Do all versions of GCC C++ standard library on 64-bit Ubuntu implement this bitset and vector bool specialization with bit packing?

I have a special bit data structure that I want to ensure is as space efficient as possible (up to padding for alignment). It needs to hold up to 2^20 bits. I'm considering implementing it manually on uint64 words with bit operations.

Do all versions of GCC C++ standard library on 64-bit Ubuntu implement this bitset and vector bool specialization with bit packing?

I have a special bit data structure that I want to ensure is as space efficient as possible (up to padding for alignment). It needs to hold up to 2^20 bits. I'm considering implementing it manually on uint64 words with bit operations.

• ubuntu
• gcc
• std-bitset

• Given that bit packing is the entire point of std::bitset, I'd be extremely surprised if it has ever not been the case. – Nate Eldredge Commented Feb 2 at 5:10
• @NateEldredge Do you know if GCC will do the same for vector<bool>? – qwr Commented Feb 2 at 5:56

1 Answer 1

The standard says:

There is no requirement that the data be stored as a contiguous allocation of bool values. A space-optimized representation of bits is recommended instead.

Although that's only a recommendation, it's natural to use similar representations for both std::vector<bool> and std::bitset.

With std::bitset, you could use static_assert() with sizeof to help ensure the size is minimal; that's more difficult with std::vector<bool>.