_iv_null_gen(struct crypt_config *cc, u8 *iv, sector_t sector)
+{
+ memset(iv, 0, cc->iv_size);
return 0;
}
.generator = crypt_iv_essiv_gen
};
+static struct crypt_iv_operations crypt_iv_benbi_ops = {
+ .ctr = crypt_iv_benbi_ctr,
+ .dtr = crypt_iv_benbi_dtr,
+ .generator = crypt_iv_benbi_gen
+};
+
+static struct crypt_iv_operations crypt_iv_null_ops = {
+ .generator = crypt_iv_null_gen
+};
-static inline int
+static int
crypt_convert_scatterlist(struct crypt_config *cc, struct scatterlist *out,
struct scatterlist *in, unsigned int length,
int write, sector_t sector)
{
- u8 iv[cc->iv_size];
+ u8 iv[cc->iv_size] __attribute__ ((aligned(__alignof__(u64))));
+ struct blkcipher_desc desc = {
+ .tfm = cc->tfm,
+ .info = iv,
+ .flags = CRYPTO_TFM_REQ_MAY_SLEEP,
+ };
int r;
if (cc->iv_gen_ops) {
return r;
if (write)
- r = crypto_cipher_encrypt_iv(cc->tfm, out, in, length, iv);
+ r = crypto_blkcipher_encrypt_iv(&desc, out, in, length);
else
- r = crypto_cipher_decrypt_iv(cc->tfm, out, in, length, iv);
+ r = crypto_blkcipher_decrypt_iv(&desc, out, in, length);
} else {
if (write)
- r = crypto_cipher_encrypt(cc->tfm, out, in, length);
+ r = crypto_blkcipher_encrypt(&desc, out, in, length);
else
- r = crypto_cipher_decrypt(cc->tfm, out, in, length);
+ r = crypto_blkcipher_decrypt(&desc, out, in, length);
}
return r;
}
-static void
-crypt_convert_init(struct crypt_config *cc, struct convert_context *ctx,
- struct bio *bio_out, struct bio *bio_in,
- sector_t sector, int write)
+static void crypt_convert_init(struct crypt_config *cc,
+ struct convert_context *ctx,
+ struct bio *bio_out, struct bio *bio_in,
+ sector_t sector)
{
ctx->bio_in = bio_in;
ctx->bio_out = bio_out;
ctx->idx_in = bio_in ? bio_in->bi_idx : 0;
ctx->idx_out = bio_out ? bio_out->bi_idx : 0;
ctx->sector = sector + cc->iv_offset;
- ctx->write = write;
}
/*
* Encrypt / decrypt data from one bio to another one (can be the same one)
*/
static int crypt_convert(struct crypt_config *cc,
- struct convert_context *ctx)
+ struct convert_context *ctx)
{
int r = 0;
ctx->idx_out < ctx->bio_out->bi_vcnt) {
struct bio_vec *bv_in = bio_iovec_idx(ctx->bio_in, ctx->idx_in);
struct bio_vec *bv_out = bio_iovec_idx(ctx->bio_out, ctx->idx_out);
- struct scatterlist sg_in = {
- .page = bv_in->bv_page,
- .offset = bv_in->bv_offset + ctx->offset_in,
- .length = 1 << SECTOR_SHIFT
- };
- struct scatterlist sg_out = {
- .page = bv_out->bv_page,
- .offset = bv_out->bv_offset + ctx->offset_out,
- .length = 1 << SECTOR_SHIFT
- };
+ struct scatterlist sg_in, sg_out;
+
+ sg_init_table(&sg_in, 1);
+ sg_set_page(&sg_in, bv_in->bv_page, 1 << SECTOR_SHIFT, bv_in->bv_offset + ctx->offset_in);
+
+ sg_init_table(&sg_out, 1);
+ sg_set_page(&sg_out, bv_out->bv_page, 1 << SECTOR_SHIFT, bv_out->bv_offset + ctx->offset_out);
ctx->offset_in += sg_in.length;
if (ctx->offset_in >= bv_in->bv_len) {
}
r = crypt_convert_scatterlist(cc, &sg_out, &sg_in, sg_in.length,
- ctx->write, ctx->sector);
+ bio_data_dir(ctx->bio_in) == WRITE, ctx->sector);
if (r < 0)
break;
return r;
}
+static void dm_crypt_bio_destructor(struct bio *bio)
+{
+ struct dm_crypt_io *io = bio->bi_private;
+ struct crypt_config *cc = io->target->private;
+
+ bio_free(bio, cc->bs);
+}
+
/*
* Generate a new unfragmented bio with the given size
* This should never violate the device limitations
* May return a smaller bio when running out of pages
*/
-static struct bio *
-crypt_alloc_buffer(struct crypt_config *cc, unsigned int size,
- struct bio *base_bio, unsigned int *bio_vec_idx)
+static struct bio *crypt_alloc_buffer(struct dm_crypt_io *io, unsigned size)
{
- struct bio *bio;
+ struct crypt_config *cc = io->target->private;
+ struct bio *clone;
unsigned int nr_iovecs = (size + PAGE_SIZE - 1) >> PAGE_SHIFT;
gfp_t gfp_mask = GFP_NOIO | __GFP_HIGHMEM;
- unsigned int i;
+ unsigned i, len;
+ struct page *page;
- /*
- * Use __GFP_NOMEMALLOC to tell the VM to act less aggressively and
- * to fail earlier. This is not necessary but increases throughput.
