libfirm
16 years ago- optimize Tuple nodes away before calling scalar replacement
Michael Beck [Mon, 7 Apr 2008 11:11:07 +0000 (11:11 +0000)]
- optimize Tuple nodes away before calling scalar replacement

[r19163]

16 years ago- Do not set phase_low here. currently ssa-construction (in scalar_replacement) wants...
Michael Beck [Mon, 7 Apr 2008 11:10:10 +0000 (11:10 +0000)]
- Do not set phase_low here. currently ssa-construction (in scalar_replacement) wants phase high ...

[r19162]

16 years ago- uses new SSA constructor interface, shortening and clering the code
Michael Beck [Mon, 7 Apr 2008 11:08:54 +0000 (11:08 +0000)]
- uses new SSA constructor interface, shortening and clering the code
- bugfix: really set current_ir_graph
- return non-zero if replacements were found

[r19161]

16 years agoupdated
Michael Beck [Sat, 5 Apr 2008 19:42:46 +0000 (19:42 +0000)]
updated

[r19147]

16 years agocompletely removed old verbosity based debug output
Michael Beck [Sat, 5 Apr 2008 19:19:42 +0000 (19:19 +0000)]
completely removed old verbosity based debug output

[r19146]

16 years agoswitch scalar replacement to new style debug output
Michael Beck [Sat, 5 Apr 2008 18:38:22 +0000 (18:38 +0000)]
switch scalar replacement to new style debug output

[r19145]

16 years agoafter inlining, the loop-info is inconsistent
Michael Beck [Sat, 5 Apr 2008 18:37:30 +0000 (18:37 +0000)]
after inlining, the loop-info is inconsistent

[r19144]

16 years ago- init function for inlining added
Michael Beck [Sat, 5 Apr 2008 18:18:00 +0000 (18:18 +0000)]
- init function for inlining added

[r19143]

16 years agoImproved inliner: rerun scalar_replacement() if addresses of local variables were...
Michael Beck [Fri, 4 Apr 2008 21:21:59 +0000 (21:21 +0000)]
Improved inliner: rerun scalar_replacement() if addresses of local variables were inlined

[r19139]

16 years agoImproved param weights: handle Id and Tuple nodes
Michael Beck [Fri, 4 Apr 2008 21:19:27 +0000 (21:19 +0000)]
Improved param weights: handle Id and Tuple nodes

[r19138]

16 years agoBugFix:
Michael Beck [Fri, 4 Apr 2008 19:01:48 +0000 (19:01 +0000)]
BugFix:
- node were allocated on the wrong obstack when creating a graph copy

[r19137]

16 years agoBugFix: store the weigths to the right location
Michael Beck [Fri, 4 Apr 2008 18:49:18 +0000 (18:49 +0000)]
BugFix: store the weigths to the right location

[r19135]

16 years agofixed obstack initialization, removed unused variable
Michael Beck [Fri, 4 Apr 2008 18:43:42 +0000 (18:43 +0000)]
fixed obstack initialization, removed unused variable

[r19134]

16 years ago- only local visible methods get the big benefice
Michael Beck [Fri, 4 Apr 2008 18:26:24 +0000 (18:26 +0000)]
- only local visible methods get the big benefice
- calculate benefice when transmitting addresses of local variables

[r19133]

16 years agofixed indentation, deleted useless comment
Michael Beck [Fri, 4 Apr 2008 17:36:32 +0000 (17:36 +0000)]
fixed indentation, deleted useless comment

[r19132]

16 years agocalculate benefice on real callee
Michael Beck [Fri, 4 Apr 2008 16:55:30 +0000 (16:55 +0000)]
calculate benefice on real callee

[r19130]

16 years agono bonus for recursive inlining
Michael Beck [Fri, 4 Apr 2008 16:45:04 +0000 (16:45 +0000)]
no bonus for recursive inlining

[r19129]

16 years agofixed warnings
Michael Beck [Fri, 4 Apr 2008 16:30:56 +0000 (16:30 +0000)]
fixed warnings

[r19128]

16 years agonew benefice based heuristic inliner added
Michael Beck [Fri, 4 Apr 2008 16:26:49 +0000 (16:26 +0000)]
new benefice based heuristic inliner added

[r19127]

