nik/lush
1
1 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
3,955 Commits 1 Branch 0 Tags
686e57cf9c61070ff961407e3304071e171542f4
Commit Graph
100 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
Roberto Ierusalimschy 686e57cf9c GC local pause configurable 2013-09-13 13:21:52 -03:00
Roberto Ierusalimschy 06156e7575 detail (setmetatable do not need to use a back GC barrier) 2013-09-11 17:15:31 -03:00
Roberto Ierusalimschy 2e2c109daf type of 'refcount' changed to lu_mem (to ensure there will be no 
overflows)
 2013-09-11 12:17:00 -03:00
Roberto Ierusalimschy 6ca7b63bce check for shrinking string table done only at the end of a GC cycle 2013-09-11 11:56:15 -03:00
Roberto Ierusalimschy 4eeb1831be new names and better order for GC states (sweep first lists that 
can have dead objects)
 2013-09-11 11:47:08 -03:00
Roberto Ierusalimschy dd373a8f66 threads are kept in a separated GC list, linked after the main thread 2013-09-11 11:09:55 -03:00
Roberto Ierusalimschy 7061fe1d56 detail: 'sweepstep' checks end of phase after calling 'sweeplist', so 
that phases with small lists return 0 at the first call to 'sweepstep'
 2013-09-11 10:24:55 -03:00
Roberto Ierusalimschy d8aa8dd97e objects in list 'tobefnz' have a GC life-cycle like all others 
(specifically they are cleaned during sweep phase)
 2013-09-11 09:47:48 -03:00
Roberto Ierusalimschy 1150873447 'luaC_newobj' does not handle special cases; only special case 
now is threads, which do not use 'luaC_newobj' anymore.
 2013-09-11 09:26:14 -03:00
Roberto Ierusalimschy 79ab21be90 more invariants added to memory check 2013-09-11 09:26:14 -03:00
Roberto Ierusalimschy d3bbb34c24 back to open hashing for the string table (but with a different 
'hnext' field, to strings are still collected like all other
objects)
 2013-09-05 16:31:49 -03:00
Roberto Ierusalimschy 0ad15fc100 better (and correct!) control of 'maybedead' 2013-09-04 12:34:24 -03:00
Roberto Ierusalimschy aeff4f79fa local collection now calls finalizers 2013-09-03 12:37:10 -03:00
Roberto Ierusalimschy 1bf4faec64 new GC state to sweep 'localgc' list + small changes in sweep control 2013-08-30 16:14:26 -03:00
Roberto Ierusalimschy 8ef9e8460e bug (GC can collect long identifier during parser) + change (using 
a single constant table for all functions in a chunk)
 2013-08-30 13:01:37 -03:00
Roberto Ierusalimschy 4f292d753c bug: GC can collect a long string still in use during parser 2013-08-30 12:51:12 -03:00
Roberto Ierusalimschy 26629d0af1 details (a few casts moved from macro invocation to macro definition) 2013-08-29 10:49:57 -03:00
Roberto Ierusalimschy b5e75fde4e bug: local collection must clear rest of stack 2013-08-29 10:34:16 -03:00
Roberto Ierusalimschy 9a871dd3db tables and userdata all go to local list, too 2013-08-28 15:30:26 -03:00
Roberto Ierusalimschy 90972ff136 LOCALBLACK changed to LOCALMARK and used also to control whether object 
is in 'localgc' list + luaC_newobj by default puts object in 'localgc'
list
 2013-08-27 17:04:00 -03:00
Roberto Ierusalimschy af35c7f398 upvalues collected by reference count 2013-08-27 15:53:35 -03:00
Roberto Ierusalimschy 742b7377d3 Lua closures go to local, too 2013-08-26 09:41:10 -03:00
Roberto Ierusalimschy 50955e27f5 C functions and strings now go to the local list; first version 
of the local collector
 2013-08-23 10:34:54 -03:00
Roberto Ierusalimschy 33c49f7fa0 some details over new implementation of string table 2013-08-22 12:21:48 -03:00
Roberto Ierusalimschy 0df6635711 "fixed" objects kept in a separated list (instead of being kept in 
'allgc' list with a bit marking them)
 2013-08-21 17:09:51 -03:00
Roberto Ierusalimschy ae800656c9 change in string table: string table is now independent of GC lists; all 