- * FIXME: Is this really intelligent?
- */
- if (base_bio)
- bio = bio_clone(base_bio, GFP_NOIO|__GFP_NOMEMALLOC);
- else
- bio = bio_alloc(GFP_NOIO|__GFP_NOMEMALLOC, nr_iovecs);
- if (!bio)
+ clone = bio_alloc_bioset(GFP_NOIO, nr_iovecs, cc->bs);
+ if (!clone)
return NULL;
- /* if the last bio was not complete, continue where that one ended */
- bio->bi_idx = *bio_vec_idx;
- bio->bi_vcnt = *bio_vec_idx;
- bio->bi_size = 0;
- bio->bi_flags &= ~(1 << BIO_SEG_VALID);
-
- /* bio->bi_idx pages have already been allocated */
- size -= bio->bi_idx * PAGE_SIZE;
-
- for(i = bio->bi_idx; i < nr_iovecs; i++) {
- struct bio_vec *bv = bio_iovec_idx(bio, i);
+ clone_init(io, clone);
- bv->bv_page = mempool_alloc(cc->page_pool, gfp_mask);
- if (!bv->bv_page)
+ for (i = 0; i < nr_iovecs; i++) {
+ page = mempool_alloc(cc->page_pool, gfp_mask);
+ if (!page)
break;
/*
* return a partially allocated bio, the caller will then try
* to allocate additional bios while submitting this partial bio
*/
- if ((i - bio->bi_idx) == (MIN_BIO_PAGES - 1))
+ if (i == (MIN_BIO_PAGES - 1))
gfp_mask = (gfp_mask | __GFP_NOWARN) & ~__GFP_WAIT;
- bv->bv_offset = 0;
- if (size > PAGE_SIZE)
- bv->bv_len = PAGE_SIZE;
- else
- bv->bv_len = size;
+ len = (size > PAGE_SIZE) ? PAGE_SIZE : size;
- bio->bi_size += bv->bv_len;
- bio->bi_vcnt++;
- size -= bv->bv_len;
+ if (!bio_add_page(clone, page, len, 0)) {
+ mempool_free(page, cc->page_pool);
+ break;
+ }
+
+ size -= len;
}
- if (!bio->bi_size) {
- bio_put(bio);
+ if (!clone->bi_size) {
+ bio_put(clone);
return NULL;
}
- /*
- * Remember the last bio_vec allocated to be able
- * to correctly continue after the splitting.
- */
- *bio_vec_idx = bio->bi_vcnt;
-
- return bio;
+ return clone;
}
-static void crypt_free_buffer_pages(struct crypt_config *cc,
- struct bio *bio, unsigned int bytes)
+static void crypt_free_buffer_pages(struct crypt_config *cc, struct bio *clone)
{
- unsigned int i, start, end;
+ unsigned int i;
struct bio_vec *bv;
- /*
- * This is ugly, but Jens Axboe thinks that using bi_idx in the
- * endio function is too dangerous at the moment, so I calculate the
- * correct position using bi_vcnt and bi_size.
- * The bv_offset and bv_len fields might already be modified but we
- * know that we always allocated whole pages.
- * A fix to the bi_idx issue in the kernel is in the works, so
- * we will hopefully be able to revert to the cleaner solution soon.
- */
- i = bio->bi_vcnt - 1;
- bv = bio_iovec_idx(bio, i);
- end = (i << PAGE_SHIFT) + (bv->bv_offset + bv->bv_len) - bio->bi_size;
- start = end - bytes;
-
- start >>= PAGE_SHIFT;
- if (!bio->bi_size)
- end = bio->bi_vcnt;
- else
- end >>= PAGE_SHIFT;
-
- for(i = start; i < end; i++) {
- bv = bio_iovec_idx(bio, i);
+ for (i = 0; i < clone->bi_vcnt; i++) {
+ bv = bio_iovec_idx(clone, i);
BUG_ON(!bv->bv_page);
mempool_free(bv->bv_page, cc->page_pool);
bv->bv_page = NULL;
* One of the bios was finished. Check for completion of
* the whole request and correctly clean up the buffer.
