diff mbox

[2/2] md: dm-crypt: Optimize the dm-crypt for XTS mode

Message ID 02be0f42bf2d3c3d27b43bc050a783582b7af733.1450080755.git.baolin.wang@linaro.org (mailing list archive)
State New, archived
Headers show

Commit Message

(Exiting) Baolin Wang Dec. 14, 2015, 8:23 a.m. UTC
In now dm-crypt code, it is ineffective to map one bio with just only
one scatterlist at one time for XTS mode. We can use multiple scatterlists
to map the whole bio and send all scatterlists of one bio to crypto engine
to encrypt or decrypt, which can improve the hardware engine's efficiency.

With this optimization, On my test setup (beaglebone black board) using 64KB
I/Os on an eMMC storage device I saw about 60% improvement in throughput for
encrypted writes, and about 100% improvement for encrypted reads. But this
is not fit for other modes which need different IV for each sector.

Signed-off-by: Baolin Wang <baolin.wang@linaro.org>
---
 drivers/md/dm-crypt.c |  315 ++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 309 insertions(+), 6 deletions(-)

Comments

Milan Broz Dec. 14, 2015, 11:31 p.m. UTC | #1
On 12/14/2015 09:23 AM, Baolin Wang wrote:
> In now dm-crypt code, it is ineffective to map one bio with just only
> one scatterlist at one time for XTS mode. We can use multiple scatterlists
> to map the whole bio and send all scatterlists of one bio to crypto engine
> to encrypt or decrypt, which can improve the hardware engine's efficiency.

...

> +	/*
> +	 * Here we need to check if it can be encrypted or decrypted with
> +	 * bulk block, which means these encryption modes don't need IV or
> +	 * just need one initial IV. For bulk mode, we can expand the
> +	 * scatterlist entries to map the bio, then send all the scatterlists
> +	 * to the hardware engine at one time to improve the crypto engine
> +	 * efficiency. But it does not fit for other encryption modes, it has
> +	 * to do encryption and decryption sector by sector because every
> +	 * sector has different IV.
> +	 */
> +	if (!strcmp(chainmode, "ecb") || !strcmp(chainmode, "xts"))
> +		cc->bulk_crypto = 1;
> +	else
> +		cc->bulk_crypto = 0;n

It is perfectly fine to use another IV even for XTS mode (despite it is not needed).
You can use ESSIV for example, or benbi (big-endian variant of plain IV).
(And it is really used for some LUKS devices.)

How it is supposed to work in this case?
If I read this patch correctly, it will completely corrupt data in this case because
it expects plain (consecutive) IV...

Also how it handles 32bit plain IV (that restart after 2TB)? (IOW plain IV, not plain64).

Milan

--
To unsubscribe from this list: send the line "unsubscribe linux-block" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
(Exiting) Baolin Wang Dec. 15, 2015, 2:56 a.m. UTC | #2
>> +     /*
>> +      * Here we need to check if it can be encrypted or decrypted with
>> +      * bulk block, which means these encryption modes don't need IV or
>> +      * just need one initial IV. For bulk mode, we can expand the
>> +      * scatterlist entries to map the bio, then send all the scatterlists
>> +      * to the hardware engine at one time to improve the crypto engine
>> +      * efficiency. But it does not fit for other encryption modes, it has
>> +      * to do encryption and decryption sector by sector because every
>> +      * sector has different IV.
>> +      */
>> +     if (!strcmp(chainmode, "ecb") || !strcmp(chainmode, "xts"))
>> +             cc->bulk_crypto = 1;
>> +     else
>> +             cc->bulk_crypto = 0;n
>
> It is perfectly fine to use another IV even for XTS mode (despite it is not needed).
> You can use ESSIV for example, or benbi (big-endian variant of plain IV).
> (And it is really used for some LUKS devices.)
>
> How it is supposed to work in this case?
> If I read this patch correctly, it will completely corrupt data in this case because
> it expects plain (consecutive) IV...

