: Johannes Goetzfried
---
arch/x86/crypto/Makefile |2 +
arch/x86/crypto/cast6-avx-x86_64-asm_64.S | 336 +++
arch/x86/crypto/cast6_avx_glue.c | 648 +
crypto/Kconfig| 17 +
crypto/testmgr.c
Rename cast6 module to cast6_generic to allow autoloading of optimized
implementations. Generic functions and s-boxes are exported to be able to use
them within optimized implementations.
Signed-off-by: Johannes Goetzfried
---
crypto/Makefile|2 +-
crypto/cast6.c | 547
New ECB, CBC and CTR testvectors for cast5. We need larger testvectors to check
parallel code paths in the optimized implementation. Tests have also been added
to the tcrypt module.
Signed-off-by: Johannes Goetzfried
---
crypto/tcrypt.c | 32 +++
crypto/tcrypt.h |1 +
crypto/testmgr.c
2.32x 0.95x 2.35x 2.39x 2.39x
Signed-off-by: Johannes Goetzfried
---
arch/x86/crypto/Makefile |2 +
arch/x86/crypto/cast5-avx-x86_64-asm_64.S | 323 ++
arch/x86/crypto/cast5_avx_glue.c | 530 +
crypto/Kconfig
This patchset adds x86_64/avx assembler implementations of the Cast5 and the
Cast6 cipher.
Johannes Goetzfried (6):
crypto: cast5 - prepare generic module for optimized implementations
crypto: testmgr - add larger cast5 testvectors
crypto: cast5 - add x86_64/avx assembler implementation
On Thu, Jul 05, 2012 at 08:46:12PM -0300, Henrique de Moraes Holschuh wrote:
> Is it useless, or is it there to harden against some side-channel
> attack?
When writing this implementation, I first had a different approach with more
data getting read in and I just forgot to delete this instruction
The register %rdx is written, but never read till the end of the encryption
routine. Therefore let's delete the useless instruction.
Signed-off-by: Johannes Goetzfried
---
arch/x86/crypto/twofish-avx-x86_64-asm_64.S |1 -
1 files changed, 0 insertions(+), 1 deletions(-)
diff --git a
On Wed, Jun 13, 2012 at 08:03:17PM +0300, Jussi Kivilinna wrote:
> >arch/x86/crypto/serpent-avx-x86_64.o: In function `serpent_xts_decrypt':
> >(.text+0x856c): multiple definition of `serpent_xts_decrypt'
> >arch/x86/crypto/serpent-sse2-x86_64.o:(.text+0xae8): first defined here
>
> Reverting comm
On Wed, May 30, 2012 at 06:26:45PM +0200, Andi Kleen wrote:
> > I tried to be consistent with the existing ciphers in arch/x86/crypto.
> > In serpent_sse2_glue.c and sha1_ssse3_glue.c it is done exactly that way,
> > so if
> > it will be decided to remove the printk in this patch it probably shoul
On Wed, May 30, 2012 at 05:39:49PM +0200, Andi Kleen wrote:
> > I agree with that. Currently when I boot my PC with a new 3.4 kernel all the
> > ciphers from the intel-aesni module get loaded whether I need them or not.
> > As
> > Jussi stated most people using distros probably won't need the
> >
On Tue, May 29, 2012 at 07:27:43PM -0700, Andi Kleen wrote:
> Also drivers should never print anything when they cannot find hardware.
> Remove that printk.
I tried to be consistent with the existing ciphers in arch/x86/crypto.
In serpent_sse2_glue.c and sha1_ssse3_glue.c it is done exactly that w
On Wed, May 30, 2012 at 10:30:25AM +0300, Jussi Kivilinna wrote:
> Quoting Andi Kleen :
>
> >
> >The driver needs CPUID annotations now (since 3.3), so that it can be
> >autoloaded. Something like:
> >
>
> Is it really good idea to autoload crypto modules? Currently loading
> different cipher im
):
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
---
arch/x86
Hello,
> Should be serpent_sse2_glue.c?
Yeah, same error as in the other patch :-)
> >+}, {
> >+.cra_name = "ecb(serpent)",
> >+.cra_driver_name= "ecb-serpent-avx",
> >+.cra_priority = 400,
>
> serpent_sse2_glue.c has priority 400 too, so you should i
The AVX implementation of the twofish cipher processes 8 blocks parallel, so we
need to make test vectors larger to check parallel code paths. Test vectors are
also large enough to deal with 16 block parallel implementations which may occur
in the future.
Signed-off-by: Johannes Goetzfried
-dec 1.16x 1.56x
Signed-off-by: Johannes Goetzfried
---
arch/x86/crypto/Makefile|2 +
arch/x86/crypto/twofish-avx-x86_64-asm_64.S | 302
arch/x86/crypto/twofish_avx_glue.c | 1086 +++
arch/x86/crypto/twofish_glue_3way.c
Hello,
> It would be benefical to expand the twofish vectors in
> crypto/testmgr.h from 3 blocks
> to 8 blocks so that 8-way algorithm(s) can be checked runtime. And
> while at expanding
> test-vectors, why not just expand to 16 blocks... AVX2 is just one year away:
> https://github.com/jkivilin/c
):
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
---
arch/x86
-dec 1.16x 1.56x
Signed-off-by: Johannes Goetzfried
---
arch/x86/crypto/Makefile|2 +
arch/x86/crypto/twofish-avx-x86_64-asm_64.S | 302
arch/x86/crypto/twofish_avx_glue.c | 1086 +++
arch/x86/crypto/twofish_glue_3way.c
19 matches
Mail list logo