*/
-static void dec_pending(struct crypt_io *io, int error)
+static void crypt_dec_pending(struct dm_crypt_io *io)
{
- struct crypt_config *cc = (struct crypt_config *) io->target->private;
-
- if (error < 0)
- io->error = error;
+ struct crypt_config *cc = io->target->private;
if (!atomic_dec_and_test(&io->pending))
return;
- if (io->first_clone)
- bio_put(io->first_clone);
-
- bio_endio(io->bio, io->bio->bi_size, io->error);
-
+ bio_endio(io->base_bio, io->error);
mempool_free(io, cc->io_pool);
}
/*
- * kcryptd:
+ * kcryptd/kcryptd_io:
*
* Needed because it would be very unwise to do decryption in an
- * interrupt context, so bios returning from read requests get
- * queued here.
+ * interrupt context.
+ *
+ * kcryptd performs the actual encryption or decryption.
+ *
+ * kcryptd_io performs the IO submission.
+ *
+ * They must be separated as otherwise the final stages could be
+ * starved by new requests which can block in the first stages due
+ * to memory allocation.
*/
-static struct workqueue_struct *_kcryptd_workqueue;
+static void kcryptd_io(struct work_struct *work);
+static void kcryptd_crypt(struct work_struct *work);
-static void kcryptd_do_work(void *data)
+static void kcryptd_queue_io(struct dm_crypt_io *io)
{
- struct crypt_io *io = (struct crypt_io *) data;
- struct crypt_config *cc = (struct crypt_config *) io->target->private;
- struct convert_context ctx;
- int r;
+ struct crypt_config *cc = io->target->private;
+
+ INIT_WORK(&io->work, kcryptd_io);
+ queue_work(cc->io_queue, &io->work);
+}
+
+static void kcryptd_queue_crypt(struct dm_crypt_io *io)
+{
+ struct crypt_config *cc = io->target->private;
+
+ INIT_WORK(&io->work, kcryptd_crypt);
+ queue_work(cc->crypt_queue, &io->work);
+}
+
+static void crypt_endio(struct bio *clone, int error)
+{
+ struct dm_crypt_io *io = clone->bi_private;
+ struct crypt_config *cc = io->target->private;
+ unsigned rw = bio_data_dir(clone);
+
+ if (unlikely(!bio_flagged(clone, BIO_UPTODATE) && !error))
+ error = -EIO;
+
+ /*
+ * free the processed pages
+ */
+ if (rw == WRITE)
+ crypt_free_buffer_pages(cc, clone);
+
+ bio_put(clone);
+
+ if (rw == READ && !error) {
+ kcryptd_queue_crypt(io);
+ return;
+ }
+
+ if (unlikely(error))
+ io->error = error;
+
+ crypt_dec_pending(io);
+}
+
+static void clone_init(struct dm_crypt_io *io, struct bio *clone)
+{
+ struct crypt_config *cc = io->target->private;
+
+ clone->bi_private = io;
+ clone->bi_end_io = crypt_endio;
+ clone->bi_bdev = cc->dev->bdev;
+ clone->bi_rw = io->base_bio->bi_rw;
+ clone->bi_destructor = dm_crypt_bio_destructor;
+}
+
+static void kcryptd_io_read(struct dm_crypt_io *io)
+{
+ struct crypt_config *cc = io->target->private;
+ struct bio *base_bio = io->base_bio;
+ struct bio *clone;
+ sector_t sector = base_bio->bi_sector - io->target->begin;
+
+ atomic_inc(&io->pending);
+
+ /*
+ * The block layer might modify the bvec array, so always
+ * copy the required bvecs because we need the original
+ * one in order to decrypt the whole bio data *afterwards*.
+ */
+ clone = bio_alloc_bioset(GFP_NOIO, bio_segments(base_bio), cc->bs);
+ if (unlikely(!clone)) {
+ io->error = -ENOMEM;
+ crypt_dec_pending(io);
+ return;
+ }
+
+ clone_init(io, clone);
+ clone->bi_idx = 0;
+ clone->bi_vcnt = bio_segments(base_bio);
+ clone->bi_size = base_bio->bi_size;
+ clone->bi_sector = cc->start + sector;
+ memcpy(clone->bi_io_vec, bio_iovec(base_bio),
+ sizeof(struct bio_vec) * clone->bi_vcnt);
+
+ generic_make_request(clone);
+}
+
+static void kcryptd_io_write(struct dm_crypt_io *io)
+{
+}
+
+static void kcryptd_crypt_write_io_submit(struct dm_crypt_io *io, int error)
+{
+}
+
+static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)
+{
+ struct crypt_config *cc = io->target->private;
+ struct bio *base_bio = io->base_bio;
+ struct bio *clone;
+ unsigned remaining = base_bio->bi_size;
+ sector_t sector = base_bio->bi_sector - io->target->begin;
+
+ atomic_inc(&io->pending);
+
+ crypt_convert_init(cc, &io->ctx, NULL, base_bio, sector);
+
+ /*
+ * The allocated buffers can be smaller than the whole bio,
+ * so repeat the whole process until all the data can be handled.
