[crypto] Update relic to 9206ae5

[huitseeker]

The diff contains:

• a lot of changes that concern parts of the library we don't use (other curves a la BN, BLS24-X, BLS12-383 ...), integer protocols (ETRS), field extension machinery ...
• otherwise irrelevant changes, e.g. CI/CD
• some memory bug fixing

Formatted Log
Full Changeset

Benchmark Diff:

benchstat ~/old_relic.txt ~/new_relic.txt

name                           old time/op  new time/op  delta
ScalarMult/G1-8                 144µs ± 0%   142µs ± 0%   ~     (p=1.000 n=1+1)
ScalarMult/G2-8                 329µs ± 0%   326µs ± 0%   ~     (p=1.000 n=1+1)
HashToG1-8                      231µs ± 0%   230µs ± 0%   ~     (p=1.000 n=1+1)
CheckG1-8                       210µs ± 0%   209µs ± 0%   ~     (p=1.000 n=1+1)
BLSBLS12381Sign-8               477µs ± 0%   473µs ± 0%   ~     (p=1.000 n=1+1)
BLSBLS12381Verify-8            2.71ms ± 0%  2.64ms ± 0%   ~     (p=1.000 n=1+1)
BatchVerify/happy_path-8       81.5ms ± 0%  81.5ms ± 0%   ~     (p=1.000 n=1+1)
BatchVerify/unhappy_path-8      117ms ± 0%   115ms ± 0%   ~     (p=1.000 n=1+1)
VerifySignatureManyMessages-8  3.14ms ± 0%  3.12ms ± 0%   ~     (p=1.000 n=1+1)
Aggregate/PrivateKeys-8        71.6µs ± 0%  70.1µs ± 0%   ~     (p=1.000 n=1+1)
Aggregate/PublicKeys-8         2.85ms ± 0%  2.81ms ± 0%   ~     (p=1.000 n=1+1)
Aggregate/Signatures-8         47.9ms ± 0%  48.0ms ± 0%   ~     (p=1.000 n=1+1)
SimpleKeyGen-8                 20.2ms ± 0%  20.3ms ± 0%   ~     (p=1.000 n=1+1)

Fixed bugs:

• Unexpected failure of ep2_mul[_lwnaf] above the prime group order #64

Closed issues:

• Other way to construct towered extension fields #203
• blake2.h:101:5: error: size of array element is not a multiple of its alignment #202
• ECIES 160bit #201
• Compilation with "ARITH gmp" fails #200
• Support for armv8-a ? #198
• Function name bn_init conflicts with OpenSSL when used in tandem #196
• 16-bit MSP430 #193
• Modular exponentiation returns 1 if exponent is 0 and modulo is 1 #185
• Compilation of RELIC with bls12-446 and bls12-455 fails #182
• test_bn fails with BLS12-381 preset #181
• [BUG] undefined reference to bench_init', bench_clean' #180
• Tests FTBFS because of missing symbol in header #179
• Builds are broken #178
• compile error inlining failed in call to always_inline ‘_mm_alignr_epi8’ on unbantu20.04 gcc9 #177
• bn_write_str buffer overflow #176
• ECDSA verify succeeds when it should fail #175
• ec_mul_gen hangs with curve SECG_K256 #174
• Wrong square root computation #173
• Out-of-bounds read via bn_sqr_basic #172
• OSS-Fuzz integration #171
• Building Relic with Curve NIST_P256 throws FATAL ERROR in relic_fp_prime.c:120 #170
• Compressing (packing) a point to binary array does not comply with X9.62 standard #169
• ‘ctx_t’ {aka ‘struct _ctx_t’} has no member named ‘total’ #168
• relic does not work with C++ #167
• Memory leak in ep2_curve_init/clean with ALLOC=DYNAMIC #166
• *_is_valid() functions produce false negative for not normalized points #147
• Bench and Test doesnt build #122

Merged pull requests:

• Add pairing delegation protocols #199 (dfaranha)
• Fix support for Win64/MSVC targets. #197 (dfaranha)
• Simplify generator getting for Gt. #194 (luozejiaqun)
• cmake: Always use user defined CFLAGS, not only for release builds #187 (xdustinface)
• Fix MinGW build #186 (xdustinface)
• Remove debug printf in bn_mxp_slide #184 (guidovranken)
• Remove ALLOC = STACK to simplify memory allocation. #183 (dfaranha)
• Update relic_alloc.h #165 (aguycalled)
• Add correct support for FreeBSD and NetBSD #164 (hoffmang9)

[tarakby]

This is amazing! Thank you @huitseeker 🙏
I can continue looking at these changes and fix the failing unit tests 👌

[huitseeker]

@tarakby I'm looking at the failing test, which is a change in the semantics of interpreting a zero slice as zero since relic-toolkit/relic@16f3c55 (see the last file)

[codecov-commenter]

Codecov Report

Merging #846 (f80a224) into master (f4a99c7) will increase coverage by 0.38%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #846      +/-   ##
==========================================
+ Coverage   56.51%   56.89%   +0.38%     
==========================================
  Files         427      429       +2     
  Lines       25126    25652     +526     
==========================================
+ Hits        14199    14595     +396     
- Misses       9008     9083      +75     
- Partials     1919     1974      +55     

Flag	Coverage Δ
unittests	56.89% <ø> (+0.38%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
model/flow/service_event.go	40.00% <0.00%> (-16.72%)	⬇️
model/flow/identifier.go	75.00% <0.00%> (-5.00%)	⬇️
engine/execution/state/delta/view.go	73.07% <0.00%> (-3.76%)	⬇️
engine/execution/state/delta/delta.go	66.66% <0.00%> (-3.61%)	⬇️
engine/common/rpc/convert/convert.go	31.31% <0.00%> (-3.04%)	⬇️
model/flow/role.go	89.36% <0.00%> (-1.95%)	⬇️
engine/consensus/sealing/core.go	60.98% <0.00%> (-0.76%)	⬇️
engine/verification/verifier/engine.go	52.97% <0.00%> (-0.64%)	⬇️
cmd/util/ledger/migrations/storage_v4.go	41.56% <0.00%> (-0.61%)	⬇️
model/flow/ledger.go	12.50% <0.00%> (-0.55%)	⬇️
... and 36 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ce77c63...f80a224. Read the comment docs.

[tarakby]

Thank you for compiling all the relevant commits and updates 🏄

I have made a suggestion branch to compile some changes. It mainly adds two points compared to your branch:

• COMP is not taken into account by the cmake pre-build, I updated it to CFLAGS.
• I added initializations for other bn_st variables used in the package. They need to be properly initialized after relic trimming was updated).

Feel free to cherry-pick the commits if you agree to them 🙏

[huitseeker]

@tarakby I think I merged all the relic update related changes from your branch, thanks for the commits!

[tarakby]

Thanks @huitseeker for adding the commits!

Everything looks good, we just need to delete the 2 lines commented here and here since the initializations are now happening before calling the functions.

[tarakby]

🏄 🏄‍♀️

[Kay-Zee]

I mostly focused my attention on the two new go helper functions and their usages, making sure those made sense.

Did not fully follow the C changes but nothing obvious jumped out at me looking through them.