Bug: restart from charge density cannot work as expected with symmetry=0

[kirk0830]

Describe the bug

if restart from a converged charge density, the step of SCF should not be more than 1, otherwise there must be something wrong.
What I find is:

symmetry (out)	symmetry (restart)	scf steps
1	1	1
1	0	>1
0	1	1
0	0	>1

I output the charge density in real space before the further construction of operators such as hartree (which needs first call real2recip function to get "porter"), vxc, etc. I find the combination of cases restart from (rho(r) or rho(G)), symmetry set as (0 or 1), in total four cases, the charge densities before operator construction are totally identical. Therefore, it is a not-well designed behavior of symmetry, which now impacts the efficiency of SCF convergence.

I post my test case in the following.

INPUT_PARAMETERS
calculation scf
basis_type pw
pseudo_dir .
nspin 1
init_chg file
ecutwfc 60
symmetry 0
nbands 4
smearing_method fixed
ks_solver cg
out_chg 1 10

ATOMIC_SPECIES
Si 28.0855 Si_ONCV_PBE-1.0.upf

LATTICE_CONSTANT
1.8897259886

LATTICE_VECTORS
2.7023226613 0.0000000000 0.0000000000
1.3511605841 2.3402795831 0.0000000000
1.3511603830 0.7800934082 2.2064371988

ATOMIC_POSITIONS
Cartesian
Si
0.00
1
0 0 0 1 1 1

K_POINTS
0
Gamma
5 5 5 0 0 0

I will check in details.

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).