The XTS mode can limit maximum size of each encrypted data unit
(typically a sector or disk block) to 2^20 AES blocks, so we can use
one bio as one encrypted data unit (we don't do it sector by sector).
It can generate one IV for each separate encrypted data unit. Please
correct me if I misunderstand something. Thanks.

>
> Also how it handles 32bit plain IV (that restart after 2TB)? (IOW plain IV, not plain64).
>
> Milan
>
Milan Broz Dec. 15, 2015, 8:20 a.m. UTC | #3
On 12/15/2015 03:56 AM, Baolin Wang wrote:
>>> +     /*
>>> +      * Here we need to check if it can be encrypted or decrypted with
>>> +      * bulk block, which means these encryption modes don't need IV or
>>> +      * just need one initial IV. For bulk mode, we can expand the
>>> +      * scatterlist entries to map the bio, then send all the scatterlists
>>> +      * to the hardware engine at one time to improve the crypto engine
>>> +      * efficiency. But it does not fit for other encryption modes, it has
>>> +      * to do encryption and decryption sector by sector because every
>>> +      * sector has different IV.
>>> +      */
>>> +     if (!strcmp(chainmode, "ecb") || !strcmp(chainmode, "xts"))
>>> +             cc->bulk_crypto = 1;
>>> +     else
>>> +             cc->bulk_crypto = 0;n
>>
>> It is perfectly fine to use another IV even for XTS mode (despite it is not needed).
>> You can use ESSIV for example, or benbi (big-endian variant of plain IV).
>> (And it is really used for some LUKS devices.)
>>
>> How it is supposed to work in this case?
>> If I read this patch correctly, it will completely corrupt data in this case because
>> it expects plain (consecutive) IV...
> 
> The XTS mode can limit maximum size of each encrypted data unit
> (typically a sector or disk block) to 2^20 AES blocks, so we can use
> one bio as one encrypted data unit (we don't do it sector by sector).
> It can generate one IV for each separate encrypted data unit. Please
> correct me if I misunderstand something. Thanks.

How this will help XTS-ESSIV, if you have to recalculate IV on every fixed
encrypted block?

TBH I think the whole patch here is doing something more than optimization
and seriously touches cryptography part.

Isn't in de-facto reinventing how the full disk encryption works today?

All currently used systems use always disk encrypted block size (usually fixed
to 512 bytes sectors like in dmcrypt) and these are encrypted independently with own IV.
(We can talk about supporting larger disk encrypted block sizes but the principle
it is still the same.
And we played with it before already  http://www.saout.de/pipermail/dm-crypt/2013-January/003125.html)

Every encrypted disk block here has independent initialization vector - IV (or tweak).
For some modes (like XTS, LRW) this vector can produce just predictable linear offset,
usually just block number (it must not repeat though; exceptions are legacy/compatible IVs).
But it can be also something else - pseudorandom sequence like in ESSIV or so.

If I understand your patch correctly, you are trying to use the XTS mode for
the whole bulk request (there is only one IV for the bulk request and this bulk request
is larger than currently fixed disk block size).

Isn't it even incompatible with the current XTS (aes-xts-plain64) encrypted disk
per-sector where for every sectors new IV initialization is used?
In fact I think that your approach would need to implement some different
IV name (aes-xts-bulk) or something like that so userspace (cryptsetup)
can be backward compatible.

BTW the 2^20 block limit requirement (the XTS block size) is strict limit in some specifications.
For example, NIST SP800-38E (XTS-AES) says:
"The length of the data unit for any instance of an implementation of XTS-AES shall not exceed
2^20 AES blocks."

There are a lot of badly implemented hw crypto accelerators where initialization cost
is quite big and that performs better if it encrypts large blocks of data
(it applies even for CBC mode). These are simply not designed for FDE use case...

But IMHO this is not correct reason to patch it in kernel, moreover on dmcrypt layer
with side effect of hardcoding another kind of crypto logic into dmcrypt.

DMcrypt should be completely independent of used symmetric block encryption mode
(that's the crypto API job). There will be new modes in future, there can be requests
to switch to different IVs (similar to preventing CBC watermarking attack with
predictable IV) etc.

