Multi-core results of dav_subspace are incorrect

[YuLiu98]

Describe the bug

As the title said, multi-core results of dav_subspace are incorrect, which are different from those of cg and dav as well as the single-core results.

I found this bug in the integrated test 006_PW_UPF201_Eu, in which ks_solver can be set to dav_subspace, and the force and stress calculation should be open.

Expected behavior

No response

To Reproduce

No response

Environment

No response

Additional Context

No response

Task list for Issue attackers (only for developers)

• [ ] Verify the issue is not a duplicate.
• [ ] Describe the bug.
• [ ] Steps to reproduce.
• [ ] Expected behavior.
• [ ] Error message.
• [ ] Environment details.
• [ ] Additional context.
• [ ] Assign a priority level (low, medium, high, urgent).
• [ ] Assign the issue to a team member.
• [ ] Label the issue with relevant tags.
• [ ] Identify possible related issues.
• [ ] Create a unit test or automated test to reproduce the bug (if applicable).
• [ ] Fix the bug.
• [ ] Test the fix.
• [ ] Update documentation (if necessary).
• [ ] Close the issue and inform the reporter (if applicable).

[dyzheng]

When I compare the results of the single-process and four-processes, the ecutwfc of 20 will be different, while 100 will remain the same.