Fatal error in PMPI_Wait

Dear all,
I got a error when I run a simulation using multi-cores (like 10 cores or 24 cores):
Fatal error in PMPI_Wait: Message truncated, error stack:
PMPI_Wait(197)…: MPI_Wait(request=0x7fffe6a88b90, status=0x1) failed
MPIR_Wait_impl(79)…:
MPIDI_CH3U_Receive_data_found(131): Message from rank 4 and tag 0 truncated; 232 bytes received but buffer size is 208
When I used 1 or 2 cores, no such error appeared. What’s the reasons and how to solve the problem? Thanks.

Here is my input:
units real

atom_style full

read_data data.PtCO

pair_style hybrid eam lj/cut 10.0 lj/cut/coul/cut 10.0 morse 2.5
pair_coeff 1 1 eam Pt_u3_real.eam
pair_coeff 1 2 lj/cut 52.25 1.68
pair_coeff 3 3 lj/cut/coul/cut 0.1591 3.12
pair_coeff 2 2 lj/cut/coul/cut 0.0262 3.83

pair_coeff 2 3 lj/cut/coul/cut 0.0001 3.8
pair_coeff 2 4 lj/cut/coul/cut 0.0001 3.8
pair_coeff 3 4 lj/cut/coul/cut 0.0001 3.8
pair_coeff 4 4 lj/cut/coul/cut 0.0001 3.8
pair_coeff 1 4 lj/cut/coul/cut 0.0001 3.8

pair_coeff 1 3 morse 0.03 1.7 5.15

#neigh_modify exclude type 2 3
#neigh_modify exclude type 2 4
#neigh_modify exclude type 3 4

neighbor 0.3 bin
neigh_modify delay 10

group Pt type 1
group CO type 2 3 4

timestep 1

fix 1 CO rigid/nvt/small molecule temp 1000.0 1000.0 10
neigh_modify exclude molecule/intra CO
fix 2 Pt nvt temp 1000.0 1000.0 10

thermo_style custom step elapsed elaplong dt time temp epair ke pe etotal press
thermo 1

dump 1 all custom 100 dump.custom id mol element x y z
dump_modify 1 sort id element Pt C O M format float %12.8f

#velocity CO create 1000.0 123456
#run 0
#velocity CO scale 1000.0

run_style verlet
run 1000000

unfix 1
unfix 2
fix 3 CO rigid/nve/small molecule
fix 4 Pt nve

restart 10000000 restart.*.PtCO
run 100000000
write_data data-simu.PtCO
write_restart restart-simu.PtCO

I set the masses in the data file:
1 195.09
2 12.0000
3 15.9990
4 0.00000001

What is the LAMMPS version?

LAMMPS (23 Jun 2022 - Update 2)

Please try the most recent version from last month. If that still fails in the same way, please provide a complete input deck so we can do a detailed debugging.

Thanks, I’ll try the most recent version.

Additionally, in this simulation, I want to use the repulsive Morse potential only, but the Morse potential in the manual includes both attractive and repulsive sections. Need I modify the pair_morse.cpp file to meet my requirements?

Just set the cutoff so that it is in the minimum.