Milan
--
To unsubscribe from this list: send the line "unsubscribe linux-block" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
(Exiting) Baolin Wang Dec. 15, 2015, 11:35 a.m. UTC | #4
On 15 December 2015 at 16:20, Milan Broz <gmazyland@gmail.com> wrote:
> On 12/15/2015 03:56 AM, Baolin Wang wrote:
>>>> +     /*
>>>> +      * Here we need to check if it can be encrypted or decrypted with
>>>> +      * bulk block, which means these encryption modes don't need IV or
>>>> +      * just need one initial IV. For bulk mode, we can expand the
>>>> +      * scatterlist entries to map the bio, then send all the scatterlists
>>>> +      * to the hardware engine at one time to improve the crypto engine
>>>> +      * efficiency. But it does not fit for other encryption modes, it has
>>>> +      * to do encryption and decryption sector by sector because every
>>>> +      * sector has different IV.
>>>> +      */
>>>> +     if (!strcmp(chainmode, "ecb") || !strcmp(chainmode, "xts"))
>>>> +             cc->bulk_crypto = 1;
>>>> +     else
>>>> +             cc->bulk_crypto = 0;n
>>>
>>> It is perfectly fine to use another IV even for XTS mode (despite it is not needed).
>>> You can use ESSIV for example, or benbi (big-endian variant of plain IV).
>>> (And it is really used for some LUKS devices.)
>>>
>>> How it is supposed to work in this case?
>>> If I read this patch correctly, it will completely corrupt data in this case because
>>> it expects plain (consecutive) IV...
>>
>> The XTS mode can limit maximum size of each encrypted data unit
>> (typically a sector or disk block) to 2^20 AES blocks, so we can use
>> one bio as one encrypted data unit (we don't do it sector by sector).
>> It can generate one IV for each separate encrypted data unit. Please
>> correct me if I misunderstand something. Thanks.
>
> How this will help XTS-ESSIV, if you have to recalculate IV on every fixed
> encrypted block?
>
> TBH I think the whole patch here is doing something more than optimization
> and seriously touches cryptography part.
>
> Isn't in de-facto reinventing how the full disk encryption works today?
>
> All currently used systems use always disk encrypted block size (usually fixed
> to 512 bytes sectors like in dmcrypt) and these are encrypted independently with own IV.
> (We can talk about supporting larger disk encrypted block sizes but the principle
> it is still the same.
> And we played with it before already  http://www.saout.de/pipermail/dm-crypt/2013-January/003125.html)
>
> Every encrypted disk block here has independent initialization vector - IV (or tweak).
> For some modes (like XTS, LRW) this vector can produce just predictable linear offset,
> usually just block number (it must not repeat though; exceptions are legacy/compatible IVs).
> But it can be also something else - pseudorandom sequence like in ESSIV or so.
>
> If I understand your patch correctly, you are trying to use the XTS mode for
> the whole bulk request (there is only one IV for the bulk request and this bulk request
> is larger than currently fixed disk block size).

That's right.

>
> Isn't it even incompatible with the current XTS (aes-xts-plain64) encrypted disk
> per-sector where for every sectors new IV initialization is used?
> In fact I think that your approach would need to implement some different
> IV name (aes-xts-bulk) or something like that so userspace (cryptsetup)
> can be backward compatible.

Yes, that sounds reasonable.

>
> BTW the 2^20 block limit requirement (the XTS block size) is strict limit in some specifications.
> For example, NIST SP800-38E (XTS-AES) says:
> "The length of the data unit for any instance of an implementation of XTS-AES shall not exceed
> 2^20 AES blocks."

OK.

>
> There are a lot of badly implemented hw crypto accelerators where initialization cost
> is quite big and that performs better if it encrypts large blocks of data
> (it applies even for CBC mode). These are simply not designed for FDE use case...
>
> But IMHO this is not correct reason to patch it in kernel, moreover on dmcrypt layer
> with side effect of hardcoding another kind of crypto logic into dmcrypt.