+ */
+ while (remaining) {
+ clone = crypt_alloc_buffer(io, remaining);
+ if (unlikely(!clone)) {
+ io->error = -ENOMEM;
+ crypt_dec_pending(io);
+ return;
+ }
+
+ io->ctx.bio_out = clone;
+ io->ctx.idx_out = 0;
+
+ if (unlikely(crypt_convert(cc, &io->ctx) < 0)) {
+ crypt_free_buffer_pages(cc, clone);
+ bio_put(clone);
+ io->error = -EIO;
+ crypt_dec_pending(io);
+ return;
+ }
+
+ /* crypt_convert should have filled the clone bio */
+ BUG_ON(io->ctx.idx_out < clone->bi_vcnt);
+
+ clone->bi_sector = cc->start + sector;
+ remaining -= clone->bi_size;
+ sector += bio_sectors(clone);
+
+ /* Grab another reference to the io struct
+ * before we kick off the request */
+ if (remaining)
+ atomic_inc(&io->pending);
+
+ generic_make_request(clone);
+
+ /* Do not reference clone after this - it
+ * may be gone already. */
+
+ /* out of memory -> run queues */
+ if (remaining)
+ congestion_wait(WRITE, HZ/100);
+ }
+}
+
+static void kcryptd_crypt_read_done(struct dm_crypt_io *io, int error)
+{
+ if (unlikely(error < 0))
+ io->error = -EIO;
+
+ crypt_dec_pending(io);
+}
+
+static void kcryptd_crypt_read_convert(struct dm_crypt_io *io)
+{
+ struct crypt_config *cc = io->target->private;
+ int r = 0;
+
+ crypt_convert_init(cc, &io->ctx, io->base_bio, io->base_bio,
+ io->base_bio->bi_sector - io->target->begin);
+
+ r = crypt_convert(cc, &io->ctx);
- crypt_convert_init(cc, &ctx, io->bio, io->bio,
- io->bio->bi_sector - io->target->begin, 0);
- r = crypt_convert(cc, &ctx);
+ kcryptd_crypt_read_done(io, r);
+}
+
+static void kcryptd_io(struct work_struct *work)
+{
+ struct dm_crypt_io *io = container_of(work, struct dm_crypt_io, work);
- dec_pending(io, r);
+ if (bio_data_dir(io->base_bio) == READ)
+ kcryptd_io_read(io);
+ else
+ kcryptd_io_write(io);
}
-static void kcryptd_queue_io(struct crypt_io *io)
+static void kcryptd_crypt(struct work_struct *work)
{
- INIT_WORK(&io->work, kcryptd_do_work, io);
- queue_work(_kcryptd_workqueue, &io->work);
+ struct dm_crypt_io *io = container_of(work, struct dm_crypt_io, work);
+
+ if (bio_data_dir(io->base_bio) == READ)
+ kcryptd_crypt_read_convert(io);
+ else
+ kcryptd_crypt_write_convert(io);
}
/*
buffer[2] = '\0';
- for(i = 0; i < size; i++) {
+ for (i = 0; i < size; i++) {
buffer[0] = *hex++;
buffer[1] = *hex++;
{
unsigned int i;
- for(i = 0; i < size; i++) {
+ for (i = 0; i < size; i++) {
sprintf(hex, "%02x", *key);
hex += 2;
key++;
}
}
+static int crypt_set_key(struct crypt_config *cc, char *key)
+{
+ unsigned key_size = strlen(key) >> 1;
+
+ if (cc->key_size && cc->key_size != key_size)
+ return -EINVAL;
+
+ cc->key_size = key_size; /* initial settings */
+
+ if ((!key_size && strcmp(key, "-")) ||
+ (key_size && crypt_decode_key(cc->key, key, key_size) < 0))
+ return -EINVAL;
+
+ set_bit(DM_CRYPT_KEY_VALID, &cc->flags);
+
+ return 0;
+}
+
+static int crypt_wipe_key(struct crypt_config *cc)
+{
+ clear_bit(DM_CRYPT_KEY_VALID, &cc->flags);
+ memset(&cc->key, 0, cc->key_size * sizeof(u8));
+ return 0;
+}
+
/*
* Construct an encryption mapping:
* <cipher> <key> <iv_offset> <dev_path> <start>
static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
{
struct crypt_config *cc;
- struct crypto_tfm *tfm;
+ struct crypto_blkcipher *tfm;
char *tmp;
char *cipher;
char *chainmode;
char *ivmode;
