75a97f2fa0
compiler.cfg.dependence: Add ##allot to control dependency chain, so that we don't re-order an allocation and initialization with an alien call. The alien call might GC, exposing uninitialized memory to the GC. Fixes #53 , fixes #58 , fixes #20 .
2011-08-28 21:25:39 -07:00
1f6469878a
compiler.cfg.debugger: add random-scheduling word.
...
Run 'random-scheduling "compiler" test' to find bugs in compiler.cfg.dependencies.
2011-08-28 21:25:37 -07:00
4aa5bbdc79
compiler.cfg.scheduling: remove old scheduling test code, adding a better way of testing scheduling
2011-08-28 21:25:35 -07:00
1ed2bc6231
32 and 64 bit Linux PPC support
2011-05-23 23:36:14 -04:00
1d6f318a1a
compiler.cfg.branch-splitting was totally broken
2011-02-27 16:43:26 -08:00
35e9f6f072
compiler.cfg.ssa.construction: update unit tests for nths change
2011-02-25 21:05:38 -08:00
78d529c7c4
compiler.cfg.linear-scan: extend lifetime intervals of base pointers correctly (reported by Anton Gorenko)
2010-09-30 20:49:03 -07:00
4877d32cb3
compiler: preliminary implementation of tracking derived pointers in GC maps
2010-09-27 20:12:44 -07:00
c43255712e
compiler.cfg.alias-analysis: fix bug in lazy alias class instantiation
2010-09-27 20:12:43 -07:00
0fb2cbd3a6
compiler.cfg.scheduling: always run scheduling because heuristic was broken
2010-09-25 16:31:42 -07:00
1ddf4bef8c
compiler.cfg.alias-analysis: don't need to do a local live-in calculation anymore
2010-09-25 16:13:17 -07:00
526dca873d
compiler.cfg.liveness: merge in compiler.cfg.liveness.ssa and simplify the code, since we don't compute live sets before SSA construction anymore
2010-09-25 14:36:58 -07:00
581bbe5278
compiler.cfg.ssa.construction: use the baller method for pruned SSA
2010-09-20 21:36:22 -07:00
bbaf070bd9
compiler.cfg.write-barrier: fix soundness issue with ##copy that never came up but probably should've been fixed anyway
2010-09-06 17:01:44 -07:00
cf327bbc41
compiler.cfg.alias-analysis: dead store elimination was too aggressive, can't eliminate dead stores across a GC call even for fresh allocations because GC will see uninitialized data
2010-09-06 16:57:56 -07:00
4190040611
compiler.cfg.write-barrier: handle ##copy instructions
2010-09-05 23:07:30 -07:00
06c7424f3d
compiler.cfg.dependence: compute dependency graph for write barrier instructions
2010-09-05 22:51:17 -07:00
8c5a22e2ca
compiler.cfg: fix major facepalm with write barrier elimination
2010-09-05 21:39:45 -07:00
f3dfae04ba
compiler: now that FFI has been deconcatenatized, we no longer need the special ##unary-float-function and ##binary-float-function fastpaths
2010-08-13 23:59:19 -07:00
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
31ef4ad844
math.libm: fix regression: fsqrt intrinsic was not working ever since change was made to inline FUNCTION: bodies
2010-08-12 21:41:57 -07:00
179f32cdeb
compiler.cfg.linear-scan: fix bad interaction between inactive intervals and sync points
2010-07-29 21:06:48 -04:00
7d792778f5
FFI rewrite part 7: compile callback bodies with the optimizing compiler
2010-07-28 00:49:26 -04:00
10d25cc0ad
Revert "compiler.cfg: change linear ordering to place GC call blocks at the end"
...
This reverts commit 544a288192
2010-07-27 23:58:41 -04:00
544a288192
compiler.cfg: change linear ordering to place GC call blocks at the end
2010-07-27 13:00:28 -04:00
f69411fcfb
compiler.cfg: now that kill-blocks cannot contain instructions that define vregs we can skip them all
2010-07-27 12:40:31 -04:00
08a139bb64
compiler.cfg.alias-analysis: factor-call-insns which defined values were not handled properly
2010-07-19 19:56:00 -04:00
0abdfbfc1b
compiler.cfg: open-code parameter boxing and unboxing for certain C types
2010-07-19 10:25:13 -04:00
f70b01e6f2
compiler.cfg.save-contexts: don't insert ##save-context in front of ##phi
2010-07-19 09:27:10 -04:00
641c973eb2
compiler.cfg.linear-scan.allocation: fix wrong order
2010-07-19 06:46:50 -04:00
b2d2507721
compiler.cfg.linear-scan: fix bad interaction between split position calculation and register-clobbering instructions
2010-07-18 17:38:52 -04:00
80b0ada324
compiler.cfg.intrinsics.allot: clean up
2010-07-16 20:04:03 -04:00
333b9a34cc
compiler.cfg: nuke ##allot-byte-array instruction
2010-07-16 19:57:45 -04:00
028faf0f1b
compiler.cfg.scheduling: ##dispatch must go at the end since its a control flow transfer
2010-07-15 20:55:31 -04:00
bbcffab3fa
compiler.cfg: ##unbox-long-long can have multiple outputs now, clean up long long parameter passing code using this
2010-07-15 19:49:29 -04:00
be6bb12bbd
compiler.cfg.scheduling: update to support multiple-output instructions
2010-07-15 17:38:34 -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
8fd5d3c86a
FFI rewrite part 6: deconcatenatize
2010-07-02 15:44:12 -04:00
eb61f5f4fe
Whitespace fix
2010-06-23 03:56:17 -04:00
58441d5e2e
compiler.cfg.builder: if there is a conditional where every branch is known to throw an error, don't process any of the (unreachable) code after the conditional, since this can confuse analysis (reported by Daniel Ehrenberg)
2010-06-23 03:33:15 -04:00
51d16c7a2b
compiler.cfg.height: clean it up a bit
2010-06-22 04:14:03 -04:00
5c7aaa0473
compiler.cfg.ssa: fixing test failures
2010-06-15 17:18:51 -04:00
fb4de47abf
Add GC maps to ##box, ##box-long-long, ##alien-invoke, ##alien-indirect and ##call-gc; remove ##gc-map instruction
2010-06-14 19:39:46 -04:00
a08295d9b7
GC maps for more compact inline GC checks
2010-06-11 20:06:00 -04:00
ba143db496
compiler.cfg.intrinsics.simd: intrinsic support for double-2>float-4 conversion
2010-06-01 00:34:50 -07:00
b8ec5c6176
Stack allocation improvements
...
- New with-out-parameters combinator
- Inhibit tail call optimization in frames with local allocation, to ensure that passing a stack allocated value to the last word in the quotation works
- local allocations are now aligned properly
- spill slots are now aligned properly aligned in frames which have parameter and local allocation areas
2010-05-22 02:37:00 -04:00
0f935bfbde
compiler.cfg.builder.alien: use ##local-allot to fix value struct passing on Win64
2010-05-19 01:53:32 -04:00
bcf7a5e15e
Use ##local-allot to simplify longlong unboxing
2010-05-19 01:07:22 -04:00
84e7076474
alien.data: add with-scoped-allocation combinator for stack-allocating C data
2010-05-19 00:33:36 -04:00
5dd2115665
compiler.cfg: generalize ##prepare-struct-caller instruction to alloca()-like ##local-allot
2010-05-19 00:33:15 -04:00
Slava Pestov
Erik Charlebois
Slava Pestov
Slava Pestov
Slava Pestov
Slava Pestov
Joe Groff