summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
* Add table of IR type sizes.Mike Pall2012-08-273-6/+21
|
* ARM, PPC, MIPS: Improve XLOAD operand fusion and register hinting.Mike Pall2012-08-274-5/+9
|
* Fix underflow handling in builtin string to number conversion.Mike Pall2012-08-271-1/+1
|
* Replace some trivial uses of fprintf() with fputs.Mike Pall2012-08-262-5/+9
|
* Replace divisions with simpler code.Mike Pall2012-08-262-11/+14
|
* Fix compilation with FFI disabled.Mike Pall2012-08-262-3/+5
|
* Remove old Makefile rules.Mike Pall2012-08-262-9/+2
|
* Replace strtod() with builtin string to number conversion.Mike Pall2012-08-2524-272/+660
|
* Add support for '%a' and '%A' to string.format.Mike Pall2012-08-251-3/+3
|
* ARM: Fix base register coalescing.Mike Pall2012-08-252-3/+7
|
* ARM: Fix commit ceaa60c0.Mike Pall2012-08-251-2/+5
|
* Minor fixes to FOLD rules.Mike Pall2012-08-201-8/+20
|
* ARM: Fix hard-float lj_vm_trunc() (used by compiled math.modf).Mike Pall2012-08-191-2/+2
|
* MIPS: Don't use argument GPRs to rematerialize FPR arguments.Mike Pall2012-08-191-11/+20
|
* ARM: Don't use argument GPRs to rematerialize FPR arguments.Mike Pall2012-08-191-4/+15
|
* ARM, MIPS, PPC: Rematerialize FPR invariants first.Mike Pall2012-08-191-2/+11
|
* MIPS: Fix calls to floor/ceil/trunc.Mike Pall2012-08-192-8/+6
|
* ARM: Fix calls to hard-float floor/ceil/trunc.Mike Pall2012-08-191-1/+2
|
* Fix shrinking of direct mapped block in builtin allocator.Mike Pall2012-08-161-1/+1
|
* Fix last commit.Mike Pall2012-08-151-3/+3
|
* FOLD x / 2^k ==> x * 2^-k.Mike Pall2012-08-151-0/+9
|
* x86: Simplify error message for CPUs lacking CMOV.Mike Pall2012-08-141-1/+1
|
* FFI: Don't keep CType * across call. Callback may reallocate table.Mike Pall2012-08-121-0/+2
|
* Do not use DWARF unwinder on Windows.Mike Pall2012-08-111-45/+0
|
* Update cross-compilation docs.Mike Pall2012-08-091-63/+63
|
* ARM: Enable VFP and hard-float ABI support.Mike Pall2012-08-091-5/+2
|
* ARM: Add VFP and hard-float ABI variants to JIT compiler.Mike Pall2012-08-098-113/+760
|
* Never compile DWARF unwinder on Windows.Mike Pall2012-08-091-1/+1
|
* FFI: Compile ffi.typeof(cdata).Mike Pall2012-08-083-2/+16
| | | | Thanks to Robert G. Jakabosky.
* FFI: Compile assignments from enums.Mike Pall2012-08-081-2/+5
| | | | Thanks to Robert G. Jakabosky.
* Add missing change for last commit.Mike Pall2012-08-051-0/+1
|
* FFI: Improve error message for bad indexing attempts.Mike Pall2012-08-051-3/+7
|
* FFI: Initialize all fields of standard types.Mike Pall2012-08-051-0/+2
|
* FFI: Fix alignment of C call structure.Mike Pall2012-08-021-1/+5
|
* ARM: Add hard-float ABI support to the FFI (interpreter).Mike Pall2012-07-305-19/+193
|
* ARM: Add VFP and hard-float ABI variants to interpreter.Mike Pall2012-07-303-28/+434
|
* Only replay sunk stores up to the snapshot reference.Mike Pall2012-07-291-2/+1
|
* Need to handle TBAR in SINK pass.Mike Pall2012-07-291-1/+1
| | | | Can only happen with -O-dse.
* Handle cross-compiles with FPU/no-FPU or hard-fp/soft-fp ABI mismatch.Mike Pall2012-07-292-6/+16
|
* ARM: Add VFP instructions to ARM disassembler.Mike Pall2012-07-291-9/+154
|
* DynASM/ARM: Add VFP instructions.Mike Pall2012-07-292-49/+235
|
* Don't treat all constified cdata content as constant.Mike Pall2012-07-243-2/+16
|
* Turn loads from immutable upvalues into constants.Mike Pall2012-07-206-24/+72
|
* Fix slot flags in snapshot replay.Mike Pall2012-07-201-1/+1
|
* Fix restore of sunk CNEWI with implicit P32 to P64 conversion.Mike Pall2012-07-181-0/+4
|
* FFI: Box all accessed or returned enums.Mike Pall2012-07-1711-60/+121
|
* Fix FOLD rule (i-j)-i => 0-j.Mike Pall2012-07-171-1/+1
| | | | Thanks to Thomas Schilling.
* Workaround for MSVC conversion bug (double -> uint32_t -> int32_t).Mike Pall2012-07-171-0/+7
|
* x86/x64: Fix fusion of unsigned byte comparisons with swapped ops.Mike Pall2012-07-161-2/+2
|
* ARM: Invoke SPLIT pass for leftover IR_TOBIT.Mike Pall2012-07-161-0/+2
| | | | Can only happen with -O-fold for TOBIT(constant).