One simple question about zk_trie_impl

[hsyodyssey]

Hi, may I ask what the purpose of this line in TryUpdate function is?

The value of dbKeyRootNode is not changed in TryUpdate process . So this function call will always insert the KV with the same Key into the DB, right?

Please correct me if I have a misunderstanding.

[shu-unifra]

The value of dbKeyRootNode is (DBEntryTypeRoot+mt.rootKey) and mt.rootKey is changed in TryUpdate, so......

[hsyodyssey]

The value of dbKeyRootNode is (DBEntryTypeRoot+mt.rootKey) and mt.rootKey is changed in TryUpdate, so......

Yeah, but the Key of dbKeyRootNode is constant (dbKeyRootNode = []byte("currentroot")), right? The function directly inserts KVs with the same Key (sure different Value) to underlying database.

My question is, if more than one zkTries exist at the same time, will there be conflict or overwriting in their Root updates?

[shu-unifra]

The value of dbKeyRootNode is (DBEntryTypeRoot+mt.rootKey) and mt.rootKey is changed in TryUpdate, so......

Yeah, but the Key of dbKeyRootNode is constant (dbKeyRootNode = []byte("currentroot")), right? The function directly insert KVs with the same Key (sure different Value) to underlying database.

My question is, if more than one zkTries exist at the same time, will there be conflict or overwriting in their Root updates?

OK, I got your question, but I don't know why~

[hsyodyssey]

Haha, there seems to be no conflict problem encountered now. So I'm curious if there are other mechanisms to avoid read conflicts.

[hsyodyssey]

Having understood that, I think I know why there is no conflict at the current codebase.

[shu-unifra]

Having understood that, I think I know why there is no conflict at the current codebase.

Can you tell me about it please?

[hsyodyssey]

Having understood that, I think I know why there is no conflict at the current codebase.

Can you tell me about it please?

Basically, I think this function call is redundant since it just inserts a KV with a constant key. There is no other function to query this KV and the Root value of each contract's Storage Trie does not point to this KV. Therefore, this function call is just writing continuously and does not cause any conflicts. Maybe could optimize this area in the future 😊.