aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* full GC must run finalizers only after finishing everything elseRoberto Ierusalimschy2010-05-041-5/+5
| | | | (including a return to generational mode if needed)
* details (comments)Roberto Ierusalimschy2010-05-041-3/+3
|
* missing parentheses in declarationRoberto Ierusalimschy2010-05-041-2/+2
|
* no need to handle '\0' differently from other control chars inRoberto Ierusalimschy2010-05-041-2/+2
| | | | format '%q'
* no need of a KGC_FORCED collection kind; it has the same behavior ofRoberto Ierusalimschy2010-05-032-9/+7
| | | | a normal collection
* no more 'finalize' phase in GC; finalizers are called along theRoberto Ierusalimschy2010-05-033-31/+33
| | | | entire cycle
* items in 'tobefnz' are kept black (as before recent change) and changedRoberto Ierusalimschy2010-05-032-9/+15
| | | | | to white only when needed (being moved to 'allgc' when not keeping invariant).
* invariant must be kept in atomic 'phase' tooRoberto Ierusalimschy2010-05-031-9/+9
|
* "gray lists" only need to be valid when 'keepinvariant' is trueRoberto Ierusalimschy2010-04-302-15/+11
|
* added comment explaining a bit about the invariants of the collectorRoberto Ierusalimschy2010-04-301-1/+14
|
* detailRoberto Ierusalimschy2010-04-301-3/+2
|
* wrong commentRoberto Ierusalimschy2010-04-301-2/+2
|
* default PAUSE should be 200 (differences may be corrected inRoberto Ierusalimschy2010-04-301-6/+6
| | | | | internal multiplers) + check for unbilt stack equal the one used by lgc.c
* nasty GC bug: upvalue must be turned white when not keeping invariant,Roberto Ierusalimschy2010-04-293-9/+26
| | | | but barrier was not being called when uv->v were already white.
* removed commented-out debugging codeRoberto Ierusalimschy2010-04-291-3/+1
|
* new way to control GC speedRoberto Ierusalimschy2010-04-292-16/+26
|
* 'luaC_linkupval' embedded into call siteRoberto Ierusalimschy2010-04-291-4/+10
|
* improved 'lua_checkmemory', with better control over gray objectsRoberto Ierusalimschy2010-04-291-24/+45
|
* 'luaC_linkupval' moved into 'lfunc.c' code + new way to control GC speedRoberto Ierusalimschy2010-04-292-80/+80
|
* new way to control GC speed (keeping a 'debt' counter)Roberto Ierusalimschy2010-04-293-10/+11
|
* lots of new comments + small changes in loop controls + other smallRoberto Ierusalimschy2010-04-261-117/+161
| | | | "janitor work"
* "to-be-finalized" objects marked and sweeped like all other objectsRoberto Ierusalimschy2010-04-201-13/+13
|
* upvalue barriers for Lua functions must act on the upvalue itself,Roberto Ierusalimschy2010-04-201-6/+10
| | | | not on its closure
* missing parentheses around 'luaL_pushresultsize' declarationRoberto Ierusalimschy2010-04-191-2/+2
|
* bug with io.read(op, "*n")Roberto Ierusalimschy2010-04-191-4/+14
|
* 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...)
* ensure that 'luai_userstatethread' is always called (even ifRoberto Ierusalimschy2010-04-193-11/+12
| | | | 'stack_init' throws a memory error)
* 'gcinfo' was deprecated in version 5.0.Roberto Ierusalimschy2010-04-191-8/+1
|
* 'string.format' may get buffer as an argument when there areRoberto Ierusalimschy2010-04-191-2/+14
| | | | missing arguments and format string is too long
* complete control over number of each kind of object allocatedRoberto Ierusalimschy2010-04-191-46/+53
|
* with light C functions, 'pairs' does not need to keep 'next' as anRoberto Ierusalimschy2010-04-191-18/+8
| | | | upvalue.
* tells the allocation function that the first block is a threadRoberto Ierusalimschy2010-04-191-2/+2
|
* in lua_gc/step, 'luaC_step' changes GCthreshold, so there was littleRoberto Ierusalimschy2010-04-191-4/+2
| | | | control over the real step size.
* "light C function" is a better name than "C-function pointer"Roberto Ierusalimschy2010-04-186-19/+19
|
* comparsion tag methods follow the same rule as other binary TMsRoberto Ierusalimschy2010-04-181-20/+12
|
* macro LUA_USE_LONGJMP allows choosing longjmp when compiling Lua asRoberto Ierusalimschy2010-04-181-2/+2
| | | | C++ code
* better line numbers for function calls and unary/binary operators +Roberto Ierusalimschy2010-04-173-20/+27
| | | | null statement + no more "ambiguous syntax" restriction
* 'gcstate' now also runs collector until given state + small changesRoberto Ierusalimschy2010-04-161-21/+22
| | | | in 'testC' to test 'lua_topointer' and 'lua_tocfunction'
* new escape sequence '\*' + several comments + moving options fromRoberto Ierusalimschy2010-04-161-42/+62
| | | | switch default into cases (as now locale is fixed)
* no need to avoid calling ctype functions as now they are implementedRoberto Ierusalimschy2010-04-151-4/+2
| | | | by us (no inefficiencies due to accessing locale information)
* invalid instructions "cannot" happenRoberto Ierusalimschy2010-04-151-3/+2
|
* no more 'ccall' nor 'cpcall' functions. (With light C functions theyRoberto Ierusalimschy2010-04-145-40/+8
| | | | are obsolete.)
* first implementation of light C functionsRoberto Ierusalimschy2010-04-1411-88/+141
|
* BUG: 'string.format' may get buffer when there are missing argumentsRoberto Ierusalimschy2010-04-121-4/+15
|
* better control in 'totalmem' over choosing countersRoberto Ierusalimschy2010-04-121-8/+18
|
* new macro LUA_NUMTAGSRoberto Ierusalimschy2010-04-126-19/+15
|
* check memory allows strings to live in the main GC list (it shouldRoberto Ierusalimschy2010-04-121-1/+2
| | | | work ok).
* patterns now accept '\0' as a regular characterRoberto Ierusalimschy2010-04-121-28/+38
|
* new implementation for Generic Buffer manipulation (using userdata asRoberto Ierusalimschy2010-04-093-103/+93
| | | | temporary buffer space)
* keep memory-error message in the global state, so that its useRoberto Ierusalimschy2010-04-084-9/+13
| | | | | does not depend on Lua internalizing strings to avoid a string creation on memory errors
generated by cgit v1.2.3-55-g6feb (git 2.39.1) at 2025-03-21 00:46:17 +0000