The problem is that constructing a float16_t, int8_t, uint8_t, int16_t, or uint16_t with a non 32-bit argument generates an aggregate with the wrong construction op. For int8_t and uint8_t, this resulted in a crash. |
||
|---|---|---|
| .. | ||
| GenericCodeGen | ||
| Include | ||
| MachineIndependent | ||
| OSDependent | ||
| Public | ||
| CMakeLists.txt | ||
| updateGrammar | ||