Revert "sched/wqueue: some minor improve to reduce sched_lock range"

[hujun260]

Summary

This reverts commit

reason:
may cause dead lock

Impact

wqueue

Testing

sim

[nuttxpr]

[Experimental Bot, please feedback here]

No. This PR does not meet the NuttX requirements. It is missing critical information.

Here's why:

• Insufficient Summary: Simply stating "reverts commit ..." and providing a reason of "may cause deadlock" is inadequate. The summary needs to explain what the original commit changed, what the deadlock was, how this revert fixes it, and which functional area of the code is affected (e.g., work queues). It needs to be understandable without having to look up the original commit.
• Vague Impact: "wqueue" is not sufficient. It needs to specify the impact on the user. Will applications using work queues behave differently? Will any APIs change? It also needs to specify the impact on the build, hardware, documentation, security, and compatibility, even if the answer is "NO" for each.
• Missing Testing Details: "sim" is inadequate. What simulator? What test case demonstrated the deadlock before the revert and the correct behavior after the revert? Actual log output demonstrating the issue and the fix is required, not just screenshots. Build host information is also missing.

The PR author needs to provide significantly more detail for this to be considered.

[anchao]

let me fix this issue

[hujun260]

let me fix this issue

ok

[anchao]

@hujun260 spin_lock already has the function of sched_lock. could the sched_lock/unlock here be deleted directly?

[xiaoxiang781216]

let's remove it