zfs/module/icp/algs/modes/gcm.c

1576 lines
42 KiB
C
Raw Normal View History

/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2008, 2010, Oracle and/or its affiliates. All rights reserved.
*/
#include <sys/zfs_context.h>
#include <modes/modes.h>
#include <sys/crypto/common.h>
#include <sys/crypto/icp.h>
#include <sys/crypto/impl.h>
#include <sys/byteorder.h>
#include <sys/simd.h>
#include <modes/gcm_impl.h>
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
#include <aes/aes_impl.h>
#endif
#define GHASH(c, d, t, o) \
xor_block((uint8_t *)(d), (uint8_t *)(c)->gcm_ghash); \
(o)->mul((uint64_t *)(void *)(c)->gcm_ghash, (c)->gcm_H, \
(uint64_t *)(void *)(t));
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
/* Select GCM implementation */
#define IMPL_FASTEST (UINT32_MAX)
#define IMPL_CYCLE (UINT32_MAX-1)
#ifdef CAN_USE_GCM_ASM
#define IMPL_AVX (UINT32_MAX-2)
#endif
#define GCM_IMPL_READ(i) (*(volatile uint32_t *) &(i))
static uint32_t icp_gcm_impl = IMPL_FASTEST;
static uint32_t user_sel_impl = IMPL_FASTEST;
#ifdef CAN_USE_GCM_ASM
/* Does the architecture we run on support the MOVBE instruction? */
boolean_t gcm_avx_can_use_movbe = B_FALSE;
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
/*
* Whether to use the optimized openssl gcm and ghash implementations.
* Set to true if module parameter icp_gcm_impl == "avx".
*/
static boolean_t gcm_use_avx = B_FALSE;
#define GCM_IMPL_USE_AVX (*(volatile boolean_t *)&gcm_use_avx)
static inline boolean_t gcm_avx_will_work(void);
static inline void gcm_set_avx(boolean_t);
static inline boolean_t gcm_toggle_avx(void);
extern boolean_t atomic_toggle_boolean_nv(volatile boolean_t *);
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
static int gcm_mode_encrypt_contiguous_blocks_avx(gcm_ctx_t *, char *, size_t,
crypto_data_t *, size_t);
static int gcm_encrypt_final_avx(gcm_ctx_t *, crypto_data_t *, size_t);
static int gcm_decrypt_final_avx(gcm_ctx_t *, crypto_data_t *, size_t);
static int gcm_init_avx(gcm_ctx_t *, unsigned char *, size_t, unsigned char *,
size_t, size_t);
#endif /* ifdef CAN_USE_GCM_ASM */
/*
* Encrypt multiple blocks of data in GCM mode. Decrypt for GCM mode
* is done in another function.
*/
int
gcm_mode_encrypt_contiguous_blocks(gcm_ctx_t *ctx, char *data, size_t length,
crypto_data_t *out, size_t block_size,
int (*encrypt_block)(const void *, const uint8_t *, uint8_t *),
void (*copy_block)(uint8_t *, uint8_t *),
void (*xor_block)(uint8_t *, uint8_t *))
{
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
if (ctx->gcm_use_avx == B_TRUE)
return (gcm_mode_encrypt_contiguous_blocks_avx(
ctx, data, length, out, block_size));
#endif
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
const gcm_impl_ops_t *gops;
size_t remainder = length;
size_t need = 0;
uint8_t *datap = (uint8_t *)data;
uint8_t *blockp;
uint8_t *lastp;
void *iov_or_mp;
offset_t offset;
uint8_t *out_data_1;
uint8_t *out_data_2;
size_t out_data_1_len;
uint64_t counter;
uint64_t counter_mask = ntohll(0x00000000ffffffffULL);
if (length + ctx->gcm_remainder_len < block_size) {
/* accumulate bytes here and return */
bcopy(datap,
(uint8_t *)ctx->gcm_remainder + ctx->gcm_remainder_len,
length);
ctx->gcm_remainder_len += length;
if (ctx->gcm_copy_to == NULL) {
ctx->gcm_copy_to = datap;
}
return (CRYPTO_SUCCESS);
}
lastp = (uint8_t *)ctx->gcm_cb;
if (out != NULL)
crypto_init_ptrs(out, &iov_or_mp, &offset);
gops = gcm_impl_get_ops();
do {
/* Unprocessed data from last call. */
if (ctx->gcm_remainder_len > 0) {
need = block_size - ctx->gcm_remainder_len;
if (need > remainder)
return (CRYPTO_DATA_LEN_RANGE);
bcopy(datap, &((uint8_t *)ctx->gcm_remainder)
[ctx->gcm_remainder_len], need);
blockp = (uint8_t *)ctx->gcm_remainder;
} else {
blockp = datap;
}
/*
* Increment counter. Counter bits are confined
* to the bottom 32 bits of the counter block.
*/
counter = ntohll(ctx->gcm_cb[1] & counter_mask);
counter = htonll(counter + 1);
counter &= counter_mask;
ctx->gcm_cb[1] = (ctx->gcm_cb[1] & ~counter_mask) | counter;
encrypt_block(ctx->gcm_keysched, (uint8_t *)ctx->gcm_cb,
(uint8_t *)ctx->gcm_tmp);
xor_block(blockp, (uint8_t *)ctx->gcm_tmp);
lastp = (uint8_t *)ctx->gcm_tmp;
ctx->gcm_processed_data_len += block_size;
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
/*
* The following copies a complete GCM block back to where it
* came from if there was a remainder in the last call and out
* is NULL. That doesn't seem to make sense. So we assert this
* can't happen and leave the code in for reference.
* See https://github.com/zfsonlinux/zfs/issues/9661
*/
ASSERT(out != NULL);
if (out == NULL) {
if (ctx->gcm_remainder_len > 0) {
bcopy(blockp, ctx->gcm_copy_to,
ctx->gcm_remainder_len);
bcopy(blockp + ctx->gcm_remainder_len, datap,
need);
}
} else {
crypto_get_ptrs(out, &iov_or_mp, &offset, &out_data_1,
&out_data_1_len, &out_data_2, block_size);
/* copy block to where it belongs */
if (out_data_1_len == block_size) {
copy_block(lastp, out_data_1);
} else {
bcopy(lastp, out_data_1, out_data_1_len);
if (out_data_2 != NULL) {
bcopy(lastp + out_data_1_len,
out_data_2,
block_size - out_data_1_len);
}
}
/* update offset */
out->cd_offset += block_size;
}
/* add ciphertext to the hash */
GHASH(ctx, ctx->gcm_tmp, ctx->gcm_ghash, gops);
/* Update pointer to next block of data to be processed. */
if (ctx->gcm_remainder_len != 0) {
datap += need;
ctx->gcm_remainder_len = 0;
} else {
datap += block_size;
}
remainder = (size_t)&data[length] - (size_t)datap;
/* Incomplete last block. */
if (remainder > 0 && remainder < block_size) {
bcopy(datap, ctx->gcm_remainder, remainder);
ctx->gcm_remainder_len = remainder;
ctx->gcm_copy_to = datap;
goto out;
}
ctx->gcm_copy_to = NULL;
} while (remainder > 0);
out:
return (CRYPTO_SUCCESS);
}
/* ARGSUSED */
int
gcm_encrypt_final(gcm_ctx_t *ctx, crypto_data_t *out, size_t block_size,
int (*encrypt_block)(const void *, const uint8_t *, uint8_t *),
void (*copy_block)(uint8_t *, uint8_t *),
void (*xor_block)(uint8_t *, uint8_t *))
{
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
if (ctx->gcm_use_avx == B_TRUE)
return (gcm_encrypt_final_avx(ctx, out, block_size));
#endif
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
const gcm_impl_ops_t *gops;
uint64_t counter_mask = ntohll(0x00000000ffffffffULL);
uint8_t *ghash, *macp = NULL;
int i, rv;
if (out->cd_length <
(ctx->gcm_remainder_len + ctx->gcm_tag_len)) {
return (CRYPTO_DATA_LEN_RANGE);
}
gops = gcm_impl_get_ops();
ghash = (uint8_t *)ctx->gcm_ghash;
if (ctx->gcm_remainder_len > 0) {
uint64_t counter;
uint8_t *tmpp = (uint8_t *)ctx->gcm_tmp;
/*
* Here is where we deal with data that is not a
* multiple of the block size.
