fix: erroneous nonce updates

[nbaztec]

What 💻

• Fixes erroneous nonce updates
  • We now revert tx nonce update for CALL (on tx caller) while leaving the nonce update for CREATE as we are manually dispatching the CALLs to zkEVM which does not constitute a real transaction.
• Support zksync-specific nonce cheatcodes
• Make nonces u128 as that's their theoretical maximum.

Why ✋

• Nonces were erroneously being skipped or updated that was inconsistent with EVM logic
• getNonce cheatcode was previously only returning tx nonce causing confusion, with no way to obtain deployment nonce.
• Avoid inconsistency of nonce bit sizes. Theoretical maximum 128-bit, server accepts 32-bit.

Evidence 📷

Tests pass

Documentation 📚

Please ensure the following before submitting your PR:

• Check if these changes affect any documented features or workflows.
• Update the book if these changes affect any documented features or workflows.

Notes 📝

• Add test

[elfedy]

Did I get correctly that tx_nonce stays consistent with evm and deploy_nonce stays consistent with zkEVM? If so (and if not, then even more) maybe we should document the nonce expected flow somewhere so it does not get lost in ancient lore.