mirrors/lua
1
0
Fork 0
mirror of https://github.com/lua/lua synced 2024-11-28 23:53:16 +03:00
Code Issues Packages Projects Releases Wiki Activity
5,599 Commits 4 Branches 53 Tags 16 MiB
d51022bf9e
Commit Graph

362 Commits

Author SHA1 Message Date
Roberto Ierusalimschy
d39ea8b3ce Make sure that main thread is non yieldable 
Main thread must be non yieldable even at "level 0" (bare API), outside
the 'pcall' from 'lua.c'.
 2020-07-06 13:54:01 -03:00
Roberto Ierusalimschy
a304199836 Detail in 'lua_resetthread' 
'lua_resetthread' should reset the CallInfo list before calling
'luaF_close'. luaF_close can call functions, and those functions
should not run with dead functions still in the CallInfo list.
 2020-06-17 10:36:42 -03:00
Roberto Ierusalimschy
d49b288728 'luaE_shrinkCI' shouldn't remove first free CallInfo 
Due to emergency collections, 'luaE_shrinkCI' can be called while Lua
is building a new CallInfo, which for a while is still a free CallInfo.
 2020-06-15 11:58:59 -03:00
Roberto Ierusalimschy
17dbaa8639 Improvements in the handling of signals 
Added 'volatile' to 'l_signalT' variables plus some minor changes.
 2020-05-22 11:40:34 -03:00
Roberto Ierusalimschy
7288528a1e Short strings always use all bytes in the hash 
Collisions in short strings occurr just by their existence, when
internalizing them. (Collisions in long strings is caused/controlled
by the program, when adding them as keys to the same table.)
 2020-04-01 10:52:41 -03:00
Roberto Ierusalimschy
46c3587a6f Clearer distinction between types and tags 
LUA_T* represents only types; tags (types + Variants) are represented
by LUA_V* constants.
 2020-01-31 11:09:53 -03:00
Roberto Ierusalimschy
6e1aec7a67 Larger C-stack limits for new threads 
New threads were being created with very small C-stack limits.
This is not a problem for coroutines, because 'lua_resume' sets
a new limit, but not all threads are coroutines.
 2019-10-17 13:09:17 -03:00
Roberto Ierusalimschy
03cde80b58 'setCstacklimit' renamed to 'setcstacklimit' 
Function names in the API use only lowercase letters.
 2019-09-24 14:31:06 -03:00
Roberto Ierusalimschy
ca13be9af7 Supressed errors in '__close' generate warnings 2019-08-16 09:51:54 -03:00
Roberto Ierusalimschy
8b7cfee26b Small changes around C-stack limit 
- Better documentation in 'testes/cstack.lua' about using
'debug.setCstacklimit' to find a good limit.

- Constant LUAI_MAXCSTACK gets added CSTACKERR (extra stack for
error handling), so that it is compatible with the argument to
'debug.setCstacklimit'.
 2019-06-26 13:26:36 -03:00
Roberto Ierusalimschy
be73f72fcc New function 'setCstacklimit' 
Added new functions to dynamically set the C-stack limit
('lua_setCstacklimit' in the C-API, 'debug.setCstacklimit' in Lua).
 2019-06-18 16:52:22 -03:00
Roberto Ierusalimschy
3cd9b56ae6 Revamp around 'L->nCcalls' count 
The field 'L->nCcalls' now counts downwards, so that the C-stack
limits do not depend on the stack size.
 2019-06-12 10:31:38 -03:00
Roberto Ierusalimschy
969b8c1f14 Fixed bug with to-be-closed variables in base C level 
To-be-closed variables in C use 'ci.nresults' to code that there is
a variable to be closed in that function. The intialization of the
base C level (the one "running" when calling API functions outside
any Lua call) did not initialize 'ci.nresults', creating (correct)
warnings in valgrind.
 2019-04-26 11:13:01 -03:00
Roberto Ierusalimschy
0443ad9e28 LUAI_MAXCCALLS renamed LUAI_MAXCSTACK 
The limit LUAI_MAXCCALLS was renamed LUAI_MAXCSTACK, which better
represents its meaning. Moreover, its definition was moved to
'luaconf.h', given its importance now that Lua does not use
a "stackless" implementation.
 2019-03-25 14:12:06 -03:00
Roberto Ierusalimschy
b56d4e570a Changes in the warning system 
- The warning functions get an extra parameter that tells whether
message is to be continued (instead of using end-of-lines as a signal).