*/
/*
* Increment counter.
*/
counter = ntohll(ctx->gcm_cb[1] & counter_mask);
counter = htonll(counter + 1);
counter &= counter_mask;
ctx->gcm_cb[1] = (ctx->gcm_cb[1] & ~counter_mask) | counter;
encrypt_block(ctx->gcm_keysched, (uint8_t *)ctx->gcm_cb,
(uint8_t *)ctx->gcm_tmp);
macp = (uint8_t *)ctx->gcm_remainder;
bzero(macp + ctx->gcm_remainder_len,
block_size - ctx->gcm_remainder_len);
/* XOR with counter block */
for (i = 0; i < ctx->gcm_remainder_len; i++) {
macp[i] ^= tmpp[i];
}
/* add ciphertext to the hash */
GHASH(ctx, macp, ghash, gops);
ctx->gcm_processed_data_len += ctx->gcm_remainder_len;
}
ctx->gcm_len_a_len_c[1] =
htonll(CRYPTO_BYTES2BITS(ctx->gcm_processed_data_len));
GHASH(ctx, ctx->gcm_len_a_len_c, ghash, gops);
encrypt_block(ctx->gcm_keysched, (uint8_t *)ctx->gcm_J0,
(uint8_t *)ctx->gcm_J0);
xor_block((uint8_t *)ctx->gcm_J0, ghash);
if (ctx->gcm_remainder_len > 0) {
rv = crypto_put_output_data(macp, out, ctx->gcm_remainder_len);
if (rv != CRYPTO_SUCCESS)
return (rv);
}
out->cd_offset += ctx->gcm_remainder_len;
ctx->gcm_remainder_len = 0;
rv = crypto_put_output_data(ghash, out, ctx->gcm_tag_len);
if (rv != CRYPTO_SUCCESS)
return (rv);
out->cd_offset += ctx->gcm_tag_len;
return (CRYPTO_SUCCESS);
}
/*
* This will only deal with decrypting the last block of the input that
* might not be a multiple of block length.
*/
static void
gcm_decrypt_incomplete_block(gcm_ctx_t *ctx, size_t block_size, size_t index,
int (*encrypt_block)(const void *, const uint8_t *, uint8_t *),
void (*xor_block)(uint8_t *, uint8_t *))
{
uint8_t *datap, *outp, *counterp;
uint64_t counter;
uint64_t counter_mask = ntohll(0x00000000ffffffffULL);
int i;
/*
* Increment counter.
* Counter bits are confined to the bottom 32 bits
*/
counter = ntohll(ctx->gcm_cb[1] & counter_mask);
counter = htonll(counter + 1);
counter &= counter_mask;
ctx->gcm_cb[1] = (ctx->gcm_cb[1] & ~counter_mask) | counter;
datap = (uint8_t *)ctx->gcm_remainder;
outp = &((ctx->gcm_pt_buf)[index]);
counterp = (uint8_t *)ctx->gcm_tmp;
/* authentication tag */
bzero((uint8_t *)ctx->gcm_tmp, block_size);
bcopy(datap, (uint8_t *)ctx->gcm_tmp, ctx->gcm_remainder_len);
/* add ciphertext to the hash */
GHASH(ctx, ctx->gcm_tmp, ctx->gcm_ghash, gcm_impl_get_ops());
/* decrypt remaining ciphertext */
encrypt_block(ctx->gcm_keysched, (uint8_t *)ctx->gcm_cb, counterp);
/* XOR with counter block */
for (i = 0; i < ctx->gcm_remainder_len; i++) {
outp[i] = datap[i] ^ counterp[i];
}
}
/* ARGSUSED */
int
gcm_mode_decrypt_contiguous_blocks(gcm_ctx_t *ctx, char *data, size_t length,
crypto_data_t *out, size_t block_size,
int (*encrypt_block)(const void *, const uint8_t *, uint8_t *),
void (*copy_block)(uint8_t *, uint8_t *),
void (*xor_block)(uint8_t *, uint8_t *))
{
size_t new_len;
uint8_t *new;
/*
* Copy contiguous ciphertext input blocks to plaintext buffer.
* Ciphertext will be decrypted in the final.
*/
if (length > 0) {
new_len = ctx->gcm_pt_buf_len + length;
new = vmem_alloc(new_len, ctx->gcm_kmflag);
if (new == NULL) {
vmem_free(ctx->gcm_pt_buf, ctx->gcm_pt_buf_len);
ctx->gcm_pt_buf = NULL;
return (CRYPTO_HOST_MEMORY);
}
bcopy(ctx->gcm_pt_buf, new, ctx->gcm_pt_buf_len);
vmem_free(ctx->gcm_pt_buf, ctx->gcm_pt_buf_len);
ctx->gcm_pt_buf = new;
ctx->gcm_pt_buf_len = new_len;
bcopy(data, &ctx->gcm_pt_buf[ctx->gcm_processed_data_len],
length);
ctx->gcm_processed_data_len += length;
}
ctx->gcm_remainder_len = 0;
return (CRYPTO_SUCCESS);
}
int
gcm_decrypt_final(gcm_ctx_t *ctx, crypto_data_t *out, size_t block_size,
int (*encrypt_block)(const void *, const uint8_t *, uint8_t *),
void (*xor_block)(uint8_t *, uint8_t *))
{
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
if (ctx->gcm_use_avx == B_TRUE)
return (gcm_decrypt_final_avx(ctx, out, block_size));
#endif
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
const gcm_impl_ops_t *gops;
size_t pt_len;
size_t remainder;
uint8_t *ghash;
uint8_t *blockp;
uint8_t *cbp;
uint64_t counter;
uint64_t counter_mask = ntohll(0x00000000ffffffffULL);
int processed = 0, rv;
ASSERT(ctx->gcm_processed_data_len == ctx->gcm_pt_buf_len);
gops = gcm_impl_get_ops();
pt_len = ctx->gcm_processed_data_len - ctx->gcm_tag_len;
ghash = (uint8_t *)ctx->gcm_ghash;
blockp = ctx->gcm_pt_buf;
remainder = pt_len;
while (remainder > 0) {
/* Incomplete last block */
if (remainder < block_size) {
bcopy(blockp, ctx->gcm_remainder, remainder);
ctx->gcm_remainder_len = remainder;
/*
* not expecting anymore ciphertext, just
* compute plaintext for the remaining input
*/
gcm_decrypt_incomplete_block(ctx, block_size,
processed, encrypt_block, xor_block);
ctx->gcm_remainder_len = 0;
goto out;
}
/* add ciphertext to the hash */
GHASH(ctx, blockp, ghash, gops);
/*
* Increment counter.
