LP64 defines a C environment with 64bit pointers and char/short/int/long
are 8/16/32/64bit. char/short/int/long have defined sizes in D. We need
to look only at the pointer size.
There are now 64bit environments with 32bit pointers. Examples are
x32 in Intel and N32 on MIPS64. For these environments D_LP64 should
not defined but the global.param.is64bit is still true. The definition
of size_t is also affected. If only 32bit are addressable then size_t
should also be a 32bit type. (This is not required by C standard but
common practize.)
The net result is that not only the definition of D_LP64 must be changed
but also DtoSize_t().
First merge of current development stream. Asm blocks are known to be broken.
DMD: 0c9f437bc24015707130ba42dc434d9cd58282fb
druntime: 86d49cfb3670904603df0cfdfe44c6fff565c0fc
Phobos: a8ca4f7964becac680af0eadbde05aa7d10fc338
Since the change to the symbol emission strategy, a lot of the
previous attempts to derive the linkage for a symbol from its
AST node has been superfluous. This cleans up all the related
code, as it is really not doing much anymore.
The code to opportunistically internalize symbols is now gone
entirely. It was commented out anyway due to problems with
making it work reliably in the face of the not-so-stable
structure of the AST we get from the frontend (regarding the
modules certain symbols end up in). But in any case internal
linkage woudl also be troublesome for cross-module inlining,
so I gave up on making this work for the time being.
The only (minor) change in the emitted code should be in
.offTi generation, which is untested and disabled by default
anyway (GENERATE_OFFTI).
Turns out that since the change to use DMD's symbol emission
logic a few releases back we didn't actually ever emit code
for the extra modules. Thus, all the effort for the extra
semantic passes went to waste.
If we want to implement DMD-style cross-module inlining, we
need to add the extra semantic3s back in, and then iterate
over all the non-root modules with a codegen visitor that
only emits symbols as available_externally.
We probably won't need the availableExternally members if
we start with a clean design in the new symbol emission
framework. Thus, remove everything in the meantime. This
will make some release builds a bit faster, and should not
negatively affect generated code.
This works around linking problems such as rejectedsoftware/vibe.d#338,
caused by the frontend appending template instances to the wrong
module.
GitHub: Fixes#558.