16 years agoadd missing include <config.h>
Andreas Zwinkau [Fri, 4 Apr 2008 14:58:35 +0000 (14:58 +0000)]
add missing include <config.h>

[r19126]

16 years agoadd missing else for data in data section
Matthias Braun [Fri, 4 Apr 2008 13:47:21 +0000 (13:47 +0000)]
add missing else for data in data section

[r19125]

16 years ago- improced param weight calculation
Michael Beck [Fri, 4 Apr 2008 13:31:04 +0000 (13:31 +0000)]
- improced param weight calculation
- indentation changed

[r19122]

16 years agocosmetic
Michael Beck [Fri, 4 Apr 2008 13:30:14 +0000 (13:30 +0000)]
cosmetic

[r19121]

16 years agoMoved inlining and DCE (which is implemented like inlinung) into opt_inline.c
Michael Beck [Fri, 4 Apr 2008 11:46:22 +0000 (11:46 +0000)]
Moved inlining and DCE (which is implemented like inlinung) into opt_inline.c

[r19119]

16 years ago- code cleanup
Michael Beck [Fri, 4 Apr 2008 11:32:32 +0000 (11:32 +0000)]
- code cleanup
- moved code placement and critical edges removement to own files

[r19117]

16 years agoC99 features removed
Michael Beck [Fri, 4 Apr 2008 11:10:48 +0000 (11:10 +0000)]
C99 features removed

[r19113]

16 years agohandle the macro block properly
Michael Beck [Fri, 4 Apr 2008 11:10:17 +0000 (11:10 +0000)]
handle the macro block properly

[r19112]

16 years agosome cosmetic
Michael Beck [Fri, 4 Apr 2008 11:09:46 +0000 (11:09 +0000)]
some cosmetic

[r19111]

16 years agostarted implementing Mac Style PIC
Matthias Braun [Thu, 3 Apr 2008 21:08:42 +0000 (21:08 +0000)]
started implementing Mac Style PIC

[r19102]

16 years agomake mangle3 public
Matthias Braun [Thu, 3 Apr 2008 21:08:21 +0000 (21:08 +0000)]
make mangle3 public

[r19101]

16 years agoshouldn't be here anymore
Matthias Braun [Thu, 3 Apr 2008 19:09:46 +0000 (19:09 +0000)]
shouldn't be here anymore

[r19100]

16 years agoshouldn't be here anymore
Matthias Braun [Thu, 3 Apr 2008 19:09:29 +0000 (19:09 +0000)]
shouldn't be here anymore

[r19099]

16 years agoremove img_section stuff and create a new constructors_type
Matthias Braun [Thu, 3 Apr 2008 18:56:41 +0000 (18:56 +0000)]
remove img_section stuff and create a new constructors_type

[r19098]

16 years agofix warnings
Matthias Braun [Thu, 3 Apr 2008 18:54:29 +0000 (18:54 +0000)]
fix warnings

[r19097]

16 years agoimproved Cmp(x, 0) optimizations
Michael Beck [Thu, 3 Apr 2008 13:12:57 +0000 (13:12 +0000)]
improved Cmp(x, 0) optimizations

[r19089]

16 years agofix warning
Matthias Braun [Wed, 2 Apr 2008 22:02:41 +0000 (22:02 +0000)]
fix warning

[r19079]

16 years agohelper function for creating unique identifiers
Matthias Braun [Wed, 2 Apr 2008 22:01:24 +0000 (22:01 +0000)]
helper function for creating unique identifiers

[r19078]

16 years agoswitch off usage of global_pointer again: globals are only SymConsts. -fpic will...
Michael Beck [Wed, 2 Apr 2008 18:34:01 +0000 (18:34 +0000)]
switch off usage of global_pointer again: globals are only SymConsts. -fpic will be completely handled by the beabi

[r19073]

16 years agoSkip Projs in cost calculation.
Christoph Mallon [Wed, 2 Apr 2008 14:34:23 +0000 (14:34 +0000)]
Skip Projs in cost calculation.

