Calling conventions must specify how to handle arguments smaller than a
machine word. Should the caller pad them up to a machine word size with
predictable values (zero/sign-extended), or should the callee tolerate
an arbitrary bit pattern?
Annoyingly, I found no text in either SysV or Win64 ABI documentation
describing any of this and resorted to experiment. The short answer is
that callees must tolerate an arbitrary bit pattern on x86_64, which
means we must test this. See the comment in abi_test::internal::ToWord
for the long answer.
CHECK_ABI now, if the type of the parameter is smaller than
crypto_word_t, fills the remaining bytes with 0xaa. This is so the
number is out of bounds for code expecting either zero or sign
extension. (Not that crypto assembly has any business seeing negative
numbers.)
Doing so reveals a bug in ecp_nistz256_ord_sqr_mont. The rep parameter
is typed int, but the code expected uint64_t. In practice, the compiler
will always compile this correctly because:
- On both Win64 and SysV, rep is a register parameter.
- The rep parameter is always a constant, so the compiler has no reason
to leave garbage in the upper half.
However, I was indeed able to get a bug out of GCC via:
uint64_t foo = (1ull << 63) | 2; // Some global the compiler can't
// prove constant.
ecp_nistz256_ord_sqr_mont(res, a, foo >> 1);
Were ecp_nistz256_ord_sqr_mont a true int-taking function, this would
act like ecp_nistz256_ord_sqr_mont(res, a, 1). Instead, it hung. Fix
this by having it take a full-width word.
This mess has several consequences:
- ABI testing now ideally needs a functional testing component to fully cover
this case. A bad input might merely produce the wrong answer. Still,
this is fairly effective as it will cause most code to either segfault
or loop forever. (Not the enc parameter to AES however...)
- We cannot freely change the type of assembly function prototypes. If the
prototype says int or unsigned, it must be ignoring the upper half and
thus "fixing" it to size_t cannot have handled the full range. (Unless
it was simply wrong of the parameter is already bounded.) If the
prototype says size_t, switching to int or unsigned will hit this type
of bug. The former is a safer failure mode though.
- The simplest path out of this mess: new assembly code should *only*
ever take word-sized parameters. This is not a tall order as the bad
parameters are usually ints that should have been size_t.
Calling conventions are hard.
Change-Id: If8254aff8953844679fbce4bd3e345e5e2fa5213
Reviewed-on: https://boringssl-review.googlesource.com/c/34627
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: Adam Langley <agl@google.com>
This doesn't cover all the functions used by Node, but it's the easy
bits. (EVP_PKEY_paramgen will be done separately as its a non-trivial
bit of machinery.)
Change-Id: I6501e99f9239ffcdcc57b961ebe85d0ad3965549
Reviewed-on: https://boringssl-review.googlesource.com/c/34544
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: Adam Langley <agl@google.com>
We expect the table to have a slightly nested structure, so just
generate it that way. Avoid risking strict aliasing problems. Thanks to
Brian Smith for pointing this out.
Change-Id: Ie21610c4afab07a610d914265079135dba17b3b7
Reviewed-on: https://boringssl-review.googlesource.com/c/34264
Commit-Queue: Adam Langley <agl@google.com>
Reviewed-by: Adam Langley <agl@google.com>
Unwind testing will make CHECK_ABI much slower. The original
ptrace-based design is some 10,000x slower. I've found an alternate
design that's a mere 1,000x slower, but this probably warrants being
more straightforward. It also removes the weirdness where NDEBUG
controlled which tests were run.
While it does mean we need to write some extra tests for p256-x86_64.pl,
we otherwise do not directly unit test our assembly anyway. Usually we
test the public crypto APIs themselves. So, for most files, this isn't
actually extra work.
Bug: 181
Change-Id: I7cbb7f930c2ea6ae32a201da503dcd36844704f0
Reviewed-on: https://boringssl-review.googlesource.com/c/33965
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: Adam Langley <agl@google.com>
While gdb can figure it out, libunwind requires CFI directives to
unwind a leaf function, even though the directives are trivial.
Adding them matches what GCC outputs, and likely gdb has many
heuristics that less complex tools (e.g. profilers) may not.
