|
|
|
--- !ditz.rubyforge.org,2008-03-06/issue
|
|
|
|
title: remove 64-bit type detection altogether for non-C99 and non-major compilers?
|
|
|
|
desc: |-
|
|
|
|
64-bit detection is a major hassle:
|
|
|
|
|
|
|
|
* Even if the compiler supports 64-bit types and provides defines for it,
|
|
|
|
it might not support the *format* of those defines itself. For instance,
|
|
|
|
the compiler may define ULLONG_MAX but when you use it, it may generate
|
|
|
|
warnings.
|
|
|
|
|
|
|
|
* Similarly, the compiler may support a type (e.g. unsigned long long) but
|
|
|
|
will warn if you use it.
|
|
|
|
|
|
|
|
It's next to impossible to do warning-free detection and usage of these
|
|
|
|
types.
|
|
|
|
|
|
|
|
Perhaps the cleanest approach is to use 64-bit arithmetic only on C99/C++11
|
|
|
|
platforms and on specific compilers where the support is known to exist.
|
|
|
|
This would remove the need for 64-bit detection and avoid 64-bit types
|
|
|
|
when they cannot be reliably detected/used.
|
|
|
|
type: :task
|
|
|
|
component: duk
|
|
|
|
release: v1.1
|
|
|
|
reporter: Sami Vaarala <sami.vaarala@iki.fi>
|
|
|
|
status: :unstarted
|
|
|
|
disposition:
|
|
|
|
creation_time: 2014-07-09 11:40:31.830356 Z
|
|
|
|
references: []
|
|
|
|
|
|
|
|
id: a9c50f83253ff47e03ad931de7f973a5c6ed319a
|
|
|
|
log_events:
|
|
|
|
- - 2014-07-09 11:40:32.022198 Z
|
|
|
|
- Sami Vaarala <sami.vaarala@iki.fi>
|
|
|
|
- created
|
|
|
|
- ""
|
|
|
|
- - 2014-09-09 11:31:10.758381 Z
|
|
|
|
- Sami Vaarala <sami.vaarala@iki.fi>
|
|
|
|
- assigned to release v1.1 from v0.12
|
|
|
|
- ""
|