You're right, this patch is used for hw crypto acceleration. As you
know some hardware engine may handle the IV for one bulk data, so the
software just generate one initial IV. Qualcomm's implementation for
xts mode just use one initial IV for one bulk data to improve the hw
engine efficiency.
(https://github.com/major91/Zeta-Chromium-N5/blob/master/drivers/md/dm-req-crypt.c)
So like you said I can change the IV name as 'aes-xts-bulk' or
something else to avoid the ESSIV, cause this can really gain big
improvements for hw engine.

>
> DMcrypt should be completely independent of used symmetric block encryption mode
> (that's the crypto API job). There will be new modes in future, there can be requests
> to switch to different IVs (similar to preventing CBC watermarking attack with
> predictable IV) etc.
>
> Milan
diff mbox

Patch

diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index 917d47e..9f6f131 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -32,6 +32,7 @@ 
 #include <linux/device-mapper.h>
 
 #define DM_MSG_PREFIX "crypt"
+#define DM_MAX_SG_LIST	1024
 
 /*
  * context holding the current state of a multi-part conversion
@@ -68,6 +69,8 @@  struct dm_crypt_request {
 	struct convert_context *ctx;
 	struct scatterlist sg_in;
 	struct scatterlist sg_out;
+	struct sg_table sgt_in;
+	struct sg_table sgt_out;
 	sector_t iv_sector;
 };
 
@@ -140,6 +143,7 @@  struct crypt_config {
 	char *cipher;
 	char *cipher_string;
 
+	int bulk_crypto;
 	struct crypt_iv_operations *iv_gen_ops;
 	union {
 		struct iv_essiv_private essiv;
@@ -833,6 +837,11 @@  static u8 *iv_of_dmreq(struct crypt_config *cc,
 		crypto_ablkcipher_alignmask(any_tfm(cc)) + 1);
 }
 
+static int crypt_is_bulk_mode(struct crypt_config *cc)
+{
+	return cc->bulk_crypto;
+}
+
 static int crypt_convert_block(struct crypt_config *cc,
 			       struct convert_context *ctx,
 			       struct ablkcipher_request *req)
@@ -881,24 +890,40 @@  static int crypt_convert_block(struct crypt_config *cc,
 
 static void kcryptd_async_done(struct crypto_async_request *async_req,
 			       int error);
+static void kcryptd_async_all_done(struct crypto_async_request *async_req,
+				   int error);
 
 static void crypt_alloc_req(struct crypt_config *cc,
 			    struct convert_context *ctx)
 {
 	unsigned key_index = ctx->cc_sector & (cc->tfms_count - 1);
+	struct dm_crypt_request *dmreq;
 
 	if (!ctx->req)
 		ctx->req = mempool_alloc(cc->req_pool, GFP_NOIO);
 
+	dmreq = dmreq_of_req(cc, ctx->req);
+	dmreq->sgt_in.orig_nents = 0;
+	dmreq->sgt_out.orig_nents = 0;
+
 	ablkcipher_request_set_tfm(ctx->req, cc->tfms[key_index]);
 
 	/*
 	 * Use REQ_MAY_BACKLOG so a cipher driver internally backlogs
 	 * requests if driver request queue is full.
 	 */
-	ablkcipher_request_set_callback(ctx->req,
-	    CRYPTO_TFM_REQ_MAY_BACKLOG | CRYPTO_TFM_REQ_MAY_SLEEP,
-	    kcryptd_async_done, dmreq_of_req(cc, ctx->req));
+	if (crypt_is_bulk_mode(cc))
+		ablkcipher_request_set_callback(ctx->req,
+						CRYPTO_TFM_REQ_MAY_BACKLOG
+						| CRYPTO_TFM_REQ_MAY_SLEEP,
+						kcryptd_async_all_done,
+						dmreq_of_req(cc, ctx->req));
+	else
+		ablkcipher_request_set_callback(ctx->req,
+						CRYPTO_TFM_REQ_MAY_BACKLOG
+						| CRYPTO_TFM_REQ_MAY_SLEEP,
+						kcryptd_async_done,
+						dmreq_of_req(cc, ctx->req));
 }
 
 static void crypt_free_req(struct crypt_config *cc,
@@ -911,6 +936,221 @@  static void crypt_free_req(struct crypt_config *cc,
 }
 
