Age | Commit message (Collapse) | Author | Files | Lines |
|
Move all users of ablk_helper under x86/ to the generic version
and delete the x86 specific version.
Acked-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Commit 3d387ef08c4 (Revert "crypto: blowfish - add AVX2/x86_64 implementation
of blowfish cipher") reverted too much as it removed the 'assembler supports
AVX2' check and therefore disabled remaining AVX2 implementations of Camellia
and Serpent. Patch restores the check and enables these implementations.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
transform framework"
This patch reinstates commits
67822649d7305caf3dd50ed46c27b99c94eff996
39761214eefc6b070f29402aa1165f24d789b3f7
0b95a7f85718adcbba36407ef88bba0a7379ed03
31d939625a9a20b1badd2d4e6bf6fd39fa523405
2d31e518a42828df7877bca23a958627d60408bc
Now that module softdeps are in the kernel we can use that to resolve
the boot issue which cause the revert.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
transform framework"
This reverts commits
67822649d7305caf3dd50ed46c27b99c94eff996
39761214eefc6b070f29402aa1165f24d789b3f7
0b95a7f85718adcbba36407ef88bba0a7379ed03
31d939625a9a20b1badd2d4e6bf6fd39fa523405
2d31e518a42828df7877bca23a958627d60408bc
Unfortunately this change broke boot on some systems that used an
initrd which does not include the newly created crct10dif modules.
As these modules are required by sd_mod under certain configurations
this is a serious problem.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
twofish cipher"
This reverts commit cf1521a1a5e21fd1e79a458605c4282fbfbbeee2.
Instruction (vpgatherdd) that this implementation relied on turned out to be
slow performer on real hardware (i5-4570). The previous 8-way twofish/AVX
implementation is therefore faster and this implementation should be removed.
Converting this implementation to use the same method as in twofish/AVX for
table look-ups would give additional ~3% speed up vs twofish/AVX, but would
hardly be worth of the added code and binary size.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This reverts commit 604880107010a1e5794552d184cd5471ea31b973.
Instruction (vpgatherdd) that this implementation relied on turned out to be
slow performer on real hardware (i5-4570). The previous 4-way blowfish
implementation is therefore faster and this implementation should be removed.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
crypto transform
Glue code that plugs the PCLMULQDQ accelerated CRC T10 DIF hash into the
crypto framework. The config CRYPTO_CRCT10DIF_PCLMUL should be turned
on to enable the feature. The crc_t10dif crypto library function will
use this faster algorithm when crct10dif_pclmul module is loaded.
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
camellia cipher
Patch adds AVX2/AES-NI/x86-64 implementation of Camellia cipher, requiring
32 parallel blocks for input (512 bytes). Compared to AVX implementation, this
version is extended to use the 256-bit wide YMM registers. For AES-NI
instructions data is split to two 128-bit registers and merged afterwards.
Even with this additional handling, performance should be higher compared
to the AES-NI/AVX implementation.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Patch adds AVX2/x86-64 implementation of Serpent cipher, requiring 16 parallel
blocks for input (256 bytes). Implementation is based on the AVX implementation
and extends to use the 256-bit wide YMM registers. Since serpent does not use
table look-ups, this implementation should be close to two times faster than
the AVX implementation.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Patch adds AVX2/x86-64 implementation of Twofish cipher, requiring 16 parallel
blocks for input (256 bytes). Table look-ups are performed using vpgatherdd
instruction directly from vector registers and thus should be faster than
earlier implementations. Implementation also uses 256-bit wide YMM registers,
which should give additional speed up compared to the AVX implementation.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Patch adds AVX2/x86-64 implementation of Blowfish cipher, requiring 32 parallel
blocks for input (256 bytes). Table look-ups are performed using vpgatherdd
instruction directly from vector registers and thus should be faster than
earlier implementations.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
SSSE3, AVX or AVX2 instructions.
We added glue code and config options to create crypto
module that uses SSE/AVX/AVX2 optimized SHA512 x86_64 assembly routines.
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
SSSE3, AVX or AVX2 instructions.
