nir: Add a fquantize2f16 opcode
authorJason Ekstrand <jason.ekstrand@intel.com>
Tue, 12 Jan 2016 23:32:51 +0000 (15:32 -0800)
committerJason Ekstrand <jason.ekstrand@intel.com>
Tue, 12 Jan 2016 23:33:02 +0000 (15:33 -0800)
commit15a56459d7562c02fb288f1b9a5a77f41c2d3ee1
tree773b9177abbd690d070a1fa3a00b5c760c697c3d
parentaee970c844f85c00dcb5d0df85ef56981c255291
nir: Add a fquantize2f16 opcode

This opcode simply takes a 32-bit floating-point value and reduces its
effective precision to 16 bits.
src/glsl/nir/nir_opcodes.py