refactor: Update authwit computation by LHerskind · Pull Request #2651 · AztecProtocol/aztec-packages

LHerskind · 2023-10-03T18:27:12Z

Fixes #2448

Checklist:

Remove the checklist to signal you've completed it. Enable auto-merge if the PR is ready to merge.

If the pull request requires a cryptography review (e.g. cryptographic algorithm implementations) I have added the 'crypto' tag.
I have reviewed my diff in github, line by line and removed unexpected formatting changes, testing logs, or commented-out code.
Every change is related to the PR description.
I have linked this pull request to relevant issues (if any exist).

spalladino

Love this. Some nitpicks, which may be pending because this is a draft, but sharing just in case:

computeAuthWitHash should be in aztec.js, it's super useful for devs
assert_current_call_is_valid_for should probably be assert_valid_current_call_for if we want to follow the same naming as we had, or we should rename assert_valid_message_for to assert_message_is_valid_for
compute_message_hash could be used within the assert_* function in that same module

LHerskind · 2023-10-03T18:43:05Z

compute_message_hash could be used within the assert_* function in that same module

Not sure you are right here. compute_message_hash is computing the hash of the args passed and then the other hash, so h(...,h(...)) where the assert_* ones are doing h(...). The intention is that you can use it when you don't already got the args_hash, such as when uniswap one is compute message_hsah for a different call than itself.

rahul-kothari · 2023-10-04T09:35:30Z

Had a high level look - looks very very clean!

Agree that computeAuthWitness should be in aztec.js.
I don't think assert_current_call_is_valid_for is the best name as it doesn't really reveal what exactly it does. The current function name makes me think "oh so it is checking if the current call is valid?" but what does "is valid" even mean? What exactly "is valid" - the function simulation? May I suggest the name assert_auth_witness_is_valid? Something that explicitly says what we are doing here?
One of my other problems with the name is - on the JS side we call it "computeAuthWitness" and expose it to devs. But on noir side, we hide it away and don't use that term. Either we use "authWitness" externally in both places or don't.

rahul-kothari

minor nits on naming but otherwise

rahul-kothari · 2023-10-04T11:51:04Z

+// Compute the message hash to be used by an authentication witness 
+fn compute_authwit_message_hash<N>(caller: AztecAddress, target: AztecAddress, selector: Field, args: [Field; N]) -> Field {
+  let args_hash = hash_args(args);
+  pedersen_with_separator([caller.address, target.address, selector, args_hash], GENERATOR_INDEX__SIGNATURE_PAYLOAD)[0]


previously we had a TODO here reminding us to reconsider if this is the best pederson generator. Is that not the case anymore?

There was a todo in the tracking issue, have created #2676 now, and will ref that.

rahul-kothari · 2023-10-04T11:52:52Z

+
+/**
+ * Compute an authentication witness message hash from a caller and a request
+ * H(caller, target, selector, args_hash)


can you rename target -> target_contract_address

hahaha why? Cos these can be than just contract addresses?

caller and target are both addresses, and requests can only really be targeted to addresses so seemed unnecessary.

rahul-kothari · 2023-10-04T11:54:47Z

+    await ownerWallet.setPublicAuth(swapMessageHash, true).send().wait();
+
+    // 4.2 Call swap_public from user2 on behalf of owner
+    const withdrawReceipt = await action.send().wait();


this is so neat!

rahul-kothari · 2023-10-04T11:54:58Z

-
-    // 4.2 Call swap_public from user2 on behalf of owner
-    const withdrawReceipt = await uniswapL2Contract
+    const action = uniswapL2Contract


rename action to functionCall?

We have something that is a FunctionCall, the return value from the request xD so think that might be more confusing.

You guys discuss the names of local variables in tests..? Wow, that's a thorough review indeed.

spalladino

The one thing I'm not sold on is moving entrypoint and account to the new authwit library. It feels odd having the entire account "base" contract within an authwit library.