We added glue code and config options to create crypto
module that uses SSE/AVX/AVX2 optimized SHA256 x86_64 assembly routines.
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
supports AVX instructions
These modules require AVX support in assembler, so add new check to Makefile
for this.
Other option would be to use CONFIG_AS_AVX inside source files, but that would
result dummy/empty/no-fuctionality modules being created.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This bool option can never be set to anything other than y. So
let's just kill it.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
implementation
This patch adds crc32 algorithms to shash crypto api. One is wrapper to
gerneric crc32_le function. Second is crc32 pclmulqdq implementation. It
use hardware provided PCLMULQDQ instruction to accelerate the CRC32 disposal.
This instruction present from Intel Westmere and AMD Bulldozer CPUs.
For intel core i5 I got 450MB/s for table implementation and 2100MB/s
for pclmulqdq implementation.
Signed-off-by: Alexander Boyko <alexander_boyko@xyratex.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
camellia cipher
This patch adds AES-NI/AVX/x86_64 assembler implementation of Camellia block
cipher. Implementation process data in sixteen block chunks, which are
byte-sliced and AES SubBytes is reused for Camellia s-box with help of pre-
and post-filtering.
Patch has been tested with tcrypt and automated filesystem tests.
tcrypt test results:
Intel Core i5-2450M:
camellia-aesni-avx vs camellia-asm-x86_64-2way:
128bit key: (lrw:256bit) (xts:256bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 0.98x 0.96x 0.99x 0.96x 0.96x 0.95x 0.95x 0.94x 0.97x 0.98x
64B 0.99x 0.98x 1.00x 0.98x 0.98x 0.99x 0.98x 0.93x 0.99x 0.98x
256B 2.28x 2.28x 1.01x 2.29x 2.25x 2.24x 1.96x 1.97x 1.91x 1.90x
1024B 2.57x 2.56x 1.00x 2.57x 2.51x 2.53x 2.19x 2.17x 2.19x 2.22x
8192B 2.49x 2.49x 1.00x 2.53x 2.48x 2.49x 2.17x 2.17x 2.22x 2.22x
256bit key: (lrw:384bit) (xts:512bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 0.97x 0.98x 0.99x 0.97x 0.97x 0.96x 0.97x 0.98x 0.98x 0.99x
64B 1.00x 1.00x 1.01x 0.99x 0.98x 0.99x 0.99x 0.99x 0.99x 0.99x
256B 2.37x 2.37x 1.01x 2.39x 2.35x 2.33x 2.10x 2.11x 1.99x 2.02x
1024B 2.58x 2.60x 1.00x 2.58x 2.56x 2.56x 2.28x 2.29x 2.28x 2.29x
8192B 2.50x 2.52x 1.00x 2.56x 2.51x 2.51x 2.24x 2.25x 2.26x 2.29x
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This patch adds the crc_pcl function that calculates CRC32C checksum using the
PCLMULQDQ instruction on processors that support this feature. This will
provide speedup over using CRC32 instruction only.
The usage of PCLMULQDQ necessitate the invocation of kernel_fpu_begin and
kernel_fpu_end and incur some overhead. So the new crc_pcl function is only
invoked for buffer size of 512 bytes or more. Larger sized
buffers will expect to see greater speedup. This feature is best used coupled
with eager_fpu which reduces the kernel_fpu_begin/end overhead. For
buffer size of 1K the speedup is around 1.6x and for buffer size greater than
4K, the speedup is around 3x compared to original implementation in crc32c-intel
module. Test was performed on Sandy Bridge based platform with constant frequency
set for cpu.
A white paper detailing the algorithm can be found here:
http://download.intel.com/design/intarch/papers/323405.pdf
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This patch renames the crc32c-intel.c file to crc32c-intel_glue.c file
in preparation for linking with the new crc32c-pcl-intel-asm.S file,
which contains optimized crc32c calculation based on PCLMULQDQ
instruction.
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This patch adds a x86_64/avx assembler implementation of the Cast6 block
cipher. The implementation processes eight blocks in parallel (two 4 block
chunk AVX operations). The table-lookups are done in general-purpose registers.
