-
Notifications
You must be signed in to change notification settings - Fork 236
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feat: Changing finite field arithmetic in wasm to 29 bits for multiplications #5435
Conversation
Benchmark resultsMetrics with a significant change:
Detailed resultsAll benchmarks are run on txs on the This benchmark source data is available in JSON format on S3 here. L2 block published to L1Each column represents the number of txs on an L2 block published to L1.
L2 chain processingEach column represents the number of blocks on the L2 chain where each block has 16 txs.
Circuits statsStats on running time and I/O sizes collected for every circuit run across all benchmarks.
Tree insertion statsThe duration to insert a fixed batch of leaves into each tree type.
MiscellaneousTransaction sizes based on how many contract classes are registered in the tx.
Transaction size based on fee payment method
Transaction processing duration by data writes.
|
10ee871
to
ad6e000
Compare
tmp_1 *= (delta_1 + minus_one); | ||
tmp_1 *= (delta_1 + minus_two); | ||
tmp_1 *= (delta_1 + minus_three); | ||
auto tmp_1 = (delta_1 + minus_one).sqr() - one; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
why not +minus_one
rather than -one
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
maybe i'm brainded but i computed by hand the diff and i don't get the same result, can you please assure me I'm brainded?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Replaced. wdym the result is not the same? (a+1)*(a-1)= a^2 - 1. We have a(a-1)(a-2)(a-3)= ((a-1)^2-1)((a-2)^2-1)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I can't maths :)
@@ -8,7 +8,7 @@ namespace bb::secp256r1 { | |||
struct FqParams { | |||
static constexpr uint64_t modulus_0 = 0xFFFFFFFFFFFFFFFFULL; | |||
static constexpr uint64_t modulus_1 = 0x00000000FFFFFFFFULL; | |||
static constexpr uint64_t modulus_2 = 0X0000000000000000ULL; | |||
static constexpr uint64_t modulus_2 = 0x0000000000000000ULL; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
this is not a diff?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
x was capital X
@@ -67,10 +67,14 @@ TEST(secp256k1, TestToMontgomeryForm) | |||
{ | |||
const size_t n = 10; | |||
for (size_t i = 0; i < n; ++i) { | |||
uint256_t a_raw = get_fq_element(); | |||
uint256_t a_raw = get_fq_element() - 1; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
why?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
No idea. Removing
@@ -42,6 +42,48 @@ struct FqParams { | |||
static constexpr uint64_t primitive_root_1 = 0UL; | |||
static constexpr uint64_t primitive_root_2 = 0UL; | |||
static constexpr uint64_t primitive_root_3 = 0UL; | |||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
have been derived manually or from a known source? i think we should have some scripts or something to verify them, wdyt?
fr a{ 0x192f9ddc938ea63, 0x1db93d61007ec4fe, 0xc89284ec31fa49c0, 0x2478d0ff12b04f0f }; | ||
fr b{ 0x7aade4892631231c, 0x8e7515681fe70144, 0x98edb76e689b6fd8, 0x5d0886b15fc835fa }; | ||
fr expected{ 0xab961ef46b4756b6, 0xbc6b636fc29678c8, 0xd247391ed6b5bd16, 0x12e8538b3bde6784 }; | ||
auto a_uint = uint256_t{ 0x192f9ddc938ea63, 0x1db93d61007ec4fe, 0xc89284ec31fa49c0, 0x2478d0ff12b04f0f }; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
can you explain why uint form is tested?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
At this point I still had hopes that I'd transform all tests into tests based on uint forms first. The problem was that all the tests had predefined field values in INTERNAL representation. Since Montgomery form changed for WASM, so did the internal representation. However, if we define uint256 and then convert to field, then the form will automatically change to whichever we are using now.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for the docs! Did a pass through all the commits succeeding my last review and things look good, just requests for clarifying stuff
1. \f$d=29⋅9=261\f$ for builds that don't support the uint128_t type, for example, for WASM build | ||
2. \f$d=64⋅4=256\f$ for standard builds (x86_64). | ||
|
||
The goal of using Montgomery form is to avoid heavy division modulo \f$p\f$. To compute a representative of element $$c = a⋅b\ mod\ p$$ we compute $$c⋅R = (a⋅R)⋅(b⋅R) / R\ mod\ p$$, but we use an efficient division trick to avoid straight modular division. Let's look into the standard 4⋅64 case: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
what do you mean by representative?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Internally an element c is represented by
The goal of using Montgomery form is to avoid heavy division modulo \f$p\f$. To compute a representative of element $$c = a⋅b\ mod\ p$$ we compute $$c⋅R = (a⋅R)⋅(b⋅R) / R\ mod\ p$$, but we use an efficient division trick to avoid straight modular division. Let's look into the standard 4⋅64 case: | ||
1. First, we compute the value $$c_r=c⋅R⋅R = aR⋅bR$$ in integers and get a value with 8 64-bit limbs | ||
2. Then we take the lowest limb of \f$c_r\f$ (\f$c_r[0]\f$) and multiply it by a special value $$r_{inv} = -1 ⋅ p^{-1}\ mod\ 2^{64}$$ As a result we get $$k = r_{inv}⋅ c_r[0]\ mod\ 2^{64}$$ | ||
3. Next we update \f$c_r\f$ in integers by adding a value \f$k⋅p\f$: $$c_r = c_{r\_prev} + k⋅p$$ You might notice that the value of \f$c_r\ mod\ p\f$ hasn't changed, since we've added a multiple of the modulus. A the same time, if we look at the expression modulo \f$2^{64}\f$: $$c_r + k⋅p = c_r + c_r⋅r_{inv}⋅p = c_r + c_r⋅ (-1)⋅p^{-1}⋅p = c_r - c_r = 0\ mod\ 2^{64}$$ The result is equivalent modulo \f$p\f$, but we zeroed out the lowest limb |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
this renders weirdly, also can you define c_r_{prev} somewhere?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What exactly renders weirdly?
|
||
Why does this work? Originally both \f$aR\f$ and \f$bR\f$ are less than the modulus \f$p\f$ in integers, so $$aR\cdot bR <= (p-1)^2$$ During each of the \f$k\cdot p\f$ addition rounds we can add at most \f$(2^{64}-1)p\f$ to corresponding digits, so at most we add \f$(2^{256}-1)p\f$ and the total is $$aR\cdot bR + k_{0,1,2,3}p \le (p-1)^2+(2^{256}-1)p < 2\cdot 2^{256}p \Rightarrow c_{r.high} = \frac{aR\cdot bR + k_{0,1,2,3}p}{2^{256}} < 2p$$. | ||
|
||
For bn254 scalar and base fields we can do even better by employing a simple trick. The moduli of both fields are 254 bits, while 4 64-bit limbs allow 256 bits of storage. We relax the internal representation to use values in range \f$[0,2p)\f$. The addition, negation and subtraction operation logic doesn't change, we simply replace the modulus \f$p\f$ with \f$2p\f$, but the mutliplication becomes more efficient. The multiplicands are in range \f$[0,2p)\f$, but we add multiples of modulus \f$p\f$ to reduce limbs, not \f$2p\f$. If we revisit the \f$c_r\f$ formula: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
can you add a sentence that it's because we represent limbs in a lazy manner that we have only 256 bits (it's because of the lazy representation right?)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Only 256 rather than 254?
