CODEWITHSUNDEEP
c++
octopus
octopus

Reputation: 528

The arithmetic types (in C++) are defined to cater to the peculiarities of various kinds of hardware

I read this sentence, while reading the "C++ Primer Book". What exactly does this sentence mean? And what are examples of those various kinds of hardware? How were the arithmetic types designed to do that? (at least consideration that were made to achieve that). Thanks in advance.

Upvotes: 3

Views: 87

Answers (2)

Paul Evans
Paul Evans

Reputation: 27577

Care is taken in the C++ standard not to restrict arithmetic types in such a way as to make them "unnatural" to any given hardware. Most implementations, for example, use 2's complement integers and use IEEE Standard 754 for floating point numbers. But they don't have to in order to be compliant with the standards. They could be 1's complement integers and not use IEEE Standard 754 for floating point types if that suited the hardware better and still be fully compliant with the C++ standard.

Upvotes: 3

Pete Becker
Pete Becker

Reputation: 76438

In C and C++ a char has to be at least 8 bits wide. A short has to be at least 16 bits wide and at least as wide as a char. An int has to be at least 16 bits wide and at least as wide as a short. A long has to be at least 32 bits wide and at least as wide as an int. A long long has to be at least 64 bits wide and at least as wide as a long.

Note all those "at least"s. Each of those types can be wider than the minimum that's allowed, and when one of those types is wider it ripples through the rest of the types.

There are DSP chips out in the wild today that only address 32-bit values, so char is 32 bits wide, short is 32 bits wide, and int is 32 bits wide. That's an example of catering to the peculiarities of that hardware.

Upvotes: 2

Related Questions