2009-09-21 |
Matthias Braun | fix emitting jne, now queens works :)
|
commit | commitdiff | tree |
2009-09-21 |
Matthias Braun | implement PopMem, PopEbp, fix push register, fix some...
|
commit | commitdiff | tree |
2009-09-21 |
Matthias Braun | fix binary emitter for cmp with addressmode and immediate
|
commit | commitdiff | tree |
2009-09-21 |
Matthias Braun | added jump emitters (only far jump at the moment)
|
commit | commitdiff | tree |
2009-09-21 |
Matthias Braun | fix strange bug introduced in refactoring (a half-written if)
|
commit | commitdiff | tree |
2009-09-21 |
Matthias Braun | fix bug introduced in refactoring: displacement only...
|
commit | commitdiff | tree |
2009-09-21 |
Matthias Braun | fix cltd
|
commit | commitdiff | tree |
2009-09-21 |
Matthias Braun | fix relative entity emitter
|
commit | commitdiff | tree |
2009-09-21 |
Matthias Braun | do creation and merge of affinity classes in 1 pass
|
commit | commitdiff | tree |
2009-09-21 |
Matthias Braun | emit statev times in usec
|
commit | commitdiff | tree |
2009-09-18 |
Matthias Braun | - further refactoring and finally eliminated the callback...
|
commit | commitdiff | tree |
2009-09-18 |
Matthias Braun | - Reworked backends to put out register_requirements...
|
commit | commitdiff | tree |
2009-09-17 |
Matthias Braun | - Bigger refactoring and cleanup in backend:
|
commit | commitdiff | tree |
2009-09-17 |
Matthias Braun | experimental beginning of a binary emitter I had lying...
|
commit | commitdiff | tree |
2009-09-17 |
Matthias Braun | rename benode_t.h to benode.h, remove some unused code
|
commit | commitdiff | tree |
2009-09-17 |
Matthias Braun | unnecessary map
|
commit | commitdiff | tree |
2009-09-16 |
Matthias Braun | small optimisation (avoid some arch_irn_consider_in_regalloc
|
commit | commitdiff | tree |
2009-09-11 |
Matthias Braun | inline arch_get_register_req_out to make the backend...
|
commit | commitdiff | tree |
2009-09-11 |
Matthias Braun | LC_OPT_ENT_BOOL needs an int
|
commit | commitdiff | tree |
2009-09-10 |
Matthias Braun | we need last_uses even without preferences
|
commit | commitdiff | tree |
2009-09-10 |
Matthias Braun | small bugfix/cleanup of phi preference propagation
|
commit | commitdiff | tree |
2009-09-10 |
Matthias Braun | add commandline options to disable preference, congruence...
|
commit | commitdiff | tree |
2009-09-09 |
Matthias Braun | don't classify Proj nodes
|
commit | commitdiff | tree |
2009-09-09 |
Matthias Braun | multi level optimistic split
|
commit | commitdiff | tree |
2009-09-07 |
Matthias Braun | return 0 for number of outputs for nodes without backend...
|
commit | commitdiff | tree |
2009-09-07 |
Matthias Braun | give fewer penalties to neighbors of uses with multiple...
|
commit | commitdiff | tree |
2009-09-05 |
Matthias Braun | allow final_cost pointer to be NULL
|
commit | commitdiff | tree |
2009-09-05 |
Matthias Braun | avoid putting interfering stuff into the same congruence...
|
commit | commitdiff | tree |
2009-09-04 |
Matthias Braun | new heursitic for good coloring order; add utility...
|
commit | commitdiff | tree |
2009-09-04 |
Matthias Braun | only compile pbqp stuff if pbqp defines are enabled
|
commit | commitdiff | tree |
2009-09-03 |
Matthias Braun | tweak optimistic split heuristic to take execfreq of...
|
commit | commitdiff | tree |
2009-09-03 |
Matthias Braun | align stackframe to 4 bytes to avoid bad esp alignments
|
commit | commitdiff | tree |
2009-09-03 |
Matthias Braun | improve hack to also work when the value has been ...
