fix:Unblock RPC/SDIO threads for graceful shutdown (EHM-194)

[Kevincoooool]

Unblock and allow orderly suspension of RPC and SDIO worker threads during deinit to avoid heap corruption and dangling waits. rpc_core: include FreeRTOS headers, post RPC/serial semaphores, call serial_post_read_semaphore and delay before cancelling threads so they can observe INACTIVE state. serial_drv: add serial_post_read_semaphore to wake the serial read semaphore. sdio_drv: add sdio_exit_requested flag, post SDIO-related semaphores on bus deinit, wait before cancelling threads, make tasks check the exit flag and call vTaskSuspend(NULL) instead of being abruptly deleted, use a timed wait for slave interrupt to allow periodic exit checks, reorder mempool/bus deinit and buffer free to avoid DMA writing to freed memory, and reset counters/flags on init/deinit. These changes make shutdowns safer and prevent buffer/heap races.

Description

Related

Testing

---

Checklist

Before submitting a Pull Request, please ensure the following:

• 🚨 This PR does not introduce breaking changes.
• All CI checks (GH Actions) pass.
• Documentation is updated as needed.
• Tests are updated or added as necessary.
• Code is well-commented, especially in complex areas.
• Git history is clean — commits are squashed to the minimum necessary.

[ThunderDai]

host_shuts_down_slave_to_power_save.zip
hosted_heap_issues.txt
hi, this is test project and log, please check, mainly enable HEAP_POISONING_COMPREHENSIVE

[mantriyogesh]

Thanks a lot @Kevincoooool , for your contribution and time.
We appreciate it much.

We will have a review and merge the changes.

@SohKamYung-Espressif Can you please review?

Some generic comments:

1. Host side core code to be written agnostic of OS. Any port level stuff can go inside the port layer.
2. Is the suspend mandatory, before deinit? Need to check how we can minimise the change.

[mantriyogesh]

@ThunderDai ,

#173 (comment)

Yes these would be really useful, for co-relating the case. thank you.