For small blocksizes the functions from the generic module are called. A good
performance increase is provided for blocksizes greater or equal to 128B.
Patch has been tested with tcrypt and automated filesystem tests.
Tcrypt benchmark results:
Intel Core i5-2500 CPU (fam:6, model:42, step:7)
cast6-avx-x86_64 vs. cast6-generic
128bit key: (lrw:256bit) (xts:256bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 0.97x 1.00x 1.01x 1.01x 0.99x 0.97x 0.98x 1.01x 0.96x 0.98x
64B 0.98x 0.99x 1.02x 1.01x 0.99x 1.00x 1.01x 0.99x 1.00x 0.99x
256B 1.77x 1.84x 0.99x 1.85x 1.77x 1.77x 1.70x 1.74x 1.69x 1.72x
1024B 1.93x 1.95x 0.99x 1.96x 1.93x 1.93x 1.84x 1.85x 1.89x 1.87x
8192B 1.91x 1.95x 0.99x 1.97x 1.95x 1.91x 1.86x 1.87x 1.93x 1.90x
256bit key: (lrw:384bit) (xts:512bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 0.97x 0.99x 1.02x 1.01x 0.98x 0.99x 1.00x 1.00x 0.98x 0.98x
64B 0.98x 0.99x 1.01x 1.00x 1.00x 1.00x 1.01x 1.01x 0.97x 1.00x
256B 1.77x 1.83x 1.00x 1.86x 1.79x 1.78x 1.70x 1.76x 1.71x 1.69x
1024B 1.92x 1.95x 0.99x 1.96x 1.93x 1.93x 1.83x 1.86x 1.89x 1.87x
8192B 1.94x 1.95x 0.99x 1.97x 1.95x 1.95x 1.87x 1.87x 1.93x 1.91x
Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This patch adds a x86_64/avx assembler implementation of the Cast5 block
cipher. The implementation processes sixteen blocks in parallel (four 4 block
chunk AVX operations). The table-lookups are done in general-purpose registers.
For small blocksizes the functions from the generic module are called. A good
performance increase is provided for blocksizes greater or equal to 128B.
Patch has been tested with tcrypt and automated filesystem tests.
Tcrypt benchmark results:
Intel Core i5-2500 CPU (fam:6, model:42, step:7)
cast5-avx-x86_64 vs. cast5-generic
64bit key:
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec
16B 0.99x 0.99x 1.00x 1.00x 1.02x 1.01x
64B 1.00x 1.00x 0.98x 1.00x 1.01x 1.02x
256B 2.03x 2.01x 0.95x 2.11x 2.12x 2.13x
1024B 2.30x 2.24x 0.95x 2.29x 2.35x 2.35x
8192B 2.31x 2.27x 0.95x 2.31x 2.39x 2.39x
128bit key:
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec
16B 0.99x 0.99x 1.00x 1.00x 1.01x 1.01x
64B 1.00x 1.00x 0.98x 1.01x 1.02x 1.01x
256B 2.17x 2.13x 0.96x 2.19x 2.19x 2.19x
1024B 2.29x 2.32x 0.95x 2.34x 2.37x 2.38x
8192B 2.35x 2.32x 0.95x 2.35x 2.39x 2.39x
Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Now that serpent-sse2 glue code has been made generic, it can be split to
separate module.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
to shared module
Move ablk-* functions to separate module to share common code between cipher
implementations.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This patch adds a x86_64/avx assembler implementation of the Serpent block
cipher. The implementation is very similar to the sse2 implementation and
processes eight blocks in parallel. Because of the new non-destructive three
operand syntax all move-instructions can be removed and therefore a little
performance increase is provided.
Patch has been tested with tcrypt and automated filesystem tests.