- The user data for the warning function is a regular value, instead
of a writable slot inside the Lua state.
 2019-03-14 15:30:54 -03:00
Roberto Ierusalimschy
2c32bff609 After a "bad collections", avoid switching back back to generational 
After a major bad collection (one that collects too few objects),
next collection will be major again. In that case, avoid switching
back to generational mode (as it will have to switch again to
incremental to do next major collection).
 2019-01-30 11:44:42 -02:00
Roberto Ierusalimschy
c6f7181e91 No more LUA_ERRGCMM errors 
Errors in finalizers (__gc metamethods) are never propagated.
Instead, they generate a warning.
 2019-01-01 12:14:56 -02:00
Roberto Ierusalimschy
437a5b07d4 Added a warning system to Lua 
The warning system is just a way for Lua to emit warnings, messages
to the programmer that do not interfere with the running program.
 2018-12-28 15:42:34 -02:00
Roberto Ierusalimschy
ba7da13ec5 Changes in the control of C-stack overflow 
* unification of the 'nny' and 'nCcalls' counters;
  * external C functions ('lua_CFunction') count more "slots" in
    the C stack (to allow for their possible use of buffers)
  * added a new test script specific for C-stack overflows. (Most
    of those tests were already present, but concentrating them
    in a single script easies the task of checking whether
    'LUAI_MAXCCALLS' is adequate in a system.)
 2018-12-27 14:32:29 -02:00
Roberto Ierusalimschy
fdc25a1ebf New functions 'lua_resetthread' and 'coroutine.kill' 
New functions to reset/kill a thread/coroutine, mainly (only?) to
close any pending to-be-closed variable. ('lua_resetthread' also
allows a thread to be reused...)
 2018-12-13 13:07:53 -02:00
Roberto Ierusalimschy
e8c779736f Removed internal cache for closures 
The mechanism of "caching the last closure created for a prototype to
try to reuse it the next time a closure for that prototype is created"
was removed. There are several reasons:

- It is hard to find a natural example where this cache has a measurable
impact on performance.

- Programmers already perceive closure creation as something slow,
so they tend to avoid it inside hot paths. (Any case where the cache
could reuse a closure can be rewritten predefining the closure in some
variable and using that variable.)