char *ivopts;
- unsigned int crypto_flags;
unsigned int key_size;
+ unsigned long long tmpll;
if (argc != 5) {
- ti->error = PFX "Not enough arguments";
+ ti->error = "Not enough arguments";
return -EINVAL;
}
ivmode = strsep(&ivopts, ":");
if (tmp)
- DMWARN(PFX "Unexpected additional cipher options");
+ DMWARN("Unexpected additional cipher options");
key_size = strlen(argv[1]) >> 1;
- cc = kmalloc(sizeof(*cc) + key_size * sizeof(u8), GFP_KERNEL);
+ cc = kzalloc(sizeof(*cc) + key_size * sizeof(u8), GFP_KERNEL);
if (cc == NULL) {
ti->error =
- PFX "Cannot allocate transparent encryption context";
+ "Cannot allocate transparent encryption context";
return -ENOMEM;
}
- cc->key_size = key_size;
- if ((!key_size && strcmp(argv[1], "-") != 0) ||
- (key_size && crypt_decode_key(cc->key, argv[1], key_size) < 0)) {
- ti->error = PFX "Error decoding key";
- goto bad1;
+ if (crypt_set_key(cc, argv[1])) {
+ ti->error = "Error decoding key";
+ goto bad_cipher;
}
/* Compatiblity mode for old dm-crypt cipher strings */
ivmode = "plain";
}
- /* Choose crypto_flags according to chainmode */
- if (strcmp(chainmode, "cbc") == 0)
- crypto_flags = CRYPTO_TFM_MODE_CBC;
- else if (strcmp(chainmode, "ecb") == 0)
- crypto_flags = CRYPTO_TFM_MODE_ECB;
- else {
- ti->error = PFX "Unknown chaining mode";
- goto bad1;
+ if (strcmp(chainmode, "ecb") && !ivmode) {
+ ti->error = "This chaining mode requires an IV mechanism";
+ goto bad_cipher;
}
- if (crypto_flags != CRYPTO_TFM_MODE_ECB && !ivmode) {
- ti->error = PFX "This chaining mode requires an IV mechanism";
- goto bad1;
+ if (snprintf(cc->cipher, CRYPTO_MAX_ALG_NAME, "%s(%s)",
+ chainmode, cipher) >= CRYPTO_MAX_ALG_NAME) {
+ ti->error = "Chain mode + cipher name is too long";
+ goto bad_cipher;
}
- tfm = crypto_alloc_tfm(cipher, crypto_flags | CRYPTO_TFM_REQ_MAY_SLEEP);
- if (!tfm) {
- ti->error = PFX "Error allocating crypto tfm";
- goto bad1;
- }
- if (crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_CIPHER) {
- ti->error = PFX "Expected cipher algorithm";
- goto bad2;
+ tfm = crypto_alloc_blkcipher(cc->cipher, 0, CRYPTO_ALG_ASYNC);
+ if (IS_ERR(tfm)) {
+ ti->error = "Error allocating crypto tfm";
+ goto bad_cipher;
}
+ strcpy(cc->cipher, cipher);
+ strcpy(cc->chainmode, chainmode);
cc->tfm = tfm;
/*
- * Choose ivmode. Valid modes: "plain", "essiv:<esshash>".
+ * Choose ivmode. Valid modes: "plain", "essiv:<esshash>", "benbi".
* See comments at iv code
*/
cc->iv_gen_ops = &crypt_iv_plain_ops;
else if (strcmp(ivmode, "essiv") == 0)
cc->iv_gen_ops = &crypt_iv_essiv_ops;
+ else if (strcmp(ivmode, "benbi") == 0)
+ cc->iv_gen_ops = &crypt_iv_benbi_ops;
+ else if (strcmp(ivmode, "null") == 0)
+ cc->iv_gen_ops = &crypt_iv_null_ops;
else {
- ti->error = PFX "Invalid IV mode";
- goto bad2;
+ ti->error = "Invalid IV mode";
+ goto bad_ivmode;
}
if (cc->iv_gen_ops && cc->iv_gen_ops->ctr &&
cc->iv_gen_ops->ctr(cc, ti, ivopts) < 0)
- goto bad2;
+ goto bad_ivmode;
- if (tfm->crt_cipher.cit_decrypt_iv && tfm->crt_cipher.cit_encrypt_iv)
+ cc->iv_size = crypto_blkcipher_ivsize(tfm);
+ if (cc->iv_size)
/* at least a 64 bit sector number should fit in our buffer */
- cc->iv_size = max(crypto_tfm_alg_ivsize(tfm),
- (unsigned int)(sizeof(u64) / sizeof(u8)));
+ cc->iv_size = max(cc->iv_size,