 /*
+ * Check how many sg entry numbers are needed when map one bio
+ * with scatterlist in advance.
+ */
+static unsigned int crypt_sg_entry(struct bio *bio_t)
+{
+	struct request_queue *q = bdev_get_queue(bio_t->bi_bdev);
+	int cluster = blk_queue_cluster(q);
+	struct bio_vec bvec, bvprv = { NULL };
+	struct bvec_iter biter;
+	unsigned long nbytes = 0, sg_length = 0;
+	unsigned int sg_cnt = 0;
+
+	if (bio_t->bi_rw & REQ_DISCARD) {
+		if (bio_t->bi_vcnt)
+			return 1;
+		return 0;
+	}
+
+	if (bio_t->bi_rw & REQ_WRITE_SAME)
+		return 1;
+
+	bio_for_each_segment(bvec, bio_t, biter) {
+		nbytes = bvec.bv_len;
+
+		if (!cluster) {
+			sg_cnt++;
+			continue;
+		}
+
+		if (sg_length + nbytes > queue_max_segment_size(q)) {
+			sg_length = nbytes;
+			sg_cnt++;
+			goto next;
+		}
+
+		if (!BIOVEC_PHYS_MERGEABLE(&bvprv, &bvec)) {
+			sg_length = nbytes;
+			sg_cnt++;
+			goto next;
+		}
+
+		if (!BIOVEC_SEG_BOUNDARY(q, &bvprv, &bvec)) {
+			sg_length = nbytes;
+			sg_cnt++;
+			goto next;
+		}
+
+		sg_length += nbytes;
+next:
+		memcpy(&bvprv, &bvec, sizeof(struct bio_vec));
+	}
+
+	return sg_cnt;
+}
+
+static int crypt_convert_all_blocks(struct crypt_config *cc,
+				   struct convert_context *ctx,
+				   struct ablkcipher_request *req)
+{
+	struct dm_crypt_io *io =
+		container_of(ctx, struct dm_crypt_io, ctx);
+	struct dm_crypt_request *dmreq = dmreq_of_req(cc, req);
+	u8 *iv = iv_of_dmreq(cc, dmreq);
+	struct bio *orig_bio = io->base_bio;
+	struct bio *bio_in = ctx->bio_in;
+	struct bio *bio_out = ctx->bio_out;
+	unsigned int total_bytes = orig_bio->bi_iter.bi_size;
+	struct scatterlist *sg_in = NULL;
+	struct scatterlist *sg_out = NULL;
+	struct scatterlist *sg = NULL;
+	unsigned int total_sg_len_in = 0;
+	unsigned int total_sg_len_out = 0;
+	unsigned int sg_in_max = 0, sg_out_max = 0;
+	int ret;
+
+	dmreq->iv_sector = ctx->cc_sector;
+	dmreq->ctx = ctx;
+
+	/*
+	 * Need to calculate how many sg entry need to be used
+	 * for this bio.
+	 */
+	sg_in_max = crypt_sg_entry(bio_in) + 1;
+	if (sg_in_max > DM_MAX_SG_LIST || sg_in_max <= 0) {
+		DMERR("%s sg entry too large or none %d\n",
+		      __func__, sg_in_max);
+		return -EINVAL;
+	} else if (sg_in_max == 2) {
+		sg_in = &dmreq->sg_in;
+	}
+
+	if (!sg_in) {
+		ret = sg_alloc_table(&dmreq->sgt_in, sg_in_max, GFP_KERNEL);
+		if (ret) {
+			DMERR("%s sg in allocation failed\n", __func__);
+			return -ENOMEM;
+		}
+
+		sg_in = dmreq->sgt_in.sgl;
+	}
+
+	total_sg_len_in = __blk_bios_map_sg(bdev_get_queue(bio_in->bi_bdev),
+					    bio_in, sg_in, &sg);
+	if ((total_sg_len_in <= 0)
+	    || (total_sg_len_in > sg_in_max)) {
+		DMERR("%s in sg map error %d, sg_in_max[%d]\n",
+		      __func__, total_sg_len_in, sg_in_max);
+		return -EINVAL;