I'm assuming you did this because of the dependency on IS_VALID_SELECTOR for implementing the account actions, but that seems to be the only contact point between account and authwit.

I think I'd prefer:

Keep everything mashed up together in aztec (granular dependencies are overrated, especially when you have a compiler that removes everything that's unused)
Keep account and entrypoint in aztec, and consider moving IS_VALID_SELECTOR there as well.
Rename the authwith package to auth

But I won't cry too much if you just leave it as it is now in the PR, so feel free to ignore this and merge.

spalladino · 2023-10-04T13:31:19Z

-            // if someone else is calling on swap on sender's behalf, they need to have authorisation to do so:
-            let selector = compute_selector(
-                "swap_public((Field),(Field),Field,(Field),Field,Field,Field,(Field),Field,Field,(Field),(Field),Field)"
-            );
-            let message_field = compute_message_hash([
-                context.msg_sender(),
-                context.this_address(),
-                selector,
-                sender.address,
-                input_asset_bridge.address,
-                input_amount,
-                output_asset_bridge.address,
-                nonce_for_transfer_approval,
-                uniswap_fee_tier,
-                minimum_output_amount,
-                recipient.address,
-                secret_hash_for_L1_to_l2_message,
-                deadline_for_L1_to_l2_message,
-                canceller_for_L1_to_L2_message.address,
-                caller_on_L1.address,
-                nonce_for_swap_approval,
-            ]);
-            // this also emits a nullifier for the message
-            assert_valid_public_message_for(&mut context,sender.address,message_field);


spalladino · 2023-10-04T13:32:47Z

-
-    // 4.2 Call swap_public from user2 on behalf of owner
-    const withdrawReceipt = await uniswapL2Contract
+    const action = uniswapL2Contract


You guys discuss the names of local variables in tests..? Wow, that's a thorough review indeed.

🤖 I have created a release *beep* *boop* --- <details><summary>aztec-packages: 0.8.3</summary> ## [0.8.3](aztec-packages-v0.8.2...aztec-packages-v0.8.3) (2023-10-04) ### Bug Fixes * Do not depend on npx for check rebuild script ([#2681](#2681)) ([20ffbbc](20ffbbc)) * Remove package json properties whitelist ([#2680](#2680)) ([ef499a0](ef499a0)) ### Miscellaneous * Update authwit computation ([#2651](#2651)) ([fdbe2b2](fdbe2b2)), closes [#2448](#2448) </details> <details><summary>barretenberg.js: 0.8.3</summary> ## [0.8.3](barretenberg.js-v0.8.2...barretenberg.js-v0.8.3) (2023-10-04) ### Miscellaneous * **barretenberg.js:** Synchronize aztec-packages versions </details> <details><summary>barretenberg: 0.8.3</summary> ## [0.8.3](barretenberg-v0.8.2...barretenberg-v0.8.3) (2023-10-04) ### Miscellaneous * **barretenberg:** Synchronize aztec-packages versions </details> --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please). --------- Co-authored-by: Santiago Palladino <santiago@aztecprotocol.com> Co-authored-by: Charlie Lye <karl.lye@gmail.com>

🤖 I have created a release *beep* *boop* --- <details><summary>aztec-packages: 0.8.3</summary> ## [0.8.3](AztecProtocol/aztec-packages@aztec-packages-v0.8.2...aztec-packages-v0.8.3) (2023-10-04) ### Bug Fixes * Do not depend on npx for check rebuild script ([#2681](AztecProtocol/aztec-packages#2681)) ([20ffbbc](AztecProtocol/aztec-packages@20ffbbc)) * Remove package json properties whitelist ([#2680](AztecProtocol/aztec-packages#2680)) ([ef499a0](AztecProtocol/aztec-packages@ef499a0)) ### Miscellaneous * Update authwit computation ([#2651](AztecProtocol/aztec-packages#2651)) ([fdbe2b2](AztecProtocol/aztec-packages@fdbe2b2)), closes [#2448](AztecProtocol/aztec-packages#2448) </details> <details><summary>barretenberg.js: 0.8.3</summary> ## [0.8.3](AztecProtocol/aztec-packages@barretenberg.js-v0.8.2...barretenberg.js-v0.8.3) (2023-10-04) ### Miscellaneous * **barretenberg.js:** Synchronize aztec-packages versions </details> <details><summary>barretenberg: 0.8.3</summary> ## [0.8.3](AztecProtocol/aztec-packages@barretenberg-v0.8.2...barretenberg-v0.8.3) (2023-10-04) ### Miscellaneous * **barretenberg:** Synchronize aztec-packages versions </details> --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please). --------- Co-authored-by: Santiago Palladino <santiago@aztecprotocol.com> Co-authored-by: Charlie Lye <karl.lye@gmail.com>

