Matthias Braun [Wed, 3 Feb 2010 20:03:00 +0000 (20:03 +0000)]
fix some assertions; correctly handle weak symbols
[r27031]
Matthias Braun [Wed, 3 Feb 2010 19:31:48 +0000 (19:31 +0000)]
Rework linkage types in firm.
- Merge stickyness, visibility and variability into ir_linkage
- Add proper support for common symbols, weak symbols and C++ COMDAT stuff
- Mark OO-features type_visibility, peculiarity and allocation as deprecated
OO feature will be separated from firm core into higher levels in the long run
- Rework, simplify and fix bugs in begnuas.c
[r27030]
Matthias Braun [Mon, 1 Feb 2010 20:45:23 +0000 (20:45 +0000)]
fix strangeattr.c
[r27024]
Matthias Braun [Fri, 22 Jan 2010 09:04:51 +0000 (09:04 +0000)]
ast2firm already layouts structs no need to have such code in driver too
[r26992]
Michael Beck [Sat, 9 Jan 2010 02:01:09 +0000 (02:01 +0000)]
Fixed some usages of type_name in DBG().
[r26941]
Matthias Braun [Thu, 7 Jan 2010 18:36:31 +0000 (18:36 +0000)]
fix -O0
[r26921]
Matthias Braun [Thu, 7 Jan 2010 18:30:31 +0000 (18:30 +0000)]
Liebe part3: mux lowerer (with minimal changes)
[r26919]
Matthias Braun [Tue, 5 Jan 2010 15:56:54 +0000 (15:56 +0000)]
Primitive, Pointer, Array and Method types are anonymous now
- There's a new type_dbg_info* that allows you to attach debug names to types
- get_type_name and get_type_ident are now more. The new ir_print_type
and the usual ir_printf("%+F", type) help in most usage cases.
But you should be aware that names are not guaranteed to be unique anymore
(or positively said: You don't have no trouble anymore building unique names
in code that creates types)
- No need to specify mode for new pointer types anymore (you can still do it
with set_type_mode)
[r26909]
Matthias Braun [Mon, 4 Jan 2010 02:04:21 +0000 (02:04 +0000)]
fix
[r26907]
Matthias Braun [Mon, 4 Jan 2010 01:12:46 +0000 (01:12 +0000)]
set dbg_retrieve in ast2firm where it is defined
[r26906]
Michael Beck [Mon, 4 Jan 2010 00:33:57 +0000 (00:33 +0000)]
Fixed broken r26886: bring back debug info for middle-end firm dumps.
[r26903]
Matthias Braun [Mon, 4 Jan 2010 00:28:46 +0000 (00:28 +0000)]
add a printing abstraction layer so we can print type/ast to files and memory buffers
[r26901]
Matthias Braun [Sat, 2 Jan 2010 19:30:16 +0000 (19:30 +0000)]
preliminary support for namespace mangling (I had this lying around for some weeks)
[r26889]
Matthias Braun [Sat, 2 Jan 2010 18:52:50 +0000 (18:52 +0000)]
fix for last commit
[r26888]
Matthias Braun [Sat, 2 Jan 2010 18:33:35 +0000 (18:33 +0000)]
remove gen_firm_asm because I am too lazy to fix it (it's probably outdated and I know noone who uses it)
[r26887]
Matthias Braun [Sat, 2 Jan 2010 18:31:00 +0000 (18:31 +0000)]
dbg_snprint is no more
[r26886]
Matthias Braun [Tue, 29 Dec 2009 16:23:40 +0000 (16:23 +0000)]
fix incorrect type size for wide string literals
[r26855]
Matthias Braun [Tue, 29 Dec 2009 15:38:37 +0000 (15:38 +0000)]
don't use compound_graph_path stuff anymore
[r26853]
Matthias Braun [Mon, 21 Dec 2009 18:08:14 +0000 (18:08 +0000)]
fix last commit
[r26824]
Matthias Braun [Mon, 21 Dec 2009 17:59:22 +0000 (17:59 +0000)]
use functions instead of macros, reformat/reindent
[r26823]
Matthias Braun [Mon, 21 Dec 2009 17:44:17 +0000 (17:44 +0000)]
adapt to timer changes
[r26822]
Matthias Braun [Wed, 16 Dec 2009 16:11:14 +0000 (16:11 +0000)]
most of the -fopt flags are handled now by looking at the optimisation list
[r26799]
Matthias Braun [Fri, 11 Dec 2009 16:49:57 +0000 (16:49 +0000)]
fixes for M_except changes
[r26781]
Christoph Mallon [Sat, 28 Nov 2009 09:51:34 +0000 (09:51 +0000)]
Do not use tab to align table columns.