Bug: 181
Change-Id: I25c72152de33109a29710a828aeb99c608dd0470
Reviewed-on: https://boringssl-review.googlesource.com/c/33964
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: Adam Langley <agl@google.com>
This was also caught by the in-progress unwind tester. There are two
issues here. First, .cfi_endproc must come after ret to fully cover the
function. More importantly, this function is confused about whether it
has a frame pointer or not.
It looks like it does (movq %rsp, %rbp), and annotates accordingly, but
it does not actually use the frame pointer. It cannot. $y4 is rbp and
gets clobbered immediately after the preamble!
Remove this instruction and align the CFI annotations with a
frame-pointer-less function.
Bug: 181
Change-Id: I47b5f9798b3bcee1748e537b21c173d312a14b42
Reviewed-on: https://boringssl-review.googlesource.com/c/33947
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
This was caught by in-progress work to test unwind information. It was
incorrect at two instructions: immediately before we jump to
.Lpoint_double_shortcut$x. This is needed because
ecp_nistz256_point_add$x tries to be clever about not unwinding the
stack frame in its tail call.
It's also unlikely that the SEH handlers in this file are correct at
this point, but that will be handled separately while overhauling
everything else here. (For Win64, probably the only ABI-compliant option
is to just properly unwind the stack frame. Without a custom handler,
Win64 unwind codes are very restrictive.)
Bug: 181
Change-Id: I9f576d868850312d6c14d1386f8fbfa85021b347
Reviewed-on: https://boringssl-review.googlesource.com/c/33946
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: Adam Langley <agl@google.com>
ecp_nistz256_point_add_affine does not support the doubling case and,
unlike ecp_nistz256_point_add which does a tail call, computes the wrong
answer. Note TestPointAdd in the unit tests skips this case.
This works fine because we only use ecp_nistz256_point_add_affine for
the g_scalar term, which is fully computed before the p_scalar term.
(Additionally it requires that the windowing pattern never hit the
doubling case for single multiplication.)
But this is not obvious from reading the multiplication functions, so
leave a comment at the call site to point this out.
Change-Id: I08882466d98030cdc882a5be9e702ee404e80cce
Reviewed-on: https://boringssl-review.googlesource.com/c/33945
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: Adam Langley <agl@google.com>
The old points weren't even on the curve. I probably had no clue what I
was doing at the time when I generated them. Refresh them with a
checked-in generate script.
Change-Id: Ib4613fe922edcf45fc4ea49fc4c2cc23a9a2a9bd
Reviewed-on: https://boringssl-review.googlesource.com/c/33944
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
Dear reader, I must apologize in advance. This CL contains the following:
- A new 256-line perlasm file with non-trivial perl bits and a dual-ABI
variadic function caller.
- C preprocessor gymnastics, with variadic macros and fun facts about
__VA_ARGS__'s behavior on empty argument lists.
- C++ template gymnastics, including variadic arguments, template
specialization, std::enable_if, and machinery to control template argument
deduction.
Enjoy.
This tests that our assembly functions correctly honor platform ABI
conventions. Right now this only tests callee-saved registers, but it should be
extendable to SEH/CFI unwind testing with single-step debugging APIs.
Register-checking does not involve anything funny and should be compatible with
SDE. (The future unwind testing is unlikely to be compatible.)
This CL adds support for x86_64 SysV and Win64 ABIs. ARM, AArch64, and x86 can
be added in the future. The testing is injected in two places. First, all the
assembly tests in p256-x86_64-test.cc are now instrumented. This is the
intended workflow and should capture all registers.
However, we currently do not unit-test our assembly much directly. We should do
that as follow-up work[0] but, in the meantime, I've also wrapped all of the GTest
main function in an ABI test. This is imperfect as ABI failures may be masked
by other stack frames, but it costs nothing[1] and is pretty reliable at
catching Win64 xmm register failures.
[0] An alternate strategy would be, in debug builds, unconditionally instrument
every assembly call in libcrypto. But the CHECK_ABI macro would be difficult to
replicate in pure C, and unwind testing may be too invasive for this. Still,
something to consider when we C++ libcrypto.
[1] When single-stepped unwind testing exists, it won't cost nothing. The
gtest_main.cc call will turn unwind testing off.