For bn254 scalar and base fields we can do even better by employing a simple trick. The moduli of both fields are 254 bits, while 4 64-bit limbs allow 256 bits of storage. We relax the internal representation to use values in range \f$[0,2p)\f$. The addition, negation and subtraction operation logic doesn't change, we simply replace the modulus \f$p\f$ with \f$2p\f$, but the mutliplication becomes more efficient. The multiplicands are in range \f$[0,2p)\f$, but we add multiples of modulus \f$p\f$ to reduce limbs, not \f$2p\f$. If we revisit the \f$c_r\f$ formula: | ||
$$aR\cdot bR + k_{0,1,2,3}p \le (2p-1)^2+(2^{256}-1)p = 2^{256}p+4p^2-5p+1 \Rightarrow$$ $$\Rightarrow c_{r.high} = \frac{aR\cdot bR + k_{0,1,2,3}p}{2^{256}} \le \frac{2^{256}p+4p^2-5p+1}{2^{256}}=p +\frac{4p^2 - 5p +1}{2^{256}}, 4p < 2^{256} \Rightarrow$$ $$\Rightarrow p +\frac{4p^2 - 5p +1}{2^{256}} < 2p$$ So we ended in the same range and we don't have to perform additional reductions. | ||
|
||
**N.B.** In the code we refer to this form as coarse |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
coarse is such wonderful word :)
|
||
## Field parameters | ||
|
||
The field template is instantiated with field parameter classes, for example, class bb::Bn254FqParams. Each such class contains at least the modulus (in 64-bit and 29-bit form), r_inv and 2 versions of r_squared (64-bit and WASM/29-bit version). r_squared and other parameters (such as cube_root, primitive_root and coset_generators) are defined for wasm separately, becuase the values represent an element already in Montgomery form. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can you remind the reader what r_inv
is, why we need r_square
also why we don't also store R
, R_square
since it's also constant right?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We store R_squared. We don't store R
|
||
The field template is instantiated with field parameter classes, for example, class bb::Bn254FqParams. Each such class contains at least the modulus (in 64-bit and 29-bit form), r_inv and 2 versions of r_squared (64-bit and WASM/29-bit version). r_squared and other parameters (such as cube_root, primitive_root and coset_generators) are defined for wasm separately, becuase the values represent an element already in Montgomery form. | ||
|
||
## Helpful python snippets |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
thank you for your community service :)
* master: feat: Wire AVM gas used to public kernel (#5740) fix(ci,noir-projects): bring apt-get higher in cache (#5775) fix: e2e getStack, disable failing e2e (#5768) feat: Changing finite field arithmetic in wasm to 29 bits for multiplications (#5435) feat!: trap with revert data (#5732) feat: impl of missing functionality in new key store (#5750) git subrepo push --branch=master noir-projects/aztec-nr git_subrepo.sh: Fix parent in .gitrepo file. [skip ci] chore: replace relative paths to noir-protocol-circuits git subrepo push --branch=master barretenberg fix(ci): 192 core spot runner (#5767) feat(acir)!: BrilligCall opcode (#5709) fix(ci): speculative deploy fix fix(ci): don't fail if can't prune feat(ci): turn on new CI as mandatory (#5761)
🤖 I have created a release *beep* *boop* --- <details><summary>aztec-package: 0.35.0</summary> ## [0.35.0](aztec-package-v0.34.0...aztec-package-v0.35.0) (2024-04-16) ### ⚠ BREAKING CHANGES * pay fee for account init ([#5601](#5601)) ### Features * Pay fee for account init ([#5601](#5601)) ([aca804f](aca804f)) </details> <details><summary>barretenberg.js: 0.35.0</summary> ## [0.35.0](barretenberg.js-v0.34.0...barretenberg.js-v0.35.0) (2024-04-16) ### Features * Export poseidon2_permutation and add to foundation/crypto ([#5706](#5706)) ([6b91e27](6b91e27)) ### Miscellaneous * Don't strip bb wasm ([#5743](#5743)) ([d4cb410](d4cb410)) * TS hash wrappers cleanup ([#5691](#5691)) ([7f8b09f](7f8b09f)) </details> <details><summary>aztec-cli: 0.35.0</summary> ## [0.35.0](aztec-cli-v0.34.0...aztec-cli-v0.35.0) (2024-04-16) ### ⚠ BREAKING CHANGES * pay fee for account init ([#5601](#5601)) ### Features * Pay fee for account init ([#5601](#5601)) ([aca804f](aca804f)) </details> <details><summary>aztec-packages: 0.35.0</summary> ## [0.35.0](aztec-packages-v0.34.0...aztec-packages-v0.35.0) (2024-04-16) ### ⚠ BREAKING CHANGES * Use fixed size arrays in black box functions where sizes are known ([#5620](#5620)) * trap with revert data ([#5732](#5732)) * **acir:** BrilligCall opcode ([#5709](#5709)) * rename request_max_block_number ([#5675](#5675)) * pay fee for account init ([#5601](#5601)) ### Features * **acir:** BrilligCall opcode ([#5709](#5709)) ([f06f64c](f06f64c)) * Add serialisation methods ([#5749](#5749)) ([20d290c](20d290c)) * App siloing in new key store ([#5721](#5721)) ([ae37d32](ae37d32)), closes [#5635](#5635) * **avm-simulator:** Plumb noir assertion messages ([#5774](#5774)) ([2cf11ac](2cf11ac)) * **avm:** CMOV opcode ([#5575](#5575)) ([19dbe46](19dbe46)), closes [#5557](#5557) * **avm:** Enable contract testing with bb binary ([#5584](#5584)) ([d007d79](d007d79)) * **avm:** Enable range check on the ALU registers ([#5696](#5696)) ([202fc1b](202fc1b)) * **avm:** Keccak as blackbox function ([#5722](#5722)) ([6ea677a](6ea677a)) * **avm:** Poseidon2_permutation as black box ([#5707](#5707)) ([5526b36](5526b36)) * **avm:** Sha256 as blackbox function ([#5727](#5727)) ([cac9cba](cac9cba)) * **avm:** Take sizeOffset in CALL ([#5763](#5763)) ([95eadd6](95eadd6)) * Brillig