+ (unsigned int)(sizeof(u64) / sizeof(u8)));
else {
- cc->iv_size = 0;
if (cc->iv_gen_ops) {
- DMWARN(PFX "Selected cipher does not support IVs");
+ DMWARN("Selected cipher does not support IVs");
if (cc->iv_gen_ops->dtr)
cc->iv_gen_ops->dtr(cc);
cc->iv_gen_ops = NULL;
}
}
- cc->io_pool = mempool_create(MIN_IOS, mempool_alloc_slab,
- mempool_free_slab, _crypt_io_pool);
+ cc->io_pool = mempool_create_slab_pool(MIN_IOS, _crypt_io_pool);
if (!cc->io_pool) {
- ti->error = PFX "Cannot allocate crypt io mempool";
- goto bad3;
+ ti->error = "Cannot allocate crypt io mempool";
+ goto bad_slab_pool;
}
- cc->page_pool = mempool_create(MIN_POOL_PAGES, mempool_alloc_page,
- mempool_free_page, NULL);
+ cc->page_pool = mempool_create_page_pool(MIN_POOL_PAGES, 0);
if (!cc->page_pool) {
- ti->error = PFX "Cannot allocate page mempool";
- goto bad4;
+ ti->error = "Cannot allocate page mempool";
+ goto bad_page_pool;
+ }
+
+ cc->bs = bioset_create(MIN_IOS, MIN_IOS);
+ if (!cc->bs) {
+ ti->error = "Cannot allocate crypt bioset";
+ goto bad_bs;
}
- if (tfm->crt_cipher.cit_setkey(tfm, cc->key, key_size) < 0) {
- ti->error = PFX "Error setting key";
- goto bad5;
+ if (crypto_blkcipher_setkey(tfm, cc->key, key_size) < 0) {
+ ti->error = "Error setting key";
+ goto bad_device;
}
- if (sscanf(argv[2], SECTOR_FORMAT, &cc->iv_offset) != 1) {
- ti->error = PFX "Invalid iv_offset sector";
- goto bad5;
+ if (sscanf(argv[2], "%llu", &tmpll) != 1) {
+ ti->error = "Invalid iv_offset sector";
+ goto bad_device;
}
+ cc->iv_offset = tmpll;
- if (sscanf(argv[4], SECTOR_FORMAT, &cc->start) != 1) {
- ti->error = PFX "Invalid device sector";
- goto bad5;
+ if (sscanf(argv[4], "%llu", &tmpll) != 1) {
+ ti->error = "Invalid device sector";
+ goto bad_device;
}
+ cc->start = tmpll;
if (dm_get_device(ti, argv[3], cc->start, ti->len,
- dm_table_get_mode(ti->table), &cc->dev)) {
- ti->error = PFX "Device lookup failed";
- goto bad5;
+ dm_table_get_mode(ti->table), &cc->dev)) {
+ ti->error = "Device lookup failed";
+ goto bad_device;
}
if (ivmode && cc->iv_gen_ops) {
*(ivopts - 1) = ':';
cc->iv_mode = kmalloc(strlen(ivmode) + 1, GFP_KERNEL);
if (!cc->iv_mode) {
- ti->error = PFX "Error kmallocing iv_mode string";
- goto bad5;
+ ti->error = "Error kmallocing iv_mode string";
+ goto bad_ivmode_string;
}
strcpy(cc->iv_mode, ivmode);
} else
cc->iv_mode = NULL;
+ cc->io_queue = create_singlethread_workqueue("kcryptd_io");
+ if (!cc->io_queue) {
+ ti->error = "Couldn't create kcryptd io queue";
+ goto bad_io_queue;
+ }
+
+ cc->crypt_queue = create_singlethread_workqueue("kcryptd");
+ if (!cc->crypt_queue) {
+ ti->error = "Couldn't create kcryptd queue";
+ goto bad_crypt_queue;
+ }
+
ti->private = cc;
return 0;
-bad5:
+bad_crypt_queue:
+ destroy_workqueue(cc->io_queue);
+bad_io_queue:
+ kfree(cc->iv_mode);
+bad_ivmode_string:
+ dm_put_device(ti, cc->dev);
+bad_device:
+ bioset_free(cc->bs);
+bad_bs:
mempool_destroy(cc->page_pool);
-bad4:
+bad_page_pool:
mempool_destroy(cc->io_pool);
-bad3:
+bad_slab_pool:
if (cc->iv_gen_ops && cc->iv_gen_ops->dtr)
cc->iv_gen_ops->dtr(cc);
-bad2:
- crypto_free_tfm(tfm);
-bad1:
+bad_ivmode:
+ crypto_free_blkcipher(tfm);
+bad_cipher:
+ /* Must zero key material before freeing */
+ memset(cc, 0, sizeof(*cc) + cc->key_size * sizeof(u8));
kfree(cc);
