arm64: syscall SYS_switch_context and SYS_restore_context use 0 para

[hujun260]

Summary

arm64: syscall SYS_switch_context and SYS_restore_context use 0 para
reason:
simplify context switch
sys_call0(SYS_switch_context)
sys_call0(SYS_restore_context)

size nuttx
before
text data bss dec hex filename
504644 54968 81085 640697 9c6b9 nuttx

after
text data bss dec hex filename
504360 54968 81085 640413 9c59d nuttx

size reduce -284

Impact

arm64

Testing

ostest
qemu-armv8a:nsh_smp

[nuttxpr]

[Experimental Bot, please feedback here]

Yes, this PR appears to meet the basic NuttX requirements, although some sections could be improved. Here's a breakdown:

Strengths:

• Clear Summary of "What" and "Why": The summary explains the change (using zero-parameter syscalls for context switching) and the rationale (simplification).
• Quantifiable Impact: The impact section shows a reduction in binary size, which is a concrete benefit.
• Testing Evidence: Testing logs are included (although the content is just "ostest" and "qemu-armv8a:nsh_smp," which needs more detail). The mention of the target architecture (arm64/qemu-armv8a) is good.

Weaknesses & Areas for Improvement:

• Missing Issue References: If this relates to a specific issue in the NuttX or NuttX Apps repositories, those should be linked.
• Incomplete "How" in Summary: The summary mentions what syscalls are changed but lacks detail on how the change is implemented. More details on the mechanism of switching to zero-parameter syscalls would be beneficial.
• Impact Section Needs More "YES/NO" Clarity: The impact section is mostly just "arm64" and size reduction. Explicitly stating "YES" or "NO" for each impact category (user, build, hardware, documentation, security, compatibility) with explanations would make the impact clearer. For example:
  • Impact on user: NO
  • Impact on build: NO
  • Impact on hardware: YES (affects arm64 architecture)
  • ...and so on.
• Insufficient Testing Logs: "ostest" and "qemu-armv8a:nsh_smp" are not detailed enough. The logs should show the relevant parts of the test output demonstrating correct functionality both before and after the change. Ideally, there should be tests specifically targeting the context switching functionality.
• Missing Build Host Information: The testing section should specify the build host OS, CPU, and compiler used.
• "Anything else to consider?" is unanswered: This section is meant to capture any unusual edge cases or potential issues that reviewers should be aware of.

Recommendation: Revise the PR to address the weaknesses listed above. Providing more detail and explicit answers to all sections of the template will greatly improve clarity and make it easier for reviewers to assess the change.