heterogeneous memory cells ([#5608](#5608)) ([3287aa2](3287aa2)) * Change public nullifiers api ([#5660](#5660)) ([986e7f9](986e7f9)) * Changing finite field arithmetic in wasm to 29 bits for multiplications ([#5435](#5435)) ([b2d9b9d](b2d9b9d)) * **ci:** Turn on new CI as mandatory ([#5761](#5761)) ([bebed32](bebed32)) * **docs:** Merge yellow paper into docs protocol specs section ([#5668](#5668)) ([66dc509](66dc509)) * E2e token contract can run in 2m with snapshots and test separation. ([#5526](#5526)) ([b0037dd](b0037dd)) * Export poseidon2_permutation and add to foundation/crypto ([#5706](#5706)) ([6b91e27](6b91e27)) * Get last mock oracles params (noir-lang/noir#4789) ([825c455](825c455)) * Impl of missing functionality in new key store ([#5750](#5750)) ([af49a29](af49a29)) * LT/LTE for AVM ([#5559](#5559)) ([350abeb](350abeb)) * New key store ([#5653](#5653)) ([3e44a58](3e44a58)), closes [#5607](#5607) * Pay fee for account init ([#5601](#5601)) ([aca804f](aca804f)) * Poseidon separator ([#5717](#5717)) ([d5256d2](d5256d2)) * Proving the rollup circuits ([#5599](#5599)) ([145cbcd](145cbcd)) * Public Kernel proving orchestration ([#5748](#5748)) ([2ae0ee5](2ae0ee5)) * Rename request_max_block_number ([#5675](#5675)) ([c695fcd](c695fcd)) * Separate nullfier_inclusion checks for private/public/avm ([#5657](#5657)) ([e4d2df6](e4d2df6)) * Sequencer validates setup/teardown function selectors ([#5649](#5649)) ([8f8ad56](8f8ad56)), closes [#5401](#5401) * Shared mutable storage ([#5490](#5490)) ([c4e41a9](c4e41a9)) * **simulator:** Fetch return values at circuit execution ([#5642](#5642)) ([413a4e0](413a4e0)) * Split `backend_barretenburg` into prover and verifier classes (noir-lang/noir#4769) ([825c455](825c455)) * Sync from aztec-packages (noir-lang/noir#4764) ([1eb288e](1eb288e)) * Sync from aztec-packages (noir-lang/noir#4787) ([825c455](825c455)) * Trap with revert data ([#5732](#5732)) ([f849575](f849575)) * Unroll loops iteratively (noir-lang/noir#4779) ([1eb288e](1eb288e)) * Update circuits structs with gas info ([#5677](#5677)) ([3db6dd1](3db6dd1)) * Use fixed size arrays in black box functions where sizes are known ([#5620](#5620)) ([f50b180](f50b180)) * Variable length returns ([#5633](#5633)) ([b4a6f17](b4a6f17)) * Wire AVM gas used to public kernel ([#5740](#5740)) ([4f55d10](4f55d10)) ### Bug Fixes * "feat: Changing finite field arithmetic in wasm to 29 bits for multiplications" ([#5779](#5779)) ([bcfee97](bcfee97)) * Anvil start retry in case something bad. Fix colors. ([#5673](#5673)) ([0b6b6f6](0b6b6f6)) * ArrayGet and Set are not pure (noir-lang/noir#4783) ([825c455](825c455)) * Avoid get row in databus ([#5742](#5742)) ([d67b6c8](d67b6c8)) * Avoid huge unrolling in hash_args ([#5703](#5703)) ([10d9ad9](10d9ad9)) * **ci,noir-projects:** Bring apt-get higher in cache ([#5775](#5775)) ([d37cbb9](d37cbb9)) * **ci:** 192 core spot runner ([#5767](#5767)) ([37daac6](37daac6)) * **ci:** Bigger cache disk, cache+prune docker images, disable ClientIvcTests.Full ([#5729](#5729)) ([5dcbd75](5dcbd75)) * **ci:** Builder types ([#5711](#5711)) ([b16f169](b16f169)) * **ci:** Cache size not honoured ([#5738](#5738)) ([d4ff340](d4ff340)) * **ci:** Don't fail if can't prune ([d9bb2c7](d9bb2c7)) * **ci:** Error in spot ([#5745](#5745)) ([4d754aa](4d754aa)) * **ci:** Fix arm e2e references, spot shutdown ([#5741](#5741)) ([1c4667c](1c4667c)) * **ci:** Hotfix arm ([1ddb1c7](1ddb1c7)) * **ci:** Hotfix just one ARM task ([10f27ae](10f27ae)) * **ci:** Speculative deploy fix ([9a9eab6](9a9eab6)) * **ci:** Wait for mainnet fork deployment ([#5735](#5735)) ([8f3794d](8f3794d)) * **ci:** Wait_for_fork env var ([#5780](#5780)) ([d85267b](d85267b)) * Correct ICE panic messages in brillig `convert_black_box_call` (noir-lang/noir#4761) ([1eb288e](1eb288e)) * Disable flakey vanilla recursion test ([#5672](#5672)) ([f84f7b6](f84f7b6)) * Don't run e2e tests against wrong anvil ([#5686](#5686)) ([9ff45f6](9ff45f6)) * Dont error in bench summary ([#5693](#5693)) ([470b0f3](470b0f3)) * E2e getStack, disable failing e2e ([#5768](#5768)) ([e5f3ece](e5f3ece)) * GA concurrency ([#5713](#5713)) ([eac2585](eac2585)) * Generate_aztecnr_reference.js not getting generics or multi-line params ([#5679](#5679)) ([a22bc3d](a22bc3d)) * Hotfix submodule cache ([92b92b3](92b92b3)) * Hotfix underspec'd machine ([#5710](#5710)) ([059e38e](059e38e)) * **hotfix:** CI ignore git safe.directory checks ([#5659](#5659)) ([9fc3fe3](9fc3fe3)) * Less earthly cache ([#5690](#5690)) ([8190dc7](8190dc7)) * Make earthly more parallel ([#5747](#5747)) ([9734455](9734455)) * Primary_message typo in errors.rs ([#5646](#5646)) ([1dfbe7b](1dfbe7b)) * Pull noir ([#5699](#5699)) ([bf35464](bf35464)) * REDO dont error in bench summary ([#5695](#5695)) ([8c1a7b9](8c1a7b9)) * Running e2e tests as part of build, requires forcing ip4 (not ip6) when connecting to anvil ([#5744](#5744)) ([66fc89f](66fc89f)) * Simplify ECCVM prover constructor and add a TODO ([#5681](#5681)) ([8c151ea](8c151ea)) * Spot refcount ([#5746](#5746)) ([9e18444](9e18444)) * Take a deep copy of circuit inputs for proving ([#5777](#5777)) ([785591e](785591e)) * Temporarily disable the bench tests ([#5755](#5755)) ([1d52ac5](1d52ac5)) * Update commit for noir-gates-diff (noir-lang/noir#4773) ([1eb288e](1eb288e)) * Use entrypoint instead of pay_init_fee ([#5623](#5623)) ([62ac765](62ac765)) * Watch less files. ([#5651](#5651)) ([57a1d69](57a1d69)) ### Miscellaneous * Add missing aztec-address tests ([#5674](#5674)) ([58aefba](58aefba)) * **avm:** Add a boolean to toggle proving in AVM unit tests ([#5667](#5667)) ([ec122c9](ec122c9)), closes [#5663](#5663) * **avm:** Hashing tests cleanup ([#5733](#5733)) ([53d0102](53d0102)) * **avm:** Range checks negative tests ([#5770](#5770)) ([2907142](2907142)) * **avm:** Split the negative test on range check for high 16-bit registers ([#5785](#5785)) ([8ebbe57](8ebbe57)) * **avm:** Split up AVM test contract as it was growing too large ([#5702](#5702)) ([5b8e812](5b8e812)) * **aztec-nr:** Minor public interface changes ([#5776](#5776)) ([91b8110](91b8110)) * **ci:** Break e2e-deploy into multiple test suites ([#5704](#5704)) ([2522294](2522294)) * **ci:** Earthly in spot with persistent cache ([#5644](#5644)) ([a39c2f6](a39c2f6)) * **ci:** Hotfix AMI's, workflow to stop personal spot runners ([#5712](#5712)) ([5f18139](5f18139)) * **ci:** Only run ARM on master ([#5705](#5705)) ([f77c142](f77c142)) * **ci:** Use 128 cores for x86 and add timeouts ([#5665](#5665)) ([0c5dc0a](0c5dc0a)) * Compute_note_hash_and_nullifier - improve error message ([#5671](#5671)) ([8942d69](8942d69)) * Create placeholder version of 0.26.0 docs (noir-lang/noir#4782) ([1eb288e](1eb288e)) * **doc:** Fix broken docs links (noir-lang/noir#4606) ([1eb288e](1eb288e)) * **docs:** Fix link in the Data Types page (noir-lang/noir#4527) ([1eb288e](1eb288e)) * Don't strip bb wasm ([#5743](#5743)) ([d4cb410](d4cb410)) * Fix master after merge issue related to validate_trace renaming ([#5676](#5676)) ([44e0d8a](44e0d8a)) * Fix max-block-number and auth e2e tests ([#5694](#5694)) ([f1bf314](f1bf314)) * Op queue ([#5648](#5648)) ([822c7e6](822c7e6)) * **public:** Remove getNullifierMembershipWitness ([#5715](#5715)) ([3be402c](3be402c)) * Re-enable e2e fees tests ([#5784](#5784)) ([102e8b8](102e8b8)) * Release Noir(0.27.0) (noir-lang/noir#4632) ([1eb288e](1eb288e)) * Remove the old Value struct from the oracle docs (noir-lang/noir#4738) ([1eb288e](1eb288e)) * Replace relative paths to noir-protocol-circuits ([fb2b298](fb2b298)) * Replace relative paths to noir-protocol-circuits ([e20920d](e20920d)) * Replace relative paths to noir-protocol-circuits ([6351dc5](6351dc5)) * Replace relative paths to noir-protocol-circuits ([fee13bf](fee13bf)) * Replacing unsafe::zeroed() ([#5685](#5685)) ([ea3884e](ea3884e)) * Small logging changes ([#5654](#5654)) ([25cc70d](25cc70d)) * Temporarily skip failing e2e fees test ([a3ac5ff](a3ac5ff)) * Testing that nargo fmt is idempotent (noir-lang/noir#4765) ([825c455](825c455)) * TS hash wrappers cleanup ([#5691](#5691)) ([7f8b09f](7f8b09f)) * Turn ENABLE_GAS where it is needed ([#5730](#5730)) ([30a2edd](30a2edd)) * Update noir gates diff ([#5658](#5658)) ([9816c1a](9816c1a)) * We can run 35 of our e2e tests just using jest. ([#5643](#5643)) ([4fcaeae](4fcaeae)) ### Documentation * Fix yp typo control-flow.md ([#5638](#5638)) ([363d227](363d227)) </details> <details><summary>barretenberg: 0.35.0</summary> ## [0.35.0](barretenberg-v0.34.0...barretenberg-v0.35.0) (2024-04-16) ### ⚠ BREAKING CHANGES * Use fixed size arrays in black box functions where sizes are known ([#5620](#5620)) * trap with revert data ([#5732](#5732)) * **acir:** BrilligCall opcode ([#5709](#5709)) ### Features * **acir:** BrilligCall opcode ([#5709](#5709)) ([f06f64c](f06f64c)) * **avm:** CMOV opcode ([#5575](#5575)) ([19dbe46](19dbe46)), closes [#5557](#5557) * **avm:** Enable contract testing with bb binary ([#5584](#5584)) ([d007d79](d007d79)) * **avm:** Enable range check on the ALU registers ([#5696](#5696)) ([202fc1b](202fc1b)) * Changing finite field arithmetic in wasm to 29 bits for multiplications ([#5435](#5435)) ([b2d9b9d](b2d9b9d)) * **ci:** Turn on new CI as mandatory ([#5761](#5761)) ([bebed32](bebed32)) * Export poseidon2_permutation and add to foundation/crypto ([#5706](#5706)) ([6b91e27](6b91e27)) * LT/LTE for AVM ([#5559](#5559)) ([350abeb](350abeb)) * Trap with revert data ([#5732](#5732)) ([f849575](f849575)) * Use fixed size arrays in black box functions where sizes are known ([#5620](#5620)) ([f50b180](f50b180)) ### Bug Fixes * "feat: Changing finite field arithmetic in wasm to 29 bits for multiplications" ([#5779](#5779)) ([bcfee97](bcfee97)) * Avoid get row in databus ([#5742](#5742)) ([d67b6c8](d67b6c8)) * **ci:** Bigger cache disk, cache+prune docker images, disable ClientIvcTests.Full ([#5729](#5729)) ([5dcbd75](5dcbd75)) * Disable flakey vanilla recursion test ([#5672](#5672)) ([f84f7b6](f84f7b6)) * Less earthly cache ([#5690](#5690)) ([8190dc7](8190dc7)) * Make earthly more parallel ([#5747](#5747)) ([9734455](9734455)) * Simplify ECCVM prover constructor and add a TODO ([#5681](#5681)) ([8c151ea](8c151ea)) ### Miscellaneous * **avm:** Add a boolean to toggle proving in AVM unit tests ([#5667](#5667)) ([ec122c9](ec122c9)), closes [#5663](#5663) * **avm:** Range checks negative tests ([#5770](#5770)) ([2907142](2907142)) * **avm:** Split the negative test on range check for high 16-bit registers ([#5785](#5785)) ([8ebbe57](8ebbe57)) * **ci:** Use 128 cores for x86 and add timeouts ([#5665](#5665)) ([0c5dc0a](0c5dc0a)) * Don't strip bb wasm ([#5743](#5743)) ([d4cb410](d4cb410)) * Fix master after merge issue related to validate_trace renaming ([#5676](#5676)) ([44e0d8a](44e0d8a)) * Op queue ([#5648](#5648)) ([822c7e6](822c7e6)) </details> --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please).