return -EINVAL;
}
{
struct crypt_config *cc = (struct crypt_config *) ti->private;
+ destroy_workqueue(cc->io_queue);
+ destroy_workqueue(cc->crypt_queue);
+
+ bioset_free(cc->bs);
mempool_destroy(cc->page_pool);
mempool_destroy(cc->io_pool);
kfree(cc->iv_mode);
if (cc->iv_gen_ops && cc->iv_gen_ops->dtr)
cc->iv_gen_ops->dtr(cc);
- crypto_free_tfm(cc->tfm);
+ crypto_free_blkcipher(cc->tfm);
dm_put_device(ti, cc->dev);
- kfree(cc);
-}
-
-static int crypt_endio(struct bio *bio, unsigned int done, int error)
-{
- struct crypt_io *io = (struct crypt_io *) bio->bi_private;
- struct crypt_config *cc = (struct crypt_config *) io->target->private;
-
- if (bio_data_dir(bio) == WRITE) {
- /*
- * free the processed pages, even if
- * it's only a partially completed write
- */
- crypt_free_buffer_pages(cc, bio, done);
- }
- if (bio->bi_size)
- return 1;
-
- bio_put(bio);
-
- /*
- * successful reads are decrypted by the worker thread
- */
- if ((bio_data_dir(bio) == READ)
- && bio_flagged(bio, BIO_UPTODATE)) {
- kcryptd_queue_io(io);
- return 0;
- }
-
- dec_pending(io, error);
- return error;
-}
-
-static inline struct bio *
-crypt_clone(struct crypt_config *cc, struct crypt_io *io, struct bio *bio,
- sector_t sector, unsigned int *bvec_idx,
- struct convert_context *ctx)
-{
- struct bio *clone;
-
- if (bio_data_dir(bio) == WRITE) {
- clone = crypt_alloc_buffer(cc, bio->bi_size,
- io->first_clone, bvec_idx);
- if (clone) {
- ctx->bio_out = clone;
- if (crypt_convert(cc, ctx) < 0) {
- crypt_free_buffer_pages(cc, clone,
- clone->bi_size);
- bio_put(clone);
- return NULL;
- }
- }
- } else {
- /*
- * The block layer might modify the bvec array, so always
- * copy the required bvecs because we need the original
- * one in order to decrypt the whole bio data *afterwards*.
- */
- clone = bio_alloc(GFP_NOIO, bio_segments(bio));
- if (clone) {
- clone->bi_idx = 0;
- clone->bi_vcnt = bio_segments(bio);
- clone->bi_size = bio->bi_size;
- memcpy(clone->bi_io_vec, bio_iovec(bio),
- sizeof(struct bio_vec) * clone->bi_vcnt);
- }
- }
-
- if (!clone)
- return NULL;
-
- clone->bi_private = io;
- clone->bi_end_io = crypt_endio;
- clone->bi_bdev = cc->dev->bdev;
- clone->bi_sector = cc->start + sector;
- clone->bi_rw = bio->bi_rw;
-
- return clone;
+ /* Must zero key material before freeing */
+ memset(cc, 0, sizeof(*cc) + cc->key_size * sizeof(u8));
+ kfree(cc);
}
static int crypt_map(struct dm_target *ti, struct bio *bio,
union map_info *map_context)
{
- struct crypt_config *cc = (struct crypt_config *) ti->private;
- struct crypt_io *io = mempool_alloc(cc->io_pool, GFP_NOIO);
- struct convert_context ctx;
- struct bio *clone;
- unsigned int remaining = bio->bi_size;
- sector_t sector = bio->bi_sector - ti->begin;
- unsigned int bvec_idx = 0;
+ struct crypt_config *cc = ti->private;
+ struct dm_crypt_io *io;
+ io = mempool_alloc(cc->io_pool, GFP_NOIO);
io->target = ti;
- io->bio = bio;
- io->first_clone = NULL;
+ io->base_bio = bio;
io->error = 0;
- atomic_set(&io->pending, 1); /* hold a reference */
+ atomic_set(&io->pending, 0);
- if (bio_data_dir(bio) == WRITE)
- crypt_convert_init(cc, &ctx, NULL, bio, sector, 1);
-
- /*
- * The allocated buffers can be smaller than the whole bio,
- * so repeat the whole process until all the data can be handled.