- The implementation was somewhat complex, due to a bad interaction
with the generational collector. (Typically, new closures are new,
while prototypes are old. So, the cache breaks the invariant that
old objects should not point to new ones.)
 2018-11-01 13:21:00 -03:00
Roberto Ierusalimschy
bd96330d03 First "complete" implementation of to-be-closed variables 
Still missing:
- handling of memory errors when creating upvalue (must run closing
method all the same)
- interaction with coroutines
 2018-10-17 10:44:42 -03:00
Roberto Ierusalimschy
f99509581e Removed extra information from RCS keyword strings 
Version numbers and dates (mostly wrong) from RCS keyword strings
removed from all source files; only the file name are kept.
 2018-08-23 14:26:12 -03:00
Roberto Ierusalimschy
af70905246 no need to check whether libraries and host use the same kernel; 
Lua should work correctly with several copies of the kernel
 2018-06-18 09:08:10 -03:00
Roberto Ierusalimschy
b95e466218 new field 'nilvalue' in struct 'global_State' to avoid the use of 
addresses of static variables
 2018-06-15 16:31:22 -03:00
Roberto Ierusalimschy
505fc91222 no more 'luaO_nilobject' to avoid comparison of global variable addresses 
(now uses static variables)
 2018-06-01 14:40:38 -03:00
Roberto Ierusalimschy
97e394ba18 macro 'luai_makeseed' now controls the whole process of making the seed 2018-05-29 15:02:51 -03:00
Roberto Ierusalimschy
022abc301b 'gcemergency' should be zero before any memory allocation 2018-02-05 15:11:37 -02:00
Roberto Ierusalimschy
e2b15aa21d janitor work on casts 2018-01-28 13:13:26 -02:00
Roberto Ierusalimschy
4dc0be950a new macro 'isLuacode' (to distinguish regular Lua code from 
hooks, where C code can run inside a Lua function).
 2017-12-19 14:40:17 -02:00
Roberto Ierusalimschy
196c87c9ce no more 'stackless' implementation; 'luaV_execute' calls itself 
recursively to execute function calls. 'unroll' continues all
executions suspended by an yield (through a long jump)
 2017-11-23 14:41:16 -02:00
Roberto Ierusalimschy
5440b42f43 using 'trap' to stop 'luaV_execute' when necessary (tracing and 
to update its copy of 'base' when the stack is reallocated)
 2017-11-13 13:36:52 -02:00
Roberto Ierusalimschy
ad0704e40c back to 'CallInfo' (no gains with its removal) 2017-11-07 11:25:26 -02:00
Roberto Ierusalimschy
93fd67b793 no more 'CallInfo' structure 2017-11-04 10:57:02 -02:00
Roberto Ierusalimschy
54eb35a8aa more fields moved out of 'CallInfo' 2017-11-03 10:12:30 -02:00
Roberto Ierusalimschy
c5482468fd baby steps to remove 'CallInfo': keeping 'L->func' correct 2017-10-31 15:54:35 -02:00
Roberto Ierusalimschy
1d8920dd7f some cleaning in GC parameters 2017-10-11 09:38:45 -03:00
Roberto Ierusalimschy
f96497397a new type 'StackValue' for stack elements 
(we may want to put extra info there in the future)
 2017-06-29 12:06:44 -03:00
Roberto Ierusalimschy
72d82a296c revamping the incremental collector 
Some simplifications (not counting bytes, couting only slots visited;
no more 'gcfinnum'); more GC parameters; using vararg in 'lua_gc' to
set parameters in different GC modes
 2017-05-26 16:14:29 -03:00
Roberto Ierusalimschy
2376eb6347 barrier for prototype's cache (with new gray list 'protogray' to keep 
prototypes to have their caches visited again) + constant 'MAXMISS'
 2017-05-04 10:32:01 -03:00
Roberto Ierusalimschy
69371c4b84 'KGC_NORMAL' -> 'KGC_INC' + emergency GC signalled by flag (instead 
of mode)
 2017-04-24 13:59:26 -03:00
Roberto Ierusalimschy
f74b87c3c2 removed initialization of 'GCestimate' (it is initialized during 
a GC cycle, when it start counting)
 2017-04-12 15:56:25 -03:00
Roberto Ierusalimschy
0c8a7e071b 'mainthread' lives in 'allgc' list, like everybody else 2017-04-11 16:00:27 -03:00
Roberto Ierusalimschy
e4287da3a6 generational collector (still not complete) 2017-04-05 13:50:51 -03:00
Roberto Ierusalimschy
f5f3df3bd1 generational collection: new attempt (still incomplete) 2017-02-23 18:07:34 -03:00
Roberto Ierusalimschy
89e3a84344 removed field 'n' from 'CallInfo' (not being used right now) 2015-11-13 10:16:51 -02:00
Roberto Ierusalimschy
33b366ec32 added counters for total and individual CallInfo entries (to allow 
better syncronization between CallInfo size and stack size)
 2015-11-02 14:01:41 -02:00
Roberto Ierusalimschy
75d5a8924c macro 'addbuff' was using external variable name, instead of using 
its parameter name inside the macro.
 2015-10-20 11:11:05 -02:00
Roberto Ierusalimschy
41964648ee long strings are created directly in final position when possible 
(instead of using an auxiliar buffer to first create the string
and then allocate the final string and copy result there)
 2015-09-08 12:41:05 -03:00
Roberto Ierusalimschy
2b61360d82 avoid overflows (detected with 'clang -ftrapv') 2015-07-13 10:30:03 -03:00
Roberto Ierusalimschy
a80cada914 new cache for interning strings 2015-03-04 10:31:21 -03:00
Roberto Ierusalimschy
e75c0148c3 comments (references to "ANSI C" changed to "ISO C", which is the 