[r19071]

16 years agoadd list_head typedef
Michael Beck [Wed, 2 Apr 2008 14:09:23 +0000 (14:09 +0000)]
add list_head typedef

[r19067]

16 years agoused new is_Global() and get_Global_entity() instead of SymConst calls
Michael Beck [Wed, 2 Apr 2008 14:08:02 +0000 (14:08 +0000)]
used new is_Global() and get_Global_entity() instead of SymConst calls

[r19066]

16 years agoRemove dead code.
Christoph Mallon [Wed, 2 Apr 2008 14:05:47 +0000 (14:05 +0000)]
Remove dead code.

[r19065]

16 years agoadd is_Global() and get_Global_ent()
Michael Beck [Wed, 2 Apr 2008 13:51:39 +0000 (13:51 +0000)]
add is_Global() and get_Global_ent()

[r19064]

16 years agoC99 features removed
Michael Beck [Wed, 2 Apr 2008 09:53:48 +0000 (09:53 +0000)]
C99 features removed

[r19060]

16 years agosome comments added and fixed
Michael Beck [Wed, 2 Apr 2008 09:53:14 +0000 (09:53 +0000)]
some comments added and fixed

[r19059]

16 years agoC99 features removed
Michael Beck [Wed, 2 Apr 2008 09:51:15 +0000 (09:51 +0000)]
C99 features removed

[r19058]

16 years agofixed Win32 build
Michael Beck [Wed, 2 Apr 2008 09:50:40 +0000 (09:50 +0000)]
fixed Win32 build

[r19057]

16 years agoskip Bitfield Sels when doing memory disambiguation
Michael Beck [Wed, 2 Apr 2008 09:49:17 +0000 (09:49 +0000)]
skip Bitfield Sels when doing memory disambiguation

[r19056]

16 years agoMake Ld/St parallelisation work (but seems to be broken with bit fields).
Christoph Mallon [Wed, 2 Apr 2008 09:06:53 +0000 (09:06 +0000)]
Make Ld/St parallelisation work (but seems to be broken with bit fields).

[r19052]

16 years agoMove grgen stuff to seprarate directory (will adapt firm build process in a few moments)
Matthias Braun [Tue, 1 Apr 2008 12:26:42 +0000 (12:26 +0000)]
Move grgen stuff to seprarate directory (will adapt firm build process in a few moments)

[r19038]

16 years agoinserted pbqp transformation step
Andreas Zwinkau [Tue, 1 Apr 2008 12:04:51 +0000 (12:04 +0000)]
inserted pbqp transformation step

[r19037]

16 years agomoved firmext code into the backend dir
Andreas Zwinkau [Tue, 1 Apr 2008 11:40:24 +0000 (11:40 +0000)]
moved firmext code into the backend dir

[r19036]

16 years agoAllow exchanging nodes with oparity_dynamic and delete their in-array.
Christoph Mallon [Tue, 1 Apr 2008 10:41:23 +0000 (10:41 +0000)]
Allow exchanging nodes with oparity_dynamic and delete their in-array.

[r19035]

16 years agoImprove equivalent_node_Sync() and transform_node_Sync() so they remove Bad and dupli...
Christoph Mallon [Tue, 1 Apr 2008 10:38:08 +0000 (10:38 +0000)]
Improve equivalent_node_Sync() and transform_node_Sync() so they remove Bad and duplicate predecessors.

[r19034]

16 years agoAdd del_Sync_n().
Christoph Mallon [Tue, 1 Apr 2008 10:32:37 +0000 (10:32 +0000)]
Add del_Sync_n().

[r19033]

16 years agorenamed functions to clarify its usage
Michael Beck [Tue, 1 Apr 2008 08:49:03 +0000 (08:49 +0000)]
renamed functions to clarify its usage

[r19032]

16 years agoMove Syncs out of the way before removing a block.
Christoph Mallon [Mon, 31 Mar 2008 13:47:15 +0000 (13:47 +0000)]
Move Syncs out of the way before removing a block.

[r18307]

16 years agoLet the Syncs float!.
Christoph Mallon [Mon, 31 Mar 2008 13:46:17 +0000 (13:46 +0000)]
Let the Syncs float!.

[r18306]

16 years agoDo not dump block edges for blocks.
Christoph Mallon [Mon, 31 Mar 2008 13:14:18 +0000 (13:14 +0000)]
Do not dump block edges for blocks.

