make progress on stack depth issues when running coverage

[0age]

The goal here is to run forge coverage which compiles the code without the IR optimization step (note that forge coverage --ir-minimum does not cover the inline assembly so isn't going to work).

There are a few places where the code gets uglier for sure, but the actual functionality and efficiency should be intact with the exception of a few gas here and there for an extra calldataload / pop / etc. I think the value of having coverage in place is worth it (with the caveat that once we are comfortable with the coverage in place we can always change things back).

however I'm now hitting a blocker on this dreadful error:

Error: Compiler error (/solidity/libyul/backends/evm/AsmCodeGen.cpp:68):Stack too deep. Try compiling with --via-ir (cli) or the equivalent viaIR: true (standard JSON) while enabling the optimizer. Otherwise, try removing local variables. When compiling inline assembly: Variable headStart is 1 slot(s) too deep inside the stack. Stack too deep. Try compiling with --via-ir (cli) or the equivalent viaIR: true (standard JSON) while enabling the optimizer. Otherwise, try removing local variables.

best I can gather is that this means there's some function that has too many arguments which means the function can't even be entered or called at all; if anyone has thoughts on how we might track it down and make progress here it'd be much appreciated!

[0age]

ok so this error is originating from both of the permit2 depositAndRegister functions

even with no named variables and no function body, the error still gets tripped

we have a number of options for these:

• (obviously) change the function interface a bit, e.g. replace some distinct args with a struct or pack a few arguments
• work out some minimal set of steps we can take to address this at the compiler level while still keeping coverage (everything I've tried so far has come up short here)
• mine a function name that takes fewer arguments and calldatacopy the others where relevant; relatedly, would need to drop these from the interface (this feels really hacky to me)

will plan to modify the interface to get this working for now, which seems fine since we already changed it for the lockTag adjustment; open to other suggestions though

[codecov-commenter]

Welcome to Codecov 🎉

Once you merge this PR into your default branch, you're all set! Codecov will compare coverage reports and display results in all future pull requests.

Thanks for integrating Codecov - We've got you covered ☂️

[reednaa]

lgtm

[reednaa]

Note: We may be looping over idsAndAmounts twice in this function. Once in _enforceConsistentAllocators and once in .hasValidSponsor if the emissary is checked. Because of stack issues, I don't believe we want to fix.