🤖 I have created a release *beep* *boop* --- <details><summary>aztec-package: 0.35.0</summary> ## [0.35.0](AztecProtocol/aztec-packages@aztec-package-v0.34.0...aztec-package-v0.35.0) (2024-04-16) ### ⚠ BREAKING CHANGES * pay fee for account init ([#5601](AztecProtocol/aztec-packages#5601)) ### Features * Pay fee for account init ([#5601](AztecProtocol/aztec-packages#5601)) ([aca804f](AztecProtocol/aztec-packages@aca804f)) </details> <details><summary>barretenberg.js: 0.35.0</summary> ## [0.35.0](AztecProtocol/aztec-packages@barretenberg.js-v0.34.0...barretenberg.js-v0.35.0) (2024-04-16) ### Features * Export poseidon2_permutation and add to foundation/crypto ([#5706](AztecProtocol/aztec-packages#5706)) ([6b91e27](AztecProtocol/aztec-packages@6b91e27)) ### Miscellaneous * Don't strip bb wasm ([#5743](AztecProtocol/aztec-packages#5743)) ([d4cb410](AztecProtocol/aztec-packages@d4cb410)) * TS hash wrappers cleanup ([#5691](AztecProtocol/aztec-packages#5691)) ([7f8b09f](AztecProtocol/aztec-packages@7f8b09f)) </details> <details><summary>aztec-cli: 0.35.0</summary> ## [0.35.0](AztecProtocol/aztec-packages@aztec-cli-v0.34.0...aztec-cli-v0.35.0) (2024-04-16) ### ⚠ BREAKING CHANGES * pay fee for account init ([#5601](AztecProtocol/aztec-packages#5601)) ### Features * Pay fee for account init ([#5601](AztecProtocol/aztec-packages#5601)) ([aca804f](AztecProtocol/aztec-packages@aca804f)) </details> <details><summary>aztec-packages: 0.35.0</summary> ## [0.35.0](AztecProtocol/aztec-packages@aztec-packages-v0.34.0...aztec-packages-v0.35.0) (2024-04-16) ### ⚠ BREAKING CHANGES * Use fixed size arrays in black box functions where sizes are known ([#5620](AztecProtocol/aztec-packages#5620)) * trap with revert data ([#5732](AztecProtocol/aztec-packages#5732)) * **acir:** BrilligCall opcode ([#5709](AztecProtocol/aztec-packages#5709)) * rename request_max_block_number ([#5675](AztecProtocol/aztec-packages#5675)) * pay fee for account init ([#5601](AztecProtocol/aztec-packages#5601)) ### Features * **acir:** BrilligCall opcode ([#5709](AztecProtocol/aztec-packages#5709)) ([f06f64c](AztecProtocol/aztec-packages@f06f64c)) * Add serialisation methods ([#5749](AztecProtocol/aztec-packages#5749)) ([20d290c](AztecProtocol/aztec-packages@20d290c)) * App siloing in new key store ([#5721](AztecProtocol/aztec-packages#5721)) ([ae37d32](AztecProtocol/aztec-packages@ae37d32)), closes [#5635](AztecProtocol/aztec-packages#5635) * **avm-simulator:** Plumb noir assertion messages ([#5774](AztecProtocol/aztec-packages#5774)) ([2cf11ac](AztecProtocol/aztec-packages@2cf11ac)) * **avm:** CMOV opcode ([#5575](AztecProtocol/aztec-packages#5575)) ([19dbe46](AztecProtocol/aztec-packages@19dbe46)), closes [#5557](AztecProtocol/aztec-packages#5557) * **avm:** Enable contract testing with bb binary ([#5584](AztecProtocol/aztec-packages#5584)) ([d007d79](AztecProtocol/aztec-packages@d007d79)) * **avm:** Enable range check on the ALU registers ([#5696](AztecProtocol/aztec-packages#5696)) ([202fc1b](AztecProtocol/aztec-packages@202fc1b)) * **avm:** Keccak as blackbox function ([#5722](AztecProtocol/aztec-packages#5722)) ([6ea677a](AztecProtocol/aztec-packages@6ea677a)) * **avm:** Poseidon2_permutation as black box ([#5707](AztecProtocol/aztec-packages#5707)) ([5526b36](AztecProtocol/aztec-packages@5526b36)) * **avm:** Sha256 as blackbox function ([#5727](AztecProtocol/aztec-packages#5727)) ([cac9cba](AztecProtocol/aztec-packages@cac9cba)) * **avm:** Take sizeOffset in CALL ([#5763](AztecProtocol/aztec-packages#5763)) ([95eadd6](AztecProtocol/aztec-packages@95eadd6)) * Brillig heterogeneous memory cells ([#5608](AztecProtocol/aztec-packages#5608)) ([3287aa2](AztecProtocol/aztec-packages@3287aa2)) * Change public nullifiers api ([#5660](AztecProtocol/aztec-packages#5660)) ([986e7f9](AztecProtocol/aztec-packages@986e7f9)) * Changing finite field arithmetic in wasm to 29 bits for multiplications ([#5435](AztecProtocol/aztec-packages#5435)) ([b2d9b9d](AztecProtocol/aztec-packages@b2d9b9d)) * **ci:** Turn on new CI as mandatory ([#5761](AztecProtocol/aztec-packages#5761)) ([bebed32](AztecProtocol/aztec-packages@bebed32)) * **docs:** Merge yellow paper into docs protocol specs section ([#5668](AztecProtocol/aztec-packages#5668)) ([66dc509](AztecProtocol/aztec-packages@66dc509)) * E2e token contract can run in 2m with snapshots and test separation. ([#5526](AztecProtocol/aztec-packages#5526)) ([b0037dd](AztecProtocol/aztec-packages@b0037dd)) * Export poseidon2_permutation and add to foundation/crypto ([#5706](AztecProtocol/aztec-packages#5706)) ([6b91e27](AztecProtocol/aztec-packages@6b91e27)) * Get last mock oracles params (noir-lang/noir#4789) ([825c455](AztecProtocol/aztec-packages@825c455)) * Impl of missing functionality in new key store ([#5750](AztecProtocol/aztec-packages#5750)) ([af49a29](AztecProtocol/aztec-packages@af49a29)) * LT/LTE for AVM ([#5559](AztecProtocol/aztec-packages#5559)) ([350abeb](AztecProtocol/aztec-packages@350abeb)) * New key store ([#5653](AztecProtocol/aztec-packages#5653)) ([3e44a58](AztecProtocol/aztec-packages@3e44a58)), closes [#5607](AztecProtocol/aztec-packages#5607) * Pay fee for account init ([#5601](AztecProtocol/aztec-packages#5601)) ([aca804f](AztecProtocol/aztec-packages@aca804f)) * Poseidon separator ([#5717](AztecProtocol/aztec-packages#5717)) ([d5256d2](AztecProtocol/aztec-packages@d5256d2)) * Proving the rollup circuits ([#5599](AztecProtocol/aztec-packages#5599)) ([145cbcd](AztecProtocol/aztec-packages@145cbcd)) * Public Kernel proving orchestration ([#5748](AztecProtocol/aztec-packages#5748)) ([2ae0ee5](AztecProtocol/aztec-packages@2ae0ee5)) * Rename request_max_block_number ([#5675](AztecProtocol/aztec-packages#5675)) ([c695fcd](AztecProtocol/aztec-packages@c695fcd)) * Separate nullfier_inclusion checks for private/public/avm ([#5657](AztecProtocol/aztec-packages#5657)) ([e4d2df6](AztecProtocol/aztec-packages@e4d2df6)) * Sequencer validates setup/teardown function selectors ([#5649](AztecProtocol/aztec-packages#5649)) ([8f8ad56](AztecProtocol/aztec-packages@8f8ad56)), closes [#5401](AztecProtocol/aztec-packages#5401) * Shared mutable storage ([#5490](AztecProtocol/aztec-packages#5490)) ([c4e41a9](AztecProtocol/aztec-packages@c4e41a9)) * **simulator:** Fetch return values at circuit execution ([#5642](AztecProtocol/aztec-packages#5642)) ([413a4e0](AztecProtocol/aztec-packages@413a4e0)) * Split `backend_barretenburg` into prover and verifier classes (noir-lang/noir#4769) ([825c455](AztecProtocol/aztec-packages@825c455)) * Sync from aztec-packages (noir-lang/noir#4764) ([1eb288e](AztecProtocol/aztec-packages@1eb288e)) * Sync from aztec-packages (noir-lang/noir#4787) ([825c455](AztecProtocol/aztec-packages@825c455)) * Trap with revert data ([#5732](AztecProtocol/aztec-packages#5732)) ([f849575](AztecProtocol/aztec-packages@f849575)) * Unroll loops iteratively (noir-lang/noir#4779) ([1eb288e](AztecProtocol/aztec-packages@1eb288e)) * Update circuits structs with gas info ([#5677](AztecProtocol/aztec-packages#5677)) ([3db6dd1](AztecProtocol/aztec-packages@3db6dd1)) * Use fixed size arrays in black box functions where sizes are known ([#5620](AztecProtocol/aztec-packages#5620)) ([f50b180](AztecProtocol/aztec-packages@f50b180)) * Variable length returns ([#5633](AztecProtocol/aztec-packages#5633)) ([b4a6f17](AztecProtocol/aztec-packages@b4a6f17)) * Wire AVM gas used to public kernel ([#5740](AztecProtocol/aztec-packages#5740)) ([4f55d10](AztecProtocol/aztec-packages@4f55d10)) ### Bug Fixes * "feat: Changing finite field arithmetic in wasm to 29 bits for multiplications" ([#5779](AztecProtocol/aztec-packages#5779)) ([bcfee97](AztecProtocol/aztec-packages@bcfee97)) * Anvil start retry in case something bad. Fix colors. ([#5673](AztecProtocol/aztec-packages#5673)) ([0b6b6f6](AztecProtocol/aztec-packages@0b6b6f6)) * ArrayGet and Set are not pure (noir-lang/noir#4783) ([825c455](AztecProtocol/aztec-packages@825c455)) * Avoid get row in databus ([#5742](AztecProtocol/aztec-packages#5742)) ([d67b6c8](AztecProtocol/aztec-packages@d67b6c8)) * Avoid huge unrolling in hash_args ([#5703](AztecProtocol/aztec-packages#5703)) ([10d9ad9](AztecProtocol/aztec-packages@10d9ad9)) * **ci,noir-projects:** Bring apt-get higher in cache ([#5775](AztecProtocol/aztec-packages#5775)) ([d37cbb9](AztecProtocol/aztec-packages@d37cbb9)) * **ci:** 192 core spot runner ([#5767](AztecProtocol/aztec-packages#5767)) ([37daac6](AztecProtocol/aztec-packages@37daac6)) * **ci:** Bigger cache disk, cache+prune docker images, disable ClientIvcTests.Full ([#5729](AztecProtocol/aztec-packages#5729)) ([5dcbd75](AztecProtocol/aztec-packages@5dcbd75)) * **ci:** Builder types ([#5711](AztecProtocol/aztec-packages#5711)) ([b16f169](AztecProtocol/aztec-packages@b16f169)) * **ci:** Cache size not honoured ([#5738](AztecProtocol/aztec-packages#5738)) ([d4ff340](AztecProtocol/aztec-packages@d4ff340)) * **ci:** Don't fail if can't prune ([d9bb2c7](AztecProtocol/aztec-packages@d9bb2c7)) * **ci:** Error in spot ([#5745](AztecProtocol/aztec-packages#5745)) ([4d754aa](AztecProtocol/aztec-packages@4d754aa)) * **ci:** Fix arm e2e references, spot shutdown ([#5741](AztecProtocol/aztec-packages#5741)) ([1c4667c](AztecProtocol/aztec-packages@1c4667c)) * **ci:** Hotfix arm ([1ddb1c7](AztecProtocol/aztec-packages@1ddb1c7)) * **ci:** Hotfix just one ARM task ([10f27ae](AztecProtocol/aztec-packages@10f27ae)) * **ci:** Speculative deploy fix ([9a9eab6](AztecProtocol/aztec-packages@9a9eab6)) * **ci:** Wait for mainnet fork deployment ([#5735](AztecProtocol/aztec-packages#5735)) ([8f3794d](AztecProtocol/aztec-packages@8f3794d)) * **ci:** Wait_for_fork env var ([#5780](AztecProtocol/aztec-packages#5780)) ([d85267b](AztecProtocol/aztec-packages@d85267b)) * Correct ICE panic messages in brillig `convert_black_box_call` (noir-lang/noir#4761) ([1eb288e](AztecProtocol/aztec-packages@1eb288e)) * Disable flakey vanilla recursion test ([#5672](AztecProtocol/aztec-packages#5672)) ([f84f7b6](AztecProtocol/aztec-packages@f84f7b6)) * Don't run e2e tests against wrong anvil ([#5686](AztecProtocol/aztec-packages#5686)) ([9ff45f6](AztecProtocol/aztec-packages@9ff45f6)) * Dont error in bench summary ([#5693](AztecProtocol/aztec-packages#5693)) ([470b0f3](AztecProtocol/aztec-packages@470b0f3)) * E2e getStack, disable failing e2e ([#5768](AztecProtocol/aztec-packages#5768)) ([e5f3ece](AztecProtocol/aztec-packages@e5f3ece)) * GA concurrency ([#5713](AztecProtocol/aztec-packages#5713)) ([eac2585](AztecProtocol/aztec-packages@eac2585)) * Generate_aztecnr_reference.js not getting generics or multi-line params ([#5679](AztecProtocol/aztec-packages#5679)) ([a22bc3d](AztecProtocol/aztec-packages@a22bc3d)) * Hotfix submodule cache ([92b92b3](AztecProtocol/aztec-packages@92b92b3)) * Hotfix underspec'd machine ([#5710](AztecProtocol/aztec-packages#5710)) ([059e38e](AztecProtocol/aztec-packages@059e38e)) * **hotfix:** CI