strings live in 'normal' GC lists
 2013-08-21 16:21:16 -03:00
Roberto Ierusalimschy 8c68863960 initial size of string table (MINSTRTABSIZE) should be large enough 
to hold all strings needed by an empty Lua state
 2013-08-21 16:02:30 -03:00
Roberto Ierusalimschy 8f6b80aa1d GC bits SEPARATED and FINALIZEDBIT mixed in FINALIZEDBIT (with simpler 
control)
 2013-08-20 14:46:34 -03:00
Roberto Ierusalimschy 9eff921f8f "barrier" for link prototype->cache changed to be consistent with 
GC behavior (link is cleared to preserve invariant)
 2013-08-19 11:18:43 -03:00
Roberto Ierusalimschy 8e6b7ef9ab new function 'isgclocal' + test for GC consistency of prototype's caches 2013-08-19 11:16:33 -03:00
Roberto Ierusalimschy caceeab750 'next' field for tables changed from pointer to integer (for better 
alignment on 64-bit machines)
 2013-08-18 13:12:18 -03:00
Roberto Ierusalimschy 3991312b94 details 2013-08-16 16:02:31 -03:00
Roberto Ierusalimschy 439d74e29f added 'local' bit (true => object is only refered by local variables) 2013-08-16 15:55:49 -03:00
Roberto Ierusalimschy 3679d33b02 barrier conditions rewritten to test first 'isblack' and then 'iswhite' 
(during a pause all objects are white, so 'isblack' fails much more
often than 'iswhite')
 2013-08-13 14:36:44 -03:00
Roberto Ierusalimschy 78b941039d change to allow collector to stop when it goes to the atomic phase 
(just for debugging)
 2013-08-07 12:39:09 -03:00
Roberto Ierusalimschy 623e388bb4 double-linked list of all upvalues elliminated and changed to a 
traversal of all non-marked threads
 2013-08-07 09:18:11 -03:00
Roberto Ierusalimschy 677d90165f no more generational collection !!! 2013-08-05 13:58:28 -03:00
Roberto Ierusalimschy 4244da96bf 'debug.subtype' -> 'math.type' (to distinguish integers and floats) 2013-07-22 13:05:53 -03:00
Roberto Ierusalimschy fcf99bf788 'math.isfloat' replaced by 'debug.subtype' 2013-07-10 17:57:05 -03:00
Roberto Ierusalimschy 98d76cdcae bug: Wrong error message in some short-cut expressions 2013-07-10 17:32:36 -03:00
Roberto Ierusalimschy 2f8c51a552 use unsigneds for unary minus, too 2013-07-10 14:15:12 -03:00
Roberto Ierusalimschy b5f5fcd782 use different constants (instead of different values for the same constant) 
to define number sizes (LUA_INT_INT/LUA_INT_LONG/LUA_INT_LONGLONG and
LUA_REAL_FLOAT/LUA_REAL_DOUBLE/LUA_REAL_LONGDOUBLE) + use __int64 and
I64 instead of long long (and ll) for Windows
 2013-07-09 15:31:35 -03:00
Roberto Ierusalimschy 2ef9bcfd11 avoid undefined shift of LUA_NBITS in rotate operation 2013-07-09 15:31:01 -03:00
Roberto Ierusalimschy 5fa680d47f no need for type 'b_uint' (lua_Unsigned must have at least 32 bits) 2013-07-09 14:49:50 -03:00
Roberto Ierusalimschy 48735da0d0 When loading a file, Lua may call the reader function again after 
it returned end of input + luac listings choke on long strings
 2013-07-05 15:02:28 -03:00
Roberto Ierusalimschy 2b4bd21585 'tonumber': base must be a number 2013-07-05 11:39:15 -03:00
Roberto Ierusalimschy beff5b078a 'tonumber': does not accept numbers when given a base + convertion from 
strings to numbers with a base moved to a separated function (more
modular)
 2013-07-05 11:35:49 -03:00
Roberto Ierusalimschy 5519c98655 'lua_cvtonum' -> 'lua_strtonum'; converts only strings to numbers 2013-07-05 11:29:51 -03:00
Roberto Ierusalimschy cbe164191c new function 'ifloor' 2013-07-03 14:23:19 -03:00
Roberto Ierusalimschy 66d046833d no need to check "bad conversion number->int;" in luaL_checkversion, 
as now Lua does not use tricks for the conversion, but there is a
need to check the sizes of number types, as they can be different
in two modules
 2013-06-27 15:32:33 -03:00
Roberto Ierusalimschy 15fdbd26fe 'numbits' moved from 'math' to 'debug' library 2013-06-25 16:37:00 -03:00