[r26761]
Christoph Mallon [Sat, 28 Nov 2009 09:51:27 +0000 (09:51 +0000)]
Unbreak build: loop_optimization() is undeclared.
[r26760]
Christian Helmer [Mon, 23 Nov 2009 15:26:11 +0000 (15:26 +0000)]
Options for loop optimization added.
[r26738]
Christoph Mallon [Fri, 20 Nov 2009 12:05:27 +0000 (12:05 +0000)]
Slightly simplify daisy chaining several lists.
[r26735]
Christoph Mallon [Fri, 20 Nov 2009 12:05:23 +0000 (12:05 +0000)]
Replace if-cascade by switch.
[r26734]
Christoph Mallon [Fri, 20 Nov 2009 12:05:20 +0000 (12:05 +0000)]
Add next_if(), which consumes the next token, if it is of the desired type.
[r26733]
Christoph Mallon [Fri, 20 Nov 2009 12:05:16 +0000 (12:05 +0000)]
Reduce variable scope.
[r26732]
Christoph Mallon [Fri, 20 Nov 2009 12:05:12 +0000 (12:05 +0000)]
Correct indentation.
[r26731]
Michael Beck [Tue, 3 Nov 2009 23:19:57 +0000 (23:19 +0000)]
Fixed doxygen comments.
[r26710]
Matthias Braun [Thu, 24 Sep 2009 13:42:30 +0000 (13:42 +0000)]
allow overriding of assembler, linker, preprocessor with environment variables
[r26650]
Christoph Mallon [Tue, 15 Sep 2009 07:50:34 +0000 (07:50 +0000)]
Correct off-by-one error of source position in parse_qualified_identifier().
[r26530]
Christoph Mallon [Fri, 11 Sep 2009 11:39:10 +0000 (11:39 +0000)]
Small simplification for lexing wide strings/char literals.
[r26523]
Christoph Mallon [Fri, 11 Sep 2009 11:33:57 +0000 (11:33 +0000)]
Fix C/should_fail/init2.c: if there is no outer_type (error case), just go on.
[r26522]
Christoph Mallon [Fri, 11 Sep 2009 11:33:04 +0000 (11:33 +0000)]
Remove return at end of function.
[r26521]
Matthias Braun [Wed, 9 Sep 2009 13:39:19 +0000 (13:39 +0000)]
write an error message if architecture is unknown
[r26506]
Christoph Mallon [Wed, 2 Sep 2009 06:21:16 +0000 (06:21 +0000)]
Minor simplification.
[r26466]
Michael Beck [Sat, 29 Aug 2009 13:28:02 +0000 (13:28 +0000)]
- Bugfix: missing '\0' added
[r26445]
Christoph Mallon [Thu, 27 Aug 2009 13:27:55 +0000 (13:27 +0000)]
Correct typo in commment.
[r26439]
Matthias Braun [Mon, 17 Aug 2009 12:45:38 +0000 (12:45 +0000)]
when redeclaring stuff with different attributes merge the attributes and don't display the redundant declaration warning
[r26373]
Matthias Braun [Fri, 7 Aug 2009 15:27:39 +0000 (15:27 +0000)]
extend/fix timing stuff
[r26322]
Matthias Braun [Sun, 2 Aug 2009 15:31:32 +0000 (15:31 +0000)]
build plugin specification for cparser (doesn't completely work yet - you have to create builtins.h and revision.h manually)
[r26309]
Matthias Braun [Thu, 16 Jul 2009 11:41:33 +0000 (11:41 +0000)]
it's called thread-jumps now
[r26283]
Sebastian Buchwald [Tue, 14 Jul 2009 16:10:38 +0000 (16:10 +0000)]
condeval is called jump threading now
[r26272]
Sebastian Buchwald [Tue, 14 Jul 2009 15:21:16 +0000 (15:21 +0000)]
condeval is called jump threading now
[r26270]
Sebastian Buchwald [Tue, 14 Jul 2009 15:18:25 +0000 (15:18 +0000)]
condeval is called jump threading now
[r26269]
Matthias Braun [Tue, 14 Jul 2009 15:16:02 +0000 (15:16 +0000)]
condeval is called jump threading now
[r26268]
Matthias Braun [Tue, 14 Jul 2009 15:15:59 +0000 (15:15 +0000)]
remove old/unnecessary files
[r26267]
Michael Beck [Mon, 6 Jul 2009 20:39:03 +0000 (20:39 +0000)]
- icc 11 defines __ptr32
[r26242]
Michael Beck [Sun, 5 Jul 2009 17:21:40 +0000 (17:21 +0000)]
- updated for libFirm w/O irg parameters
[r26238]
Michael Beck [Thu, 25 Jun 2009 23:02:46 +0000 (23:02 +0000)]
- moved ansi alias to optimization level 2 (like gcc), effectively removing
level O4. We loose some benchmarks due to -O3 usage ...