* Counter bits are confined to the bottom 32 bits
*/
counter = ntohll(ctx->gcm_cb[1] & counter_mask);
counter = htonll(counter + 1);
counter &= counter_mask;
ctx->gcm_cb[1] = (ctx->gcm_cb[1] & ~counter_mask) | counter;
cbp = (uint8_t *)ctx->gcm_tmp;
encrypt_block(ctx->gcm_keysched, (uint8_t *)ctx->gcm_cb, cbp);
/* XOR with ciphertext */
xor_block(cbp, blockp);
processed += block_size;
blockp += block_size;
remainder -= block_size;
}
out:
ctx->gcm_len_a_len_c[1] = htonll(CRYPTO_BYTES2BITS(pt_len));
GHASH(ctx, ctx->gcm_len_a_len_c, ghash, gops);
encrypt_block(ctx->gcm_keysched, (uint8_t *)ctx->gcm_J0,
(uint8_t *)ctx->gcm_J0);
xor_block((uint8_t *)ctx->gcm_J0, ghash);
/* compare the input authentication tag with what we calculated */
if (bcmp(&ctx->gcm_pt_buf[pt_len], ghash, ctx->gcm_tag_len)) {
/* They don't match */
return (CRYPTO_INVALID_MAC);
} else {
rv = crypto_put_output_data(ctx->gcm_pt_buf, out, pt_len);
if (rv != CRYPTO_SUCCESS)
return (rv);
out->cd_offset += pt_len;
}
return (CRYPTO_SUCCESS);
}
static int
gcm_validate_args(CK_AES_GCM_PARAMS *gcm_param)
{
size_t tag_len;
/*
* Check the length of the authentication tag (in bits).
*/
tag_len = gcm_param->ulTagBits;
switch (tag_len) {
case 32:
case 64:
case 96:
case 104:
case 112:
case 120:
case 128:
break;
default:
return (CRYPTO_MECHANISM_PARAM_INVALID);
}
if (gcm_param->ulIvLen == 0)
return (CRYPTO_MECHANISM_PARAM_INVALID);
return (CRYPTO_SUCCESS);
}
static void
gcm_format_initial_blocks(uchar_t *iv, ulong_t iv_len,
gcm_ctx_t *ctx, size_t block_size,
void (*copy_block)(uint8_t *, uint8_t *),
void (*xor_block)(uint8_t *, uint8_t *))
{
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
const gcm_impl_ops_t *gops;
uint8_t *cb;
ulong_t remainder = iv_len;
ulong_t processed = 0;
uint8_t *datap, *ghash;
uint64_t len_a_len_c[2];
gops = gcm_impl_get_ops();
ghash = (uint8_t *)ctx->gcm_ghash;
cb = (uint8_t *)ctx->gcm_cb;
if (iv_len == 12) {
bcopy(iv, cb, 12);
cb[12] = 0;
cb[13] = 0;
cb[14] = 0;
cb[15] = 1;
/* J0 will be used again in the final */
copy_block(cb, (uint8_t *)ctx->gcm_J0);
} else {
/* GHASH the IV */
do {
if (remainder < block_size) {
bzero(cb, block_size);
bcopy(&(iv[processed]), cb, remainder);
datap = (uint8_t *)cb;
remainder = 0;
} else {
datap = (uint8_t *)(&(iv[processed]));
processed += block_size;
remainder -= block_size;
}
GHASH(ctx, datap, ghash, gops);
} while (remainder > 0);
len_a_len_c[0] = 0;
len_a_len_c[1] = htonll(CRYPTO_BYTES2BITS(iv_len));
GHASH(ctx, len_a_len_c, ctx->gcm_J0, gops);
/* J0 will be used again in the final */
copy_block((uint8_t *)ctx->gcm_J0, (uint8_t *)cb);
}
}
/*
* The following function is called at encrypt or decrypt init time
* for AES GCM mode.
*/
int
gcm_init(gcm_ctx_t *ctx, unsigned char *iv, size_t iv_len,
unsigned char *auth_data, size_t auth_data_len, size_t block_size,
int (*encrypt_block)(const void *, const uint8_t *, uint8_t *),
void (*copy_block)(uint8_t *, uint8_t *),
void (*xor_block)(uint8_t *, uint8_t *))
{
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
const gcm_impl_ops_t *gops;
uint8_t *ghash, *datap, *authp;
size_t remainder, processed;
/* encrypt zero block to get subkey H */
bzero(ctx->gcm_H, sizeof (ctx->gcm_H));
encrypt_block(ctx->gcm_keysched, (uint8_t *)ctx->gcm_H,
(uint8_t *)ctx->gcm_H);
gcm_format_initial_blocks(iv, iv_len, ctx, block_size,
copy_block, xor_block);
gops = gcm_impl_get_ops();
authp = (uint8_t *)ctx->gcm_tmp;
ghash = (uint8_t *)ctx->gcm_ghash;
bzero(authp, block_size);
bzero(ghash, block_size);
processed = 0;
remainder = auth_data_len;
do {
if (remainder < block_size) {
/*
* There's not a block full of data, pad rest of
* buffer with zero
*/
bzero(authp, block_size);
bcopy(&(auth_data[processed]), authp, remainder);
datap = (uint8_t *)authp;
remainder = 0;
} else {
datap = (uint8_t *)(&(auth_data[processed]));
processed += block_size;
remainder -= block_size;
}
/* add auth data to the hash */
GHASH(ctx, datap, ghash, gops);
} while (remainder > 0);
return (CRYPTO_SUCCESS);
}
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
/*
* Init the GCM context struct. Handle the cycle and avx implementations here.
*/
int
gcm_init_ctx(gcm_ctx_t *gcm_ctx, char *param, size_t block_size,
int (*encrypt_block)(const void *, const uint8_t *, uint8_t *),
void (*copy_block)(uint8_t *, uint8_t *),
void (*xor_block)(uint8_t *, uint8_t *))
{
int rv;
CK_AES_GCM_PARAMS *gcm_param;
if (param != NULL) {
gcm_param = (CK_AES_GCM_PARAMS *)(void *)param;
if ((rv = gcm_validate_args(gcm_param)) != 0) {
return (rv);
}
gcm_ctx->gcm_tag_len = gcm_param->ulTagBits;
gcm_ctx->gcm_tag_len >>= 3;
gcm_ctx->gcm_processed_data_len = 0;
/* these values are in bits */
gcm_ctx->gcm_len_a_len_c[0]
= htonll(CRYPTO_BYTES2BITS(gcm_param->ulAADLen));
rv = CRYPTO_SUCCESS;
gcm_ctx->gcm_flags |= GCM_MODE;
} else {
return (CRYPTO_MECHANISM_PARAM_INVALID);
}
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
if (GCM_IMPL_READ(icp_gcm_impl) != IMPL_CYCLE) {
gcm_ctx->gcm_use_avx = GCM_IMPL_USE_AVX;
} else {
/*
* Handle the "cycle" implementation by creating avx and
* non-avx contexts alternately.
*/
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
gcm_ctx->gcm_use_avx = gcm_toggle_avx();
/*
* We don't handle byte swapped key schedules in the avx
* code path.
