Add support for basic type min/max values

[rkaric]

In C/C++ it's possible to use values such as INT_MAX or FLT_MAX by including a standard library header such as <limits.h> or <float.h>. It would be nice if there was some way for ISPC to offer those as well. Perhaps through predefined preprocessor symbols or a standard library header?

https://en.cppreference.com/w/c/types/limits

[aneshlya]

ISPC has limits for integral types in stdlib:
INT8_MAX/INT8_MIN, INT16_MAX/INT16_MIN, INT32_MAX/INT32_MIN, INT64_MAX/INT64_MIN, UINT8_MAX, UINT16_MAX, UINT32_MAX, UINT64_MAX .
But doesn't have for floating point types.
Plus, no mentions of it in documentation.

[nurmukhametov]

ISPC has limits for integral types in stdlib: INT8_MAX/INT8_MIN, INT16_MAX/INT16_MIN, INT32_MAX/INT32_MIN, INT64_MAX/INT64_MIN, UINT8_MAX, UINT16_MAX, UINT32_MAX, UINT64_MAX .

They are internal stdlib.ispc definitions that are unavailable for users.

So, this example

uniform int foo() {
   return INT32_MIN;
}

results to

<source>:4:11: Error: Undeclared symbol "INT32_MIN". 
   return INT32_MIN;
          ^^^^^^^^^