libfirm
18 years agochecked for dead blocks before getting dom depht.
Daniel Grund [Thu, 1 Sep 2005 15:35:34 +0000 (15:35 +0000)]
checked for dead blocks before getting dom depht.

[r6567]

18 years ago*** empty log message ***
Daniel Grund [Thu, 1 Sep 2005 15:13:55 +0000 (15:13 +0000)]
*** empty log message ***

[r6566]

18 years agoget_irn_generic_attr() added, needed for backends that use it's own
Michael Beck [Thu, 1 Sep 2005 15:13:47 +0000 (15:13 +0000)]
get_irn_generic_attr() added, needed for backends that use it's own
opcodes

[r6565]

18 years agoused PTR_TO_INT() and INT_TO_PTR() macros for 64bit safety
Michael Beck [Thu, 1 Sep 2005 09:36:27 +0000 (09:36 +0000)]
used PTR_TO_INT() and INT_TO_PTR() macros for 64bit safety

[r6561]

18 years agoremoved wrong const
Michael Beck [Thu, 1 Sep 2005 09:35:10 +0000 (09:35 +0000)]
removed wrong const
some comments added

[r6560]

18 years agoed PTR_TO_INT() and INT_TO_PTR() macros for 64bit safety
Michael Beck [Thu, 1 Sep 2005 09:32:38 +0000 (09:32 +0000)]
ed PTR_TO_INT() and INT_TO_PTR() macros for 64bit safety

[r6559]

18 years agoused HASH_PTR() now
Michael Beck [Thu, 1 Sep 2005 09:32:06 +0000 (09:32 +0000)]
used HASH_PTR() now

[r6558]

18 years agorenamed function (clashed with buildin)
Michael Beck [Thu, 1 Sep 2005 09:31:07 +0000 (09:31 +0000)]
renamed function (clashed with buildin)

[r6557]

18 years agobitfields of size 1 should be unsigned if 1 is stored there
Michael Beck [Thu, 1 Sep 2005 09:27:22 +0000 (09:27 +0000)]
bitfields of size 1 should be unsigned if 1 is stored there

[r6556]

18 years agoused HASHPTR() now
Michael Beck [Thu, 1 Sep 2005 09:17:25 +0000 (09:17 +0000)]
used HASHPTR() now

[r6555]

18 years agoRemoved dominance computation if irg_vrfy
Sebastian Hack [Wed, 31 Aug 2005 17:39:47 +0000 (17:39 +0000)]
Removed dominance computation if irg_vrfy

[r6554]

18 years agonewer version that can handle x86_64
Michael Beck [Wed, 31 Aug 2005 17:26:07 +0000 (17:26 +0000)]
newer version that can handle x86_64

[r6553]

18 years agoirtools.h is not installed anymore
Michael Beck [Wed, 31 Aug 2005 17:24:18 +0000 (17:24 +0000)]
irtools.h is not installed anymore

[r6552]

18 years agoINT_TO_PTR() and PTR_TO_INT() macros added
Michael Beck [Wed, 31 Aug 2005 17:24:03 +0000 (17:24 +0000)]
INT_TO_PTR() and PTR_TO_INT() macros added

[r6551]

18 years agoAdded private data facility for edges
Sebastian Hack [Wed, 31 Aug 2005 13:47:57 +0000 (13:47 +0000)]
Added private data facility for edges

[r6546]

18 years agoremoved warnings: single bit type should be unsigned if 0 and 1 should be stored...
Michael Beck [Wed, 31 Aug 2005 09:38:21 +0000 (09:38 +0000)]
removed warnings: single bit type should be unsigned if 0 and 1 should be stored there

[r6543]

18 years agoFixes for GCC 4.0: void * and pointer to function are not compatible anymore
Michael Beck [Wed, 31 Aug 2005 09:36:58 +0000 (09:36 +0000)]
Fixes for GCC 4.0: void * and pointer to function are not compatible anymore

[r6542]

18 years agofixes for gcc 4.0: void * and pointer to function are not compatible anymore
Michael Beck [Wed, 31 Aug 2005 09:36:08 +0000 (09:36 +0000)]
fixes for gcc 4.0: void * and pointer to function are not compatible anymore

[r6541]

18 years agoFixes for gcc4.0: void * and pointer to functions are NOT compatible anymore
Michael Beck [Wed, 31 Aug 2005 09:34:57 +0000 (09:34 +0000)]
Fixes for gcc4.0: void * and pointer to functions are NOT compatible anymore

[r6540]

18 years agoC99 feature removed
Michael Beck [Wed, 31 Aug 2005 09:25:48 +0000 (09:25 +0000)]
C99 feature removed