Tcrypt benchmark results:
Intel Core i5-2500 CPU (fam:6, model:42, step:7)
serpent-avx-x86_64 vs. serpent-sse2-x86_64
128bit key: (lrw:256bit) (xts:256bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x
64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x
256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x
1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x
8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x
256bit key: (lrw:384bit) (xts:512bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x
64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x
256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x
1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x
8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x
serpent-avx-x86_64 vs aes-asm (8kB block):
128bit 256bit
ecb-enc 1.26x 1.73x
ecb-dec 1.20x 1.64x
cbc-enc 0.33x 0.45x
cbc-dec 1.24x 1.67x
ctr-enc 1.32x 1.76x
ctr-dec 1.32x 1.76x
lrw-enc 1.20x 1.60x
lrw-dec 1.15x 1.54x
xts-enc 1.22x 1.64x
xts-dec 1.17x 1.57x
Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This patch adds a x86_64/avx assembler implementation of the Twofish block
cipher. The implementation processes eight blocks in parallel (two 4 block
chunk AVX operations). The table-lookups are done in general-purpose registers.
For small blocksizes the 3way-parallel functions from the twofish-x86_64-3way
module are called. A good performance increase is provided for blocksizes
greater or equal to 128B.
Patch has been tested with tcrypt and automated filesystem tests.
Tcrypt benchmark results:
Intel Core i5-2500 CPU (fam:6, model:42, step:7)
twofish-avx-x86_64 vs. twofish-x86_64-3way
128bit key: (lrw:256bit) (xts:256bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 0.96x 0.97x 1.00x 0.95x 0.97x 0.97x 0.96x 0.95x 0.95x 0.98x
64B 0.99x 0.99x 1.00x 0.99x 0.98x 0.98x 0.99x 0.98x 0.99x 0.98x
256B 1.20x 1.21x 1.00x 1.19x 1.15x 1.14x 1.19x 1.20x 1.18x 1.19x
1024B 1.29x 1.30x 1.00x 1.28x 1.23x 1.24x 1.26x 1.28x 1.26x 1.27x
8192B 1.31x 1.32x 1.00x 1.31x 1.25x 1.25x 1.28x 1.29x 1.28x 1.30x
256bit key: (lrw:384bit) (xts:512bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 0.96x 0.96x 1.00x 0.96x 0.97x 0.98x 0.95x 0.95x 0.95x 0.96x
64B 1.00x 0.99x 1.00x 0.98x 0.98x 1.01x 0.98x 0.98x 0.98x 0.98x
256B 1.20x 1.21x 1.00x 1.21x 1.15x 1.15x 1.19x 1.20x 1.18x 1.19x
1024B 1.29x 1.30x 1.00x 1.28x 1.23x 1.23x 1.26x 1.27x 1.26x 1.27x
8192B 1.31x 1.33x 1.00x 1.31x 1.26x 1.26x 1.29x 1.29x 1.28x 1.30x
twofish-avx-x86_64 vs aes-asm (8kB block):
128bit 256bit
ecb-enc 1.19x 1.63x
ecb-dec 1.18x 1.62x
cbc-enc 0.75x 1.03x
cbc-dec 1.23x 1.67x
ctr-enc 1.24x 1.65x
ctr-dec 1.24x 1.65x
lrw-enc 1.15x 1.53x
lrw-dec 1.14x 1.52x
xts-enc 1.16x 1.56x
xts-dec 1.16x 1.56x
Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Commit ea4d26ae ("raid5: add AVX optimized RAID5 checksumming")
introduced x86/ arch wide defines for AFLAGS and CFLAGS indicating AVX
support in binutils based on the same test we have in x86/crypto/ right
now. To minimize duplication drop our implementation in favour to the
one in x86/.
Signed-off-by: Mathias Krause <minipli@googlemail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Patch adds x86_64 assembler implementation of Camellia block cipher. Two set of
functions are provided. First set is regular 'one-block at time' encrypt/decrypt
functions. Second is 'two-block at time' functions that gain performance increase
on out-of-order CPUs. Performance of 2-way functions should be equal to 1-way
functions with in-order CPUs.
Patch has been tested with tcrypt and automated filesystem tests.