- */
- while (remaining) {
- clone = crypt_clone(cc, io, bio, sector, &bvec_idx, &ctx);
- if (!clone)
- goto cleanup;
-
- if (!io->first_clone) {
- /*
- * hold a reference to the first clone, because it
- * holds the bio_vec array and that can't be freed
- * before all other clones are released
- */
- bio_get(clone);
- io->first_clone = clone;
- }
- atomic_inc(&io->pending);
-
- remaining -= clone->bi_size;
- sector += bio_sectors(clone);
-
- generic_make_request(clone);
-
- /* out of memory -> run queues */
- if (remaining)
- blk_congestion_wait(bio_data_dir(clone), HZ/100);
- }
-
- /* drop reference, clones could have returned before we reach this */
- dec_pending(io, 0);
- return 0;
-
-cleanup:
- if (io->first_clone) {
- dec_pending(io, -ENOMEM);
- return 0;
- }
+ if (bio_data_dir(io->base_bio) == READ)
+ kcryptd_queue_io(io);
+ else
+ kcryptd_queue_crypt(io);
- /* if no bio has been dispatched yet, we can directly return the error */
- mempool_free(io, cc->io_pool);
- return -ENOMEM;
+ return DM_MAPIO_SUBMITTED;
}
static int crypt_status(struct dm_target *ti, status_type_t type,
char *result, unsigned int maxlen)
{
struct crypt_config *cc = (struct crypt_config *) ti->private;
- const char *cipher;
- const char *chainmode = NULL;
unsigned int sz = 0;
switch (type) {
break;
case STATUSTYPE_TABLE:
- cipher = crypto_tfm_alg_name(cc->tfm);
-
- switch(cc->tfm->crt_cipher.cit_mode) {
- case CRYPTO_TFM_MODE_CBC:
- chainmode = "cbc";
- break;
- case CRYPTO_TFM_MODE_ECB:
- chainmode = "ecb";
- break;
- default:
- BUG();
- }
-
if (cc->iv_mode)
- DMEMIT("%s-%s-%s ", cipher, chainmode, cc->iv_mode);
+ DMEMIT("%s-%s-%s ", cc->cipher, cc->chainmode,
+ cc->iv_mode);
else
- DMEMIT("%s-%s ", cipher, chainmode);
+ DMEMIT("%s-%s ", cc->cipher, cc->chainmode);
if (cc->key_size > 0) {
if ((maxlen - sz) < ((cc->key_size << 1) + 1))
result[sz++] = '-';
}
- DMEMIT(" " SECTOR_FORMAT " %s " SECTOR_FORMAT,
- cc->iv_offset, cc->dev->name, cc->start);
+ DMEMIT(" %llu %s %llu", (unsigned long long)cc->iv_offset,
+ cc->dev->name, (unsigned long long)cc->start);
break;
}
return 0;
}
+static void crypt_postsuspend(struct dm_target *ti)
+{
+ struct crypt_config *cc = ti->private;
+
+ set_bit(DM_CRYPT_SUSPENDED, &cc->flags);
+}
+
+static int crypt_preresume(struct dm_target *ti)
+{
+ struct crypt_config *cc = ti->private;
+
+ if (!test_bit(DM_CRYPT_KEY_VALID, &cc->flags)) {
+ DMERR("aborting resume - crypt key is not set.");
+ return -EAGAIN;
+ }
+
+ return 0;
+}
+
+static void crypt_resume(struct dm_target *ti)
+{
+ struct crypt_config *cc = ti->private;
+
+ clear_bit(DM_CRYPT_SUSPENDED, &cc->flags);
+}
+
+/* Message interface
+ * key set <key>
+ * key wipe
+ */
+static int crypt_message(struct dm_target *ti, unsigned argc, char **argv)
+{
+ struct crypt_config *cc = ti->private;
+
+ if (argc < 2)
+ goto error;
+
+ if (!strnicmp(argv[0], MESG_STR("key"))) {
+ if (!test_bit(DM_CRYPT_SUSPENDED, &cc->flags)) {
+ DMWARN("not suspended during key manipulation.");
+ return -EINVAL;
+ }
+ if (argc == 3 && !strnicmp(argv[1], MESG_STR("set")))
+ return crypt_set_key(cc, argv[2]);
+ if (argc == 2 && !strnicmp(argv[1], MESG_STR("wipe")))
+ return crypt_wipe_key(cc);
+ }
+
+error:
+ DMWARN("unrecognised message received.");
+ return -EINVAL;
+}
+
static struct target_type crypt_target = {
.name = "crypt",
- .version= {1, 1, 0},
+ .version= {1, 5, 0},
.module = THIS_MODULE,
.ctr = crypt_ctr,
.dtr = crypt_dtr,
.map = crypt_map,
.status = crypt_status,
+ .postsuspend = crypt_postsuspend,
+ .preresume = crypt_preresume,
+ .resume = crypt_resume,
+ .message = crypt_message,
};
static int __init dm_crypt_init(void)
{
int r;
- _crypt_io_pool = kmem_cache_create("dm-crypt_io",
- sizeof(struct crypt_io),
- 0, 0, NULL, NULL);
+ _crypt_io_pool = KMEM_CACHE(dm_crypt_io, 0);
if (!_crypt_io_pool)
return -ENOMEM;
- _kcryptd_workqueue = create_workqueue("kcryptd");
- if (!_kcryptd_workqueue) {
- r = -ENOMEM;
- DMERR(PFX "couldn't create kcryptd");
- goto bad1;
- }
-
r = dm_register_target(&crypt_target);
if (r < 0) {
- DMERR(PFX "register failed %d", r);
- goto bad2;
+ DMERR("register failed %d", r);
+ kmem_cache_destroy(_crypt_io_pool);
}
- return 0;
-
-bad2:
- destroy_workqueue(_kcryptd_workqueue);
-bad1:
- kmem_cache_destroy(_crypt_io_pool);
return r;
}
int r = dm_unregister_target(&crypt_target);
if (r < 0)
- DMERR(PFX "unregister failed %d", r);
+ DMERR("unregister failed %d", r);
- destroy_workqueue(_kcryptd_workqueue);
kmem_cache_destroy(_crypt_io_pool);
}
Code Review / linux-2.6.git / blobdiff