[r6539]

18 years agoFixed a bug
Sebastian Hack [Mon, 29 Aug 2005 16:41:03 +0000 (16:41 +0000)]
Fixed a bug

[r6536]

18 years agoAdded little more debugging output
Sebastian Hack [Mon, 29 Aug 2005 16:37:05 +0000 (16:37 +0000)]
Added little more debugging output

[r6535]

18 years agoA little better debug output
Sebastian Hack [Mon, 29 Aug 2005 16:32:37 +0000 (16:32 +0000)]
A little better debug output

[r6534]

18 years ago*** empty log message ***
Daniel Grund [Mon, 29 Aug 2005 16:28:48 +0000 (16:28 +0000)]
*** empty log message ***

[r6533]

18 years agoFixed a bug again
Sebastian Hack [Mon, 29 Aug 2005 14:53:33 +0000 (14:53 +0000)]
Fixed a bug again

[r6531]

18 years agoFixed a bug
Sebastian Hack [Mon, 29 Aug 2005 14:37:23 +0000 (14:37 +0000)]
Fixed a bug

[r6527]

18 years agoAdded dominance check
Sebastian Hack [Mon, 29 Aug 2005 14:30:34 +0000 (14:30 +0000)]
Added dominance check

[r6526]

18 years agoAdded mod and mux
Daniel Grund [Mon, 29 Aug 2005 11:18:33 +0000 (11:18 +0000)]
Added mod and mux

[r6524]

18 years agoAdded get_irn_edge()
Sebastian Hack [Mon, 29 Aug 2005 10:00:42 +0000 (10:00 +0000)]
Added get_irn_edge()

[r6523]

18 years agoremoved wrong INLINE spec
Michael Beck [Fri, 26 Aug 2005 15:01:15 +0000 (15:01 +0000)]
removed wrong INLINE spec

[r6510]

18 years agoBugFix:
Michael Beck [Fri, 26 Aug 2005 12:17:55 +0000 (12:17 +0000)]
BugFix:
  new_rd_*() functions that call optimize_node() did not
  work if current_ir_graph was not set!
  So, the current_ir_graph is now set and restored.
  Moreover, to speed up the default new_*() and new_d_*()
  cases, new_bd_*() functions have been introduced which use
  current_ir_graph, but not current_block and replace
  the use of _rd_ functions in the comfortable i/f

[r6509]

18 years agoBugFix: in place_late, do NOT count keep-alive edges as consuming ones,
Michael Beck [Fri, 26 Aug 2005 12:02:46 +0000 (12:02 +0000)]
BugFix: in place_late, do NOT count keep-alive edges as consuming ones,
or nodes might be placed wrong!

[r6508]

18 years agoBugFix: predecessors of End nodes might be Blocks
Michael Beck [Thu, 25 Aug 2005 12:42:12 +0000 (12:42 +0000)]
BugFix: predecessors of End nodes might be Blocks

[r6495]

18 years agoBugFix:
Michael Beck [Wed, 24 Aug 2005 13:59:29 +0000 (13:59 +0000)]
BugFix:
  When entities are freed, type information is used, so we cannot
  free one type before all entities are freed

[r6485]

18 years agodel_pset in now called correctly
Daniel Grund [Tue, 23 Aug 2005 20:27:13 +0000 (20:27 +0000)]
del_pset in now called correctly

[r6482]

18 years agoSet array border to 0, needed for fixed add_compound_ent_value(), one node more
Michael Beck [Tue, 23 Aug 2005 18:14:02 +0000 (18:14 +0000)]
Set array border to 0, needed for fixed add_compound_ent_value(), one node more

[r6481]

18 years agoSet array border to 0, needed for fixed add_compound_ent_value()
Michael Beck [Tue, 23 Aug 2005 18:13:51 +0000 (18:13 +0000)]
Set array border to 0, needed for fixed add_compound_ent_value()

[r6480]

18 years agoBugfixes
Daniel Grund [Tue, 23 Aug 2005 18:03:21 +0000 (18:03 +0000)]
Bugfixes

[r6479]

18 years agoBugFix: add_compound_ent_value() did not generate right indexes for compound pathes
Michael Beck [Tue, 23 Aug 2005 16:32:01 +0000 (16:32 +0000)]
BugFix: add_compound_ent_value() did not generate right indexes for compound pathes

[r6478]

18 years agoallow character values to be converted into long
Michael Beck [Tue, 23 Aug 2005 13:22:36 +0000 (13:22 +0000)]
allow character values to be converted into long
some cleanup