ignore git safe.directory checks ([#5659](AztecProtocol/aztec-packages#5659)) ([9fc3fe3](AztecProtocol/aztec-packages@9fc3fe3)) * Less earthly cache ([#5690](AztecProtocol/aztec-packages#5690)) ([8190dc7](AztecProtocol/aztec-packages@8190dc7)) * Make earthly more parallel ([#5747](AztecProtocol/aztec-packages#5747)) ([9734455](AztecProtocol/aztec-packages@9734455)) * Primary_message typo in errors.rs ([#5646](AztecProtocol/aztec-packages#5646)) ([1dfbe7b](AztecProtocol/aztec-packages@1dfbe7b)) * Pull noir ([#5699](AztecProtocol/aztec-packages#5699)) ([bf35464](AztecProtocol/aztec-packages@bf35464)) * REDO dont error in bench summary ([#5695](AztecProtocol/aztec-packages#5695)) ([8c1a7b9](AztecProtocol/aztec-packages@8c1a7b9)) * Running e2e tests as part of build, requires forcing ip4 (not ip6) when connecting to anvil ([#5744](AztecProtocol/aztec-packages#5744)) ([66fc89f](AztecProtocol/aztec-packages@66fc89f)) * Simplify ECCVM prover constructor and add a TODO ([#5681](AztecProtocol/aztec-packages#5681)) ([8c151ea](AztecProtocol/aztec-packages@8c151ea)) * Spot refcount ([#5746](AztecProtocol/aztec-packages#5746)) ([9e18444](AztecProtocol/aztec-packages@9e18444)) * Take a deep copy of circuit inputs for proving ([#5777](AztecProtocol/aztec-packages#5777)) ([785591e](AztecProtocol/aztec-packages@785591e)) * Temporarily disable the bench tests ([#5755](AztecProtocol/aztec-packages#5755)) ([1d52ac5](AztecProtocol/aztec-packages@1d52ac5)) * Update commit for noir-gates-diff (noir-lang/noir#4773) ([1eb288e](AztecProtocol/aztec-packages@1eb288e)) * Use entrypoint instead of pay_init_fee ([#5623](AztecProtocol/aztec-packages#5623)) ([62ac765](AztecProtocol/aztec-packages@62ac765)) * Watch less files. ([#5651](AztecProtocol/aztec-packages#5651)) ([57a1d69](AztecProtocol/aztec-packages@57a1d69)) ### Miscellaneous * Add missing aztec-address tests ([#5674](AztecProtocol/aztec-packages#5674)) ([58aefba](AztecProtocol/aztec-packages@58aefba)) * **avm:** Add a boolean to toggle proving in AVM unit tests ([#5667](AztecProtocol/aztec-packages#5667)) ([ec122c9](AztecProtocol/aztec-packages@ec122c9)), closes [#5663](AztecProtocol/aztec-packages#5663) * **avm:** Hashing tests cleanup ([#5733](AztecProtocol/aztec-packages#5733)) ([53d0102](AztecProtocol/aztec-packages@53d0102)) * **avm:** Range checks negative tests ([#5770](AztecProtocol/aztec-packages#5770)) ([2907142](AztecProtocol/aztec-packages@2907142)) * **avm:** Split the negative test on range check for high 16-bit registers ([#5785](AztecProtocol/aztec-packages#5785)) ([8ebbe57](AztecProtocol/aztec-packages@8ebbe57)) * **avm:** Split up AVM test contract as it was growing too large ([#5702](AztecProtocol/aztec-packages#5702)) ([5b8e812](AztecProtocol/aztec-packages@5b8e812)) * **aztec-nr:** Minor public interface changes ([#5776](AztecProtocol/aztec-packages#5776)) ([91b8110](AztecProtocol/aztec-packages@91b8110)) * **ci:** Break e2e-deploy into multiple test suites ([#5704](AztecProtocol/aztec-packages#5704)) ([2522294](AztecProtocol/aztec-packages@2522294)) * **ci:** Earthly in spot with persistent cache ([#5644](AztecProtocol/aztec-packages#5644)) ([a39c2f6](AztecProtocol/aztec-packages@a39c2f6)) * **ci:** Hotfix AMI's, workflow to stop personal spot runners ([#5712](AztecProtocol/aztec-packages#5712)) ([5f18139](AztecProtocol/aztec-packages@5f18139)) * **ci:** Only run ARM on master ([#5705](AztecProtocol/aztec-packages#5705)) ([f77c142](AztecProtocol/aztec-packages@f77c142)) * **ci:** Use 128 cores for x86 and add timeouts ([#5665](AztecProtocol/aztec-packages#5665)) ([0c5dc0a](AztecProtocol/aztec-packages@0c5dc0a)) * Compute_note_hash_and_nullifier - improve error message ([#5671](AztecProtocol/aztec-packages#5671)) ([8942d69](AztecProtocol/aztec-packages@8942d69)) * Create placeholder version of 0.26.0 docs (noir-lang/noir#4782) ([1eb288e](AztecProtocol/aztec-packages@1eb288e)) * **doc:** Fix broken docs links (noir-lang/noir#4606) ([1eb288e](AztecProtocol/aztec-packages@1eb288e)) * **docs:** Fix link in the Data Types page (noir-lang/noir#4527) ([1eb288e](AztecProtocol/aztec-packages@1eb288e)) * Don't strip bb wasm ([#5743](AztecProtocol/aztec-packages#5743)) ([d4cb410](AztecProtocol/aztec-packages@d4cb410)) * Fix master after merge issue related to validate_trace renaming ([#5676](AztecProtocol/aztec-packages#5676)) ([44e0d8a](AztecProtocol/aztec-packages@44e0d8a)) * Fix max-block-number and auth e2e tests ([#5694](AztecProtocol/aztec-packages#5694)) ([f1bf314](AztecProtocol/aztec-packages@f1bf314)) * Op queue ([#5648](AztecProtocol/aztec-packages#5648)) ([822c7e6](AztecProtocol/aztec-packages@822c7e6)) * **public:** Remove getNullifierMembershipWitness ([#5715](AztecProtocol/aztec-packages#5715)) ([3be402c](AztecProtocol/aztec-packages@3be402c)) * Re-enable e2e fees tests ([#5784](AztecProtocol/aztec-packages#5784)) ([102e8b8](AztecProtocol/aztec-packages@102e8b8)) * Release Noir(0.27.0) (noir-lang/noir#4632) ([1eb288e](AztecProtocol/aztec-packages@1eb288e)) * Remove the old Value struct from the oracle docs (noir-lang/noir#4738) ([1eb288e](AztecProtocol/aztec-packages@1eb288e)) * Replace relative paths to noir-protocol-circuits ([fb2b298](AztecProtocol/aztec-packages@fb2b298)) * Replace relative paths to noir-protocol-circuits ([e20920d](AztecProtocol/aztec-packages@e20920d)) * Replace relative paths to noir-protocol-circuits ([6351dc5](AztecProtocol/aztec-packages@6351dc5)) * Replace relative