Tomas Lindquist Olsen
c7936c2ace
!ClassInfo instances were not mutable, this is necessary for .classinfo based locking to work. !ModuleInfo generation was commented out as well.
2009-04-17 00:54:20 +02:00
Tomas Lindquist Olsen
803fbf4882
Make !TypeInfo output lazy unless we need to provide full definitions in general.
2009-04-16 22:20:26 +02:00
Tomas Lindquist Olsen
e01917bc0d
Make typeinfo for struct
2009-04-16 22:19:36 +02:00
Tomas Lindquist Olsen
8feabfc5c6
Make sure aggregates are resolved before outputting their debug info.
2009-04-16 21:15:42 +02:00
Tomas Lindquist Olsen
414e58a318
Fixed forward referenced structs.
2009-04-16 20:41:27 +02:00
Tomas Lindquist Olsen
1437bd6b7e
Merge.
2009-04-16 20:31:21 +02:00
Tomas Lindquist Olsen
02466de996
Removed invalid assertion.
2009-04-16 20:30:30 +02:00
Tomas Lindquist Olsen
9a86f9c0b5
Fixed some minitest regressions.
2009-04-16 19:21:30 +02:00
Frits van Bommel
7782d80919
Fix typo in comment
2009-04-16 12:01:28 +02:00
Frits van Bommel
5ea141e919
Tweak some optimizations.
...
Delegates passed to inlined functions now also stand a chance of being inlined.
This should make opApply as efficient as a regular loop, as long as both opApply
and the foreachbody are eligible for inlining; which is to say most non-virtual
opApply invocations will likely get fully inlined now.
(Note: above requires -O2 -enable-inlining or -O3)
2009-04-16 11:58:43 +02:00
Tomas Lindquist Olsen
bb98d68f64
Now compiles again, and fixed type of the Interface[N] ClassInfo symbol.
2009-04-16 11:01:49 +02:00
Frits van Bommel
1ef9941daf
Name some types.
2009-04-15 21:37:01 +02:00
Frits van Bommel
8c3cc1467f
Remove 'noalias' and 'nocapture' from %.nest_arg. They no longer apply now that
...
it may point to a struct containing variables from outer functions.
2009-04-15 20:59:21 +02:00
Frits van Bommel
c19ab2e539
Unify %.frames_list and %.frame into a single data structure, generalizing r1212
...
to all frames instead of just the outer-most one.
2009-04-15 20:59:19 +02:00
Tomas Lindquist Olsen
37cf5a5789
Added Doxygen file.
...
Completely seperated type and symbol generation. Should fix a lot of bugs, but is not yet 100% complete.
2009-04-15 20:06:25 +02:00
Frits van Bommel
c44c52296f
Give all symbols nested in functions internal linkage, unless it's one of the
...
other special cases.
(for example: this shouldn't be done if the symbol in question is also nested
in a template; such symbols should get template-like linkage)
2009-04-13 16:16:03 +02:00
Frits van Bommel
376a825075
Fix a bug in nested context code that occured when calling a function nested in
...
the outermost scope with a context frame from a function using a more nested
context frame.
2009-04-13 12:19:18 +02:00
Frits van Bommel
5a04c4e9a1
For the outermost function needing a context frame, use the address of that
...
frame as the nest argument instead of the address of a single-element list
containing only that frame address.
This saves some stack space and reduces memory accesses.
2009-04-13 04:09:08 +02:00
Frits van Bommel
cc492ecc82
Merge
2009-04-12 22:22:15 +02:00
Frits van Bommel
071bad95dc
Add some alignment info where LLVM might otherwise be more pessimistic.
...
In particular, %.nest_arg is always aligned even though it's bitcast from i8*.
Pointers in vtables are also guaranteed to be stored at aligned addresses.
2009-04-12 21:56:43 +02:00
Frits van Bommel
751f528969
Implement -nested-ctx=hybrid
2009-04-12 20:23:00 +02:00
Christian Kamm
893e17c259
Simplify array casts and implement dyn array -> static array.
2009-04-12 19:56:03 +02:00
Frits van Bommel
8820024070
Add an option to change the way nested variables are handled.
...
Only one value is implemented, which is the old way.
2009-04-12 16:22:21 +02:00
Christian Kamm
d251e5d1f4
Implement static array -> static array cast. Fixes #223 .
...
dynamic array -> static array is still misssing!
2009-04-12 16:18:02 +02:00
Christian Kamm
b8b9c41827
Introduce checks to fix #173 .
2009-04-12 14:58:18 +02:00
Frits van Bommel
b5af30636e
Isolate all knowledge of what a function's nested context looks like in a
...
single place. No functional change.
2009-04-12 13:08:24 +02:00
Frits van Bommel
605593bcef
Reduce include-order dependencies
2009-04-12 12:52:01 +02:00
Frits van Bommel
35a3d86cc6
Remove reference to llvm::cl::AllowInverse enum value which was removed in
...
LLVM trunk.
2009-04-08 23:05:30 +02:00
Christian Kamm
2347526ca8
Disable frame pointer elimination when -g or -gc is passed.
2009-04-08 21:12:08 +02:00
Christian Kamm
ba200214d4
Make sure -run doesn't consume the config file flags.
2009-04-05 15:58:28 +02:00
Tomas Lindquist Olsen
9c4b2b4036
Started seperating type resolution from the rest of codegen again, the merge had too many regressions.
2009-04-03 16:34:11 +02:00
Kelly Wilson
7a1052a2fc
Some more fixups for mingw -- missing underscores
2009-04-02 17:27:53 -06:00
Christian Kamm
04f3c61b97
Make sure debug info is only declared once per alloca.
2009-04-02 20:56:39 +02:00
Frits van Bommel
ff72b1f640
Make sure -defaultlib and -debuglib don't get cut off if longer than 63 chars,
...
and clean up some overly verbose code.
2009-04-01 00:52:31 +02:00
Frits van Bommel
c6af051d82
Automated merge with http://hg.dsource.org/projects/ldc
2009-04-01 00:13:21 +02:00
Frits van Bommel
3f49ddb6d5
Fix nested references to 'ref' foreach variables.
...
These "walk around" the array being iterated over, so they're a bit trickier
than other variables to get right.
2009-04-01 00:01:44 +02:00
Tomas Lindquist Olsen
c6e9f7f9e2
Moved ir/irtype.* into ir/irdtype.*.
...
Renamed IrType to IrDType, analogous to IrDSymbol.
2009-03-31 20:21:28 +02:00
Kelly Wilson
e2d6a94271
Fix one more naked asm output problem for mingw32 and remove some
...
duplicate code
2009-03-30 18:22:30 -06:00
Kelly Wilson
aae3d56e32
Fix up some naked asm output for mingw32
2009-03-30 18:18:23 -06:00
Frits van Bommel
df8547a40f
Remove some redundancy from the DValue hierarchy: all subclasses had a 'type'
...
field and identical implementations of virtual function getType().
Move these to DValue itself, and make getType() non-virtual.
2009-03-31 01:18:35 +02:00
Frits van Bommel
b602a13811
Give function & delegate literals, foreach bodies and lazy parameters internal
...
linkage.
2009-03-30 16:34:36 +02:00
Frits van Bommel
b7ca040f7c
Fix a bug I noticed. Varargs were broken if preceded by tuple parameters.
2009-03-30 00:00:43 +02:00
Frits van Bommel
b490d80c07
Automated merge with http://hg.dsource.org/projects/ldc
2009-03-29 19:46:37 +02:00
Frits van Bommel
424bb88da0
Enable inlining by default for -O3+.
2009-03-29 19:38:59 +02:00
Christian Kamm
c9b82af9f2
Add newlines to end of file.
2009-03-29 19:19:32 +02:00
Frits van Bommel
aa508e792a
Forgot to commit this file, sorry
2009-03-29 18:50:41 +02:00
Frits van Bommel
6135e4201a
Small cleanup: remove mars.h #include
2009-03-29 16:47:19 +02:00
Frits van Bommel
257d305032
Change meaning of optimization levels: -O0 now means 'no optimization' like with
...
other compilers.
2009-03-29 15:51:14 +02:00
Frits van Bommel
e129494812
Allow specific optimization passes to be requested from the command line.
...
Now you can run "`ldc test.d -c -mem2reg -simplifycfg`" if you feel the urge.
The -O<N> options are still supported, and are inserted in the passes list in
the position where they appear on the command line.
(so -simplifycfg -O1 -instcombine does the "right thing")
One small change: -inline is renamed to -enable-inlining due to a naming
conflict with the option to add the -inline pass. -inline now inserts the
inlining pass in the position specified, not in the middle of -O<N>.
(ldmd has been updated to translate -inline to -enable-inlining)
2009-03-29 15:46:55 +02:00
Frits van Bommel
9af4c078a0
Another big-endian fix. (Don't lie to [s]printf...)
2009-03-29 03:53:22 +02:00