[r6476]

18 years agomoke mode sort of character modes irms_twos_complement
Michael Beck [Tue, 23 Aug 2005 13:20:35 +0000 (13:20 +0000)]
moke mode sort of character modes irms_twos_complement

[r6475]

18 years agoadded %P for compound-graph-paths
Michael Beck [Mon, 22 Aug 2005 16:43:00 +0000 (16:43 +0000)]
added %P for compound-graph-paths

[r6470]

18 years agoAdded reflection for mulh
Daniel Grund [Thu, 18 Aug 2005 16:58:50 +0000 (16:58 +0000)]
Added reflection for mulh

[r6463]

18 years agonode function copy_irn extracted from firm_copy_node()
Michael Beck [Thu, 18 Aug 2005 12:57:45 +0000 (12:57 +0000)]
node function copy_irn extracted from firm_copy_node()

[r6461]

18 years agoremoved warning
Michael Beck [Thu, 18 Aug 2005 11:49:28 +0000 (11:49 +0000)]
removed warning

[r6458]

18 years agoenabled type_visited() function again
Michael Beck [Thu, 18 Aug 2005 11:47:58 +0000 (11:47 +0000)]
enabled type_visited() function again

[r6457]

18 years agoremoved definition of old type_visited flag
Michael Beck [Thu, 18 Aug 2005 11:44:25 +0000 (11:44 +0000)]
removed definition of old type_visited flag
restored type_visited() function declaration

[r6456]

18 years agoresreuctures compound path to make allocation/deallocation simplier
Michael Beck [Wed, 17 Aug 2005 14:13:45 +0000 (14:13 +0000)]
resreuctures compound path to make allocation/deallocation simplier

[r6455]

18 years agofixed copy'n'paste error
Michael Beck [Tue, 16 Aug 2005 13:00:55 +0000 (13:00 +0000)]
fixed copy'n'paste error
nd fixed

[r6439]

18 years agochanged option %g to %G (floating point uses %g)
Michael Beck [Tue, 16 Aug 2005 12:34:01 +0000 (12:34 +0000)]
changed option %g to %G (floating point uses %g)

[r6438]

18 years agoChanged %g to %G to enable float dumping
Sebastian Hack [Tue, 16 Aug 2005 12:23:34 +0000 (12:23 +0000)]
Changed %g to %G to enable float dumping

[r6436]

18 years agoAdded Minus for Ints
Sebastian Hack [Tue, 16 Aug 2005 12:12:38 +0000 (12:12 +0000)]
Added Minus for Ints

[r6435]

18 years agocreate_abstraction() now returns an error code if the file could not
Michael Beck [Tue, 16 Aug 2005 10:18:35 +0000 (10:18 +0000)]
create_abstraction() now returns an error code if the file could not
be opened.

[r6434]

18 years agobugfix
Götz Lindenmaier [Sun, 14 Aug 2005 11:22:51 +0000 (11:22 +0000)]
bugfix

[r6419]

18 years agonew_rd_Unknown sets the current_ir_graph right.
FIRM Projekt Account [Fri, 12 Aug 2005 16:31:20 +0000 (16:31 +0000)]
new_rd_Unknown sets the current_ir_graph right.
This fixes crashes in jack, but a general solution is needed: all
rd functions calling optimize_node must do that!

Restructure IR construction!

[r6418]

18 years agoinstall scalar_replace.h
Michael Beck [Fri, 12 Aug 2005 13:11:02 +0000 (13:11 +0000)]
install scalar_replace.h

[r6415]

18 years agochanged new_type_pointer() to new semantic
Michael Beck [Fri, 12 Aug 2005 10:28:51 +0000 (10:28 +0000)]
changed new_type_pointer() to new semantic

[r6413]

18 years agoUnknown nodes are optimized (CSE) now.
Michael Beck [Fri, 12 Aug 2005 09:42:23 +0000 (09:42 +0000)]
Unknown nodes are optimized (CSE) now.
This was not done earlier, why?

[r6412]

18 years agochanged back to old visit order
Michael Beck [Fri, 12 Aug 2005 09:41:48 +0000 (09:41 +0000)]
changed back to old visit order
removed all calles form for () headers

[r6411]

18 years agocurrent_ir_graph must be set EVEN iff the raw interface is used!!!
Michael Beck [Fri, 12 Aug 2005 09:40:28 +0000 (09:40 +0000)]
current_ir_graph must be set EVEN iff the raw interface is used!!!

