transaction_status_service: remove get_account_locks_unchecked

[apfitzge]

Problem

• validate_account_locks function got pulled into account_locks from sdk
• following that change, get_account_locks and get_account_locks_unchecked will be deprecated
• transaction_status_service uses get_account_locks_unchecked

Summary of Changes

• Replace use of get_account_locks_unchecked with more direct access
• Refactor blockstore::write_transaction_status to take a single iterator (simplifies code)

Fixes #

[apfitzge]

refactor of this function makes its' internal code slightly simpler

[apfitzge]

removing this also removes a pair of allocations

[steviez]

In general, the code looks good to me and save some temporary allocations as you mentioned.

@CriesofCarrots - In the past, I believe we have been more cautious with adjusting pub function in Blockstore than some other crates, as Blockstore is a little internal facing than some of the other stuff.

Do you know of anything that might be broken specifically if we adjust this, or just trying to be friendly to anyone downstream who might be using Blockstore ?

[steviez]

nit: On line 167, we call the same function. The underlying function just returns a ref so not a big deal, but if message() got refactored to do something more meaningful, would want to avoid duplicate work:

agave/sdk/src/transaction/sanitized.rs

Lines 164 to 166 in 094a634

	pub fn message(&self) -> &SanitizedMessage {
	&self.message
	}

[apfitzge]

eventually message() will go away entirely 😉 . or at least be abstracted by a trait (SVM{Transaction|Message}?)

[CriesofCarrots]

In the past, I believe we have been more cautious with adjusting pub function in Blockstore than some other crates, as Blockstore is a little internal facing than some of the other stuff.

Do you know of anything that might be broken specifically if we adjust this, or just trying to be friendly to anyone downstream who might be using Blockstore ?

@steviez , I think your first sentence might be missing a word? (Is it more or less internal-facing?)
But I don't know of anything specifically that would be broken by this change. The methods that write to columns are used pretty exclusively by blockstore internals and/or agave validator supplementary metadata services.

[steviez]

@steviez , I think your first sentence might be missing a word? (Is it more or less internal-facing?)

Dang it yep, I meant as Blockstore is a little more external internal facing than some of the other stuff.

But I don't know of anything specifically that would be broken by this change. The methods that write to columns are used pretty exclusively by blockstore internals and/or agave validator supplementary metadata services.

👍