Add tests for confidential transfer instruction wrappers

[cryptopapi997]

The README says this program supports all token-2022 extensions, however doesn't explicitly test all of them, including confidential transfers. Seeing as confidential transfers often turn out to be a special case, having a few tests for this would be good.

[grod220]

Good suggestion! I have a refactor (#143) of the test helpers coming down the pipeline. After that, I should be able to add a test for confidential transfer rather easily. Stay tuned 👍

[grod220]

@cryptopapi997, thinking about this more, I want to be careful/clear on how the current token-wrap program works with the confidential transfer extension.

Short answer: token-wrap can already wrap and unwrap SPL-Token-2022 mints that carry the extension, but the wrapped-side mint will not automatically copy over the extension. To do so, you'd need to fork the program and customize the CreateMint instruction.

This means the wrapped tokens do not inherit privacy guarantees. The wrapped tokens are transparent and linked to the escrow ATA.

What were the usecases you had in mind?

[cryptopapi997]

Thanks for the quick followup!

Short answer: token-wrap can already wrap and unwrap SPL-Token-2022 mints that carry the extension, but the wrapped-side mint will not automatically copy over the extension. To do so, you'd need to fork the program and customize the CreateMint instruction.
This means the wrapped tokens do not inherit privacy guarantees. The wrapped tokens are transparent and linked to the escrow ATA.

Thanks for the extra info on this! Am I correctly understanding that you're saying unless I explicitly create it that way, creating a wrapper for confidential tokens won't automatically be confidential too and so in wrapping I will leak privacy (since the wrapped token amount will be in plaintext)?

If so this works, the usecases I had in mind was more the other way around; I have a regular SPL token A and I use this program to allow for wrapping it into confidential_A. I would obv lose privacy when wrapping and unwrapping but not more than that e.g. if I wrap 100 A into confidential_A obviously the confidential amount is known at the start, but if I then transfer some confidential_A to someone else the amounts from then on are confidential.

[grod220]

unless I explicitly create it that way, creating a wrapper for confidential tokens won't automatically be confidential too and so in wrapping I will leak privacy (since the wrapped token amount will be in plaintext)?

Yes that is right. The current code is only creating standard, non-confidential mints for the wrapped version.

I have a regular SPL token A and I use this program to allow for wrapping it into confidential_A

The program currently does not support wrapping into a mint with the confidential transfers extension. That's where you'll need to fork and amend the CreateMint instruction to add that functionality. Specifically invoking the confidential transfers initialize fn.

[cryptopapi997]

I see. Sorry one last question - can the idempotent field be used to circumvent that? i.e. I create my wrapped confidential mint in advance? Then I wouldn't need to call the conf extension initialize fn as it would already have been called previously. I'm guessing not, as minting for confidential transfers is done with ConfidentialMintBurn and not the normal minting, but wanted to have asked just in case.

If not, adding this functionality to token-wrap would be quite beneficial as otherwise we'll have n token-wrap programs flying around every time someone wants to confidentially wrap a spl token. Ofc I realize this goes beyond the scope of this issue and since the tests are implemented already this issue can be closed with its merging imo :)

I'm not sure 100% how this program is managed, is the best way toward getting this implemented a SIMD? A new issue? A PR?

[grod220]

can the idempotent field be used to circumvent that? i.e. I create my wrapped confidential mint in advance?

No because a user cannot initialize the wrapped mint PDA owned by the token-wrap program.

Also, there are a number of things that would need to be updated to support wrapping into a confidential transfer mints:

• Initializing the mint with ConfidentialTransferMint::initialize
• Replacing mint_to, burn, transfer_checked (maybe more?) with their confidential versions (confidential_transfer_mint, confidential_transfer_burn, etc)

is the best way toward getting this implemented a SIMD? A new issue? A PR?

Recently a new instruction was added to the program. So audits are concluding next week. After a bit of time for remediations + doc updates, it should be ready to be deployed to mainnet. There are discussions in progress for the entity that will do this.