Change-Id: I6643b26445891fd46abfacac52bc024024c8d7f6
Reviewed-on: https://boringssl-review.googlesource.com/c/33764
Reviewed-by: Adam Langley <agl@google.com>
Reviewed-by: Adam Langley <alangley@gmail.com>
Commit-Queue: David Benjamin <davidben@google.com>
This can break delocate with certain compiler settings.
Change-Id: I76cf0f780d0e967390feed754e39b0ab25068f42
Reviewed-on: https://boringssl-review.googlesource.com/c/33485
Commit-Queue: Adam Langley <alangley@gmail.com>
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: David Benjamin <davidben@google.com>
Contraction was always and only done immediately prior to calling
|p224_felem_to_generic| so merge it into that function.
Change-Id: If4fb46c6305ba724dfff15e8362a094c599f3f2c
Reviewed-on: https://boringssl-review.googlesource.com/c/33165
Commit-Queue: Adam Langley <agl@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: David Benjamin <davidben@google.com>
cfd50c63 switched to using the add/dbl of p224_64.c, but the outputs
weren't contracted before being returned and could be out of range,
giving invalid results.
Change-Id: I3cc295c7ddbff43375770dbafe73b37a668e4e6b
Reviewed-on: https://boringssl-review.googlesource.com/c/33184
Reviewed-by: David Benjamin <davidben@google.com>
Commit-Queue: Adam Langley <agl@google.com>
C99 added macros such as PRIu64 to inttypes.h, but it said to exclude them from
C++ unless __STDC_FORMAT_MACROS or __STDC_CONSTANT_MACROS was defined. This
text was never incorporated into any C++ standard and explicitly overruled in
C++11.
Some libc headers followed C99. Notably, glibc prior to 2.18
(https://sourceware.org/bugzilla/show_bug.cgi?id=15366) and old versions of the
Android NDK.
In the NDK, although it was fixed some time ago (API level 20), the NDK used to
use separate headers per API level. Only applications using minSdkVersion >= 20
would get the fix. Starting NDK r14, "unified" headers are available which,
among other things, make the fix available (opt-in) independent of
minSdkVersion. In r15, unified headers are opt-out, and in r16 they are
mandatory.
Try removing these and see if anyone notices. The former is past our five year
watermark. The latter is not and Android has hit
https://boringssl-review.googlesource.com/c/boringssl/+/32686 before, but
unless it is really widespread, it's probably simpler to ask consumers to
define __STDC_CONSTANT_MACROS and __STDC_FORMAT_MACROS globally.
Update-Note: If you see compile failures relating to PRIu64, UINT64_MAX, and
friends, update your glibc or NDK. As a short-term fix, add
__STDC_CONSTANT_MACROS and __STDC_FORMAT_MACROS to your build, but get in touch
so we have a sense of how widespread it is.
Bug: 198
Change-Id: I56cca5f9acdff803de1748254bc45096e4c959c2
Reviewed-on: https://boringssl-review.googlesource.com/c/33146
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: Adam Langley <agl@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
The change seems to have stuck, so bring us closer to C/++11 static asserts.
(If we later find we need to support worse toolchains, we can always use
__LINE__ or __COUNTER__ to avoid duplicate typedef names and just punt on
embedding the message into the type name.)
Change-Id: I0e5bb1106405066f07740728e19ebe13cae3e0ee
Reviewed-on: https://boringssl-review.googlesource.com/c/33145
Commit-Queue: Adam Langley <agl@google.com>
Reviewed-by: Adam Langley <agl@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
With the allocations and BN_CTX gone, ECDH and point2oct are much, much
shorter.
Bug: 242
Change-Id: I3421822e94100f7eb2f5f2373df7fb3b3311365e
Reviewed-on: https://boringssl-review.googlesource.com/c/33071
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
(Otherwise the individual-file build breaks.)
Change-Id: Id3defd08cd2b49af1d8eb6890bd8454332c1aa1e
Reviewed-on: https://boringssl-review.googlesource.com/c/33124
Commit-Queue: Adam Langley <agl@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
These are used for field elements too.
Change-Id: I74e3dbcafdce34ad507f64a0718e0420b56b51ae
Reviewed-on: https://boringssl-review.googlesource.com/c/33070
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
Now the only allocations in ECDSA are the ECDSA_SIG input and output.