Roberto Ierusalimschy 2b4975dca7 LUA_SMALL_* changed to LUA_*SIZE + added support for long double + better 
support for ANSI systems
 2013-06-25 16:04:40 -03:00
Roberto Ierusalimschy e39da66c8c macro 'lua_strx2number' does not need to be global 2013-06-25 15:58:32 -03:00
Roberto Ierusalimschy 483c2ccd0f avoid using raw representation of numbers for hashing, as number may 
not use all its space (e.g., long double can have sizeof=12 but
use only 10 bytes)
 2013-06-25 15:57:18 -03:00
Roberto Ierusalimschy 982aa7058f 'luaL_len' returns lua_Integer instead of 'int' 2013-06-25 11:05:26 -03:00
Roberto Ierusalimschy 80cdf39d0e avoid using 'modf' in the implementation of 'math.modf', to avoid 
problems with 'double*'. (When using 'float' as lua_Number,
'float*' is not compatible with 'double*'...)
 2013-06-25 11:02:18 -03:00
Roberto Ierusalimschy 45f3797a5b 'long' may have more than 32 bits in some machines; use 'int' for 
small lua_Integer unless necessary to use 'long'
 2013-06-21 14:42:28 -03:00
Roberto Ierusalimschy de6e5a5919 small bugs uncovered in a 64-bit machine (where 'long' has 64 bits) 2013-06-21 14:27:24 -03:00
Roberto Ierusalimschy 81102d3689 try to avoid warnings when converting int to pointer 2013-06-20 18:59:13 -03:00
Roberto Ierusalimschy 6579bd4190 reorganization of warnings + update of explicit dependencies (gcc -MM) 2013-06-20 15:56:51 -03:00
Roberto Ierusalimschy f45177f2d3 In the table that hashes constants, use a light userdata as keys 
to integer values to avoid collisions with floats with the same
numerical value
 2013-06-20 14:37:31 -03:00
Roberto Ierusalimschy d6f5fb2d2c avoid warning for unary minus over an unsigned value 2013-06-20 12:12:43 -03:00
Roberto Ierusalimschy 7c4cc505db added "reasonable" limit for 'string.rep' (otherwise it is too easy 
to crash the machine)
 2013-06-20 12:06:51 -03:00
Roberto Ierusalimschy 453450d687 new macro 'l_floor' (allows 'floorf' even when other math operations 
do not have an 'f' variant)
 2013-06-20 12:02:49 -03:00
Roberto Ierusalimschy 55f566bd22 use lua_Integer for integer parameters to avoid truncation 2013-06-19 11:29:01 -03:00
Roberto Ierusalimschy 130c0e40e0 new constant 'MAX_SIZE', distinct from 'MAX_SIZET', for sizes visible 
from Lua; these must fit in a lua_Integer
 2013-06-19 11:27:00 -03:00
Roberto Ierusalimschy c72fb1cf8e detail (removed 'assert') 2013-06-14 17:46:40 -03:00
Roberto Ierusalimschy 138e69cc55 correct error message for floating-point values out of (integer) range 2013-06-14 15:34:49 -03:00
Roberto Ierusalimschy 53210d7e5b correct way (I hope) to convert floats to unsigned int 2013-06-14 15:32:45 -03:00
Roberto Ierusalimschy 0394314c7a avoid using a negative value to test 'lua_tounsigned' 2013-06-14 15:12:53 -03:00
Roberto Ierusalimschy 0beeb4f6fa configuration of number sizes goes "mainstream" 2013-06-13 16:35:08 -03:00
Roberto Ierusalimschy 788b251157 'math.random' operates with integers when used for integer results 
(1 or 2 parameters)
 2013-06-13 16:32:52 -03:00
Roberto Ierusalimschy 318575627f avoid adding '.0' to "numbers" like "nan" and "inf" 2013-06-07 16:02:05 -03:00
Roberto Ierusalimschy a14992992a new version (5.3) 2013-06-07 16:01:50 -03:00
Roberto Ierusalimschy 48adb6984c 'io.write' writes integers directly (and correctly) 2013-06-07 16:01:35 -03:00
Roberto Ierusalimschy c5069528e1 details ('Type* id' -> 'Type *id') 2013-06-07 11:51:10 -03:00
Roberto Ierusalimschy 6fb0b11350 string contatenation handles conversion of integers to strings + 
floats always format as floats (with decimal dot or exponent)
 2013-06-04 16:36:42 -03:00
Roberto Ierusalimschy 932e7fb0e1 'lua_tounsigned' takes number modulo 2^numbits as a result when 
number is a float
(That may change...)
