feat: prune if needed

[just-mitch]

Fix #8572
Fix #8608

[just-mitch]

Ordering of functions changed to satisfy linter.

[LHerskind]

I don't think this interface will work, unless you also do a lot of the same accounting in here as you do in the rollup. We should do them in either one or the other.

Specifically, look at the stakeBond and unstakeBond the stake gets a lot of information, some useful some not, but it knows who is bonding etc. But at the unstake it won't be told who is it that is unstaking which makes it weird here.

In the likely case we have, where we are building the escrow specifically for the rollup, you can have the rollup provide enough information for the escrow to know what to do without doubling the accounting.

[just-mitch]

I agree that one escrow per rollup is simpler- that is a deviation from what I was thinking.

[LHerskind]

The unstakeBond is still not identifying who, so it still needs to store a bunch of storage on its own. I suggest making the "mock" without validation but still doing the transfers, such that you can more easily get an idea around the information that needs to be passed along for the minimum.

Also, it is not enough to give a signature, to verify it we need to know the message etc. Either have that logic in the rollup, and then the escrow becomes just bond(address _bonder, uint256 _amount) onlyRollup or have it deal with the full thing.

Feedback has been addressed.

[LHerskind]

The unstakeBond is still not identifying who, so it still needs to store a bunch of storage on its own. I suggest making the "mock" without validation but still doing the transfers, such that you can more easily get an idea around the information that needs to be passed along for the minimum.

Also, it is not enough to give a signature, to verify it we need to know the message etc. Either have that logic in the rollup, and then the escrow becomes just bond(address _bonder, uint256 _amount) onlyRollup or have it deal with the full thing.

[github-actions]

Changes to circuit sizes

Generated at commit: d9416d467ad36390cd17dbfe52caf8d4b82deb60, compared to commit: 9a1b5b5fdd3194f4e7833aacbca4f48aadafbd74

🧾 Summary (100% most significant diffs)

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
rollup_block_merge	-563 ✅	-3.00%	-1,580 ✅	-0.06%
rollup_root	-563 ✅	-3.00%	-1,580 ✅	-0.06%
rollup_block_root	-1,116 ✅	-58.16%	-3,144 ✅	-0.08%
parity_root	-1,669 ✅	-81.69%	-4,707 ✅	-0.09%
rollup_merge	-1,117 ✅	-77.79%	-3,145 ✅	-0.11%
rollup_base	-74,503 ✅	-17.65%	-90,296 ✅	-2.42%
parity_base	-1,669 ✅	-82.71%	-2,487 ✅	-4.65%

---

Full diff report 👇

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
rollup_block_merge	18,204 (-563)	-3.00%	2,785,943 (-1,580)	-0.06%
rollup_root	18,188 (-563)	-3.00%	2,785,911 (-1,580)	-0.06%
rollup_block_root	803 (-1,116)	-58.16%	4,117,849 (-3,144)	-0.08%
parity_root	374 (-1,669)	-81.69%	5,481,787 (-4,707)	-0.09%
rollup_merge	319 (-1,117)	-77.79%	2,748,150 (-3,145)	-0.11%
rollup_base	347,688 (-74,503)	-17.65%	3,642,822 (-90,296)	-2.42%
parity_base	349 (-1,669)	-82.71%	50,957 (-2,487)	-4.65%

[LHerskind]

Thanks 👍 The rest can be addressed when there is logic 👍