Change-Id: If1fcde6dc2ee2c53f5adc16a7f692e22e9c238de
Reviewed-on: https://boringssl-review.googlesource.com/c/33069
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
For simplicity, punt order > field or width mismatches. Analogous
optimizations are possible, but the generic path works fine and no
commonly-used curve looks hits those cases.
Before:
Did 5888 ECDSA P-384 verify operations in 3094535us (1902.7 ops/sec)
After [+6.7%]:
Did 6107 ECDSA P-384 verify operations in 3007515us (2030.6 ops/sec)
Also we can fill in p - order generically and avoid extra copies of some
constants.
Change-Id: I38e1b6d51b28ed4f8cb74697b00a4f0fbc5efc3c
Reviewed-on: https://boringssl-review.googlesource.com/c/33068
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
This removes the failure cases for cmp_x_coordinate, this clearing our
earlier dilemma.
Change-Id: I057f705e49b0fb5c3fc9616ee8962a3024097b24
Reviewed-on: https://boringssl-review.googlesource.com/c/33065
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
This is in preparation for removing the BIGNUM from cmp_x_coordinate.
Change-Id: Id8394248e3019a4897c238289f039f436a13679d
Reviewed-on: https://boringssl-review.googlesource.com/c/33064
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
This function is not EC_METHOD-specific, nor is there any reason it
would be (we do not support GF2m).
Change-Id: I4896cd16a107ad6a99be445a0dc0896293e8c8f9
Reviewed-on: https://boringssl-review.googlesource.com/c/32884
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
Some of the ec files now reference ECDSA_R_BAD_SIGNATURE. Instead, lift the
error-pushing to ecdsa.c.
Change-Id: Ice3e7a22c5099756599df0ab0b215c0752ada4ee
Reviewed-on: https://boringssl-review.googlesource.com/c/32984
Reviewed-by: David Benjamin <davidben@google.com>
Commit-Queue: Adam Langley <agl@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
This reverts commit e907ed4c4b. CPUID
checks have been added so hopefully this time sticks.
Change-Id: I5e0e5b87427c1230132681f936b3c70bac8263b8
Reviewed-on: https://boringssl-review.googlesource.com/c/32924
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Now that the tuned add/dbl implementations are exposed, these can be
specific to EC_GFp_mont_method and call the felem_mul and felem_sqr
implementations directly.
felem_sqr and felem_mul are still used elsewhere in simple.c, however,
so we cannot get rid of them yet.
Change-Id: I5ea22a8815279931afc98a6fc578bc85e3f8bdcc
Reviewed-on: https://boringssl-review.googlesource.com/c/32849
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
This reverts commit 3d450d2844. It fails
SDE, looks like a missing CPUID check before using vector instructions.
Change-Id: I6b7dd71d9e5b1f509d2e018bd8be38c973476b4e
Reviewed-on: https://boringssl-review.googlesource.com/c/32864
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: Adam Langley <agl@google.com>
Some consumer stumbled upon EC_POINT_{add,dbl} being faster with a
"custom" P-224 curve than the built-in one and made "custom" clones to
work around this. Before the EC_FELEM refactor, EC_GFp_nistp224_method
used BN_mod_mul for all reductions in fallback point arithmetic (we
primarily support the multiplication functions and keep the low-level
point arithmetic for legacy reasons) which took quite a performance hit.
EC_FELEM fixed this, but standalone felem_{mul,sqr} calls out of
nistp224 perform a lot of reductions, rather than batching them up as
that implementation is intended. So it is still slightly faster to use a
"custom" curve.
Custom curves are the last thing we want to encourage, so just route the
tuned implementations out of EC_METHOD to close this gap. Now the
built-in implementation is always solidly faster than (or identical to)
the custom clone. This also reduces the number of places where we mix
up tuned vs. generic implementation, which gets us closer to making
EC_POINT's representation EC_METHOD-specific.
Change-Id: I843e1101a6208eaabb56d29d342e886e523c78b4
Reviewed-on: https://boringssl-review.googlesource.com/c/32848
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
This commit improves the performance of ECDSA signature verification
(over NIST P-256 curve) for x86 platforms. The speedup is by a factor of 1.15x.