[r6410]

18 years agonew_d_type_array now allocates only ONE Uknown node for the array
Michael Beck [Fri, 12 Aug 2005 09:39:20 +0000 (09:39 +0000)]
new_d_type_array now allocates only ONE Uknown node for the array
bounds, and Unknown nodes are optimized now, so lesser nodes

[r6409]

18 years agotypo fixed
Michael Beck [Thu, 11 Aug 2005 16:40:38 +0000 (16:40 +0000)]
typo fixed

[r6408]

18 years agoused now always the visited functions, no more access to type_visited flag directly
Michael Beck [Thu, 11 Aug 2005 16:39:57 +0000 (16:39 +0000)]
used now always the visited functions, no more access  to type_visited flag directly

[r6407]

18 years agorenamed type_visited to firm_type_visited
Michael Beck [Thu, 11 Aug 2005 16:38:32 +0000 (16:38 +0000)]
renamed type_visited to firm_type_visited
removed unused fields from union attributes
renamed new_type_pointer_mode() to new_type_pointer(), now always a
mode argument is needed

[r6406]

18 years agoBugfix: unset debug info is now initialized with NULL
Michael Beck [Thu, 11 Aug 2005 16:36:59 +0000 (16:36 +0000)]
Bugfix: unset debug info is now initialized with NULL
renamed new_type_pointer_mode() to new_type_pointer(), it's now
always necessary to give a mode when construction a pointer type
renamed type_visited to firm_type_visited and removed from public i/f
removed #if 0 code

[r6405]

18 years agorenamed global flag type_visited into firm_type_visited
Michael Beck [Thu, 11 Aug 2005 16:33:49 +0000 (16:33 +0000)]
renamed global flag type_visited into firm_type_visited

[r6404]

18 years agorenamed new_type_pointer_mode() into new_type_pointer()
Michael Beck [Thu, 11 Aug 2005 16:32:41 +0000 (16:32 +0000)]
renamed new_type_pointer_mode() into new_type_pointer()

[r6403]

18 years agoused Block_not_block_visited() instead of explicit code
Michael Beck [Thu, 11 Aug 2005 16:31:15 +0000 (16:31 +0000)]
used Block_not_block_visited() instead of explicit code

[r6402]

18 years agorenamed init_type() to firm_init_type() with additional parameter
Michael Beck [Thu, 11 Aug 2005 16:30:25 +0000 (16:30 +0000)]
renamed init_type() to firm_init_type() with additional parameter

[r6401]

18 years agofixed doxygen comments
Michael Beck [Thu, 11 Aug 2005 12:53:06 +0000 (12:53 +0000)]
fixed doxygen comments

[r6397]

18 years agofixed typos
Michael Beck [Thu, 11 Aug 2005 12:52:27 +0000 (12:52 +0000)]
fixed typos
more doxygen comments
removed useless #define

[r6396]

18 years agoscalar_replace.h added
Michael Beck [Thu, 11 Aug 2005 12:50:09 +0000 (12:50 +0000)]
scalar_replace.h added

[r6395]

18 years agoremoved C99 feature
Michael Beck [Thu, 11 Aug 2005 12:48:46 +0000 (12:48 +0000)]
removed C99 feature

[r6394]

18 years agoskip_Tuple() now skips chains of tuple, which can arise
Michael Beck [Thu, 11 Aug 2005 12:47:24 +0000 (12:47 +0000)]
skip_Tuple() now skips chains of tuple, which can arise
after inlining small functions

[r6393]

18 years agoFixed skip_Tuple(), so no need for additional loop
Michael Beck [Thu, 11 Aug 2005 12:46:24 +0000 (12:46 +0000)]
Fixed skip_Tuple(), so no need for additional loop

[r6392]

18 years agorenamed function
Michael Beck [Wed, 10 Aug 2005 15:36:23 +0000 (15:36 +0000)]
renamed function
reimplemented removement of Loads/Stores, fixed removement of to many Load/Stores
did not use loop_info/dominance info anymore

[r6389]

18 years agoBugfix in mask highest.
Daniel Grund [Wed, 10 Aug 2005 13:06:23 +0000 (13:06 +0000)]
Bugfix in mask highest.

[r6387]

18 years agoadd missing SET_LEVEL_0
Michael Beck [Wed, 10 Aug 2005 11:11:21 +0000 (11:11 +0000)]
add missing SET_LEVEL_0

[r6386]

18 years agoAdded option to enable/disable scalar replacement
Michael Beck [Wed, 10 Aug 2005 11:04:15 +0000 (11:04 +0000)]
Added option to enable/disable scalar replacement

[r6385]

18 years agoCompletely rewrote phiclass analysis.
Daniel Grund [Tue, 9 Aug 2005 17:47:49 +0000 (17:47 +0000)]
Completely rewrote phiclass analysis.

