Hi, OSQP teams !

The constrained QP problem that was originally solved normally under v0.6.2 did not converge with v1.0.0 and would exit due to exceeding the maximum number of iterations. I noticed that the main difference in v1.0.0 is that OSQPCscMatrix replaces the original csc and osqp_setup() function parameter passing. I ensured that the problem input was consistent and all parameters were at their default values, but v1.0.0 still did not converge. The following is the verbose log output:

v0.6.2

       OSQP v0.6.2  -  Operator Splitting QP Solver
          (c) Bartolomeo Stellato,  Goran Banjac
    University of Oxford  -  Stanford University 2021

problem: variables n = 84, constraints m = 152
nnz(P) + nnz(A) = 1036
settings: linear system solver = qdldl,
eps_abs = 1.0e-03, eps_rel = 1.0e-03,
eps_prim_inf = 1.0e-04, eps_dual_inf = 1.0e-04,
rho = 1.00e-01 (adaptive),
sigma = 1.00e-06, alpha = 1.60, max_iter = 4000
check_termination: on (interval 25),
scaling: on, scaled_termination: off
warm start: on, polish: off, time_limit: off

iter objective pri res dua res rho time
1 -1.2880e+02 4.80e-16 1.88e+01 1.00e-01 1.55e-04s
50 -1.7380e+02 2.68e-15 3.40e-06 1.00e-06 3.28e-04s

status: solved
number of iterations: 50
optimal objective: -173.8003
run time: 3.47e-04s
optimal rho estimate: 1.00e-06

v1.0.0

OSQP v1.0.0 - Operator Splitting QP Solver
(c) The OSQP Developer Team
problem: variables n = 84, constraints m = 152
nnz(P) + nnz(A) = 1036
settings: algebra = Built-in,
OSQPInt = 8 bytes, OSQPFloat = 8 bytes,
linear system solver = QDLDL v0.1.8,
eps_abs = 1.0e-03, eps_rel = 1.0e-03,
eps_prim_inf = 1.0e-04, eps_dual_inf = 1.0e-04,
rho = 1.00e-01 (adaptive: 50 iterations),
sigma = 1.00e-06, alpha = 1.60, max_iter = 4000
check_termination: on (interval 25, duality gap: on),
time_limit: 1.00e+10 sec,
scaling: on (10 iterations), scaled_termination: off
warm starting: on, polishing: off,
iter objective prim res dual res gap rel kkt rho time
1 -6.9344e+01 1.42e+01 7.05e+02 -nan -nan 1.00e-01 8.70e-04s
50 7.6205e+02 4.32e-03 2.54e+00 -nan -nan 1.16e-02* 1.99e-03s
200 7.6139e+02 1.02e-14 2.22e-14 -nan -nan 1.16e-02 5.51e-03s
400 7.6139e+02 1.02e-14 4.18e-14 -nan -nan 1.16e-02 9.71e-03s
600 7.6139e+02 9.05e-15 6.42e-14 -nan -nan 1.16e-02 1.31e-02s
800 7.6139e+02 1.09e-14 3.17e-14 -nan -nan 1.16e-02 1.55e-02s
1000 7.6139e+02 7.69e-15 3.17e-14 -nan -nan 1.16e-02 1.80e-02s
1200 7.6139e+02 7.24e-15 3.17e-14 -nan -nan 1.16e-02 1.99e-02s
1400 7.6139e+02 9.51e-15 3.15e-14 -nan -nan 1.16e-02 2.14e-02s
1600 7.6139e+02 7.92e-15 4.18e-14 -nan -nan 1.16e-02 2.33e-02s
1800 7.6139e+02 1.09e-14 6.42e-14 -nan -nan 1.16e-02 2.50e-02s
2000 7.6139e+02 6.45e-15 3.21e-14 -nan -nan 1.16e-02 2.64e-02s
2200 7.6139e+02 4.41e-15 4.28e-14 -nan -nan 1.16e-02 2.77e-02s
2400 7.6139e+02 6.45e-15 4.22e-14 -nan -nan 1.16e-02 2.90e-02s
2600 7.6139e+02 7.24e-15 4.23e-14 -nan -nan 1.16e-02 3.02e-02s
2800 7.6139e+02 8.60e-15 3.33e-14 -nan -nan 1.16e-02 3.12e-02s
3000 7.6139e+02 1.02e-14 3.15e-14 -nan -nan 1.16e-02 3.22e-02s
3200 7.6139e+02 7.92e-15 6.66e-14 -nan -nan 1.16e-02 3.33e-02s
3400 7.6139e+02 8.60e-15 3.33e-14 -nan -nan 1.16e-02 3.44e-02s
3600 7.6139e+02 8.60e-15 6.66e-14 -nan -nan 1.16e-02 3.54e-02s
3800 7.6139e+02 4.30e-15 8.88e-14 -nan -nan 1.16e-02 3.63e-02s
4000 7.6139e+02 9.62e-15 8.47e-14 -nan -nan 1.16e-02 3.72e-02s

status: maximum iterations reached
number of iterations: 4000
run time: 3.72e-02s
optimal rho estimate: 1.13e-02

Any feedback would be helpful, thanks for your work. By the way, I used the C interfaces.