Assertion failed with minimal quadratic model

The model is in proto format. But it is small and readable, so it should be easy to rewrite. Note I use SCIP 8.0.4 with Soplex 6.0.4 without anything else (papilo, gmp, ...).

variable {
  name: "x"
  lower_bound: 0
  upper_bound: 10
  objective_coefficient: -6.28
  is_integer: true
}
variable {
  name: "y"
  lower_bound: -10
  upper_bound: 10
  objective_coefficient: 0.2
  is_integer: true
}
objective_offset: 9.8596
quadratic_objective {
  qvar1_index: [ 0 ]
  qvar2_index: [ 0 ]
  coefficient: [ 1 ]
}

Here is the CIP file:

STATISTICS
  Problem name     : 
  Variables        : 3 (0 binary, 2 integer, 0 implicit integer, 1 continuous)
  Constraints      : 0 initial, 1 maximal
OBJECTIVE
  Sense            : minimize
  Offset           : +9.8596
VARIABLES
  [integer] <x0>: obj=-6.28, original bounds=[0,10]
  [integer] <x1>: obj=0.2, original bounds=[-10,10]
  [continuous] <x2>: obj=1, original bounds=[-inf,+inf]
CONSTRAINTS
  [nonlinear] <c0>: (<x0>)^2-<x2> == 0;
END

Running this in debug mode outputs:

feasible solution found by trivial heuristic after 0.0 seconds, objective value 9.859600e+00
presolving:
(round 1, fast)       1 del vars, 0 del conss, 0 add conss, 2 chg bounds, 0 chg sides, 0 chg coeffs, 0 upgd conss, 0 impls, 0 clqs
(round 2, fast)       1 del vars, 0 del conss, 0 add conss, 4 chg bounds, 0 chg sides, 0 chg coeffs, 0 upgd conss, 0 impls, 0 clqs
(round 3, fast)       1 del vars, 0 del conss, 0 add conss, 7 chg bounds, 0 chg sides, 0 chg coeffs, 0 upgd conss, 0 impls, 0 clqs
(round 4, fast)       1 del vars, 0 del conss, 0 add conss, 10 chg bounds, 0 chg sides, 0 chg coeffs, 0 upgd conss, 0 impls, 0 clqs
Assertion failed: (SCIPisRelEQ(scip, exprhdlrinterval.inf, SCIPexprGetActivity(expr).inf)), function forwardPropExpr, file cons_nonlinear.c, line 2402.

scipopt / scip

Assertion failed with minimal quadratic model #70