[r6384]

18 years agoFixed skip_Tuple(): Nested tuple were not optimized
Michael Beck [Tue, 9 Aug 2005 14:03:33 +0000 (14:03 +0000)]
Fixed skip_Tuple(): Nested tuple were not optimized

[r6383]

18 years agomore files added, removed be from build
Michael Beck [Mon, 8 Aug 2005 12:06:30 +0000 (12:06 +0000)]
more files added, removed be from build

[r6382]

18 years agosplit init_ir_prog() so part 1 can be called before mode initialization
Michael Beck [Wed, 3 Aug 2005 12:55:46 +0000 (12:55 +0000)]
split init_ir_prog() so part 1 can be called before mode initialization

[r6379]

18 years agosplit init_irprog() so first part ban be called before
Michael Beck [Wed, 3 Aug 2005 12:55:05 +0000 (12:55 +0000)]
split init_irprog() so first part ban be called before
mode initialization

[r6378]

18 years agoconfig.h added and other includes fixed
Michael Beck [Wed, 3 Aug 2005 12:23:35 +0000 (12:23 +0000)]
config.h added and other includes fixed

[r6375]

18 years agoused new mode_P_data instead of mode_P_mach
Michael Beck [Wed, 3 Aug 2005 12:14:50 +0000 (12:14 +0000)]
used new mode_P_data instead of mode_P_mach

[r6374]

18 years agoed new mode_P_data & mode_P_code instead of mode_P_mach
Michael Beck [Wed, 3 Aug 2005 12:13:49 +0000 (12:13 +0000)]
ed new mode_P_data & mode_P_code instead of mode_P_mach

[r6373]

18 years agobugfixed: need a cast
Michael Beck [Wed, 3 Aug 2005 12:08:19 +0000 (12:08 +0000)]
bugfixed: need a cast

[r6372]

18 years agotypo fixed
Michael Beck [Wed, 3 Aug 2005 12:05:22 +0000 (12:05 +0000)]
typo fixed

[r6371]

18 years agochanged mode_P_mach into new mode_P_data
Michael Beck [Wed, 3 Aug 2005 12:04:24 +0000 (12:04 +0000)]
changed mode_P_mach into new mode_P_data

[r6370]

18 years agochagned mode_P_mach into new mode_P_data
Michael Beck [Wed, 3 Aug 2005 12:00:53 +0000 (12:00 +0000)]
chagned mode_P_mach into new mode_P_data

[r6369]

18 years agochanged mode_P_mach into mode_P_code
Michael Beck [Wed, 3 Aug 2005 11:59:58 +0000 (11:59 +0000)]
changed mode_P_mach into mode_P_code

[r6368]

18 years agouses new irp mode list
Michael Beck [Wed, 3 Aug 2005 11:56:49 +0000 (11:56 +0000)]
uses new irp mode list

[r6367]

18 years agoremoved warnings
Michael Beck [Wed, 3 Aug 2005 11:49:11 +0000 (11:49 +0000)]
removed warnings

[r6366]

18 years agosplit modeP_mach into modeP_code and modeP_data
Michael Beck [Wed, 3 Aug 2005 11:48:28 +0000 (11:48 +0000)]
split modeP_mach into modeP_code and modeP_data
all modes are now kept in a list in the irp

[r6365]

18 years agoadd a list of modes in the irp
Michael Beck [Wed, 3 Aug 2005 11:42:51 +0000 (11:42 +0000)]
add a list of modes in the irp
renamed all __*() functions to _*()
fixed 2 small memory leaks

[r6364]

18 years agoadd magic for an program representation (irp)
Michael Beck [Wed, 3 Aug 2005 11:39:53 +0000 (11:39 +0000)]
add magic for an program representation (irp)

[r6363]

18 years agoFixed dumping with types: local nodes must be deactivated
Michael Beck [Mon, 1 Aug 2005 10:23:10 +0000 (10:23 +0000)]
Fixed dumping with types: local nodes must be deactivated

[r6339]

18 years agoFixed dumping with types by deactivation local nodes
Michael Beck [Mon, 1 Aug 2005 10:22:16 +0000 (10:22 +0000)]
Fixed dumping with types by deactivation local nodes
the visit order of types/entities changed  bit because of new type walker visiting
initialization values

[r6338]

18 years agoadded flag for frame types
Michael Beck [Mon, 1 Aug 2005 09:54:36 +0000 (09:54 +0000)]
added flag for frame types

[r6337]