[r18305]

16 years agofixing grgen dumper to autogenerate file names
Andreas Zwinkau [Mon, 31 Mar 2008 12:10:21 +0000 (12:10 +0000)]
fixing grgen dumper to autogenerate file names

[r18302]

16 years agosome mach-o related fixes
Matthias Braun [Sun, 30 Mar 2008 18:39:26 +0000 (18:39 +0000)]
some mach-o related fixes

[r18298]

16 years agomore elegant solution for tls dumping
Matthias Braun [Sun, 30 Mar 2008 16:36:00 +0000 (16:36 +0000)]
more elegant solution for tls dumping

[r18297]

16 years agosimplify begnuas.c by not sorting entities into sections
Matthias Braun [Sun, 30 Mar 2008 16:28:47 +0000 (16:28 +0000)]
simplify begnuas.c by not sorting entities into sections

[r18296]

16 years agocleanups
Matthias Braun [Sun, 30 Mar 2008 14:26:48 +0000 (14:26 +0000)]
cleanups

[r18294]

16 years ago- cleanup and rewrite some of the dbgout stuff (no passing around of handles anymore)
Matthias Braun [Sun, 30 Mar 2008 14:24:24 +0000 (14:24 +0000)]
- cleanup and rewrite some of the dbgout stuff (no passing around of handles anymore)
- port stabs debug output to beemitter
- support Mach-O in begnuas
- cleanups

[r18292]

16 years agoremove pointless continues
Matthias Braun [Sun, 30 Mar 2008 12:59:12 +0000 (12:59 +0000)]
remove pointless continues

[r18290]

16 years agoimprove debug message
Matthias Braun [Sun, 30 Mar 2008 12:58:32 +0000 (12:58 +0000)]
improve debug message

[r18289]

16 years agoreference are integermode too for the backend
Matthias Braun [Sun, 30 Mar 2008 12:58:15 +0000 (12:58 +0000)]
reference are integermode too for the backend

[r18288]

16 years agocleanup/fix arm callflags
Matthias Braun [Sun, 30 Mar 2008 12:58:01 +0000 (12:58 +0000)]
cleanup/fix arm callflags

[r18287]

16 years agosimplify testcase
Matthias Braun [Sun, 30 Mar 2008 12:57:12 +0000 (12:57 +0000)]
simplify testcase

[r18286]

16 years agoversion bump
Matthias Braun [Sun, 30 Mar 2008 12:56:31 +0000 (12:56 +0000)]
version bump

[r18285]

16 years agoremove duplicated obstack.h declare obstack_printf
Matthias Braun [Sun, 30 Mar 2008 12:18:20 +0000 (12:18 +0000)]
remove duplicated obstack.h declare obstack_printf

[r18282]

16 years agoadd missing include
Matthias Braun [Sun, 30 Mar 2008 09:47:19 +0000 (09:47 +0000)]
add missing include

[r18281]

16 years agodon't rely on obstack hacks
Matthias Braun [Sun, 30 Mar 2008 09:46:33 +0000 (09:46 +0000)]
don't rely on obstack hacks

[r18280]

16 years agofix check for posix realtime extensions
Matthias Braun [Sat, 29 Mar 2008 23:00:33 +0000 (23:00 +0000)]
fix check for posix realtime extensions

[r18279]

16 years agofix check for posix realtime extensions
Matthias Braun [Sat, 29 Mar 2008 22:59:15 +0000 (22:59 +0000)]
fix check for posix realtime extensions

[r18278]

16 years agotest commit
Matthias Braun [Sat, 22 Mar 2008 12:59:40 +0000 (12:59 +0000)]
test commit

[r18213]

16 years agotestcase for new Confirm code
Michael Beck [Thu, 20 Mar 2008 02:30:18 +0000 (02:30 +0000)]
testcase for new Confirm code

[r18198]

16 years agosmall improvements:
Michael Beck [Thu, 20 Mar 2008 02:27:20 +0000 (02:27 +0000)]
small improvements:
- value_not_zero(): safely skip Minus and Abs nodes
- classify_value_sign(): handle Minus nodes

[r18197]

16 years agofixed comment
Michael Beck [Thu, 20 Mar 2008 02:25:32 +0000 (02:25 +0000)]
fixed comment

[r18196]

