Re: constexpr w/c++ stdlib features??

[Jonathan Wakely <jwakely at redhat dot com>]

On 22/05/14 06:13 -0700, Linda A. Walsh wrote:
> Jonathan Wakely wrote:
> > On 21/05/14 22:02 -0700, Linda A. Walsh wrote:
> > > I'm trying to convert various 'C'-isms in my code.
> > >
> > > I had something that "passed" the "constexpr-test",
> > > namely, constexpr const char ** cpu_states_txt = {"USR", "Ni", "Sys"...};
> > >
> > > This compiled and worked just fine.  But I am trying to change it
> > > to a vector or valarray:
> > >
> > > constexpr const vector<const char *> const cpu_states_txt ={"Usr, 
> > > "...."};
> >
> > Why do you need this to be constexpr?
> >
> > If you really do nee it to be constexpr then you need to use a literal
> > type that can be initialized at compile-time.
> ---
> 	I thought that since vector or valarray could have dimensions
> specified at compile time,

This has dimensions specified at compile-time:

 int* p = new int[5];

That doesn't make it a compile-time constant, it uses dynamic
allocation which is inherently a run-time behaviour.

> then they might support compile time constants
> as well.  I didn't know the implementation couldn't use compile time
> dimensions to allow it to be used as constexpr.
>
> 	Is it a requirement that std::vector use runtime allocation if
> the object it is representing is fixed at compile time or is that an
> the result of implementation?

It's not possible for std::vector to tell the difference, so it's not
possible to implement it to choose to avoid run-time allocation.

As far as I'm aware the standard does not forbid the compiler from
optimising away dynamic allocation if it can statically prove there
will be no user-visible difference, but that would be a property of
the compiler not std::vector, and still wouldn't let you use
constexpr, because being able to declare the object as constexpr would
be a user-visible difference.

Optimisations are not supposed to change semantics, and declaring an
object constexpr has semantics that guarantee no non-trivial
initialization, it doesn't mean "please try to avoid non-trivial
initialization if the compiler happens to be smart enough and I enable
optimisations".

> 	It *seemed*, that by tagging things as invariant (either at compile
> or run time after some initialization), it could allow optimizations.  in this
> case, it would be the run-time code to initialize "resources" that are
> statically fixed before the program starts execution.

A const vector is invariant at run-time after some initialization.

A constexpr object cannot have any run-time initialization. At all.


> > > ....  error: the type âconst std::vector<const
> > > char*>â of constexpr variable âcpu_states_txtâ is not literal
> >
> > Obviously since std::vector uses dynamic memory allocation it can't be
> > a compile-time constant.
> >
> > Again, why do you need it to be constexpr?
> > What's wrong with just a const vector?
> ----
> 	Nothing, but since it doesn't change and is present at compile
> time, how would I let the invariant data be done at compile time
> and not run time?

I don't know what "let the invariant data be done at compile time"
means, sorry.

What's wrong with simply:

const std::vector<const char*> cpu_states_txt{"USR", "Ni", "Sys"};

Why do you need that to be "done at compile-time"?

If you need other compile-time operations to use the array elements
either leave it as an array:

constexpr const char* cpu_states_txt[] = {"USR", "Ni", "Sys"};

Or use std::array

template<typename... T>
 constexpr std::array<typename std::common_type<T...>::type,
 sizeof...(T)>
 make_array(T... t)
 {
   return { { t... } };
 }

constexpr auto cpu_states_txt = make_array("USR", "Ni", "Sys");

You can even turn the former into the latter (using
make_index_sequence from C++14):

#include <array>
#include <iterator>
#include <utility>

template<typename T, std::size_t N, std::size_t... I>
constexpr std::array<T, N>
array2array_impl(const T (&a)[N], std::index_sequence<I...>)
{
 return { { a[I]... } };
}

template<typename T, std::size_t N>
constexpr std::array<T, N>
array2array(const T (&a)[N])
{
 return array2array_impl(a, std::make_index_sequence<N>{});
}

constexpr const char* cpu_states_txt[] = {"USR", "Ni", "Sys"};
constexpr auto states = array2array(cpu_states_txt);

There are loads of ways to skin this cat, but constexpr std::vector is
not one of them.

> 	I'm trying to recode parts of this program in C++ specific
> terms where it will decomplicate the code.  Since it seems doable in 'C',
> I wanted to figure out how to have C++ do it.

You can do what you did in C, that's valid C++ too.
(If it ain't broken don't fix it.)

Adding constexpr then changing to a non-trivial type is not the C++
way to do it, because it's not valid C++!

What you're trying to do with vector is not possible (and valarray is
almost certainly not what you want to use). You can either use a
constexpr std::array or use a const std::vector.