A popular hypothesis of in situ transformation of amphibolite facies gneisses to patchy charnockites by CO₂ influx from mantle was proposed primarily from the Kabbaldurga quarries in South Karnataka and subsequently reported from several south Indian localities. However, presence of abundant mafic granulite enclaves in Kabbaldurga and its neighborhood and its implications in relation to patchy charnockite genesis were not discussed. In these quarries patchy charnockites occur in various modes and associations. Some of these patches do occupy structural weak zones, such as shear bands and fold noses in the migmatitic gneisses, but many of the patchy charnockite bodies occur as branching veins transecting the gneissic foliation and hence do not account for fluid pathways. Most importantly, charnockitic leucosomes at margins of mafic granulite enclaves and charnockitic veins within some mafic granulite enclaves indicate a close genetic link between them via dehydration partial melting. This is further corroborated by trace element distribution between them. Dehydration partial melting in mafic rocks in a migmatite terrain such as Kabbaldurga, can explain all the different modes of the patchy charnockites as various stages of segregation and mobility relative to deformation. Abundant mafic granulite enclaves and field features suggesting a relatively late origin of the patchy charnockites, are compelling evidence against the notion of a transition zone. Mantle derivation age of the mafic source rocks (protoliths of mafic granulites) at Kabbaldurga at 3.08 ± 0.08 Ga with small positive ? values is virtually identical to the source of the massive charnockite of Karnataka craton at 3.08 Ga. This could imply a widespread mafic magmatism in South India around 3.0 Ga.