Tcrypt benchmark results:
AMD Phenom II 1055T (fam:16, model:10):
camellia-asm vs camellia_generic:
128bit key: (lrw:256bit) (xts:256bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 1.27x 1.22x 1.30x 1.42x 1.30x 1.34x 1.19x 1.05x 1.23x 1.24x
64B 1.74x 1.79x 1.43x 1.87x 1.81x 1.87x 1.48x 1.38x 1.55x 1.62x
256B 1.90x 1.87x 1.43x 1.94x 1.94x 1.95x 1.63x 1.62x 1.67x 1.70x
1024B 1.96x 1.93x 1.43x 1.95x 1.98x 2.01x 1.67x 1.69x 1.74x 1.80x
8192B 1.96x 1.96x 1.39x 1.93x 2.01x 2.03x 1.72x 1.64x 1.71x 1.76x
256bit key: (lrw:384bit) (xts:512bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 1.23x 1.23x 1.33x 1.39x 1.34x 1.38x 1.04x 1.18x 1.21x 1.29x
64B 1.72x 1.69x 1.42x 1.78x 1.81x 1.89x 1.57x 1.52x 1.56x 1.65x
256B 1.85x 1.88x 1.42x 1.86x 1.93x 1.96x 1.69x 1.65x 1.70x 1.75x
1024B 1.88x 1.86x 1.45x 1.95x 1.96x 1.95x 1.77x 1.71x 1.77x 1.78x
8192B 1.91x 1.86x 1.42x 1.91x 2.03x 1.98x 1.73x 1.71x 1.78x 1.76x
camellia-asm vs aes-asm (8kB block):
128bit 256bit
ecb-enc 1.15x 1.22x
ecb-dec 1.16x 1.16x
cbc-enc 0.85x 0.90x
cbc-dec 1.20x 1.23x
ctr-enc 1.28x 1.30x
ctr-dec 1.27x 1.28x
lrw-enc 1.12x 1.16x
lrw-dec 1.08x 1.10x
xts-enc 1.11x 1.15x
xts-dec 1.14x 1.15x
Intel Core2 T8100 (fam:6, model:23, step:6):
camellia-asm vs camellia_generic:
128bit key: (lrw:256bit) (xts:256bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 1.10x 1.12x 1.14x 1.16x 1.16x 1.15x 1.02x 1.02x 1.08x 1.08x
64B 1.61x 1.60x 1.17x 1.68x 1.67x 1.66x 1.43x 1.42x 1.44x 1.42x
256B 1.65x 1.73x 1.17x 1.77x 1.81x 1.80x 1.54x 1.53x 1.58x 1.54x
1024B 1.76x 1.74x 1.18x 1.80x 1.85x 1.85x 1.60x 1.59x 1.65x 1.60x
8192B 1.77x 1.75x 1.19x 1.81x 1.85x 1.86x 1.63x 1.61x 1.66x 1.62x
256bit key: (lrw:384bit) (xts:512bit)
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B 1.10x 1.07x 1.13x 1.16x 1.11x 1.16x 1.03x 1.02x 1.08x 1.07x
64B 1.61x 1.62x 1.15x 1.66x 1.63x 1.68x 1.47x 1.46x 1.47x 1.44x
256B 1.71x 1.70x 1.16x 1.75x 1.69x 1.79x 1.58x 1.57x 1.59x 1.55x
1024B 1.78x 1.72x 1.17x 1.75x 1.80x 1.80x 1.63x 1.62x 1.65x 1.62x
8192B 1.76x 1.73x 1.17x 1.78x 1.80x 1.81x 1.64x 1.62x 1.68x 1.64x
camellia-asm vs aes-asm (8kB block):
128bit 256bit
ecb-enc 1.17x 1.21x
ecb-dec 1.17x 1.20x
cbc-enc 0.80x 0.82x
cbc-dec 1.22x 1.24x
ctr-enc 1.25x 1.26x
ctr-dec 1.25x 1.26x
lrw-enc 1.14x 1.18x
lrw-dec 1.13x 1.17x
xts-enc 1.14x 1.18x
xts-dec 1.14x 1.17x
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Patch adds i586/SSE2 assembler implementation of serpent cipher. Assembler
functions crypt data in four block chunks.
Patch has been tested with tcrypt and automated filesystem tests.
