feat: add data and into_data

[mattsse]

adds back previous Memory data functions:

https://github.com/bluealloy/revm/pull/516/files

I don't fully understand if this is correct. idk what a context is or how the shared memory is laid out.

In reth inspectors we need to get the currently active memory, and I'm not sure this is still accurate:

https://github.com/paradigmxyz/reth/blob/9dd270e4e8095a4cab1d2118f8f965eec68f4db7/crates/revm/revm-inspectors/src/tracing/types.rs#L536-L537

[rakita]

This is not okay. .context_memory() should be used to fetch current call memory frame

[mattsse]

but this is exactly what this did?

and context_memory is private

[rakita]

but this is exactly what this did?

and context_memory is private

I didn't see that context_memory is not public, it should be.

into_data would still copy the context_memory with split_off, and using into_data for consuming shared memory for its last context and disregarding previous ones could introduce footguns.

[mattsse]

into_data would still copy the context_memory with split_off, and using into_data for consuming shared memory for its last context and disregarding previous ones could introduce footguns.

I see, yeah, makes sense, we need to clone that during tracing anyway,
will open another pr that makes context_memory pub then

[rakita]

into_data would still copy the context_memory with split_off, and using into_data for consuming shared memory for its last context and disregarding previous ones could introduce footguns.

I see, yeah, makes sense, we need to clone that during tracing anyway, will open another pr that makes context_memory pub then

Sorry if i closed this abruptly, you can make the change here or a new PR sounds okay.

[mattsse]

closing in favor of #831