*/
aes_key_t *ks = (aes_key_t *)gcm_ctx->gcm_keysched;
if (ks->ops->needs_byteswap == B_TRUE) {
gcm_ctx->gcm_use_avx = B_FALSE;
}
/* Use the MOVBE and the BSWAP variants alternately. */
if (gcm_ctx->gcm_use_avx == B_TRUE &&
zfs_movbe_available() == B_TRUE) {
(void) atomic_toggle_boolean_nv(
(volatile boolean_t *)&gcm_avx_can_use_movbe);
}
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
}
/* Avx and non avx context initialization differs from here on. */
if (gcm_ctx->gcm_use_avx == B_FALSE) {
#endif /* ifdef CAN_USE_GCM_ASM */
if (gcm_init(gcm_ctx, gcm_param->pIv, gcm_param->ulIvLen,
gcm_param->pAAD, gcm_param->ulAADLen, block_size,
encrypt_block, copy_block, xor_block) != 0) {
rv = CRYPTO_MECHANISM_PARAM_INVALID;
}
#ifdef CAN_USE_GCM_ASM
} else {
if (gcm_init_avx(gcm_ctx, gcm_param->pIv, gcm_param->ulIvLen,
gcm_param->pAAD, gcm_param->ulAADLen, block_size) != 0) {
rv = CRYPTO_MECHANISM_PARAM_INVALID;
}
}
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#endif /* ifdef CAN_USE_GCM_ASM */
return (rv);
}
int
gmac_init_ctx(gcm_ctx_t *gcm_ctx, char *param, size_t block_size,
int (*encrypt_block)(const void *, const uint8_t *, uint8_t *),
void (*copy_block)(uint8_t *, uint8_t *),
void (*xor_block)(uint8_t *, uint8_t *))
{
int rv;
CK_AES_GMAC_PARAMS *gmac_param;
if (param != NULL) {
gmac_param = (CK_AES_GMAC_PARAMS *)(void *)param;
gcm_ctx->gcm_tag_len = CRYPTO_BITS2BYTES(AES_GMAC_TAG_BITS);
gcm_ctx->gcm_processed_data_len = 0;
/* these values are in bits */
gcm_ctx->gcm_len_a_len_c[0]
= htonll(CRYPTO_BYTES2BITS(gmac_param->ulAADLen));
rv = CRYPTO_SUCCESS;
gcm_ctx->gcm_flags |= GMAC_MODE;
} else {
return (CRYPTO_MECHANISM_PARAM_INVALID);
}
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
/*
* Handle the "cycle" implementation by creating avx and non avx
* contexts alternately.
*/
if (GCM_IMPL_READ(icp_gcm_impl) != IMPL_CYCLE) {
gcm_ctx->gcm_use_avx = GCM_IMPL_USE_AVX;
} else {
gcm_ctx->gcm_use_avx = gcm_toggle_avx();
}
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
/* We don't handle byte swapped key schedules in the avx code path. */
aes_key_t *ks = (aes_key_t *)gcm_ctx->gcm_keysched;
if (ks->ops->needs_byteswap == B_TRUE) {
gcm_ctx->gcm_use_avx = B_FALSE;
}
/* Avx and non avx context initialization differs from here on. */
if (gcm_ctx->gcm_use_avx == B_FALSE) {
#endif /* ifdef CAN_USE_GCM_ASM */
if (gcm_init(gcm_ctx, gmac_param->pIv, AES_GMAC_IV_LEN,
gmac_param->pAAD, gmac_param->ulAADLen, block_size,
encrypt_block, copy_block, xor_block) != 0) {
rv = CRYPTO_MECHANISM_PARAM_INVALID;
}
#ifdef CAN_USE_GCM_ASM
} else {
if (gcm_init_avx(gcm_ctx, gmac_param->pIv, AES_GMAC_IV_LEN,
gmac_param->pAAD, gmac_param->ulAADLen, block_size) != 0) {
rv = CRYPTO_MECHANISM_PARAM_INVALID;
}
}
#endif /* ifdef CAN_USE_GCM_ASM */
return (rv);
}
void *
gcm_alloc_ctx(int kmflag)
{
gcm_ctx_t *gcm_ctx;
if ((gcm_ctx = kmem_zalloc(sizeof (gcm_ctx_t), kmflag)) == NULL)
return (NULL);
gcm_ctx->gcm_flags = GCM_MODE;
return (gcm_ctx);
}
void *
gmac_alloc_ctx(int kmflag)
{
gcm_ctx_t *gcm_ctx;
if ((gcm_ctx = kmem_zalloc(sizeof (gcm_ctx_t), kmflag)) == NULL)
return (NULL);
gcm_ctx->gcm_flags = GMAC_MODE;
return (gcm_ctx);
}
void
gcm_set_kmflag(gcm_ctx_t *ctx, int kmflag)
{
ctx->gcm_kmflag = kmflag;
}
/* GCM implementation that contains the fastest methods */
static gcm_impl_ops_t gcm_fastest_impl = {
.name = "fastest"
};
/* All compiled in implementations */
const gcm_impl_ops_t *gcm_all_impl[] = {
&gcm_generic_impl,
#if defined(__x86_64) && defined(HAVE_PCLMULQDQ)
&gcm_pclmulqdq_impl,
#endif
};
/* Indicate that benchmark has been completed */
static boolean_t gcm_impl_initialized = B_FALSE;
/* Hold all supported implementations */
static size_t gcm_supp_impl_cnt = 0;
static gcm_impl_ops_t *gcm_supp_impl[ARRAY_SIZE(gcm_all_impl)];
/*
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
* Returns the GCM operations for encrypt/decrypt/key setup. When a
* SIMD implementation is not allowed in the current context, then
* fallback to the fastest generic implementation.
*/
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
const gcm_impl_ops_t *
gcm_impl_get_ops()
{
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
if (!kfpu_allowed())
return (&gcm_generic_impl);
const gcm_impl_ops_t *ops = NULL;
const uint32_t impl = GCM_IMPL_READ(icp_gcm_impl);
switch (impl) {
case IMPL_FASTEST:
ASSERT(gcm_impl_initialized);
ops = &gcm_fastest_impl;
break;
case IMPL_CYCLE:
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
/* Cycle through supported implementations */
ASSERT(gcm_impl_initialized);
ASSERT3U(gcm_supp_impl_cnt, >, 0);
static size_t cycle_impl_idx = 0;
size_t idx = (++cycle_impl_idx) % gcm_supp_impl_cnt;
ops = gcm_supp_impl[idx];
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
break;
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
case IMPL_AVX:
/*
* Make sure that we return a valid implementation while
* switching to the avx implementation since there still
* may be unfinished non-avx contexts around.
*/
ops = &gcm_generic_impl;
break;
#endif
default:
ASSERT3U(impl, <, gcm_supp_impl_cnt);
ASSERT3U(gcm_supp_impl_cnt, >, 0);
if (impl < ARRAY_SIZE(gcm_all_impl))
ops = gcm_supp_impl[impl];
break;
}
ASSERT3P(ops, !=, NULL);
return (ops);
}
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
/*
* Initialize all supported implementations.