Tcrypt benchmarks results (serpent-sse2/serpent_generic speed ratios):
Intel Atom N270:
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec
16 0.95x 1.12x 1.02x 1.07x 0.97x 0.98x
64 1.73x 1.82x 1.08x 1.82x 1.72x 1.73x
256 2.08x 2.00x 1.04x 2.07x 1.99x 2.01x
1024 2.28x 2.18x 1.05x 2.23x 2.17x 2.20x
8192 2.28x 2.13x 1.05x 2.23x 2.18x 2.20x
Full output:
http://koti.mbnet.fi/axh/kernel/crypto/atom-n270/serpent-generic.txt
http://koti.mbnet.fi/axh/kernel/crypto/atom-n270/serpent-sse2.txt
Userspace test results:
Encryption/decryption of sse2-i586 vs generic on Intel Atom N270:
encrypt: 2.35x
decrypt: 2.54x
Encryption/decryption of sse2-i586 vs generic on AMD Phenom II:
encrypt: 1.82x
decrypt: 2.51x
Encryption/decryption of sse2-i586 vs generic on Intel Xeon E7330:
encrypt: 2.99x
decrypt: 3.48x
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Patch adds x86_64/SSE2 assembler implementation of serpent cipher. Assembler
functions crypt data in eigth block chunks (two 4 block chunk SSE2 operations
in parallel to improve performance on out-of-order CPUs). Glue code is based
on one from AES-NI implementation, so requests from irq context are redirected
to cryptd.
v2:
- add missing include of linux/module.h
(appearently crypto.h used to include module.h, which changed for 3.2 by
commit 7c926402a7e8c9b279968fd94efec8700ba3859e)
Patch has been tested with tcrypt and automated filesystem tests.
Tcrypt benchmarks results (serpent-sse2/serpent_generic speed ratios):
AMD Phenom II 1055T (fam:16, model:10):
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec
16B 1.03x 1.01x 1.03x 1.05x 1.00x 0.99x
64B 1.00x 1.01x 1.02x 1.04x 1.02x 1.01x
256B 2.34x 2.41x 0.99x 2.43x 2.39x 2.40x
1024B 2.51x 2.57x 1.00x 2.59x 2.56x 2.56x
8192B 2.50x 2.54x 1.00x 2.55x 2.57x 2.57x
Intel Celeron T1600 (fam:6, model:15, step:13):
size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec
16B 0.97x 0.97x 1.01x 1.01x 1.01x 1.02x
64B 1.00x 1.00x 1.00x 1.02x 1.01x 1.01x
256B 3.41x 3.35x 1.00x 3.39x 3.42x 3.44x
1024B 3.75x 3.72x 0.99x 3.74x 3.75x 3.75x
8192B 3.70x 3.68x 0.99x 3.68x 3.69x 3.69x
Full output:
http://koti.mbnet.fi/axh/kernel/crypto/phenom-ii-1055t/serpent-generic.txt
http://koti.mbnet.fi/axh/kernel/crypto/phenom-ii-1055t/serpent-sse2.txt
http://koti.mbnet.fi/axh/kernel/crypto/celeron-t1600/serpent-generic.txt
http://koti.mbnet.fi/axh/kernel/crypto/celeron-t1600/serpent-sse2.txt
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Patch adds 3-way parallel x86_64 assembly implementation of twofish as new
module. New assembler functions crypt data in three blocks chunks, improving
cipher performance on out-of-order CPUs.
Patch has been tested with tcrypt and automated filesystem tests.
