fix concurrent issue of local account set in legacypool

[andyzhang2023]

Description

There is an concurrent map read & write issue on legacypool.locals when --txpool.nolocals=false. It causes panic.
Here's how it happend:

1. the legacypool.locals accounts map is written here, with flag --txpool.nolocals=false.
2. the legacypool.locals accounts map is then accessed by the Pending() method concurrently here, with a none-nil minTip.

Rationale

2025-03-13 09:11:46.019	fatal error: concurrent map read and map write
2025-03-13 09:11:46.022	
2025-03-13 09:11:46.022	goroutine 8573 [running]:
2025-03-13 09:11:46.022	github.com/ethereum/go-ethereum/core/txpool/legacypool.(*accountSet).contains(...)
2025-03-13 09:11:46.022		github.com/ethereum/go-ethereum/core/txpool/legacypool/legacypool.go:2080
2025-03-13 09:11:46.022	github.com/ethereum/go-ethereum/core/txpool/legacypool.(*LegacyPool).Pending(0xc000362780, {0xc00992afc0, 0xc02bfaab80, 0xc02bfaaba0, 0x1, 0x0})
2025-03-13 09:11:46.022		github.com/ethereum/go-ethereum/core/txpool/legacypool/legacypool.go:708 +0x705
2025-03-13 09:11:46.022	github.com/ethereum/go-ethereum/core/txpool.(*TxPool).Pending(0xc00736b9e0, {0xc00992afc0, 0xc02bfaab80, 0xc02bfaaba0, 0x1, 0x0})
2025-03-13 09:11:46.022		github.com/ethereum/go-ethereum/core/txpool/txpool.go:407 +0xcd
2025-03-13 09:11:46.022	github.com/ethereum/go-ethereum/miner.(*worker).fillTransactions(0xc00c887440, 0xc017ea7098, 0xc012a248f0)
2025-03-13 09:11:46.022		github.com/ethereum/go-ethereum/miner/worker.go:1293 +0x22c
2025-03-13 09:11:46.022	github.com/ethereum/go-ethereum/miner.(*worker).generateWork(0xc00c887440, 0xc0168ff080)
2025-03-13 09:11:46.022		github.com/ethereum/go-ethereum/miner/worker.go:1452 +0xb52
2025-03-13 09:11:46.022	github.com/ethereum/go-ethereum/miner.(*worker).opLoop(0xc00c887440)
2025-03-13 09:11:46.022		github.com/ethereum/go-ethereum/miner/worker.go:678 +0x7f
2025-03-13 09:11:46.022	created by github.com/ethereum/go-ethereum/miner.newWorker in goroutine 1
2025-03-13 09:11:46.022		github.com/ethereum/go-ethereum/miner/worker.go:343 +0x9f6
2025-03-13 09:11:46.022	

Example

How to replay it:

1. start a miner node with op-geth flat: --txpool.nolocals=false
2. keep sending transactions to the miner via RPC API, so the transactions will be added into the pool as local ones.
   The node should panic after a few minutes.