+	}
+
+	if (sg)
+		sg_mark_end(sg);
+
+	ctx->iter_in.bi_size -= total_bytes;
+
+	if (bio_data_dir(orig_bio) == READ)
+		goto set_crypt;
+
+	sg_out_max = crypt_sg_entry(bio_out) + 1;
+	if (sg_out_max > DM_MAX_SG_LIST || sg_out_max <= 0) {
+		DMERR("%s sg entry too large or none %d\n",
+		      __func__, sg_out_max);
+		return -EINVAL;
+	} else if (sg_out_max == 2) {
+		sg_out = &dmreq->sg_out;
+	}
+
+	if (!sg_out) {
+		ret = sg_alloc_table(&dmreq->sgt_out, sg_out_max, GFP_KERNEL);
+		if (ret) {
+			DMERR("%s sg out allocation failed\n", __func__);
+			return -ENOMEM;
+		}
+
+		sg_out = dmreq->sgt_out.sgl;
+	}
+
+	sg = NULL;
+	total_sg_len_out = __blk_bios_map_sg(bdev_get_queue(bio_out->bi_bdev),
+					     bio_out, sg_out, &sg);
+	if ((total_sg_len_out <= 0) ||
+	    (total_sg_len_out > sg_out_max)) {
+		DMERR("%s out sg map error %d, sg_out_max[%d]\n",
+		      __func__, total_sg_len_out, sg_out_max);
+		return -EINVAL;
+	}
+
+	if (sg)
+		sg_mark_end(sg);
+
+	ctx->iter_out.bi_size -= total_bytes;
+set_crypt:
+	if (cc->iv_gen_ops) {
+		ret = cc->iv_gen_ops->generator(cc, iv, dmreq);
+		if (ret < 0) {
+			DMERR("%s generator iv error %d\n", __func__, ret);
+			return ret;
+		}
+	}
+
+	if (bio_data_dir(orig_bio) == WRITE) {
+		ablkcipher_request_set_crypt(req, sg_in,
+					     sg_out, total_bytes, iv);
+
+		ret = crypto_ablkcipher_encrypt(req);
+	} else {
+		ablkcipher_request_set_crypt(req, sg_in,
+					     sg_in, total_bytes, iv);
+
+		ret = crypto_ablkcipher_decrypt(req);
+	}
+
+	if (!ret && cc->iv_gen_ops && cc->iv_gen_ops->post)
+		ret = cc->iv_gen_ops->post(cc, iv, dmreq);
+
+	return ret;
+}
+
+/*
+ * Encrypt / decrypt data from one whole bio at one time.
+ */
+static int crypt_convert_io(struct crypt_config *cc,
+			    struct convert_context *ctx)
+{
+	int r;
+
+	atomic_set(&ctx->cc_pending, 1);
+	crypt_alloc_req(cc, ctx);
+	atomic_inc(&ctx->cc_pending);
+
+	r = crypt_convert_all_blocks(cc, ctx, ctx->req);
+	switch (r) {
+	case -EBUSY:
+		/*
+		 * Lets make this synchronous bio by waiting on
+		 * in progress as well.
+		 */
+	case -EINPROGRESS:
+		wait_for_completion(&ctx->restart);
+		ctx->req = NULL;
+		break;
+	case 0:
+		atomic_dec(&ctx->cc_pending);
+		cond_resched();
+		break;
+	/* There was an error while processing the request. */
+	default:
+		atomic_dec(&ctx->cc_pending);
+		return r;
+	}
+
+	return 0;
+}
+
+/*
  * Encrypt / decrypt data from one bio to another one (can be the same one)
  */
 static int crypt_convert(struct crypt_config *cc,
@@ -1070,12 +1310,18 @@  static void crypt_dec_pending(struct dm_crypt_io *io)
 	struct crypt_config *cc = io->cc;
 	struct bio *base_bio = io->base_bio;
 	int error = io->error;
+	struct dm_crypt_request *dmreq;
 
