Add a feature that propose u32 tx_pointer in fuel-tx and fuel-vm

[AurelienFT]

Having u32 tx_pointer is mendatory to make our system scales and support more than u16::MAX tx in a block.

However, changing this in the current network is a hard breaking change and is not needed for now. That's why we decided to add this change to the VM behind a feature flag if we want to use in other system or be able to enable it in our principal one too.

Checklist

• Breaking changes are clearly marked as such in the PR description and changelog
• New behavior is reflected in tests
• If performance characteristic of an instruction change, update gas costs as well or make a follow-up PR for that
• The specification matches the implemented behavior (link update PR if changes are needed)

Before requesting review

• I have reviewed the code myself
• I have created follow-up issues caused by this PR and linked them here

[rymnc]

LGTM, but in the malleable_fields_do_not_affect_* tests if we used u16::MAX and u32::MAX to indicate change in size vs adding some zeroes OR using rng to generate those values, would be nice, but is out of scope for this PR :)

[xgreenx]

It is a big breaking change for the type system. We need to have proper testing and double check that serialization and deserialization will be not affected for mainnet/testnet/devnet blocks. It would be nice try to use extreme values and see when postcard serialization/deserialiation will be different(like from 1 to u16::MAX).

Also maybe Rollup team uses bitcode, if yes, then it also break compressed blocks for them

[AurelienFT]

@xgreenx I didn't want to activate this feature on Ignition is it something we want to do in the near future ? For xxx, it will be a new network and will be in the specs.

[xgreenx]

We can activate it on Ignition, if tests show that old database types are compatible with it.

[AurelienFT]

I have added a test to showcase that we have a difference on the postcard serialization format (it's expected because postcard use varint encoding). I tried our .to_bytes() which doesn't break because it's not using varint and was padded to a 8 bytes value so u32 fits.

I don't really understand how we could activate this on Ignition but if we want to make more tests to activate this for Ignition they should be done on the fuel-core repository