Summary of the tcrypt benchmarks:
Twofish 3-way-asm vs twofish asm (128bit 8kb block ECB)
encrypt: 1.3x speed
decrypt: 1.3x speed
Twofish 3-way-asm vs twofish asm (128bit 8kb block CBC)
encrypt: 1.07x speed
decrypt: 1.4x speed
Twofish 3-way-asm vs twofish asm (128bit 8kb block CTR)
encrypt: 1.4x speed
Twofish 3-way-asm vs AES asm (128bit 8kb block ECB)
encrypt: 1.0x speed
decrypt: 1.0x speed
Twofish 3-way-asm vs AES asm (128bit 8kb block CBC)
encrypt: 0.84x speed
decrypt: 1.09x speed
Twofish 3-way-asm vs AES asm (128bit 8kb block CTR)
encrypt: 1.15x speed
Full output:
http://koti.mbnet.fi/axh/kernel/crypto/tcrypt-speed-twofish-3way-asm-x86_64.txt
http://koti.mbnet.fi/axh/kernel/crypto/tcrypt-speed-twofish-asm-x86_64.txt
http://koti.mbnet.fi/axh/kernel/crypto/tcrypt-speed-aes-asm-x86_64.txt
Tests were run on:
vendor_id : AuthenticAMD
cpu family : 16
model : 10
model name : AMD Phenom(tm) II X6 1055T Processor
Also userspace test were run on:
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Xeon(R) CPU E7330 @ 2.40GHz
stepping : 11
Userspace test results:
Encryption/decryption of twofish 3-way vs x86_64-asm on AMD Phenom II:
encrypt: 1.27x
decrypt: 1.25x
Encryption/decryption of twofish 3-way vs x86_64-asm on Intel Xeon E7330:
encrypt: 1.36x
decrypt: 1.36x
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Patch adds x86_64 assembly implementation of blowfish. Two set of assembler
functions are provided. First set is regular 'one-block at time'
encrypt/decrypt functions. Second is 'four-block at time' functions that
gain performance increase on out-of-order CPUs. Performance of 4-way
functions should be equal to 1-way functions with in-order CPUs.
Summary of the tcrypt benchmarks:
Blowfish assembler vs blowfish C (256bit 8kb block ECB)
encrypt: 2.2x speed
decrypt: 2.3x speed
Blowfish assembler vs blowfish C (256bit 8kb block CBC)
encrypt: 1.12x speed
decrypt: 2.5x speed
Blowfish assembler vs blowfish C (256bit 8kb block CTR)
encrypt: 2.5x speed
Full output:
http://koti.mbnet.fi/axh/kernel/crypto/tcrypt-speed-blowfish-asm-x86_64.txt
http://koti.mbnet.fi/axh/kernel/crypto/tcrypt-speed-blowfish-c-x86_64.txt
Tests were run on:
vendor_id : AuthenticAMD
cpu family : 16
model : 10
model name : AMD Phenom(tm) II X6 1055T Processor
stepping : 0
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This is an assembler implementation of the SHA1 algorithm using the
Supplemental SSE3 (SSSE3) instructions or, when available, the
Advanced Vector Extensions (AVX).
Testing with the tcrypt module shows the raw hash performance is up to
2.3 times faster than the C implementation, using 8k data blocks on a
Core 2 Duo T5500. For the smalest data set (16 byte) it is still 25%
faster.
Since this implementation uses SSE/YMM registers it cannot safely be
used in every situation, e.g. while an IRQ interrupts a kernel thread.
The implementation falls back to the generic SHA1 variant, if using
the SSE/YMM registers is not possible.
With this algorithm I was able to increase the throughput of a single
IPsec link from 344 Mbit/s to 464 Mbit/s on a Core 2 Quad CPU using
the SSSE3 variant -- a speedup of +34.8%.
Saving and restoring SSE/YMM state might make the actual throughput
fluctuate when there are FPU intensive userland applications running.
For example, meassuring the performance using iperf2 directly on the
machine under test gives wobbling numbers because iperf2 uses the FPU
for each packet to check if the reporting interval has expired (in the
above test I got min/max/avg: 402/484/464 MBit/s).
Using this algorithm on a IPsec gateway gives much more reasonable and
stable numbers, albeit not as high as in the directly connected case.
Here is the result from an RFC 2544 test run with a EXFO Packet Blazer
FTB-8510:
frame size sha1-generic sha1-ssse3 delta
64 byte 37.5 MBit/s 37.5 MBit/s 0.0%
128 byte 56.3 MBit/s 62.5 MBit/s +11.0%
256 byte 87.5 MBit/s 100.0 MBit/s +14.3%
512 byte 131.3 MBit/s 150.0 MBit/s +14.2%
1024 byte 162.5 MBit/s 193.8 MBit/s +19.3%
1280 byte 175.0 MBit/s 212.5 MBit/s +21.4%
1420 byte 175.0 MBit/s 218.7 MBit/s +25.0%
1518 byte 150.0 MBit/s 181.2 MBit/s +20.8%
The throughput for the largest frame size is lower than for the
previous size because the IP packets need to be fragmented in this
case to make there way through the IPsec tunnel.