*/
void
gcm_impl_init(void)
{
gcm_impl_ops_t *curr_impl;
int i, c;
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
/* Move supported implementations into gcm_supp_impls */
for (i = 0, c = 0; i < ARRAY_SIZE(gcm_all_impl); i++) {
curr_impl = (gcm_impl_ops_t *)gcm_all_impl[i];
if (curr_impl->is_supported())
gcm_supp_impl[c++] = (gcm_impl_ops_t *)curr_impl;
}
gcm_supp_impl_cnt = c;
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This is accomplished by leveraging the fact that by definition dedicated kernel threads never need to concern themselves with saving and restoring the user FPU state. Therefore, they may use the FPU as long as we can guarantee user tasks always restore their FPU state before context switching back to user space. For the 5.0 and 5.1 kernels disabling preemption and local interrupts is sufficient to allow the FPU to be used. All non-kernel threads will restore the preserved user FPU state. For 5.2 and latter kernels the user FPU state restoration will be skipped if the kernel determines the registers have not changed. Therefore, for these kernels we need to perform the additional step of saving and restoring the FPU registers. Invalidating the per-cpu global tracking the FPU state would force a restore but that functionality is private to the core x86 FPU implementation and unavailable. In practice, restricting SIMD to kernel threads is not a major restriction for ZFS. The vast majority of SIMD operations are already performed by the IO pipeline. The remaining cases are relatively infrequent and can be handled by the generic code without significant impact. The two most noteworthy cases are: 1) Decrypting the wrapping key for an encrypted dataset, i.e. `zfs load-key`. All other encryption and decryption operations will use the SIMD optimized implementations. 2) Generating the payload checksums for a `zfs send` stream. In order to avoid making any changes to the higher layers of ZFS all of the `*_get_ops()` functions were updated to take in to consideration the calling context. This allows for the fastest implementation to be used as appropriate (see kfpu_allowed()). The only other notable instance of SIMD operations being used outside a kernel thread was at module load time. This code was moved in to a taskq in order to accommodate the new kernel thread restriction. Finally, a few other modifications were made in order to further harden this code and facilitate testing. They include updating each implementations operations structure to be declared as a constant. And allowing "cycle" to be set when selecting the preferred ops in the kernel as well as user space. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8754 Closes #8793 Closes #8965
2019-07-12 16:31:20 +00:00
/*
* Set the fastest implementation given the assumption that the
* hardware accelerated version is the fastest.
*/
#if defined(__x86_64) && defined(HAVE_PCLMULQDQ)
if (gcm_pclmulqdq_impl.is_supported()) {
memcpy(&gcm_fastest_impl, &gcm_pclmulqdq_impl,
sizeof (gcm_fastest_impl));
} else
#endif
{
memcpy(&gcm_fastest_impl, &gcm_generic_impl,
sizeof (gcm_fastest_impl));
}
strcpy(gcm_fastest_impl.name, "fastest");
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
/*
* Use the avx implementation if it's available and the implementation
* hasn't changed from its default value of fastest on module load.
*/
if (gcm_avx_will_work()) {
#ifdef HAVE_MOVBE
if (zfs_movbe_available() == B_TRUE) {
atomic_swap_32(&gcm_avx_can_use_movbe, B_TRUE);
}
#endif
if (GCM_IMPL_READ(user_sel_impl) == IMPL_FASTEST) {
gcm_set_avx(B_TRUE);
}
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
}
#endif
/* Finish initialization */
atomic_swap_32(&icp_gcm_impl, user_sel_impl);
gcm_impl_initialized = B_TRUE;
}
static const struct {
char *name;
uint32_t sel;
} gcm_impl_opts[] = {
{ "cycle", IMPL_CYCLE },
{ "fastest", IMPL_FASTEST },
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
{ "avx", IMPL_AVX },
#endif
};
/*
* Function sets desired gcm implementation.
*
* If we are called before init(), user preference will be saved in
* user_sel_impl, and applied in later init() call. This occurs when module
* parameter is specified on module load. Otherwise, directly update
* icp_gcm_impl.
*
* @val Name of gcm implementation to use
* @param Unused.
*/
int
gcm_impl_set(const char *val)
{
int err = -EINVAL;
char req_name[GCM_IMPL_NAME_MAX];
uint32_t impl = GCM_IMPL_READ(user_sel_impl);
size_t i;
/* sanitize input */
i = strnlen(val, GCM_IMPL_NAME_MAX);
if (i == 0 || i >= GCM_IMPL_NAME_MAX)
return (err);
strlcpy(req_name, val, GCM_IMPL_NAME_MAX);
while (i > 0 && isspace(req_name[i-1]))
i--;
req_name[i] = '\0';
/* Check mandatory options */
for (i = 0; i < ARRAY_SIZE(gcm_impl_opts); i++) {
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
/* Ignore avx implementation if it won't work. */
if (gcm_impl_opts[i].sel == IMPL_AVX && !gcm_avx_will_work()) {
continue;
}
#endif
if (strcmp(req_name, gcm_impl_opts[i].name) == 0) {
impl = gcm_impl_opts[i].sel;
err = 0;
break;
}
}
/* check all supported impl if init() was already called */
if (err != 0 && gcm_impl_initialized) {
/* check all supported implementations */
for (i = 0; i < gcm_supp_impl_cnt; i++) {
if (strcmp(req_name, gcm_supp_impl[i]->name) == 0) {
impl = i;
err = 0;
break;
}
}
}
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
/*
* Use the avx implementation if available and the requested one is
* avx or fastest.
*/
if (gcm_avx_will_work() == B_TRUE &&
(impl == IMPL_AVX || impl == IMPL_FASTEST)) {
gcm_set_avx(B_TRUE);
} else {
gcm_set_avx(B_FALSE);
}
#endif
if (err == 0) {
if (gcm_impl_initialized)
atomic_swap_32(&icp_gcm_impl, impl);
else
atomic_swap_32(&user_sel_impl, impl);
}
return (err);
}
#if defined(_KERNEL)
static int
icp_gcm_impl_set(const char *val, zfs_kernel_param_t *kp)
{
return (gcm_impl_set(val));
}
static int
icp_gcm_impl_get(char *buffer, zfs_kernel_param_t *kp)
{
int i, cnt = 0;
char *fmt;
const uint32_t impl = GCM_IMPL_READ(icp_gcm_impl);
ASSERT(gcm_impl_initialized);
/* list mandatory options */
for (i = 0; i < ARRAY_SIZE(gcm_impl_opts); i++) {
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#ifdef CAN_USE_GCM_ASM
/* Ignore avx implementation if it won't work. */
if (gcm_impl_opts[i].sel == IMPL_AVX && !gcm_avx_will_work()) {
continue;
}
#endif
fmt = (impl == gcm_impl_opts[i].sel) ? "[%s] " : "%s ";
cnt += sprintf(buffer + cnt, fmt, gcm_impl_opts[i].name);
}
/* list all supported implementations */
for (i = 0; i < gcm_supp_impl_cnt; i++) {
fmt = (i == impl) ? "[%s] " : "%s ";
cnt += sprintf(buffer + cnt, fmt, gcm_supp_impl[i]->name);
}
return (cnt);
}
module_param_call(icp_gcm_impl, icp_gcm_impl_set, icp_gcm_impl_get,
NULL, 0644);
MODULE_PARM_DESC(icp_gcm_impl, "Select gcm implementation.");
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
#endif /* defined(__KERNEL) */
#ifdef CAN_USE_GCM_ASM
#define GCM_BLOCK_LEN 16
/*
* The openssl asm routines are 6x aggregated and need that many bytes
* at minimum.
*/
#define GCM_AVX_MIN_DECRYPT_BYTES (GCM_BLOCK_LEN * 6)
#define GCM_AVX_MIN_ENCRYPT_BYTES (GCM_BLOCK_LEN * 6 * 3)
/*
* Ensure the chunk size is reasonable since we are allocating a
* GCM_AVX_MAX_CHUNK_SIZEd buffer and disabling preemption and interrupts.
