edquist
commented
3 years ago As lame as it is that apparently they have introduced this small bounds check for vector::operator[], technically it was still undefined behavior before (even when it still worked) to access buf[n] with n > buf.size(), and buf.reserve(n) left size at 0.
Which is to say, even though it's annoying to have to do the initialization with buf.resize(n), this change is technically more correct, even for pre-el8 compilers.
bbockelm
When using reserve for allocating memory in a vector, the size of the vector (meaning the number of elements in the vector) remains as 0. Then when the we use the "[]" operator (line 154) a new assert, introduced in el8, is done to verify that the vector isn't empty and given that it is empty the execution exits with the following error[*].
This can be fixed by using "resize" instead of "reserve" which apart from allocating memory, it sets a default value, thus affecting the size of the vector.
[*] /usr/include/c++/8/bits/stl_vector.h:932: std::vector<_Tp, _Alloc>::reference std::vector<_Tp, _Alloc>::operator[](std::vector<_Tp, _Alloc>::size_type) [with _Tp = char; _Alloc = std::allocator; std::vector<_Tp, _Alloc>::reference = char&; std::vector<_Tp, _Alloc>::size_type = long unsigned int]: Assertion '__builtin_expect(__n < this->size(), true)' failed.