It does so by:
1) Leveraging the fact that the verification does not need
to run in constant time. To this end, we implemented:
a) the function ecp_nistz256_points_mul_public in a similar way to
the current ecp_nistz256_points_mul function by removing its constant
time features.
b) the Binary Extended Euclidean Algorithm (BEEU) in x86 assembly to
replace the current modular inverse function used for the inversion.
2) The last step in the ECDSA_verify function compares the (x) affine
coordinate with the signature (r) value. Converting x from the Jacobian's
representation to the affine coordinate requires to perform one inversions
(x_affine = x * z^(-2)). We save this inversion and speed up the computations
by instead bringing r to x (r_jacobian = r*z^2) which is faster.
The measured results are:
Before (on a Kaby Lake desktop with gcc-5):
Did 26000 ECDSA P-224 signing operations in 1002372us (25938.5 ops/sec)
Did 11000 ECDSA P-224 verify operations in 1043821us (10538.2 ops/sec)
Did 55000 ECDSA P-256 signing operations in 1017560us (54050.9 ops/sec)
Did 17000 ECDSA P-256 verify operations in 1051280us (16170.8 ops/sec)
After (on a Kaby Lake desktop with gcc-5):
Did 27000 ECDSA P-224 signing operations in 1011287us (26698.7 ops/sec)
Did 11640 ECDSA P-224 verify operations in 1076698us (10810.8 ops/sec)
Did 55000 ECDSA P-256 signing operations in 1016880us (54087.0 ops/sec)
Did 20000 ECDSA P-256 verify operations in 1038736us (19254.2 ops/sec)
Before (on a Skylake server platform with gcc-5):
Did 25000 ECDSA P-224 signing operations in 1021651us (24470.2 ops/sec)
Did 10373 ECDSA P-224 verify operations in 1046563us (9911.5 ops/sec)
Did 50000 ECDSA P-256 signing operations in 1002774us (49861.7 ops/sec)
Did 15000 ECDSA P-256 verify operations in 1006471us (14903.6 ops/sec)
After (on a Skylake server platform with gcc-5):
Did 25000 ECDSA P-224 signing operations in 1020958us (24486.8 ops/sec)
Did 10373 ECDSA P-224 verify operations in 1046359us (9913.4 ops/sec)
Did 50000 ECDSA P-256 signing operations in 1003996us (49801.0 ops/sec)
Did 18000 ECDSA P-256 verify operations in 1021604us (17619.4 ops/sec)
Developers and authors:
***************************************************************************
Nir Drucker (1,2), Shay Gueron (1,2)
(1) Amazon Web Services Inc.
(2) University of Haifa, Israel
***************************************************************************
Change-Id: Idd42a7bc40626bce974ea000b61fdb5bad33851c
Reviewed-on: https://boringssl-review.googlesource.com/c/31304
Commit-Queue: Adam Langley <agl@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: David Benjamin <davidben@google.com>
Reviewed-by: Adam Langley <agl@google.com>
Although the original value of tmp does not matter, the selects
ultimately do bit operations on the uninitialized values and thus depend
on them behaving like *some* consistent concrete value. The C spec
appears to allow uninitialized values to resolve to trap
representations, which means this isn't quite valid..
(If I'm reading it wrong and the compiler must behave as if there were a
consistent value in there, it's probably fine, but there's no sense in
risking compiler bugs on a subtle corner of things.)
Change-Id: Id4547b0ec702414b387e906c4de55595e6214ddb
Reviewed-on: https://boringssl-review.googlesource.com/29124
Commit-Queue: Steven Valdez <svaldez@google.com>
Reviewed-by: Steven Valdez <svaldez@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
cryptography.io wants things exposed out of EVP_get_cipherby* including,
sadly, ECB mode.
Change-Id: I9bac46f8ffad1a79d190cee3b0c0686bf540298e
Reviewed-on: https://boringssl-review.googlesource.com/28464
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Build (and carry) issues are now resolved (as far as we know). Let's try
this again...
Measurements on a Skylake VM (so a little noisy).