 2013-06-04 16:34:51 -03:00
Roberto Ierusalimschy 188192ce9a 'luai_hashnum' "inlined" into 'hashfloat' 2013-05-29 11:05:03 -03:00
Roberto Ierusalimschy fc083f1138 removed all tricks for conversions + 'luai_hashnum' moved to ltable.c 2013-05-29 11:04:15 -03:00
Roberto Ierusalimschy ee865b9fe7 new implementation for 'lua_strx2number' to correct small 
problems with numbers like 0x10000...000p-100
 2013-05-27 14:42:38 -03:00
Roberto Ierusalimschy d630daca1a "legal" way to convert a float to an integer in C 2013-05-27 09:43:37 -03:00
Roberto Ierusalimschy 8c883cb4e8 'objlen' can return integers 2013-05-26 11:47:51 -03:00
Roberto Ierusalimschy c75cca0559 detail (case '%d' of 'luaO_pushvfstring' should use integers) 2013-05-26 11:43:35 -03:00
Roberto Ierusalimschy 1414b7123c support for the case when 'l_mathop' does not conform to lua_Number 
(problem with pointers to lua_Number solved by a typedef selected
automatically by 'l_mathop')
 2013-05-26 10:45:24 -03:00
Roberto Ierusalimschy 3563b3f519 first version of control for number sizes (still in 'local' section) 2013-05-26 10:35:52 -03:00
Roberto Ierusalimschy d46ed54646 no more IEEE tricks 2013-05-26 10:26:34 -03:00
Roberto Ierusalimschy e2c8c756ee removed 'IEEE tricks' (should not be needed in Lua with integers) 2013-05-23 18:27:06 -03:00
Roberto Ierusalimschy 5ca5086c19 'tonumber' now works with integers too 2013-05-16 15:35:57 -03:00
Roberto Ierusalimschy 2d6a0ae149 added patch to last bug 2013-05-16 13:03:50 -03:00
Roberto Ierusalimschy b5ec26101f new API function 'lua_cvtonum' to convert a value (number or string) 
to a number, following the rules of the language to create integers
or floats
 2013-05-14 13:00:11 -03:00
Roberto Ierusalimschy 36e8771076 'luaO_str2int' more generic: accepts white spaces around the numeral 
and handles signal
 2013-05-14 12:59:04 -03:00
Roberto Ierusalimschy 27f09415e3 new read format "*i" for reading integers 2013-05-14 12:57:43 -03:00
Roberto Ierusalimschy 4ad9970649 uses integers for time 2013-05-14 12:57:11 -03:00
Roberto Ierusalimschy bef345a4b8 bug: Wrong error message in some short-cut expressions 2013-05-13 13:17:47 -03:00
Roberto Ierusalimschy 0ce3018109 'pushnum' pushes a "number" (float) + new C instruction 'pushint' 2013-05-07 16:01:16 -03:00
Roberto Ierusalimschy 0233ce0815 new function 'math.numbits' (not a final decision) 2013-05-06 14:22:55 -03:00
Roberto Ierusalimschy 1f2b82bf25 correct way to avoid compile-time errors in integer divisions 2013-05-06 14:22:16 -03:00
Roberto Ierusalimschy f2043b7a58 correct error message for conversion errors from float to int 2013-05-06 14:21:59 -03:00
Roberto Ierusalimschy 2b1c2c61b0 bug: Wrong assert when reporting concatenation errors. 2013-05-06 14:21:28 -03:00