paths to noir-protocol-circuits ([fee13bf](AztecProtocol/aztec-packages@fee13bf)) * Replacing unsafe::zeroed() ([#5685](AztecProtocol/aztec-packages#5685)) ([ea3884e](AztecProtocol/aztec-packages@ea3884e)) * Small logging changes ([#5654](AztecProtocol/aztec-packages#5654)) ([25cc70d](AztecProtocol/aztec-packages@25cc70d)) * Temporarily skip failing e2e fees test ([a3ac5ff](AztecProtocol/aztec-packages@a3ac5ff)) * Testing that nargo fmt is idempotent (noir-lang/noir#4765) ([825c455](AztecProtocol/aztec-packages@825c455)) * TS hash wrappers cleanup ([#5691](AztecProtocol/aztec-packages#5691)) ([7f8b09f](AztecProtocol/aztec-packages@7f8b09f)) * Turn ENABLE_GAS where it is needed ([#5730](AztecProtocol/aztec-packages#5730)) ([30a2edd](AztecProtocol/aztec-packages@30a2edd)) * Update noir gates diff ([#5658](AztecProtocol/aztec-packages#5658)) ([9816c1a](AztecProtocol/aztec-packages@9816c1a)) * We can run 35 of our e2e tests just using jest. ([#5643](AztecProtocol/aztec-packages#5643)) ([4fcaeae](AztecProtocol/aztec-packages@4fcaeae)) ### Documentation * Fix yp typo control-flow.md ([#5638](AztecProtocol/aztec-packages#5638)) ([363d227](AztecProtocol/aztec-packages@363d227)) </details> <details><summary>barretenberg: 0.35.0</summary> ## [0.35.0](AztecProtocol/aztec-packages@barretenberg-v0.34.0...barretenberg-v0.35.0) (2024-04-16) ### ⚠ BREAKING CHANGES * Use fixed size arrays in black box functions where sizes are known ([#5620](AztecProtocol/aztec-packages#5620)) * trap with revert data ([#5732](AztecProtocol/aztec-packages#5732)) * **acir:** BrilligCall opcode ([#5709](AztecProtocol/aztec-packages#5709)) ### Features * **acir:** BrilligCall opcode ([#5709](AztecProtocol/aztec-packages#5709)) ([f06f64c](AztecProtocol/aztec-packages@f06f64c)) * **avm:** CMOV opcode ([#5575](AztecProtocol/aztec-packages#5575)) ([19dbe46](AztecProtocol/aztec-packages@19dbe46)), closes [#5557](AztecProtocol/aztec-packages#5557) * **avm:** Enable contract testing with bb binary ([#5584](AztecProtocol/aztec-packages#5584)) ([d007d79](AztecProtocol/aztec-packages@d007d79)) * **avm:** Enable range check on the ALU registers ([#5696](AztecProtocol/aztec-packages#5696)) ([202fc1b](AztecProtocol/aztec-packages@202fc1b)) * Changing finite field arithmetic in wasm to 29 bits for multiplications ([#5435](AztecProtocol/aztec-packages#5435)) ([b2d9b9d](AztecProtocol/aztec-packages@b2d9b9d)) * **ci:** Turn on new CI as mandatory ([#5761](AztecProtocol/aztec-packages#5761)) ([bebed32](AztecProtocol/aztec-packages@bebed32)) * Export poseidon2_permutation and add to foundation/crypto ([#5706](AztecProtocol/aztec-packages#5706)) ([6b91e27](AztecProtocol/aztec-packages@6b91e27)) * LT/LTE for AVM ([#5559](AztecProtocol/aztec-packages#5559)) ([350abeb](AztecProtocol/aztec-packages@350abeb)) * Trap with revert data ([#5732](AztecProtocol/aztec-packages#5732)) ([f849575](AztecProtocol/aztec-packages@f849575)) * Use fixed size arrays in black box functions where sizes are known ([#5620](AztecProtocol/aztec-packages#5620)) ([f50b180](AztecProtocol/aztec-packages@f50b180)) ### Bug Fixes * "feat: Changing finite field arithmetic in wasm to 29 bits for multiplications" ([#5779](AztecProtocol/aztec-packages#5779)) ([bcfee97](AztecProtocol/aztec-packages@bcfee97)) * Avoid get row in databus ([#5742](AztecProtocol/aztec-packages#5742)) ([d67b6c8](AztecProtocol/aztec-packages@d67b6c8)) * **ci:** Bigger cache disk, cache+prune docker images, disable ClientIvcTests.Full ([#5729](AztecProtocol/aztec-packages#5729)) ([5dcbd75](AztecProtocol/aztec-packages@5dcbd75)) * Disable flakey vanilla recursion test ([#5672](AztecProtocol/aztec-packages#5672)) ([f84f7b6](AztecProtocol/aztec-packages@f84f7b6)) * Less earthly cache ([#5690](AztecProtocol/aztec-packages#5690)) ([8190dc7](AztecProtocol/aztec-packages@8190dc7)) * Make earthly more parallel ([#5747](AztecProtocol/aztec-packages#5747)) ([9734455](AztecProtocol/aztec-packages@9734455)) * Simplify ECCVM prover constructor and add a TODO ([#5681](AztecProtocol/aztec-packages#5681)) ([8c151ea](AztecProtocol/aztec-packages@8c151ea)) ### Miscellaneous * **avm:** Add a boolean to toggle proving in AVM unit tests ([#5667](AztecProtocol/aztec-packages#5667)) ([ec122c9](AztecProtocol/aztec-packages@ec122c9)), closes [#5663](AztecProtocol/aztec-packages#5663) * **avm:** Range checks negative tests ([#5770](AztecProtocol/aztec-packages#5770)) ([2907142](AztecProtocol/aztec-packages@2907142)) * **avm:** Split the negative test on range check for high 16-bit registers ([#5785](AztecProtocol/aztec-packages#5785)) ([8ebbe57](AztecProtocol/aztec-packages@8ebbe57)) * **ci:** Use 128 cores for x86 and add timeouts ([#5665](AztecProtocol/aztec-packages#5665)) ([0c5dc0a](AztecProtocol/aztec-packages@0c5dc0a)) * Don't strip bb wasm ([#5743](AztecProtocol/aztec-packages#5743)) ([d4cb410](AztecProtocol/aztec-packages@d4cb410)) * Fix master after merge issue related to validate_trace renaming ([#5676](AztecProtocol/aztec-packages#5676)) ([44e0d8a](AztecProtocol/aztec-packages@44e0d8a)) * Op queue ([#5648](AztecProtocol/aztec-packages#5648)) ([822c7e6](AztecProtocol/aztec-packages@822c7e6)) </details> --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please).
Now when we compile for wasm we use 9 29-bit limbs to perform all multiplications. In wasmtime this results in -10% for multiplication benchmark and -35% in squaring. This makes a -7% impact on wasm client_ivc.
Other changes: