Sat, 28 Feb 2009 15:06:09 +0100 |
Christian Kamm |
Readded indentation that got lost in the dmd/mars.c -> gen/main.cpp move.
|
Sat, 28 Feb 2009 13:48:49 +0100 |
Christian Kamm |
Fix alignment info of cdouble, cfloat. They align the same as double, float now.
|
Sat, 28 Feb 2009 09:06:25 +0100 |
Christian Kamm |
Don't reset gTargetData after compiling a module. Fixes segfault when
|
Fri, 27 Feb 2009 19:41:25 +0100 |
Frits van Bommel |
Remove initialization of padding of reals, ireals and creals.
|
Fri, 27 Feb 2009 12:35:31 +0100 |
Frits van Bommel |
Fix struct field alignment on x86-64 (for align > 8, like real.alignof)
|
Thu, 26 Feb 2009 23:35:39 +0100 |
Frits van Bommel |
Fix cfloat return on x86_64: only perform ABI transformation for non-extern(D)
|
Thu, 26 Feb 2009 22:47:06 +0100 |
Frits van Bommel |
Implement support for intrinsics returning struct types
|
Thu, 26 Feb 2009 18:43:25 +0100 |
Tomas Lindquist Olsen |
Fixed druntime lib path in ldc2.conf.in, thanks Matt
|
Thu, 26 Feb 2009 18:28:50 +0100 |
Tomas Lindquist Olsen |
Added some missing version identifiers.
|
Thu, 26 Feb 2009 18:02:13 +0100 |
Tomas Lindquist Olsen |
Fixed debug info for globals in D2.
|
Thu, 26 Feb 2009 15:51:19 +0100 |
Tomas Lindquist Olsen |
Missing %s error param in main.cpp
|
Thu, 26 Feb 2009 15:23:52 +0100 |
Frits van Bommel |
Don't crash if we don't know what ABI to use. Assume some defaults instead.
|
Thu, 26 Feb 2009 14:51:05 +0100 |
Frits van Bommel |
Fix some -vv output
|
Thu, 26 Feb 2009 14:51:02 +0100 |
Frits van Bommel |
Use LLVM OStream wrapper instead of <iostream> in the logger.
|
Thu, 26 Feb 2009 15:24:20 +0100 |
Tomas Lindquist Olsen |
Fixed const/invariant struc/class fields of the form const int i = 4; they don't add a field.
|
Thu, 26 Feb 2009 14:49:32 +0100 |
Frits van Bommel |
Fix dstress runner to use ldmd
|
Thu, 26 Feb 2009 14:48:47 +0100 |
Frits van Bommel |
Fix x86_64 cfloat support to work on intended target :)
|
Thu, 26 Feb 2009 14:23:18 +0100 |
Tomas Lindquist Olsen |
oups, seems like I broke some constness in D1 last commit, compiles again!
|
Thu, 26 Feb 2009 14:13:27 +0100 |
Tomas Lindquist Olsen |
Forgot new files that were supposed to be in last commit.
|
Thu, 26 Feb 2009 14:11:49 +0100 |
Tomas Lindquist Olsen |
- Fixed LLVM style CL args for D2.
|
Wed, 25 Feb 2009 19:30:06 +0100 |
Frits van Bommel |
Call llvm_shutdown when shutting down the backend.
|
Wed, 25 Feb 2009 17:34:51 +0100 |
Frits van Bommel |
Use LLVM-style command line (instead of DMD-style)
|
Thu, 19 Feb 2009 13:51:44 +0100 |
Frits van Bommel |
Clean up regexes in .hgignore, making it more strict.
|
Thu, 19 Feb 2009 13:50:05 +0100 |
Frits van Bommel |
Removing ldc.conf. (IMPORTANT: run 'cmake .' after pull)
|
Thu, 19 Feb 2009 11:01:34 +0100 |
Frits van Bommel |
Fix out-ouf-source build for runtime as well.
|
Thu, 19 Feb 2009 00:50:35 +0100 |
Frits van Bommel |
Add explicit paths to revisions.* to unbreak out-of-source build.
|
Wed, 18 Feb 2009 23:18:34 +0100 |
Frits van Bommel |
Update some testcases now that 64-bit immediates are working.
|
Wed, 18 Feb 2009 22:50:22 +0100 |
Frits van Bommel |
Clean up the code generated when jumping out of inline asm and make label names more expressive.
|
Wed, 18 Feb 2009 21:46:14 +0100 |
Frits van Bommel |
Fix some assembler issues:
|
Wed, 18 Feb 2009 03:38:12 +0100 |
Frits van Bommel |
Fix a latent bug in the asm code.
|
Tue, 17 Feb 2009 23:14:19 +0100 |
Frits van Bommel |
Update .hgignore due to moved(?) idgen/impcnvgen locations.
|
Tue, 17 Feb 2009 19:08:38 +0100 |
Christian Kamm |
Fix for fix from [974].
|
Tue, 17 Feb 2009 18:57:11 +0100 |
Christian Kamm |
Fix #217. getInternalTypeInfo doesn't work with LDC.
|
Tue, 17 Feb 2009 18:53:15 +0100 |
Christian Kamm |
Allow 'return exp;' for void main().
|
Tue, 17 Feb 2009 18:25:34 +0100 |
Christian Kamm |
Fix #216 by implementing TypeExp::toElem. It just provides the appropriate
|
Tue, 17 Feb 2009 01:13:18 +0100 |
Frits van Bommel |
Change the way the version is printed (more verbose, with short hash)
|
Tue, 17 Feb 2009 00:08:20 +0100 |
Frits van Bommel |
Fix the problems exposed by the callingconv1.d test case.
|
Mon, 16 Feb 2009 23:56:56 +0100 |
Frits van Bommel |
Revert wilsonk's "fix" for the callingconv1.d problem.
|
Mon, 16 Feb 2009 23:56:56 +0100 |
Frits van Bommel |
Add some extra debug output that's useful in diagnosing inline assembler bugs.
|
Mon, 16 Feb 2009 19:31:10 +0100 |
Frits van Bommel |
Fix {,w,d}char handling in tests/mini/vararg6.d:print()
|
Mon, 16 Feb 2009 19:17:15 +0100 |
Christian Kamm |
Fix test on x86-32.
|
Mon, 16 Feb 2009 18:27:33 +0100 |
Christian Kamm |
Reduce size of integer constant make x86-64 asm parsing code compile on x86-32.
|
Sun, 15 Feb 2009 23:09:53 -0700 |
Kelly Wilson |
Merging fvbommels fix for asm8.d x86-64 code
|
Mon, 16 Feb 2009 01:43:14 +0100 |
Frits van Bommel |
Some changes I missed in my last commit (wound up in wrong patch file)
|
Sun, 15 Feb 2009 22:56:10 -0700 |
Kelly Wilson |
Fix for tests/mini/callingconv1.d
|
Sun, 15 Feb 2009 16:30:15 -0700 |
Kelly Wilson |
Fix 32 bit truncation of 'asm { movq RAX, 0xAA_1234_5678; }'...ie. move constant larger than 32 bits
|
Sun, 15 Feb 2009 16:26:12 -0700 |
Kelly Wilson |
Fix shl/r bad operand size error
|
Sun, 15 Feb 2009 21:44:03 +0100 |
Christian Kamm |
Add OSX version declaration for Mac OS X to improve dmd-mac compatibility.
|
Sun, 15 Feb 2009 18:04:22 +0100 |
Frits van Bommel |
Implement implicit return after inline asm on x86_64
|
Sun, 15 Feb 2009 14:54:36 +0100 |
Christian Kamm |
Static ctors/dtors added via template mixins should not be added to modules
|
Sun, 15 Feb 2009 14:50:33 +0100 |
Christian Kamm |
Backed out changeset [369996c08420].
|
Sun, 15 Feb 2009 11:46:28 +0100 |
Christian Kamm |
Change the numbering of static ctors/dtors to be module based instead of being global.
|
Sat, 14 Feb 2009 21:18:47 +0100 |
Christian Kamm |
Make sure debug info uses the right compilation units - in particular the
|
Sat, 14 Feb 2009 15:25:48 +0100 |
Tomas Lindquist Olsen |
Added support for using a temporary to implement emulated ABI return from inline asm, could be easier to use, but I think this will do. It's so extremely target dependent in any case that doing a completely generic approach seems hard.
|
Fri, 13 Feb 2009 22:20:30 +0100 |
Tomas Lindquist Olsen |
Merged a bunch recent dmd v1 frontend changes into the dmd v2 tree.
|
Thu, 12 Feb 2009 18:39:14 +0100 |
Christian Kamm |
Automated merge with http://hg.dsource.org/projects/ldc
|
Thu, 12 Feb 2009 18:39:03 +0100 |
Christian Kamm |
Now that we're using LLVM 2.5, we can stop always running the
|
Wed, 11 Feb 2009 08:32:33 -0700 |
Kelly Wilson |
Revert return value swapping on x86_64
|
Tue, 10 Feb 2009 08:02:25 -0700 |
wilsonk |
changes to get the naked asm stuff working for x64
|
Sun, 08 Feb 2009 21:44:46 +0100 |
Christian Kamm |
Revert templates to old behavior.
|
Sun, 08 Feb 2009 18:51:32 +0100 |
Christian Kamm |
Switch some getModule to getCompilationModule to make sure templates are
|
Sun, 08 Feb 2009 16:50:22 +0100 |
Christian Kamm |
Mostly rewrite debug info generation in terms of llvm/Analysis/DebugInfo.h.
|
Sun, 08 Feb 2009 05:26:54 +0100 |
Tomas Lindquist Olsen |
SWITCHED TO LLVM 2.5 !
|
Sun, 08 Feb 2009 05:14:24 +0100 |
Tomas Lindquist Olsen |
Fixed accessing function symbols from inline asm.
|
Thu, 05 Feb 2009 18:17:42 +0100 |
Tomas Lindquist Olsen |
Reverted the template instantiation changes from rev [940]. Wasn't safe it seems :(
|
Wed, 04 Feb 2009 20:47:27 +0100 |
Tomas Lindquist Olsen |
Reverted back to the LLVM_InlineAsm_X86 versions, a few more things still needs to be sorted out first.
|
Wed, 04 Feb 2009 18:51:11 +0100 |
Christian Kamm |
LDC should follow the D ABI now. Switch back from LLVM_InlineAsm_X86 to
|
Wed, 04 Feb 2009 18:48:03 +0100 |
Tomas Lindquist Olsen |
Changed the way LDC determines if a template instantiation needs to get a definition, seems to speed up compile times quite a bit in some cases.
|
Wed, 04 Feb 2009 18:39:39 +0100 |
Christian Kamm |
Automated merge with http://hg.dsource.org/projects/ldc
|
Wed, 04 Feb 2009 18:39:33 +0100 |
Christian Kamm |
Convert struct arg to integer when passing inreg to make sure LLVM doesn't
|
Wed, 04 Feb 2009 16:02:05 +0100 |
Tomas Lindquist Olsen |
Fixed warning on osx as per request from mwarning.
|
Wed, 04 Feb 2009 02:10:23 +0100 |
Tomas Lindquist Olsen |
Removed some dead code.
|
Tue, 03 Feb 2009 17:27:01 -0700 |
wilsonk |
Fixed an accidental code removal for osx
|
Wed, 04 Feb 2009 00:50:04 +0100 |
Tomas Lindquist Olsen |
Fixed problem with static ctors, in templated imported classes, not begin run. Fixes #203
|
Wed, 04 Feb 2009 00:04:09 +0100 |
Tomas Lindquist Olsen |
Replace assertion with errormessage for unsupported calling conventions. like Pascal. See dstress/run/e/extern_10_A.d
|
Tue, 03 Feb 2009 23:48:47 +0100 |
Tomas Lindquist Olsen |
Fixed dstress/run/a/array_initialization_17_A.d regression. default initialized static array elements in a constant static array initializer was getting incorrect values.
|
Tue, 03 Feb 2009 21:46:54 +0100 |
Christian Kamm |
Automated merge with http://hg.dsource.org/projects/ldc
|
Tue, 03 Feb 2009 21:46:46 +0100 |
Christian Kamm |
Follow the D ABI and pass the last arg in a register if it is a struct that fits.
|
Tue, 03 Feb 2009 12:43:03 -0700 |
wilsonk |
Changed this file to add 'naked' and a couple other missing instructions
|
Tue, 03 Feb 2009 12:35:00 -0700 |
wilsonk |
Added X86-64 version. Needed significant changes.
|
Tue, 03 Feb 2009 12:34:12 -0700 |
wilsonk |
Added X86-64 versioning
|
Tue, 03 Feb 2009 12:33:42 -0700 |
wilsonk |
Added X86-64 version
|
Tue, 03 Feb 2009 12:33:21 -0700 |
wilsonk |
Added X86-64 versioning
|
Tue, 03 Feb 2009 18:42:25 +0100 |
Tomas Lindquist Olsen |
Fixed problem accessing global variables from naked asm on osx.
|
Tue, 03 Feb 2009 18:14:30 +0100 |
Tomas Lindquist Olsen |
Added support for naked asm on OSX. (hopefully!)
|
Tue, 03 Feb 2009 18:11:39 +0100 |
Christian Kamm |
Apply naked changes of [920] to dmd2/
|
Tue, 03 Feb 2009 18:00:17 +0100 |
Christian Kamm |
Apply StaticStructInit changes from [913] to dmd2/
|
Tue, 03 Feb 2009 08:54:57 +0100 |
Tomas Lindquist Olsen |
Implemented proper support for naked asm using llvm module level asm. Still not 100% complete, but already 1000 times better that what we had before. Don's BignumX86 implementation from Tango (when turned into a standalone unittest) seems to fully work with no changes, and great performance :)
|
Mon, 02 Feb 2009 02:35:44 +0100 |
Tomas Lindquist Olsen |
Fixed assertion failure when object.d is missing the ModuleInfo class
|
Mon, 02 Feb 2009 01:44:51 +0100 |
Tomas Lindquist Olsen |
Changed templates and typeinfo to use linkonce linkage instead of weak linkage, this should fix inlining problems, fixing bug #197 . If problems show up, it's easy to change it back by changing the define in mars.h . I'm 95% sure this is safe, given how we handle templates.
|
Mon, 02 Feb 2009 00:58:36 +0100 |
Tomas Lindquist Olsen |
Removed use of TARGET_LINUX and replaced with POSIX, fixes bug [172]
|
Mon, 02 Feb 2009 00:34:22 +0100 |
Tomas Lindquist Olsen |
Renamed Rebuild profile to match DSSS and Tango convention.
|
Mon, 02 Feb 2009 00:26:32 +0100 |
Tomas Lindquist Olsen |
Fix bug #199
|
Sun, 01 Feb 2009 23:30:36 +0100 |
Tomas Lindquist Olsen |
Fixed bug #191 by rewriting DtoConstArrayInitializer, patch unfortunately caused regressions, hopefully this doesn't :P
|
Sun, 01 Feb 2009 20:20:56 +0100 |
Christian Kamm |
Fix #198 and #199 by making CTFE on static struct initializers work.
