aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* added API checks to some unprotected 'top' incrementsRoberto Ierusalimschy2015-02-111-7/+12
|
* detailRoberto Ierusalimschy2015-02-091-2/+2
|
* bug (in 5.2): Chunk with too many lines can seg. faultRoberto Ierusalimschy2015-02-091-2/+26
|
* detail (no need to write 'double' in source)Roberto Ierusalimschy2015-02-091-3/+4
|
* detail (use 'l_floor' instead of 'l_mathop(floor)')Roberto Ierusalimschy2015-02-091-2/+2
|
* detailRoberto Ierusalimschy2015-02-051-2/+2
|
* macros 'luai_num*' (for float arithmetic operations) moved toRoberto Ierusalimschy2015-02-054-46/+53
| | | | llimits.h.
* added comments documenting LUA_NOCVTN2S/LUA_NOCVTS2NRoberto Ierusalimschy2015-02-051-1/+19
|
* make buffer for format specification a little larger (length modifierRoberto Ierusalimschy2015-02-041-3/+2
| | | | can be larger than 2, e.g. "I64")
* own implementation for 'string.format("%a")' for C89 platformsRoberto Ierusalimschy2015-02-042-11/+92
|
* detail ('luaL_getmetatable' does not return a boolean, but a tag)Roberto Ierusalimschy2015-02-031-2/+2
|
* LUAL_BUFFERSIZE affects the API, so it is better not to change itRoberto Ierusalimschy2015-01-161-2/+3
| | | | | | between releases of the same version. (The long double case seems to be too rare, so that should not be a problem; moreover, that case is related to a bug)
* LUAI_MAXSHORTLEN moved from 'luaconf.h' to 'llimits.h' (does notRoberto Ierusalimschy2015-01-162-11/+12
| | | | affect the API and is seldom modified)
* size of short strings stored in a single byte, to reduce the sizeRoberto Ierusalimschy2015-01-169-45/+69
| | | | of struct 'TString'
* larger MINSTRTABSIZE (standard Lua starts with ~230 strings, afterRoberto Ierusalimschy2015-01-161-3/+8
| | | | oppening the standard libraries)
* commentsRoberto Ierusalimschy2015-01-161-2/+12
|
* removed macro LUAI_FIRSTPSEUDOIDX (that stuff does not needRoberto Ierusalimschy2015-01-132-7/+6
| | | | to be configurable)
* Long double needs a larger LUAL_BUFFERSIZE (becauseRoberto Ierusalimschy2015-01-131-3/+10
| | | | | 'string.format("%.99f", 1e4900)' can generate quite long strings), but otherwise buffer can be 1~2K
* BUG (when compiled with long double): buffer overflow when formattingRoberto Ierusalimschy2015-01-131-5/+11
| | | | string.format("%.99f", 1e4930)
* better check for overflows in 'table.move' (removes restriction thatRoberto Ierusalimschy2015-01-131-3/+5
| | | | initial position should be positive)
* definition for 'MAXUPVAL' moved for a more "private" place and itsRoberto Ierusalimschy2015-01-132-7/+9
| | | | value and comment corrected to reflect current implementation
* typo in commentRoberto Ierusalimschy2015-01-131-3/+3
|
* typo in error messageRoberto Ierusalimschy2015-01-121-2/+2
|
* 'setkey' -> 'setnodekey' (to avoid conflicts with POSIX)v5.3.0Roberto Ierusalimschy2015-01-052-4/+4
|
* new macro 'cast_func' adds '__extension__' (in gnu compilers) whenRoberto Ierusalimschy2015-01-051-2/+14
| | | | converting void* to function, to avoid warnings
* includes 'stddef.h' (as it uses NULL)Roberto Ierusalimschy2015-01-052-5/+9
|
* detail (added -Wconversion as a comment, to be used ocasionally)Roberto Ierusalimschy2015-01-021-0/+1
|
* 'lua_setlocal' should not pop value when failing (to be consistentRoberto Ierusalimschy2015-01-022-5/+11
| | | | with the manual and with 'lua_setupvalue')
* clearer(?) code (also avoids a warning about 'c' being usedRoberto Ierusalimschy2015-01-021-5/+3
| | | | without initialization)
* does not define _XOPEN_SOURCE when LUA_USE_C89 is defined + definingRoberto Ierusalimschy2014-12-291-2/+8
| | | | _XOPEN_SOURCE==0 undefines that macro
* all "divisions" (div,idiv,mod) by zero are not folded, to avoidRoberto Ierusalimschy2014-12-291-24/+10
| | | | | | | problems during compilation + does not fold zero results, as they can collapse with -0.0 and the ANSI test to distinguish both needs a division by zero (which we are trying to avoid) + removed macro 'luai_numinvalidop' (as its main use case were divisions by zero)
* detail (macro 'luai_numidiv' uses 'luai_numdiv' to compute the division)Roberto Ierusalimschy2014-12-291-2/+2
|
* bug: 'random' limit is 2^31-1, not RAND_MAXRoberto Ierusalimschy2014-12-271-3/+5
|
* detailsRoberto Ierusalimschy2014-12-271-4/+4
|
* change in macro 'vmcase', avoding code inside it. (Code inside theRoberto Ierusalimschy2014-12-271-97/+142
| | | | macro harms tools such as debuggers.)
* new year (2015)Roberto Ierusalimschy2014-12-261-3/+3
|
* detail (to avoid warnings)Roberto Ierusalimschy2014-12-261-3/+3
|
* removed macro 'luai_numinvalidop' (main motivation removed, as foldingRoberto Ierusalimschy2014-12-262-15/+2
| | | | does not handle any division by zero by default)
* removed useless initializationsRoberto Ierusalimschy2014-12-262-4/+4
|
* comments were wrong (not updated about several changes)Roberto Ierusalimschy2014-12-201-6/+6
|
* details in 'luaM_reallocvchar'Roberto Ierusalimschy2014-12-192-5/+6
|
* new macro 'luaM_reallocvchar' to allocate arrays of chars (avoidsRoberto Ierusalimschy2014-12-192-3/+9
| | | | uneeded tests and respective warnings)
* new macro 'nvalue' (to convert an object to a float when we knowRoberto Ierusalimschy2014-12-193-10/+7
| | | | object is a number)
* detail (in test for 'luai_numinvalidop', use a round float)Roberto Ierusalimschy2014-12-191-2/+2
|
* commentRoberto Ierusalimschy2014-12-191-2/+2
|
* old Visual Studio versions did not support 'noreturn' attributeRoberto Ierusalimschy2014-12-191-2/+2
|
* stupid bug in T.stacklevel (not in use by the tests)Roberto Ierusalimschy2014-12-181-2/+2
|
* Windows uses 'long long' (instead of '__int64') when present; again,Roberto Ierusalimschy2014-12-161-15/+15
| | | | uses macro LLONG_MAX as a proxy for the support for 'long long'.
* added macro LUA_USE_I64 for separate control of use of Windows-specificRoberto Ierusalimschy2014-12-161-2/+3
| | | | types
* traverse loaded modules (instead of globals) for a name for a function +Roberto Ierusalimschy2014-12-141-5/+13
| | | | removes prefix '_G.' from names (if present)