summaryrefslogtreecommitdiff
path: root/lvm.c (follow)
Commit message (Collapse)AuthorAgeFilesLines
* avoid warnings with -Wstrict-overflowRoberto Ierusalimschy2011-06-091-5/+6
|
* stricter control (using tag variants) over closure kinds (Lua x C)Roberto Ierusalimschy2011-06-021-7/+6
|
* more uniform names for 'equalobj'-related functionsRoberto Ierusalimschy2011-05-311-3/+3
|
* no need for two different implementations for equality (one raw andRoberto Ierusalimschy2011-05-311-3/+10
| | | | one with metamethods)
* avoid using expression as argument to unsafe macro 'l_isfalse'Roberto Ierusalimschy2011-05-051-2/+3
|
* change in opcode OP_LOADNIL: B is used as a counter instead of aRoberto Ierusalimschy2011-04-191-4/+4
| | | | | register. (Avoids an assignment to R(B), not present in any other instruction.)
* added macro for code checkingRoberto Ierusalimschy2011-04-181-1/+7
|
* new instruction OP_LOADKX (to replace OP_LOADK with extra argument)Roberto Ierusalimschy2011-04-071-2/+8
|
* new macro 'ttisequal'Roberto Ierusalimschy2011-04-051-2/+2
|
* 2nd operand for unary tag methods does not need to be 'nil'Roberto Ierusalimschy2011-04-051-3/+2
|
* trying to avoid assumption that sizeof(char)==1Roberto Ierusalimschy2011-02-071-3/+4
|
* no more 'OP_CLOSE' instructions (use jumps to close upvalues)Roberto Ierusalimschy2011-02-071-4/+1
|
* avoid extra increment to 'savedpc' when jumpingRoberto Ierusalimschy2011-02-011-15/+20
|
* new scheme to close upvalues in 'break'; jump instructions mayRoberto Ierusalimschy2011-02-011-11/+18
| | | | do the close, avoiding the need for a OP_CLOSE instruction
* GC does not mark dead registers inside activation recordRoberto Ierusalimschy2010-12-171-6/+21
|
* detection of erroneous numeric strings with \0 (such as "1\0")Roberto Ierusalimschy2010-12-061-2/+2
|
* comment typosRoberto Ierusalimschy2010-10-291-2/+2
|
* corrected warnings from different compilers (mostly casts and smallRoberto Ierusalimschy2010-10-251-2/+2
| | | | details)
* typo in commentsRoberto Ierusalimschy2010-06-301-3/+3
|
* better barrier for prototypesRoberto Ierusalimschy2010-06-071-4/+6
|
* Lua closures are cached for reuseRoberto Ierusalimschy2010-06-041-16/+52
|
* one more macro (mvdispatch) to easy experiments with alternativeRoberto Ierusalimschy2010-05-131-2/+3
| | | | dispatch methods
* use of macro ('vmcase') to format all cases in main switch ofRoberto Ierusalimschy2010-05-121-121/+83
| | | | 'luaV_execute' (to facilitate experiences with direct threading)
* small changes in 'luaV_execute' to make cases more regularly formattedRoberto Ierusalimschy2010-05-121-8/+7
|
* typos in commentsRoberto Ierusalimschy2010-05-051-2/+2
|
* details (comments)Roberto Ierusalimschy2010-05-041-3/+3
|
* allows thread switches (when/if there are thread switches) in the sameRoberto Ierusalimschy2010-04-191-5/+7
| | | | places that finalizers can run (so they should be safe...)
* "light C function" is a better name than "C-function pointer"Roberto Ierusalimschy2010-04-181-2/+2
|
* comparsion tag methods follow the same rule as other binary TMsRoberto Ierusalimschy2010-04-181-20/+12
|
* invalid instructions "cannot" happenRoberto Ierusalimschy2010-04-151-3/+2
|
* first implementation of light C functionsRoberto Ierusalimschy2010-04-141-3/+4
|
* new macro 'eqstr'Roberto Ierusalimschy2010-04-051-1/+2
|
* avoid using 'luaS_new' when can use 'luaS_newlstr'Roberto Ierusalimschy2010-04-021-3/+3
|
* work related to hooks moved from 'luaV_execute' to 'traceexec'Roberto Ierusalimschy2010-03-291-5/+5
|
* no more fenvs!Roberto Ierusalimschy2010-03-261-2/+2
|
* first version of _ENV; no more global variablesRoberto Ierusalimschy2010-03-121-31/+3
|
* avoid 'continue' unless necessaryRoberto Ierusalimschy2010-02-271-51/+51
|
* new instructions to optimize indexing on upvaluesRoberto Ierusalimschy2010-02-261-3/+13
|
* __unm metamethod gets nil as its 2nd parameterRoberto Ierusalimschy2010-01-151-1/+2
|
* 'lua_objlen' replaced by 'lua_rawlen', 'lua_len', and 'luaL_len'Roberto Ierusalimschy2009-12-171-3/+3
|
* new scheme for debug info about tail calls: no more 'fake' stack entries,Roberto Ierusalimschy2009-11-251-4/+4
| | | | but stack entry knows whether it was tail called
* comment explaining OP_VARARG was wrong (and corresponding code was notRoberto Ierusalimschy2009-10-281-3/+3
| | | | very clear)
* first implementation of lexical environmentsRoberto Ierusalimschy2009-09-301-1/+9
|
* information about upvalues (where they come from) kept in Proto structure,Roberto Ierusalimschy2009-09-281-16/+12
| | | | instead of sequence of pseudo-opcodes after OP_CLOSURE
* limit of constants per function changed to 2^26 using extra argumentsRoberto Ierusalimschy2009-09-231-7/+15
| | | | to opcodes LOADK, GETGLOBAL, and SETGLOBAL
* luaH_[gs]etnum renamed to luaH_[gs]etint (as they only accept integers,Roberto Ierusalimschy2009-08-071-2/+2
| | | | not generic numbers)
* tail calls do not need to move whole new frame down, only its sliceRoberto Ierusalimschy2009-07-151-8/+12
| | | | up to last parameter (the rest has not been used yet)
* bug: 'luaV_settable' may invalidate a reference to a table and tryRoberto Ierusalimschy2009-07-011-3/+6
| | | | to reuse it.
* exports luaV_arith to be used by lua_arith + uses 'luaO_arith' toRoberto Ierusalimschy2009-06-171-16/+7
| | | | perform basic arith. operations
* 'lessequal' renamed 'luaV_lessequal' and exported to be used byRoberto Ierusalimschy2009-06-171-3/+3
| | | | 'lua_compare'
generated by cgit v1.2.3-55-g6feb (git 2.39.1) at 2025-03-16 01:39:33 +0000