*/
#define GCM_AVX_MAX_CHUNK_SIZE \
(((128*1024)/GCM_AVX_MIN_DECRYPT_BYTES) * GCM_AVX_MIN_DECRYPT_BYTES)
/* Get the chunk size module parameter. */
#define GCM_CHUNK_SIZE_READ *(volatile uint32_t *) &gcm_avx_chunk_size
/* Clear the FPU registers since they hold sensitive internal state. */
#define clear_fpu_regs() clear_fpu_regs_avx()
#define GHASH_AVX(ctx, in, len) \
gcm_ghash_avx((ctx)->gcm_ghash, (const uint64_t (*)[2])(ctx)->gcm_Htable, \
in, len)
#define gcm_incr_counter_block(ctx) gcm_incr_counter_block_by(ctx, 1)
/*
* Module parameter: number of bytes to process at once while owning the FPU.
* Rounded down to the next GCM_AVX_MIN_DECRYPT_BYTES byte boundary and is
* ensured to be greater or equal than GCM_AVX_MIN_DECRYPT_BYTES.
*/
static uint32_t gcm_avx_chunk_size =
((32 * 1024) / GCM_AVX_MIN_DECRYPT_BYTES) * GCM_AVX_MIN_DECRYPT_BYTES;
extern void clear_fpu_regs_avx(void);
extern void gcm_xor_avx(const uint8_t *src, uint8_t *dst);
extern void aes_encrypt_intel(const uint32_t rk[], int nr,
const uint32_t pt[4], uint32_t ct[4]);
extern void gcm_init_htab_avx(uint64_t Htable[16][2], const uint64_t H[2]);
extern void gcm_ghash_avx(uint64_t ghash[2], const uint64_t Htable[16][2],
const uint8_t *in, size_t len);
extern size_t aesni_gcm_encrypt(const uint8_t *, uint8_t *, size_t,
const void *, uint64_t *, uint64_t *);
extern size_t aesni_gcm_decrypt(const uint8_t *, uint8_t *, size_t,
const void *, uint64_t *, uint64_t *);
static inline boolean_t
gcm_avx_will_work(void)
{
/* Avx should imply aes-ni and pclmulqdq, but make sure anyhow. */
return (kfpu_allowed() &&
zfs_avx_available() && zfs_aes_available() &&
zfs_pclmulqdq_available());
ICP: Improve AES-GCM performance Currently SIMD accelerated AES-GCM performance is limited by two factors: a. The need to disable preemption and interrupts and save the FPU state before using it and to do the reverse when done. Due to the way the code is organized (see (b) below) we have to pay this price twice for each 16 byte GCM block processed. b. Most processing is done in C, operating on single GCM blocks. The use of SIMD instructions is limited to the AES encryption of the counter block (AES-NI) and the Galois multiplication (PCLMULQDQ). This leads to the FPU not being fully utilized for crypto operations. To solve (a) we do crypto processing in larger chunks while owning the FPU. An `icp_gcm_avx_chunk_size` module parameter was introduced to make this chunk size tweakable. It defaults to 32 KiB. This step alone roughly doubles performance. (b) is tackled by porting and using the highly optimized openssl AES-GCM assembler routines, which do all the processing (CTR, AES, GMULT) in a single routine. Both steps together result in up to 32x reduction of the time spend in the en/decryption routines, leading up to approximately 12x throughput increase for large (128 KiB) blocks. Lastly, this commit changes the default encryption algorithm from AES-CCM to AES-GCM when setting the `encryption=on` property. Reviewed-By: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-By: Jason King <jason.king@joyent.com> Reviewed-By: Tom Caputi <tcaputi@datto.com> Reviewed-By: Richard Laager <rlaager@wiktel.com> Signed-off-by: Attila Fülöp <attila@fueloep.org> Closes #9749
2020-02-10 20:59:50 +00:00
}
static inline void
gcm_set_avx(boolean_t val)
{
if (gcm_avx_will_work() == B_TRUE) {
atomic_swap_32(&gcm_use_avx, val);
}
}
static inline boolean_t
gcm_toggle_avx(void)
{
if (gcm_avx_will_work() == B_TRUE) {
return (atomic_toggle_boolean_nv(&GCM_IMPL_USE_AVX));
} else {
return (B_FALSE);
}
}
/*
* Clear senssitve data in the context.
*
* ctx->gcm_remainder may contain a plaintext remainder. ctx->gcm_H and
* ctx->gcm_Htable contain the hash sub key which protects authentication.
*
* Although extremely unlikely, ctx->gcm_J0 and ctx->gcm_tmp could be used for
* a known plaintext attack, they consists of the IV and the first and last
* counter respectively. If they should be cleared is debatable.
*/
static inline void
gcm_clear_ctx(gcm_ctx_t *ctx)
{
bzero(ctx->gcm_remainder, sizeof (ctx->gcm_remainder));
bzero(ctx->gcm_H, sizeof (ctx->gcm_H));
bzero(ctx->gcm_Htable, sizeof (ctx->gcm_Htable));
bzero(ctx->gcm_J0, sizeof (ctx->gcm_J0));
bzero(ctx->gcm_tmp, sizeof (ctx->gcm_tmp));
}
/* Increment the GCM counter block by n. */
static inline void
gcm_incr_counter_block_by(gcm_ctx_t *ctx, int n)
{
uint64_t counter_mask = ntohll(0x00000000ffffffffULL);
uint64_t counter = ntohll(ctx->gcm_cb[1] & counter_mask);
counter = htonll(counter + n);
counter &= counter_mask;
ctx->gcm_cb[1] = (ctx->gcm_cb[1] & ~counter_mask) | counter;
}
/*
* Encrypt multiple blocks of data in GCM mode.
* This is done in gcm_avx_chunk_size chunks, utilizing AVX assembler routines
* if possible. While processing a chunk the FPU is "locked".
*/
static int
gcm_mode_encrypt_contiguous_blocks_avx(gcm_ctx_t *ctx, char *data,
size_t length, crypto_data_t *out, size_t block_size)
{
size_t bleft = length;
size_t need = 0;
size_t done = 0;
uint8_t *datap = (uint8_t *)data;
size_t chunk_size = (size_t)GCM_CHUNK_SIZE_READ;
const aes_key_t *key = ((aes_key_t *)ctx->gcm_keysched);
uint64_t *ghash = ctx->gcm_ghash;
uint64_t *cb = ctx->gcm_cb;
uint8_t *ct_buf = NULL;
uint8_t *tmp = (uint8_t *)ctx->gcm_tmp;
int rv = CRYPTO_SUCCESS;
ASSERT(block_size == GCM_BLOCK_LEN);
/*
* If the last call left an incomplete block, try to fill
* it first.