|
Fri, 30 Jan 2009 07:51:50 +0100 |
Christian Kamm |
_d_delarray also had one of the array consistency asserts mentioned in #184
|
Thu, 29 Jan 2009 17:59:25 +0100 |
Christian Kamm |
Fix typo in [909].
|
Thu, 29 Jan 2009 17:45:12 +0100 |
Christian Kamm |
Remove asserts on array consistancy in arraysetlength. Fixes #184, thanks Deewiant.
|
Thu, 29 Jan 2009 17:19:48 +0100 |
Christian Kamm |
Port some LDC1 portability fixes to LDC2. Fixes #196.
|
Thu, 29 Jan 2009 16:56:48 +0100 |
Christian Kamm |
Apply patch for #193. Thanks fvbommel.
|
Sun, 25 Jan 2009 12:38:29 +0100 |
Christian Kamm |
Add testcase for #189 to mini tests.
|
Sun, 25 Jan 2009 12:35:11 +0100 |
Christian Kamm |
Casts from interface to class generally need to be dynamic. Fixes #189.
|
Sat, 24 Jan 2009 00:33:06 +0100 |
Christian Kamm |
Applied doob's patch to port GDC's Dyld data segment tracking for Mac.
|
Fri, 23 Jan 2009 13:41:03 +0100 |
Christian Kamm |
Fix bug #177 by not attempting to define composite debug info for
|
Fri, 23 Jan 2009 10:13:27 +0100 |
Christian Kamm |
Further Solaris data segment detection fixes. Thanks BlueZeniX!
|
Fri, 23 Jan 2009 10:12:57 +0100 |
Christian Kamm |
Add -m64 switch to gcc calls on 64-bit platforms.
|
Thu, 22 Jan 2009 21:46:40 +0100 |
Christian Kamm |
Fix #174.
|
Thu, 22 Jan 2009 18:53:32 +0100 |
Christian Kamm |
Solaris compatibility and proc map parsing #183, thanks BlueZeniX!
|
Thu, 22 Jan 2009 18:23:30 +0100 |
Christian Kamm |
Add wilsonk's static data segment finding code for x86-64 from #181. Thanks!
|
Thu, 22 Jan 2009 17:57:38 +0100 |
Christian Kamm |
Add headers for HP's libunwind, currently unused though.
|
Thu, 22 Jan 2009 17:50:38 +0100 |
Christian Kamm |
Fix x86-64 exception handling #180. Thanks pcwalton!
|
Thu, 22 Jan 2009 17:26:50 +0100 |
Tomas Lindquist Olsen |
Makes sure only function definitions inside templates get weak linkage, llvm doesn't allow weak linkage on just declarations, so we use external linkage for those. Fixes #162
|
Mon, 19 Jan 2009 19:42:00 +0100 |
Christian Kamm |
Apply pcwalton's eh runtime portability fix.
|
Mon, 19 Jan 2009 19:34:51 +0100 |
Christian Kamm |
Don't use /proc/self/map for finding the static data segment on x86-64 for
|
Sun, 18 Jan 2009 14:14:40 +0100 |
Christian Kamm |
Remove optimizer syntax copies where they weren't proven to be necessary.
|
Sun, 18 Jan 2009 13:18:31 +0100 |
Christian Kamm |
Also apply custom allocator size_t change to DMD2 frontend.
|
Sun, 18 Jan 2009 13:10:19 +0100 |
Christian Kamm |
Change custom allocator new size argument to size_t instead of uint. See #175.
|
Sat, 17 Jan 2009 17:52:09 +0100 |
Christian Kamm |
Apply afb's patch from #168 to use $CC instead of "gcc" for toobj and linker.
|
Sat, 17 Jan 2009 17:44:54 +0100 |
Christian Kamm |
Fix #165, thanks afb.
|
Sat, 17 Jan 2009 17:41:03 +0100 |
Christian Kamm |
Fix #164.
|