|
commit | commitdiff | tree |
2009-09-03 |
Matthias Braun | don't optimistically split these strange msut_be_different...
|
commit | commitdiff | tree |
2009-09-03 |
Matthias Braun | CopyKeeps have mostly Keep-like semantics too - make...
|
commit | commitdiff | tree |
2009-09-03 |
Matthias Braun | fix (nearly) endless loops in some cases of propagate_phi
|
commit | commitdiff | tree |
2009-09-02 |
Matthias Braun | merge preferences along congruence classes
|
commit | commitdiff | tree |
2009-09-01 |
Matthias Braun | old and never used files
|
commit | commitdiff | tree |
2009-09-01 |
Matthias Braun | old and nowaday useless tool stuff
|
commit | commitdiff | tree |
2009-09-01 |
Matthias Braun | - improve block scheduling heuristic by deciding a...
|
commit | commitdiff | tree |
2009-09-01 |
Matthias Braun | improv execfreq estimation if a loop outedge leaves...
|
commit | commitdiff | tree |
2009-08-31 |
Matthias Braun | fix rematerialisation code to not ask Proj nodes for...
|
commit | commitdiff | tree |
2009-08-28 |
Matthias Braun | clarify comment
|
commit | commitdiff | tree |
2009-08-27 |
Matthias Braun | improve phi propagation
|
commit | commitdiff | tree |
2009-08-25 |
Matthias Braun | benewalloc: refactorign assignment_t struct is not...
|
commit | commitdiff | tree |
2009-08-25 |
Matthias Braun | addendum to the should_be_same fix: we also have to...
|
commit | commitdiff | tree |
2009-08-24 |
Matthias Braun | fix optimistical split
|
commit | commitdiff | tree |
2009-08-24 |
Matthias Braun | add missing NL to generated file
|
commit | commitdiff | tree |
2009-08-24 |
Matthias Braun | fix comment
|
commit | commitdiff | tree |
2009-08-24 |
Matthias Braun | - First version of optimistic splitting
|
commit | commitdiff | tree |
2009-08-24 |
Matthias Braun | fix daemel spiller accessing flags of Proj nodes
|
commit | commitdiff | tree |
2009-08-20 |
Matthias Braun | a first try to propagate phi preferences upwards
|
commit | commitdiff | tree |
2009-08-17 |
Matthias Braun | benewalloc: more special casing of unknowns as phi...
|
commit | commitdiff | tree |
2009-08-17 |
Matthias Braun | benewalloc: fix unknowns being used at phi inputs
|
commit | commitdiff | tree |
2009-08-17 |
Matthias Braun | benewalloc: fix assert hitting when dying value was...
|
commit | commitdiff | tree |
2009-08-17 |
Matthias Braun | generally rewire inputs before enforce_constraints...
|
commit | commitdiff | tree |
2009-08-17 |
Matthias Braun | fix a few errors and warnings in the new pass code...
|
commit | commitdiff | tree |
2009-08-16 |
Matthias Braun | benewalloc: fix enforce_constraints sometimes looking...
|
commit | commitdiff | tree |
2009-08-16 |
Matthias Braun | fix output constraints not always being respected
|
commit | commitdiff | tree |
2009-08-16 |
Matthias Braun | cleanup and fix permutate_values (3-cycles and values...
|
commit | commitdiff | tree |
2009-08-16 |
Matthias Braun | fix double typedef of ir_prog
|
commit | commitdiff | tree |
2009-08-14 |
Matthias Braun | free values again if we know we don't need them anymore
|
commit | commitdiff | tree |
2009-08-14 |
Matthias Braun | benewalloc: fix live-in values without phis not getting...
|
commit | commitdiff | tree |
2009-08-14 |
Matthias Braun | fix more bugs in new allocator
|
commit | commitdiff | tree |
2009-08-14 |
Matthias Braun | fix block message in register allocator verifier
|
commit | commitdiff | tree |
2009-08-14 |
Matthias Braun | fix a bug in constraint enforcement; improve phi handling...
|
commit | commitdiff | tree |
2009-08-14 |
Matthias Braun | benewalloc: cleanups; dump intermediate results; rework...