Before:
Did 3135 RSA 2048 signing operations in 3015866us (1039.5 ops/sec)
Did 89000 RSA 2048 verify (same key) operations in 3007271us (29594.9 ops/sec)
Did 66000 RSA 2048 verify (fresh key) operations in 3014363us (21895.2 ops/sec)
Did 324 RSA 4096 signing operations in 3004364us (107.8 ops/sec)
Did 23126 RSA 4096 verify (same key) operations in 3003398us (7699.9 ops/sec)
Did 21312 RSA 4096 verify (fresh key) operations in 3017043us (7063.9 ops/sec)
Did 31040 ECDH P-256 operations in 3024273us (10263.6 ops/sec)
Did 91000 ECDSA P-256 signing operations in 3019740us (30135.0 ops/sec)
Did 25678 ECDSA P-256 verify operations in 3046975us (8427.4 ops/sec)
After:
Did 3640 RSA 2048 signing operations in 3035845us (1199.0 ops/sec)
Did 129000 RSA 2048 verify (same key) operations in 3003691us (42947.2 ops/sec)
Did 105000 RSA 2048 verify (fresh key) operations in 3029935us (34654.2 ops/sec)
Did 510 RSA 4096 signing operations in 3014096us (169.2 ops/sec)
Did 38000 RSA 4096 verify (same key) operations in 3092814us (12286.5 ops/sec)
Did 34221 RSA 4096 verify (fresh key) operations in 3003817us (11392.5 ops/sec)
Did 38000 ECDH P-256 operations in 3061758us (12411.2 ops/sec)
Did 116000 ECDSA P-256 signing operations in 3001637us (38645.6 ops/sec)
Did 35100 ECDSA P-256 verify operations in 3023872us (11607.6 ops/sec)
Tested with Intel SDE.
Change-Id: Ib27c0d6012d14274e331ab03f958e5a0c8b7e885
Reviewed-on: https://boringssl-review.googlesource.com/28104
Reviewed-by: Adam Langley <agl@google.com>
Change-Id: I0674f4e9b15b546237600fb2486c46aac7cb0716
Reviewed-on: https://boringssl-review.googlesource.com/28027
Reviewed-by: David Benjamin <davidben@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Change-Id: Ib2ce220e31a4f808999934197a7f43b8723131e8
Reviewed-on: https://boringssl-review.googlesource.com/27884
Reviewed-by: David Benjamin <davidben@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
This is slower, but constant-time. It intentionally omits the signed
digit optimization because we cannot be sure the doubling case will be
unreachable for all curves. This is a fallback generic implementation
for curves which we must support for compatibility but which are not
common or important enough to justify curve-specific work.
Before:
Did 814 ECDH P-384 operations in 1085384us (750.0 ops/sec)
Did 1430 ECDSA P-384 signing operations in 1081988us (1321.6 ops/sec)
Did 308 ECDH P-521 operations in 1057741us (291.2 ops/sec)
Did 539 ECDSA P-521 signing operations in 1049797us (513.4 ops/sec)
After:
Did 715 ECDH P-384 operations in 1080161us (661.9 ops/sec)
Did 1188 ECDSA P-384 verify operations in 1069567us (1110.7 ops/sec)
Did 275 ECDH P-521 operations in 1060503us (259.3 ops/sec)
Did 506 ECDSA P-521 signing operations in 1084739us (466.5 ops/sec)
But we're still faster than the old BIGNUM implementation. EC_FELEM
more than paid for both the loss of points_make_affine and this CL.
Bug: 239
Change-Id: I65d71a731aad16b523928ee47618822d503ea704
Reviewed-on: https://boringssl-review.googlesource.com/27708
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
w=4 appears to be the correct answer for P-224 through P-521. There's
nominally some optimizations in here for 70- and 20-bit primes, but
that's absurd.
Change-Id: Id4ccec779b17e375e9258c1784e46d7d3651c59a
Reviewed-on: https://boringssl-review.googlesource.com/27707
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
EC_POINT is split into the existing public EC_POINT (where the caller is
sanity-checked about group mismatches) and the low-level EC_RAW_POINT
(which, like EC_FELEM and EC_SCALAR, assume that is your problem and is
a plain old struct). Having both EC_POINT and EC_RAW_POINT is a little
silly, but we're going to want different type signatures for functions
which return void anyway (my plan is to lift a non-BIGNUM
get_affine_coordinates up through the ECDSA and ECDH code), so I think
it's fine.