16 years agowhen propagating constants in the x == Const case, propagate them into users of x...
Michael Beck [Thu, 20 Mar 2008 02:24:42 +0000 (02:24 +0000)]
when propagating constants in the x == Const case, propagate them into users of x IFF users of users are in the effective block, see confirm.c example

[r18195]

16 years agois_Abs() added
Michael Beck [Thu, 20 Mar 2008 02:22:11 +0000 (02:22 +0000)]
is_Abs() added

[r18194]

16 years agoBugFix: scalar replacement should not remove volatile Loads/Stores
Michael Beck [Wed, 19 Mar 2008 17:35:54 +0000 (17:35 +0000)]
BugFix: scalar replacement should not remove volatile Loads/Stores

[r18184]

16 years agosetjump test
Matthias Braun [Wed, 19 Mar 2008 10:59:50 +0000 (10:59 +0000)]
setjump test

[r18173]

16 years ago- Fix comments
Christoph Mallon [Mon, 17 Mar 2008 12:58:26 +0000 (12:58 +0000)]
- Fix comments
- Remove stale assert
- Add Psi(t == f, t, f) -> f and Psi(t != f, t, f) -> t (was only implemented for f == 0)

[r18142]

16 years agoBelady2 fixes
Sebastian Hack [Sun, 16 Mar 2008 14:42:33 +0000 (14:42 +0000)]
Belady2 fixes

[r18135]

16 years agoAdded functions to set a range of bits to 1 or 0
Sebastian Hack [Sun, 16 Mar 2008 14:42:01 +0000 (14:42 +0000)]
Added functions to set a range of bits to 1 or 0

[r18134]

16 years agouse .word instead of .value, older gas don't support it
Michael Beck [Wed, 12 Mar 2008 00:08:57 +0000 (00:08 +0000)]
use .word instead of .value, older gas don't support it

[r18093]

16 years agoneed suffixes for mvf, mnf
Michael Beck [Wed, 12 Mar 2008 00:01:52 +0000 (00:01 +0000)]
need suffixes for mvf, mnf

[r18092]

16 years agoset the isa right
Michael Beck [Tue, 11 Mar 2008 23:53:54 +0000 (23:53 +0000)]
set the isa right

[r18091]

16 years agoinstructions with immediates are not commutative
Michael Beck [Tue, 11 Mar 2008 23:53:31 +0000 (23:53 +0000)]
instructions with immediates are not commutative

[r18090]

16 years agorefactored immediates:
Michael Beck [Tue, 11 Mar 2008 23:37:15 +0000 (23:37 +0000)]
refactored immediates:
 - now stored as long values (should be enough)
 - floating point values are emitted right

[r18089]

16 years agorefactored immediates:
Michael Beck [Tue, 11 Mar 2008 23:35:22 +0000 (23:35 +0000)]
refactored immediates:
 - now stored as long values (should be enough)
 - floating point values are emitted right

[r18088]

16 years agofixed CondJmp attributes
Michael Beck [Tue, 11 Mar 2008 20:11:26 +0000 (20:11 +0000)]
fixed CondJmp attributes

[r18081]

16 years agoset the global isa object
Michael Beck [Tue, 11 Mar 2008 20:10:33 +0000 (20:10 +0000)]
set the global isa object

[r18080]

16 years agouse assure_irg_outs()
Michael Beck [Tue, 11 Mar 2008 18:27:38 +0000 (18:27 +0000)]
use assure_irg_outs()

[r18074]

16 years agofix out flags for mode_T nodes
Matthias Braun [Tue, 11 Mar 2008 17:13:58 +0000 (17:13 +0000)]
fix out flags for mode_T nodes

[r18072]

16 years agoupdated be_AddSP semantics
Michael Beck [Tue, 11 Mar 2008 15:55:16 +0000 (15:55 +0000)]
updated be_AddSP semantics

[r18070]

16 years agowe do need stack:I|S...
Matthias Braun [Tue, 11 Mar 2008 15:34:50 +0000 (15:34 +0000)]
we do need stack:I|S...

[r18069]

16 years agostack:S is enough no need to specify I (explicitely)
Matthias Braun [Tue, 11 Mar 2008 15:25:35 +0000 (15:25 +0000)]
stack:S is enough no need to specify I (explicitely)

[r18068]