[r26189]
Matthias Braun [Fri, 19 Jun 2009 15:12:58 +0000 (15:12 +0000)]
fix default proj getting constructed in wrong block
[r26165]
Matthias Braun [Fri, 19 Jun 2009 11:10:06 +0000 (11:10 +0000)]
improve/fix preprocessor flag handling for some cases with assembler files
[r26161]
Matthias Braun [Fri, 19 Jun 2009 11:10:03 +0000 (11:10 +0000)]
avoid warnings in the produced jna output
[r26160]
Matthias Braun [Mon, 15 Jun 2009 12:00:46 +0000 (12:00 +0000)]
64bit compile fix
[r26140]
Matthias Braun [Sat, 13 Jun 2009 20:22:06 +0000 (20:22 +0000)]
use real (firm) output positions instead of C one (this fixes asm_test7.c
[r26137]
Matthias Braun [Thu, 11 Jun 2009 09:30:27 +0000 (09:30 +0000)]
improve handling of multiple input files
[r26130]
Matthias Braun [Thu, 11 Jun 2009 09:30:26 +0000 (09:30 +0000)]
fix cp_error059
[r26129]
Matthias Braun [Thu, 11 Jun 2009 09:30:25 +0000 (09:30 +0000)]
redefine __VERSION__ preprocessor macro
[r26128]
Christoph Mallon [Thu, 11 Jun 2009 09:29:28 +0000 (09:29 +0000)]
Handle ?: in is_address_constant().
[r26127]
Matthias Braun [Mon, 8 Jun 2009 13:23:10 +0000 (13:23 +0000)]
add a gitignore file (no direct converion between svn:ignore and .gitignore possible unfortunately)
[r26112]
Matthias Braun [Mon, 8 Jun 2009 12:29:00 +0000 (12:29 +0000)]
some more work towards (c++) namespace support
[r26111]
Christoph Mallon [Wed, 3 Jun 2009 11:01:31 +0000 (11:01 +0000)]
Declarations with empty parentheses are incompatible with prototypes with ellipsis.
[r26097]
Christoph Mallon [Wed, 3 Jun 2009 10:21:18 +0000 (10:21 +0000)]
Emit an error, if a non-variadic function contains va_start.
[r26096]
Christoph Mallon [Wed, 3 Jun 2009 10:19:58 +0000 (10:19 +0000)]
If a K&R function definition has a variadic prototype earlier, then make the function definition variadic, too.
