GCC compiler for VLIW architectures

Question

Do you know how strong support for the VLIW architecture exists in the GCC compiler infrastructure? I know that there are some VLIW architectures supported by GCC. Looking at them, it seems that pipeline optimization remains at a different level of optimization. Are there any good (non-GCC related internal documents) material on this?

Answer 1

In fact, there are VLIW goals supported by GCC: on the current trunk, this includes at least FRV , MeP, and picoChip . Support includes optimization. Since they are not “primary” goals (ie primary or secondary goals ), in some cases there may be some generation of suboptimal codes for them.