This wasn't strictly necessary, but wnaf.c is a lot tidier now. Perf is
a wash; once we get up to this layer, it's only 8 entries in the table
so not particularly interesting.
Bug: 239
Change-Id: I8ace749393d359f42649a5bb0734597bb7c07a2e
Reviewed-on: https://boringssl-review.googlesource.com/27706
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
Replace them with asserts and better justify why each of the internal
cases are not reachable. Also change the loop to count up to bits+1 so
it is obvious there is no memory error. (The previous loop shape made
more sense when ec_compute_wNAF would return a variable length
schedule.)
Change-Id: I9c7df6abac4290b7a3e545e3d4aa1462108e239e
Reviewed-on: https://boringssl-review.googlesource.com/27705
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
This introduces EC_FELEM, which is analogous to EC_SCALAR. It is used
for EC_POINT's representation in the generic EC_METHOD, as well as
random operations on tuned EC_METHODs that still are implemented
genericly.
Unlike EC_SCALAR, EC_FELEM's exact representation is awkwardly specific
to the EC_METHOD, analogous to how the old values were BIGNUMs but may
or may not have been in Montgomery form. This is kind of a nuisance, but
no more than before. (If p224-64.c were easily convertable to Montgomery
form, we could say |EC_FELEM| is always in Montgomery form. If we
exposed the internal add and double implementations in each of the
curves, we could give |EC_POINT| an |EC_METHOD|-specific representation
and |EC_FELEM| is purely a |EC_GFp_mont_method| type. I'll leave this
for later.)
The generic add and doubling formulas are aligned with the formulas
proved in fiat-crypto. Those only applied to a = -3, so I've proved a
generic one in https://github.com/mit-plv/fiat-crypto/pull/356, in case
someone uses a custom curve. The new formulas are verified,
constant-time, and swap a multiply for a square. As expressed in
fiat-crypto they do use more temporaries, but this seems to be fine with
stack-allocated EC_FELEMs. (We can try to help the compiler later,
but benchamrks below suggest this isn't necessary.)
Unlike BIGNUM, EC_FELEM can be stack-allocated. It also captures the
bounds in the type system and, in particular, that the width is correct,
which will make it easier to select a point in constant-time in the
future. (Indeed the old code did not always have the correct width. Its
point formula involved halving and implemented this in variable time and
variable width.)
Before:
Did 77274 ECDH P-256 operations in 10046087us (7692.0 ops/sec)
Did 5959 ECDH P-384 operations in 10031701us (594.0 ops/sec)
Did 10815 ECDSA P-384 signing operations in 10087892us (1072.1 ops/sec)
Did 8976 ECDSA P-384 verify operations in 10071038us (891.3 ops/sec)
Did 2600 ECDH P-521 operations in 10091688us (257.6 ops/sec)
Did 4590 ECDSA P-521 signing operations in 10055195us (456.5 ops/sec)
Did 3811 ECDSA P-521 verify operations in 10003574us (381.0 ops/sec)
After:
Did 77736 ECDH P-256 operations in 10029858us (7750.5 ops/sec) [+0.8%]
Did 7519 ECDH P-384 operations in 10068076us (746.8 ops/sec) [+25.7%]
Did 13335 ECDSA P-384 signing operations in 10029962us (1329.5 ops/sec) [+24.0%]
Did 11021 ECDSA P-384 verify operations in 10088600us (1092.4 ops/sec) [+22.6%]
Did 2912 ECDH P-521 operations in 10001325us (291.2 ops/sec) [+13.0%]
Did 5150 ECDSA P-521 signing operations in 10027462us (513.6 ops/sec) [+12.5%]
Did 4264 ECDSA P-521 verify operations in 10069694us (423.4 ops/sec) [+11.1%]
This more than pays for removing points_make_affine previously and even
speeds up ECDH P-256 slightly. (The point-on-curve check uses the
generic code.)
Next is to push the stack-allocating up to ec_wNAF_mul, followed by a
constant-time single-point multiplication.
