ed5cea57ea
compiler.cfg: Reading the return value of a float-returning function on x86-32 had a side effect of popping the x87 stack, so it was not correct for DCE to just eliminate this if the return value was not used. Fix this by adding a new dead-outputs slot to alien-call-insns and having DCE move dead returns there
2010-08-13 23:19:56 -07:00
7d792778f5
FFI rewrite part 7: compile callback bodies with the optimizing compiler
2010-07-28 00:49:26 -04:00
f133a5f2be
compiler: re-architect low-level optimizer to allow more than one output value per instruction
2010-07-13 07:40:14 -04:00
0dbc9eaae0
compiler.cfg: more silly optimizations
2010-05-14 18:37:09 -04:00
0f0571e48a
Moving new-sets to sets
2010-02-26 16:01:01 -05:00
fd7bcd35ea
Compiler.cfg.{dce,linearization} use new-sets
2010-02-16 15:48:07 -06:00
ba37429563
filter-here -> filter!
2009-10-28 00:44:05 -05:00
10ad5cad53
Working on adding support for the new write barrier to optimized code
2009-10-14 02:06:01 -05:00
775b9af2f7
compiler: eliminate boilerplate by centralizing info in declarative INSN: syntax
2009-09-02 06:22:37 -05:00
88d15dfbea
compiler.cfg: new system to track when results of analyses need to be recomputed (reverse post order, linear order, predecessors, dominance, loops). Passes can now call needs-predecessors, needs-dominance, needs-loops at the beginning, and cfg-changed, predecessors-changd at the end. Linearization order now takes loop nesting into account, and linear scan now uses linearization order instead of RPO.
2009-08-08 20:02:56 -05:00
2a63722c91
compiler.cfg: change representation of ##phi nodes. The input slot of ##phi nodes is now an assoc mapping predecessors to values, rather than a sequence of values. This ensures that optimization which delete predecessors can be implemented
2009-07-02 17:10:50 -05:00
9605ffc0dc
compiler.cfg.dce: now performs three passes over the CFG to fix a phase ordering issue
2009-07-02 03:17:19 -05:00
49250ea56b
compiler.cfg.dce: minor renamings
2009-07-02 02:41:10 -05:00
f07100739b
Dead code elimination eliminates some dead allocations
2009-07-02 02:30:46 -05:00
968a9bb666
Various codegen improvements:
...
- new-insn word to construct instructions
- cache RPO in the CFG
- re-organize low-level optimizer so that MR is built after register allocation
- register allocation now stores instruction numbers in the instructions themselves
- split defs-vregs into defs-vregs and temp-vregs
2009-05-29 13:11:34 -05:00
25805f9bfe
Refactoring low-level optimizer to support stack analysis pass
2009-05-26 19:31:19 -05:00
44437cc891
compiler.cfg.dce: new global dead code elimination pass
2009-05-26 02:58:57 -05:00
Slava Pestov
Slava Pestov
Daniel Ehrenberg
Daniel Ehrenberg