Tolman used a well known method in General Relativity setting ds to zero.
c'/c = 1- 2MG/rc2
Light is shown slowing down when approaching a large gravity source.
c is in flat space.
Dicke used a modified relativity.
c'/c = 1+ 2MG/rc2
Light is shown speeding up when approaching a large gravity source.
If Dicke is correct GR needs to be changed.
Peter Bergmann had a different opinion, that the change is apparent to some observers not real. Many researchers support this theory.
Gravitational red shift seems to change wave lengths, but does not answer the quesrion since constant light speed is assumed.
Is Variable Light Speed In Curved Space Described By Tolman Or Dicke?