Bug: 239
Change-Id: I44a2dff7c52522e491d0f8cffff64c4ab5cd353c
Reviewed-on: https://boringssl-review.googlesource.com/27668
Reviewed-by: Adam Langley <agl@google.com>
This does not appear to actually pull its weight. The purpose of this
logic is to switch some adds to the faster add_mixed in the wNAF code,
at the cost of a rather expensive inversion. This optimization kicks in
for generic curves, so P-384 and P-521:
With:
Did 32130 ECDSA P-384 signing operations in 30077563us (1068.2 ops/sec)
Did 27456 ECDSA P-384 verify operations in 30073086us (913.0 ops/sec)
Did 14122 ECDSA P-521 signing operations in 30077407us (469.5 ops/sec)
Did 11973 ECDSA P-521 verify operations in 30037330us (398.6 ops/sec)
Without:
Did 32445 ECDSA P-384 signing operations in 30069721us (1079.0 ops/sec)
Did 27056 ECDSA P-384 verify operations in 30032303us (900.9 ops/sec)
Did 13905 ECDSA P-521 signing operations in 30000430us (463.5 ops/sec)
Did 11433 ECDSA P-521 verify operations in 30021876us (380.8 ops/sec)
For single-point multiplication, the optimization is not useful. This
makes sense as we only have one table's worth of additions to convert
but still pay for the inversion. For double-point multiplication, it is
slightly useful for P-384 and very useful for P-521. However, the next
change to stack-allocate EC_FELEMs will more than compensate for
removing it. (The immediate goal here is to simplify the EC_FELEM
story.)
Additionally, that this optimization was not useful for single-point
multiplication implies that, should we wish to recover this, a modest
8-entry pre-computed (affine) base point table should have the same
effect or better.
Update-Note: I do not believe anything was calling either of these
functions. (If necessary, we can always add no-op stubs as whether a
point is affine is not visible to external code. It previously kicked in
some optimizations, but those were removed for constant-time needs
anyway.)
Bug: 239
Change-Id: Ic9c51b001c45595cfe592274c7d5d652f4234839
Reviewed-on: https://boringssl-review.googlesource.com/27667
Reviewed-by: Adam Langley <agl@google.com>
As the point may be the output of some private key operation, whether Z
accidentally hit one is secret.
Bug: 239
Change-Id: I7db34cd3b5dd5ca4b96980e8993a9b4eda49eb88
Reviewed-on: https://boringssl-review.googlesource.com/27664
Reviewed-by: Adam Langley <alangley@gmail.com>
We have little-endian BIGNUM functions now.
Change-Id: Iffc46a14e75c6bba2e170b824b1a08c69d2e9d18
Reviewed-on: https://boringssl-review.googlesource.com/27594
Reviewed-by: Adam Langley <alangley@gmail.com>
This is adapted from upstream's
eb7916960bf50f436593abe3d5f2e0592d291017.
This gives a 22% win for ECDSA signing. (Upstream cites 30-40%, but they
are unnecessarily using BN_mod_exp_mont_consttime in their generic path.
The exponent is public. I expect part of their 30-40% is just offsetting
this.)
Did 506000 ECDSA P-256 signing operations in 25044595us (20204.0 ops/sec)
Did 170506 ECDSA P-256 verify operations in 25033567us (6811.1 ops/sec)
Did 618000 ECDSA P-256 signing operations in 25031294us (24689.1 ops/sec)
Did 182240 ECDSA P-256 verify operations in 25006918us (7287.6 ops/sec)
Most of the performance win appears to be from the assembly operations
and not the addition chain. I have a CL to graft the addition chain onto
the C implementation, but it did not show measurable improvement in
ECDSA verify. ECDSA sign gets 2-4% faster, but we're more concerned
about ECDSA verify in the OPENSSL_SMALL builds.
Change-Id: Ide166f98b146c025f7f80ed7906336c16818540a
Reviewed-on: https://boringssl-review.googlesource.com/27593
Reviewed-by: Adam Langley <alangley@gmail.com>
This introduces a hook for the OpenSSL assembly.
Change-Id: I35e0588f0ed5bed375b12f738d16c9f46ceedeea
Reviewed-on: https://boringssl-review.googlesource.com/27592
Reviewed-by: Adam Langley <alangley@gmail.com>
Largely random data, but make it easy to add things in the future.
Change-Id: I30bee790bd9671b4d0327c2244fe5cd1a8954f90
Reviewed-on: https://boringssl-review.googlesource.com/27591
Reviewed-by: Adam Langley <alangley@gmail.com>
