rpc-call-max-steps flag isn't limiting RPC call steps
wojciechos opened this issue · comments
wojciechos commented
Steps to Reproduce:
- Set
rpc-call-max-steps
to 0. - Run calls that should exceed these limits. E.g(mainnet):
"jsonrpc": "2.0",
"method": "starknet_call",
"params": {
"block_id": {
"block_hash": "0x9f59dd27f387b869844285307bd1de75d970f0a69903485e164e3d1609dd42"
},
"request": {
"contract_address": "0x0772FA7347f1D6c16D7d03dEba3a0D2aDDde10c5D9259c784d0d47Fc6D231aBb",
"entry_point_selector": "0x039dd5dee115f62929cb2e75e8e4c5e964121efc81aabbd2eaf2002005ad536e",
"calldata": []
}
},
"id": 1
}
Actual Behavior: calls ignore the rpc-call-max-steps
setting and proceed without errors.
Expected Behavior: Setting rpc-call-max-steps
to low values should block calls exceeding the cairo step count. Worked fine for <4M steps before(as 4m is max limit set in blockifier), now setting it to 0 or 1 doesn’t impact multi calls - they run as if the limit's not there.
Notes: Could be related to the blockifier update?