[r26095]
Matthias Braun [Wed, 3 Jun 2009 09:28:56 +0000 (09:28 +0000)]
fix C/bool2.c
[r26094]
Matthias Braun [Wed, 3 Jun 2009 08:49:08 +0000 (08:49 +0000)]
refactor some cast code, use Convb instead of Mux to handle int->bool conversion
[r26091]
Matthias Braun [Mon, 1 Jun 2009 08:53:42 +0000 (08:53 +0000)]
adapt to latest libfirm
[r26083]
Matthias Braun [Fri, 29 May 2009 11:40:10 +0000 (11:40 +0000)]
fix int->bool conversion
[r26069]
Matthias Braun [Fri, 29 May 2009 10:42:40 +0000 (10:42 +0000)]
fix switch generation for dead case
[r26068]
Matthias Braun [Thu, 28 May 2009 10:51:29 +0000 (10:51 +0000)]
fix remaining bitfield initializer problems
[r26060]
Matthias Braun [Wed, 27 May 2009 14:47:26 +0000 (14:47 +0000)]
fix union init problems
[r26058]
Matthias Braun [Wed, 27 May 2009 08:10:26 +0000 (08:10 +0000)]
sync with edgfe
[r26055]
Matthias Braun [Fri, 15 May 2009 11:40:13 +0000 (11:40 +0000)]
NEWS update, fix type compatibility check for k&r functions
[r25970]
Matthias Braun [Thu, 14 May 2009 22:24:21 +0000 (22:24 +0000)]
don't mark functions without prototype as variadic
[r25963]
Matthias Braun [Thu, 14 May 2009 19:38:43 +0000 (19:38 +0000)]
missing skip_typerefs
[r25960]
Matthias Braun [Thu, 14 May 2009 17:45:01 +0000 (17:45 +0000)]
next iteration of kr handling after a discussion with Christoph
[r25955]
Matthias Braun [Thu, 14 May 2009 11:29:33 +0000 (11:29 +0000)]
fix fehler127 (again)
[r25946]
Matthias Braun [Mon, 11 May 2009 15:27:34 +0000 (15:27 +0000)]
unspecified_parameters, are set by kr parameter parser now (and is not set for prototyped functions)
[r25927]
Michael Beck [Fri, 8 May 2009 15:05:56 +0000 (15:05 +0000)]
new attempt to fix the prototyped kr function problem
[r25908]
Matthias Braun [Thu, 7 May 2009 14:14:05 +0000 (14:14 +0000)]
adapt to latest Builtin changes
[r25890]
Matthias Braun [Thu, 7 May 2009 11:37:17 +0000 (11:37 +0000)]
ssize_t is no C99 :-(
[r25884]
Christoph Mallon [Sun, 3 May 2009 10:49:30 +0000 (10:49 +0000)]
Remove unnecessary special case: sizeof expressions always have their type set, there is no need to look at the expression to determine the type.
[r25864]
Christoph Mallon [Sat, 2 May 2009 13:04:00 +0000 (13:04 +0000)]
When doing format string checking handle ?: as format expression, i.e. process the true- and false-expressions recursively.
[r25861]
Christoph Mallon [Sat, 2 May 2009 10:28:09 +0000 (10:28 +0000)]
Remove unnecessary special case: sizeof expressions always have their type set, there is no need to look at the expression to determine the type.
[r25860]
Michael Beck [Tue, 7 Apr 2009 22:32:08 +0000 (22:32 +0000)]
- fixed handling of variadic types/calling conventions
[r25819]
Michael Beck [Tue, 7 Apr 2009 14:36:48 +0000 (14:36 +0000)]
- 2009 patch
[r25818]
Michael Beck [Tue, 7 Apr 2009 14:24:13 +0000 (14:24 +0000)]
- implement -mrtd
- show the calling convention iff the default one is not __cdecl
- print the "bla bla free software bla" on --version
[r25817]
Michael Beck [Tue, 7 Apr 2009 12:22:09 +0000 (12:22 +0000)]
- fixed handling of calling convention in function_types_compatible()
[r25816]
Michael Beck [Tue, 7 Apr 2009 12:14:19 +0000 (12:14 +0000)]
- BugFix: use CC_DEFAULT as the default calling convention, NOT CDECL
[r25815]
Michael Beck [Tue, 7 Apr 2009 12:10:38 +0000 (12:10 +0000)]
- fixed function type calling convention
[r25814]
Michael Beck [Tue, 7 Apr 2009 12:04:51 +0000 (12:04 +0000)]
- BugFix: make_function_1_type_variadic() must hash the type itself, NOT
modify another one
- BugFix: r25810: only functions can have a prototype
- add a prototyped filed to function types
- K&R functions having a prototype are NOT variadic: this fixes
fehler166.c
[r25813]
Michael Beck [Tue, 7 Apr 2009 11:25:49 +0000 (11:25 +0000)]
- do NOT warn about "unnecessary static forward declaration" if the
definition is a K&R style function
[r25810]
Moritz Kroll [Mon, 6 Apr 2009 11:49:23 +0000 (11:49 +0000)]
Fixed assertion on C/gnu99/attributes5.c
[r25794]
Matthias Braun [Fri, 20 Mar 2009 05:32:50 +0000 (05:32 +0000)]
fix
[r25711]
Matthias Braun [Fri, 20 Mar 2009 05:21:46 +0000 (05:21 +0000)]
another bugfix
[r25709]
Matthias Braun [Fri, 20 Mar 2009 05:03:08 +0000 (05:03 +0000)]
bugfix
[r25706]
Matthias Braun [Fri, 20 Mar 2009 04:51:26 +0000 (04:51 +0000)]
another iteration of the struct layout code
[r25705]