*/
if (ctx->gcm_remainder_len > 0) {
need = block_size - ctx->gcm_remainder_len;
if (length < need) {
/* Accumulate bytes here and return. */
bcopy(datap, (uint8_t *)ctx->gcm_remainder +
ctx->gcm_remainder_len, length);
ctx->gcm_remainder_len += length;
if (ctx->gcm_copy_to == NULL) {
ctx->gcm_copy_to = datap;
}
return (CRYPTO_SUCCESS);
} else {
/* Complete incomplete block. */
bcopy(datap, (uint8_t *)ctx->gcm_remainder +
ctx->gcm_remainder_len, need);
ctx->gcm_copy_to = NULL;
}
}
/* Allocate a buffer to encrypt to if there is enough input. */
if (bleft >= GCM_AVX_MIN_ENCRYPT_BYTES) {
ct_buf = vmem_alloc(chunk_size, ctx->gcm_kmflag);
if (ct_buf == NULL) {
return (CRYPTO_HOST_MEMORY);
}
}
/* If we completed an incomplete block, encrypt and write it out. */
if (ctx->gcm_remainder_len > 0) {
kfpu_begin();
aes_encrypt_intel(key->encr_ks.ks32, key->nr,
(const uint32_t *)cb, (uint32_t *)tmp);
gcm_xor_avx((const uint8_t *) ctx->gcm_remainder, tmp);
GHASH_AVX(ctx, tmp, block_size);
clear_fpu_regs();
kfpu_end();
/*
* We don't follow gcm_mode_encrypt_contiguous_blocks() here
* but assert that out is not null.
* See gcm_mode_encrypt_contiguous_blocks() above and
* https://github.com/zfsonlinux/zfs/issues/9661
*/
ASSERT(out != NULL);
rv = crypto_put_output_data(tmp, out, block_size);
out->cd_offset += block_size;
gcm_incr_counter_block(ctx);
ctx->gcm_processed_data_len += block_size;
bleft -= need;
datap += need;
ctx->gcm_remainder_len = 0;
}
/* Do the bulk encryption in chunk_size blocks. */
for (; bleft >= chunk_size; bleft -= chunk_size) {
kfpu_begin();
done = aesni_gcm_encrypt(
datap, ct_buf, chunk_size, key, cb, ghash);
clear_fpu_regs();
kfpu_end();
if (done != chunk_size) {
rv = CRYPTO_FAILED;
goto out_nofpu;
}
if (out != NULL) {
rv = crypto_put_output_data(ct_buf, out, chunk_size);
if (rv != CRYPTO_SUCCESS) {
goto out_nofpu;
}
out->cd_offset += chunk_size;
}
datap += chunk_size;
ctx->gcm_processed_data_len += chunk_size;
}
/* Check if we are already done. */
if (bleft == 0) {
goto out_nofpu;
}
/* Bulk encrypt the remaining data. */
kfpu_begin();
if (bleft >= GCM_AVX_MIN_ENCRYPT_BYTES) {
done = aesni_gcm_encrypt(datap, ct_buf, bleft, key, cb, ghash);
if (done == 0) {
rv = CRYPTO_FAILED;
goto out;
}
if (out != NULL) {
rv = crypto_put_output_data(ct_buf, out, done);
if (rv != CRYPTO_SUCCESS) {
goto out;
}
out->cd_offset += done;
}
ctx->gcm_processed_data_len += done;
datap += done;
bleft -= done;
}
/* Less than GCM_AVX_MIN_ENCRYPT_BYTES remain, operate on blocks. */
while (bleft > 0) {
if (bleft < block_size) {
bcopy(datap, ctx->gcm_remainder, bleft);
ctx->gcm_remainder_len = bleft;
ctx->gcm_copy_to = datap;
goto out;
}
/* Encrypt, hash and write out. */
aes_encrypt_intel(key->encr_ks.ks32, key->nr,
(const uint32_t *)cb, (uint32_t *)tmp);
gcm_xor_avx(datap, tmp);
GHASH_AVX(ctx, tmp, block_size);
if (out != NULL) {
rv = crypto_put_output_data(tmp, out, block_size);
if (rv != CRYPTO_SUCCESS) {
goto out;
}
out->cd_offset += block_size;
}
gcm_incr_counter_block(ctx);
ctx->gcm_processed_data_len += block_size;
datap += block_size;
bleft -= block_size;
}
out:
clear_fpu_regs();
kfpu_end();
out_nofpu:
if (ct_buf != NULL) {
vmem_free(ct_buf, chunk_size);
}
return (rv);
}
/*
* Finalize the encryption: Zero fill, encrypt, hash and write out an eventual
* incomplete last block. Encrypt the ICB. Calculate the tag and write it out.
*/
static int
gcm_encrypt_final_avx(gcm_ctx_t *ctx, crypto_data_t *out, size_t block_size)
{
uint8_t *ghash = (uint8_t *)ctx->gcm_ghash;
uint32_t *J0 = (uint32_t *)ctx->gcm_J0;
uint8_t *remainder = (uint8_t *)ctx->gcm_remainder;
size_t rem_len = ctx->gcm_remainder_len;
const void *keysched = ((aes_key_t *)ctx->gcm_keysched)->encr_ks.ks32;
int aes_rounds = ((aes_key_t *)keysched)->nr;
int rv;
ASSERT(block_size == GCM_BLOCK_LEN);
if (out->cd_length < (rem_len + ctx->gcm_tag_len)) {
return (CRYPTO_DATA_LEN_RANGE);
}
kfpu_begin();
/* Pad last incomplete block with zeros, encrypt and hash. */
if (rem_len > 0) {
uint8_t *tmp = (uint8_t *)ctx->gcm_tmp;
const uint32_t *cb = (uint32_t *)ctx->gcm_cb;
aes_encrypt_intel(keysched, aes_rounds, cb, (uint32_t *)tmp);
bzero(remainder + rem_len, block_size - rem_len);
for (int i = 0; i < rem_len; i++) {
remainder[i] ^= tmp[i];
}
GHASH_AVX(ctx, remainder, block_size);
ctx->gcm_processed_data_len += rem_len;
/* No need to increment counter_block, it's the last block. */
}
/* Finish tag. */
ctx->gcm_len_a_len_c[1] =
htonll(CRYPTO_BYTES2BITS(ctx->gcm_processed_data_len));
GHASH_AVX(ctx, (const uint8_t *)ctx->gcm_len_a_len_c, block_size);
aes_encrypt_intel(keysched, aes_rounds, J0, J0);
gcm_xor_avx((uint8_t *)J0, ghash);
clear_fpu_regs();
kfpu_end();
/* Output remainder. */
if (rem_len > 0) {
rv = crypto_put_output_data(remainder, out, rem_len);
if (rv != CRYPTO_SUCCESS)
return (rv);
}
out->cd_offset += rem_len;
ctx->gcm_remainder_len = 0;
rv = crypto_put_output_data(ghash, out, ctx->gcm_tag_len);
if (rv != CRYPTO_SUCCESS)
return (rv);
out->cd_offset += ctx->gcm_tag_len;
/* Clear sensitive data in the context before returning. */
gcm_clear_ctx(ctx);
return (CRYPTO_SUCCESS);
}
/*
* Finalize decryption: We just have accumulated crypto text, so now we
* decrypt it here inplace.
*/
static int
gcm_decrypt_final_avx(gcm_ctx_t *ctx, crypto_data_t *out, size_t block_size)
{
ASSERT3U(ctx->gcm_processed_data_len, ==, ctx->gcm_pt_buf_len);
ASSERT3U(block_size, ==, 16);
size_t chunk_size = (size_t)GCM_CHUNK_SIZE_READ;
size_t pt_len = ctx->gcm_processed_data_len - ctx->gcm_tag_len;
uint8_t *datap = ctx->gcm_pt_buf;
const aes_key_t *key = ((aes_key_t *)ctx->gcm_keysched);
uint32_t *cb = (uint32_t *)ctx->gcm_cb;
uint64_t *ghash = ctx->gcm_ghash;
uint32_t *tmp = (uint32_t *)ctx->gcm_tmp;
int rv = CRYPTO_SUCCESS;
size_t bleft, done;
/*
* Decrypt in chunks of gcm_avx_chunk_size, which is asserted to be
* greater or equal than GCM_AVX_MIN_ENCRYPT_BYTES, and a multiple of
* GCM_AVX_MIN_DECRYPT_BYTES.