Signed-off-by: Mathias Krause <minipli@googlemail.com>
Cc: Maxim Locktyukhin <maxim.locktyukhin@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Loading fpu without aesni-intel does nothing. Loading aesni-intel
without fpu causes modes like xts to fail. (Unloading
aesni-intel will restore those modes.)
One solution would be to make aesni-intel depend on fpu, but it
seems cleaner to just combine the modules.
This is probably responsible for bugs like:
https://bugzilla.redhat.com/show_bug.cgi?id=589390
Signed-off-by: Andy Lutomirski <luto@mit.edu>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
PCLMULQDQ is used to accelerate the most time-consuming part of GHASH,
carry-less multiplication. More information about PCLMULQDQ can be
found at:
http://software.intel.com/en-us/articles/carry-less-multiplication-and-its-usage-for-computing-the-gcm-mode/
Because PCLMULQDQ changes XMM state, its usage must be enclosed with
kernel_fpu_begin/end, which can be used only in process context, the
acceleration is implemented as crypto_ahash. That is, request in soft
IRQ context will be defered to the cryptd kernel thread.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Blkcipher touching FPU need to be enclosed by kernel_fpu_begin() and
kernel_fpu_end(). If they are invoked in cipher algorithm
implementation, they will be invoked for each block, so that
performance will be hurt, because they are "slow" operations. This
patch implements "fpu" template, which makes these operations to be
invoked for each request.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Intel AES-NI is a new set of Single Instruction Multiple Data (SIMD)
instructions that are going to be introduced in the next generation of
Intel processor, as of 2009. These instructions enable fast and secure
data encryption and decryption, using the Advanced Encryption Standard
(AES), defined by FIPS Publication number 197. The architecture
introduces six instructions that offer full hardware support for
AES. Four of them support high performance data encryption and
decryption, and the other two instructions support the AES key
expansion procedure.
The white paper can be downloaded from:
http://softwarecommunity.intel.com/isn/downloads/intelavx/AES-Instructions-Set_WP.pdf
AES may be used in soft_irq context, but MMX/SSE context can not be
touched safely in soft_irq context. So in_interrupt() is checked, if
in IRQ or soft_irq context, the general x86_64 implementation are used
instead.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
From NHM processor onward, Intel processors can support hardware accelerated
CRC32c algorithm with the new CRC32 instruction in SSE 4.2 instruction set.
The patch detects the availability of the feature, and chooses the most proper
way to calculate CRC32c checksum.
Byte code instructions are used for compiler compatibility.
No MMX / XMM registers is involved in the implementation.
Signed-off-by: Austin Zhang <austin.zhang@intel.com>
Signed-off-by: Kent Liu <kent.liu@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
There is almost no difference between 32 & 64 bit glue code.
Signed-off-by: Sebastian Siewior <sebastian@breakpoint.cc>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This is the x86-64 version of the Salsa20 stream cipher algorithm. The
original assembly code came from
<http://cr.yp.to/snuffle/salsa20/amd64-3/salsa20.s>. It has been
reformatted for clarity.
Signed-off-by: Tan Swee Heng <thesweeheng@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
This patch contains the salsa20-i586 implementation. The original
assembly code came from
<http://cr.yp.to/snuffle/salsa20/x86-pm/salsa20.s>. I have reformatted
it (added indents) so that it matches the other algorithms in
arch/x86/crypto.
Signed-off-by: Tan Swee Heng <thesweeheng@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
32 bit and 64 bit glue code is using (now) the same
piece code. This patch unifies them.
Signed-off-by: Sebastian Siewior <sebastian@breakpoint.cc>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
|
|
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
|
|
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
|