MAD_SAT(3clc)
π MAD_SAT β the OpenCL manual page, but make it make sense π
okay bestie, gather round, because we're doing mad_sat from the OpenCL Manual (that's the 3clc corner of the world) and honestly? it's giving "tiny function, huge type lore." let's get into it β¨
πΈ NAME πΈ
mad_sat β does a multiply, adds a thing, and saturates the result. that's the whole vibe. π«Ά
β¨ SYNOPSIS β¨
gentype mad_sat(gentype a, gentype b, gentype c);
three args, all the same generic type, one very well-behaved little return value. no cap.
π DESCRIPTION π
so mad_sat computes a*b+c and then saturates the result. literally that's it. multiply, add, clamp the energy. iconic behavior, no notes. π₯
π NOTES π
okay here's where it gets juicy, because the built-in integer functions are not playing around:
- Built-in integer functions take scalar or vector arguments. π¦
- The vector versions of the integer functions operate component-wise. The description is per component. (so like, each lane does its own thing, independently, minding its own business π)
now, the type names. we use the generic type name gentype to say "the function can take any of these as the type for the arguments":
char, char{2|3|4|8|16}, uchar, uchar{2|3|4|8|16}, short, short{2|3|4|8|16}, ushort, ushort{2|3|4|8|16}, int, int{2|3|4|8|16}, uint, uint{2|3|4|8|16}, long, long{2|3|4|8|16}, ulong, or ulong{2|3|4|8|16}.
and we use the generic type name ugentype to refer to unsigned versions of gentype. for example: if gentype is char4, then ugentype is uchar4. easy math, bestie. π
we also use the generic type name sgentype to indicate that the function can take a scalar data type, i.e. char, uchar, short, ushort, int, uint, long, or ulong as the type for the arguments.
for built-in integer functions that take gentype and sgentype arguments, the gentype argument must be a vector or scalar version of the sgentype argument. for example: if sgentype is uchar, then gentype must be uchar or uchar{2|3|4|8|16}. and for vector versions, sgentype is implicitly widened to gentype, exactly as described in section 6.3.a of the OpenCL specification. π
and for any specific use of a function, the actual type has to be the same for all arguments and the return type, unless otherwise specified. so no mix-and-matching, babe, that's the rule. π
π SPECIFICATION π
OpenCL Specification[1]
π§ SEE ALSO π§
β AUTHORS β
The Khronos Group
π COPYRIGHT π
Copyright Β© 2007-2011 The Khronos Group Inc.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and/or associated documentation files (the "Materials"), to deal in the Materials without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Materials, and to permit persons to whom the Materials are furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Materials.
π NOTES (the footnote edition) π
- OpenCL Specification β page 256, section 6.12.3 - Integer Functions
okay and that's the whole page, we ate and left absolutely zero crumbs π β¨
<cute sign-off>sending you off with love from The Khronos Group, dated 01/14/2021 β this was MAD_SAT(3clc), and yes the manual name is right there in the title, we don't hide the lore. bye bestie, go saturate something πΈπ«Ά</cute sign-off>
π¦ where this page comes from
- ubuntu jammy: opencl-1.2-man-doc 1.0~svn33624-5 Β· other
- ubuntu noble: opencl-1.2-man-doc 1.0~svn33624-5 Β· other
- π girlified by deepseek-ai/DeepSeek-V4.1-Flash on : a language model's rewrite of the original, under its terms