Fixes #2448

🤖 I have created a release *beep* *boop* --- <details><summary>aztec-packages: 0.8.3</summary> ## [0.8.3](aztec-packages-v0.8.2...aztec-packages-v0.8.3) (2023-10-04) ### Bug Fixes * Do not depend on npx for check rebuild script ([#2681](#2681)) ([20ffbbc](20ffbbc)) * Remove package json properties whitelist ([#2680](#2680)) ([ef499a0](ef499a0)) ### Miscellaneous * Update authwit computation ([#2651](#2651)) ([fdbe2b2](fdbe2b2)), closes [#2448](#2448) </details> <details><summary>barretenberg.js: 0.8.3</summary> ## [0.8.3](barretenberg.js-v0.8.2...barretenberg.js-v0.8.3) (2023-10-04) ### Miscellaneous * **barretenberg.js:** Synchronize aztec-packages versions </details> <details><summary>barretenberg: 0.8.3</summary> ## [0.8.3](barretenberg-v0.8.2...barretenberg-v0.8.3) (2023-10-04) ### Miscellaneous * **barretenberg:** Synchronize aztec-packages versions </details> --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please). --------- Co-authored-by: Santiago Palladino <santiago@aztecprotocol.com> Co-authored-by: Charlie Lye <karl.lye@gmail.com>

spalladino reviewed Oct 3, 2023

View reviewed changes

LHerskind force-pushed the lh/authwit-refactor branch from 82c0385 to d3afb5c Compare October 4, 2023 10:47

LHerskind marked this pull request as ready for review October 4, 2023 11:33

LHerskind requested review from Maddiaa0, rahul-kothari and spalladino October 4, 2023 11:33

rahul-kothari approved these changes Oct 4, 2023

View reviewed changes

LHerskind added 7 commits October 4, 2023 14:31

refactor: Update authwit computation

fc0522a

chore: move computeAuthWitHash to aztec.js utils

636f2f4

chore: linter

7b64e57

chore: align function auth with function naming

4164e93

chore: naming improvement 🤞

1bbf745

chore: add todo and ref in auth.nr

b7dfd2b

chore: make Rahul happy

7f66228

LHerskind force-pushed the lh/authwit-refactor branch from fc1849c to 7f66228 Compare October 4, 2023 13:31

LHerskind enabled auto-merge (squash) October 4, 2023 13:31

spalladino reviewed Oct 4, 2023

View reviewed changes

LHerskind merged commit fdbe2b2 into master Oct 4, 2023

LHerskind deleted the lh/authwit-refactor branch October 4, 2023 13:48

AztecBot mentioned this pull request Oct 4, 2023

chore(master): Release 0.8.3 #2679

Merged

Maddiaa0 pushed a commit that referenced this pull request Oct 6, 2023

refactor: Update authwit computation (#2651)

65b7796

Fixes #2448

Conversation

LHerskind commented Oct 3, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Checklist:

Uh oh!

spalladino left a comment

Choose a reason for hiding this comment

Uh oh!

LHerskind commented Oct 3, 2023

Uh oh!

rahul-kothari commented Oct 4, 2023

Uh oh!

rahul-kothari left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

spalladino left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

LHerskind commented Oct 3, 2023 •

edited

Loading