*/
for (bleft = pt_len; bleft >= chunk_size; bleft -= chunk_size) {
kfpu_begin();
done = aesni_gcm_decrypt(datap, datap, chunk_size,
(const void *)key, ctx->gcm_cb, ghash);
clear_fpu_regs();
kfpu_end();
if (done != chunk_size) {
return (CRYPTO_FAILED);
}
datap += done;
}
/* Decrypt remainder, which is less then chunk size, in one go. */
kfpu_begin();
if (bleft >= GCM_AVX_MIN_DECRYPT_BYTES) {
done = aesni_gcm_decrypt(datap, datap, bleft,
(const void *)key, ctx->gcm_cb, ghash);
if (done == 0) {
clear_fpu_regs();
kfpu_end();
return (CRYPTO_FAILED);
}
datap += done;
bleft -= done;
}
ASSERT(bleft < GCM_AVX_MIN_DECRYPT_BYTES);
/*
* Now less then GCM_AVX_MIN_DECRYPT_BYTES bytes remain,
* decrypt them block by block.
*/
while (bleft > 0) {
/* Incomplete last block. */
if (bleft < block_size) {
uint8_t *lastb = (uint8_t *)ctx->gcm_remainder;
bzero(lastb, block_size);
bcopy(datap, lastb, bleft);
/* The GCM processing. */
GHASH_AVX(ctx, lastb, block_size);
aes_encrypt_intel(key->encr_ks.ks32, key->nr, cb, tmp);
for (size_t i = 0; i < bleft; i++) {
datap[i] = lastb[i] ^ ((uint8_t *)tmp)[i];
}
break;
}
/* The GCM processing. */
GHASH_AVX(ctx, datap, block_size);
aes_encrypt_intel(key->encr_ks.ks32, key->nr, cb, tmp);
gcm_xor_avx((uint8_t *)tmp, datap);
gcm_incr_counter_block(ctx);
datap += block_size;
bleft -= block_size;
}
if (rv != CRYPTO_SUCCESS) {
clear_fpu_regs();
kfpu_end();
return (rv);
}
/* Decryption done, finish the tag. */
ctx->gcm_len_a_len_c[1] = htonll(CRYPTO_BYTES2BITS(pt_len));
GHASH_AVX(ctx, (uint8_t *)ctx->gcm_len_a_len_c, block_size);
aes_encrypt_intel(key->encr_ks.ks32, key->nr, (uint32_t *)ctx->gcm_J0,
(uint32_t *)ctx->gcm_J0);
gcm_xor_avx((uint8_t *)ctx->gcm_J0, (uint8_t *)ghash);
/* We are done with the FPU, restore its state. */
clear_fpu_regs();
kfpu_end();
/* Compare the input authentication tag with what we calculated. */
if (bcmp(&ctx->gcm_pt_buf[pt_len], ghash, ctx->gcm_tag_len)) {
/* They don't match. */
return (CRYPTO_INVALID_MAC);
}
rv = crypto_put_output_data(ctx->gcm_pt_buf, out, pt_len);
if (rv != CRYPTO_SUCCESS) {
return (rv);
}
out->cd_offset += pt_len;
gcm_clear_ctx(ctx);
return (CRYPTO_SUCCESS);
}
/*
* Initialize the GCM params H, Htabtle and the counter block. Save the
* initial counter block.
*/
static int
gcm_init_avx(gcm_ctx_t *ctx, unsigned char *iv, size_t iv_len,
unsigned char *auth_data, size_t auth_data_len, size_t block_size)
{
uint8_t *cb = (uint8_t *)ctx->gcm_cb;
uint64_t *H = ctx->gcm_H;
const void *keysched = ((aes_key_t *)ctx->gcm_keysched)->encr_ks.ks32;
int aes_rounds = ((aes_key_t *)ctx->gcm_keysched)->nr;
uint8_t *datap = auth_data;
size_t chunk_size = (size_t)GCM_CHUNK_SIZE_READ;
size_t bleft;
ASSERT(block_size == GCM_BLOCK_LEN);
/* Init H (encrypt zero block) and create the initial counter block. */
bzero(ctx->gcm_ghash, sizeof (ctx->gcm_ghash));
bzero(H, sizeof (ctx->gcm_H));
kfpu_begin();
aes_encrypt_intel(keysched, aes_rounds,
(const uint32_t *)H, (uint32_t *)H);
gcm_init_htab_avx(ctx->gcm_Htable, H);
if (iv_len == 12) {
bcopy(iv, cb, 12);
cb[12] = 0;
cb[13] = 0;
cb[14] = 0;
cb[15] = 1;
/* We need the ICB later. */
bcopy(cb, ctx->gcm_J0, sizeof (ctx->gcm_J0));
} else {
/*
* Most consumers use 12 byte IVs, so it's OK to use the
* original routines for other IV sizes, just avoid nesting
* kfpu_begin calls.
*/
clear_fpu_regs();
kfpu_end();
gcm_format_initial_blocks(iv, iv_len, ctx, block_size,
aes_copy_block, aes_xor_block);
kfpu_begin();
}
/* Openssl post increments the counter, adjust for that. */
gcm_incr_counter_block(ctx);
/* Ghash AAD in chunk_size blocks. */
for (bleft = auth_data_len; bleft >= chunk_size; bleft -= chunk_size) {
GHASH_AVX(ctx, datap, chunk_size);
datap += chunk_size;
clear_fpu_regs();
kfpu_end();
kfpu_begin();
}
/* Ghash the remainder and handle possible incomplete GCM block. */
if (bleft > 0) {
size_t incomp = bleft % block_size;
bleft -= incomp;
if (bleft > 0) {
GHASH_AVX(ctx, datap, bleft);
datap += bleft;
}
if (incomp > 0) {
/* Zero pad and hash incomplete last block. */
uint8_t *authp = (uint8_t *)ctx->gcm_tmp;
bzero(authp, block_size);
bcopy(datap, authp, incomp);
GHASH_AVX(ctx, authp, block_size);
}
}
clear_fpu_regs();
kfpu_end();
return (CRYPTO_SUCCESS);
}
#if defined(_KERNEL)
static int
icp_gcm_avx_set_chunk_size(const char *buf, zfs_kernel_param_t *kp)
{
unsigned long val;
char val_rounded[16];
int error = 0;
error = kstrtoul(buf, 0, &val);
if (error)
return (error);
val = (val / GCM_AVX_MIN_DECRYPT_BYTES) * GCM_AVX_MIN_DECRYPT_BYTES;
if (val < GCM_AVX_MIN_ENCRYPT_BYTES || val > GCM_AVX_MAX_CHUNK_SIZE)
return (-EINVAL);
snprintf(val_rounded, 16, "%u", (uint32_t)val);
error = param_set_uint(val_rounded, kp);
return (error);
}
module_param_call(icp_gcm_avx_chunk_size, icp_gcm_avx_set_chunk_size,
param_get_uint, &gcm_avx_chunk_size, 0644);
MODULE_PARM_DESC(icp_gcm_avx_chunk_size,
"How many bytes to process while owning the FPU");
#endif /* defined(__KERNEL) */
#endif /* ifdef CAN_USE_GCM_ASM */