|
commit | commitdiff | tree |
2009-08-13 |
Matthias Braun | - rework backend node dumping; add a dumper for Phi...
|
commit | commitdiff | tree |
2009-08-12 |
Matthias Braun | fix rbitset_is_empty; put opening brace of functions...
|
commit | commitdiff | tree |
2009-08-11 |
Matthias Braun | - implement output constraint enforcement for new register...
|
commit | commitdiff | tree |
2009-08-11 |
Matthias Braun | SwitchJmp doesn't have 1 'none' output
|
commit | commitdiff | tree |
2009-08-11 |
Matthias Braun | add rbitsets_have_common function
|
commit | commitdiff | tree |
2009-08-06 |
Matthias Braun | fix
|
commit | commitdiff | tree |
2009-08-06 |
Matthias Braun | - Rewrite internal backend logic for querying register...
|
commit | commitdiff | tree |
2009-08-05 |
Matthias Braun | fix read from uninitialized memory
|
commit | commitdiff | tree |
2009-08-05 |
Matthias Braun | reorganize backend headers (kill some _t variants in...
|
commit | commitdiff | tree |
2009-08-04 |
Matthias Braun | fix register allocation verifier not checking output...
|
commit | commitdiff | tree |
2009-08-03 |
Matthias Braun | initial work for phi-nodes (ifs should work, loops...
|
commit | commitdiff | tree |
2009-08-03 |
Matthias Braun | more work on permutate values (not finished yet)
|
commit | commitdiff | tree |
2009-08-02 |
Matthias Braun | fix warning and compilation error in optimized mode
|
commit | commitdiff | tree |
2009-08-02 |
Matthias Braun | fix a few bugs in benewalloc, split codegen timer in 2
|
commit | commitdiff | tree |
2009-07-31 |
Matthias Braun | cleanup backend: make pre_spill_prepare_constraint...
|
commit | commitdiff | tree |
2009-07-31 |
Matthias Braun | unnecessary include
|
commit | commitdiff | tree |
2009-07-31 |
Matthias Braun | fix warning
|
commit | commitdiff | tree |
2009-07-31 |
Matthias Braun | fix warning
|
commit | commitdiff | tree |
2009-07-31 |
Matthias Braun | fix last commit
|
commit | commitdiff | tree |
2009-07-31 |
Matthias Braun | declare load/store alternative in iroptimize header
|
commit | commitdiff | tree |
2009-07-31 |
Matthias Braun | less abreviations: rebitset_cpy => rebitset_copy
|
commit | commitdiff | tree |
2009-07-10 |
Matthias Braun | condeval is called Jump Threading now
|
commit | commitdiff | tree |
2009-06-29 |
Matthias Braun | remove unused common/firmwalk.* code
|
commit | commitdiff | tree |
2009-06-29 |
Matthias Braun | remove pointless and unused common/statistics.* code
|
commit | commitdiff | tree |
2009-06-29 |
Matthias Braun | Cleanup: remove firm_common_t.h (and the PRECISE_EXC_CONTEXT...
|
commit | commitdiff | tree |
2009-06-29 |
Matthias Braun | fix bad merge in last commit
|
commit | commitdiff | tree |
2009-06-29 |
Matthias Braun | rework python generators: we can make use of the classes...
|
commit | commitdiff | tree |
2009-06-26 |
Matthias Braun | mark get_Block_irg inline to avoid warnings
|
commit | commitdiff | tree |
2009-06-19 |
Matthias Braun | another instance of wrong block in new_Proj
|
commit | commitdiff | tree |
2009-06-19 |
Matthias Braun | verify that Projs are in the same Block as their predecessor...
|
commit | commitdiff | tree |
2009-06-19 |
Matthias Braun | - Get rid of the infamous Start->ProjX->Start backedge
|
commit | commitdiff | tree |
2009-06-18 |
Matthias Braun | transform assert to an if, as some students tend to...
|
commit | commitdiff | tree |
2009-06-18 |
Matthias Braun | fix wrong assert, expor copy_node_attr function to...
|
commit | commitdiff | tree |
next |