 	if (!atomic_dec_and_test(&io->io_pending))
 		return;
 
-	if (io->ctx.req)
+	if (io->ctx.req) {
+		dmreq = dmreq_of_req(cc, io->ctx.req);
+		sg_free_table(&dmreq->sgt_out);
+		sg_free_table(&dmreq->sgt_in);
+
 		crypt_free_req(cc, io->ctx.req, base_bio);
+	}
 
 	base_bio->bi_error = error;
 	bio_endio(base_bio);
@@ -1312,7 +1558,11 @@  static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)
 	sector += bio_sectors(clone);
 
 	crypt_inc_pending(io);
-	r = crypt_convert(cc, &io->ctx);
+	if (crypt_is_bulk_mode(cc))
+		r = crypt_convert_io(cc, &io->ctx);
+	else
+		r = crypt_convert(cc, &io->ctx);
+
 	if (r)
 		io->error = -EIO;
 	crypt_finished = atomic_dec_and_test(&io->ctx.cc_pending);
@@ -1342,7 +1592,11 @@  static void kcryptd_crypt_read_convert(struct dm_crypt_io *io)
 	crypt_convert_init(cc, &io->ctx, io->base_bio, io->base_bio,
 			   io->sector);
 
-	r = crypt_convert(cc, &io->ctx);
+	if (crypt_is_bulk_mode(cc))
+		r = crypt_convert_io(cc, &io->ctx);
+	else
+		r = crypt_convert(cc, &io->ctx);
+
 	if (r < 0)
 		io->error = -EIO;
 
@@ -1387,6 +1641,40 @@  static void kcryptd_async_done(struct crypto_async_request *async_req,
 		kcryptd_crypt_write_io_submit(io, 1);
 }
 
+static void kcryptd_async_all_done(struct crypto_async_request *async_req,
+			       int error)
+{
+	struct dm_crypt_request *dmreq = async_req->data;
+	struct convert_context *ctx = dmreq->ctx;
+	struct dm_crypt_io *io = container_of(ctx, struct dm_crypt_io, ctx);
+	struct crypt_config *cc = io->cc;
+
+	if (error == -EINPROGRESS)
+		return;
+
+	if (!error && cc->iv_gen_ops && cc->iv_gen_ops->post)
+		error = cc->iv_gen_ops->post(cc, iv_of_dmreq(cc, dmreq), dmreq);
+
+	if (error < 0)
+		io->error = error;
+
+	sg_free_table(&dmreq->sgt_out);
+	sg_free_table(&dmreq->sgt_in);
+
+	crypt_free_req(cc, req_of_dmreq(cc, dmreq), io->base_bio);
+
+	if (!atomic_dec_and_test(&ctx->cc_pending)) {
+		complete(&io->ctx.restart);
+		return;
+	}
+
+	complete(&io->ctx.restart);
+	if (bio_data_dir(io->base_bio) == READ)
+		kcryptd_crypt_read_done(io);
+	else
+		kcryptd_crypt_write_io_submit(io, 1);
+}
+
 static void kcryptd_crypt(struct work_struct *work)
 {
 	struct dm_crypt_io *io = container_of(work, struct dm_crypt_io, work);
@@ -1633,6 +1921,21 @@  static int crypt_ctr_cipher(struct dm_target *ti,
 		goto bad_mem;
 	}
 
+	/*
+	 * Here we need to check if it can be encrypted or decrypted with
+	 * bulk block, which means these encryption modes don't need IV or
+	 * just need one initial IV. For bulk mode, we can expand the
+	 * scatterlist entries to map the bio, then send all the scatterlists
+	 * to the hardware engine at one time to improve the crypto engine
+	 * efficiency. But it does not fit for other encryption modes, it has
+	 * to do encryption and decryption sector by sector because every
+	 * sector has different IV.
+	 */
+	if (!strcmp(chainmode, "ecb") || !strcmp(chainmode, "xts"))
+		cc->bulk_crypto = 1;
+	else
+		cc->bulk_crypto = 0;
+
 	/* Allocate cipher */
 	ret = crypt_alloc_tfms(cc, cipher_api);
 	if (ret < 0) {