international name
 2014-11-02 17:33:33 -02:00
Roberto Ierusalimschy
28fdbcf393 added include for 'lprefix.h', for stuff that must be added before 
any other header file
 2014-11-02 17:19:04 -02:00
Roberto Ierusalimschy
57d8424146 extra space for new threads is initialized with a copy of the main 
thread, not of the creator thread.
 2014-07-24 13:17:56 -03:00
Roberto Ierusalimschy
baa0e23456 better support for extra user space associated with a Lua state 2014-07-24 11:00:16 -03:00
Roberto Ierusalimschy
ca41b43f53 type 'TString' refers directly to the structure inside the union 
(union used only for size purposes)
 2014-07-18 10:36:14 -03:00
Roberto Ierusalimschy
56137d58ff added check for conversion 'obj2gco' (and corrections for small 
problems detected by this check)
 2014-07-18 09:17:54 -03:00
Roberto Ierusalimschy
3f78de256e no need to keep threads in a different GC list, now that there is the 
'twups' list
 2014-02-18 10:46:26 -03:00
Roberto Ierusalimschy
d764cc5522 new list 'twups' to allow traversal of upvalues from dead threads 
(+ fixed some problems with cycles involving those upvalues)
 2014-02-18 10:39:37 -03:00
Roberto Ierusalimschy
de3b1c9b53 better control for number of finalizers called at each GC cycle 
(increases progressively)
 2014-02-13 12:46:38 -02:00
Roberto Ierusalimschy
733c58595b no more local collection 2014-02-13 10:11:34 -02:00
Roberto Ierusalimschy
ba3586cc90 keep a single list of objects to be finalized (with local and non-local 
objects), to ensure finalization order
 2014-02-11 10:18:12 -02:00
Roberto Ierusalimschy
de7cf8e63a bug: should call 'luai_userstateclose' only when 'luai_userstateopen' 
has been called before
 2013-11-08 15:34:22 -02:00
Roberto Ierusalimschy
4c6dfc342b CallInfo lists shrinks together with their associated stacks 2013-09-17 12:40:06 -03:00
Roberto Ierusalimschy
686e57cf9c GC local pause configurable 2013-09-13 13:21:52 -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
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
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
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
9a871dd3db tables and userdata all go to local list, too 2013-08-28 15:30:26 -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
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
439d74e29f added 'local' bit (true => object is only refered by local variables) 2013-08-16 15:55:49 -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
c6b6024a9d cast in 'luai_makeseed' now casts to needed type 2012-10-02 14:40:53 -03:00
Roberto Ierusalimschy
b77a90681e typos in comments 2012-05-30 09:33:44 -03:00
Roberto Ierusalimschy
054c20cd5b spaces -> tabs in #defines 2012-05-23 12:37:09 -03:00
Roberto Ierusalimschy
086da10dac merge of fields 'lastmajormem' (used in gen. mode) and 'estimate' 
(used in inc. mode)
 2012-05-22 14:50:39 -03:00
Roberto Ierusalimschy
b36b4b521f try to avoid sweeping new objects created with new white (and 
therefore not collectable in the current cycle)
 2012-05-22 14:32:25 -03:00
Roberto Ierusalimschy
f6a12948ca useless parameter in macro 'luai_makeseed' 2012-05-11 11:06:07 -03:00
Roberto Ierusalimschy
678c1255c9 random seed used in the hash of all strings to avoid intentional 
collisions
 2012-02-01 19:57:15 -02:00
Roberto Ierusalimschy
6d86b937e4 'luaC_separateudata' called from 'luaC_freeallobjects' + removed 
extra 'luaF_close' (already being called by 'close_state')
 2011-10-03 14:54:25 -03:00
Roberto Ierusalimschy
3dc5475e23 'nCcalls' should be local to each thread, as each thread may have its 
own C stack (with LuaThreads or something similar)
 2011-08-23 14:24:34 -03:00
Roberto Ierusalimschy
92afcf2823 no more 'luaH_setstr (used only once) + 'luaH_setint' receives value 
to be set.
 2011-08-09 17:58:29 -03:00
Roberto Ierusalimschy
551b076f1c change in the relationship between totalbytes and GCdebt - luaM_realloc_ 
is too critical to update two counters
 2010-12-20 17:40:07 -02:00
Roberto Ierusalimschy
737f119187 better control for GC running or stopped 2010-12-20 16:17:46 -02:00
Roberto Ierusalimschy
9b7a12c46d finalizers (__gc) for tables 2010-11-26 12:32:31 -02:00
Roberto Ierusalimschy
6828f6d427 new parameter 'majorinc' to control frequency of major collections 
in generational mode
 2010-09-03 11:14:01 -03:00
Roberto Ierusalimschy
dc2b8a0073 detail 2010-04-30 15:36:22 -03:00
Roberto Ierusalimschy
0062db1e06 default PAUSE should be 200 (differences may be corrected in 
internal multiplers)
+ check for unbilt stack equal the one used by lgc.c
 2010-04-30 11:22:23 -03:00
Roberto Ierusalimschy
3410dcd375 new way to control GC speed 2010-04-29 14:35:10 -03:00
Roberto Ierusalimschy
4db6f20770 ensure that 'luai_userstatethread' is always called (even if 
'stack_init' throws a memory error)
 2010-04-19 14:40:13 -03:00
Roberto Ierusalimschy
99a39f6f4a tells the allocation function that the first block is a thread 2010-04-19 13:34:46 -03:00
Roberto Ierusalimschy
afdb19ac82 no more 'ccall' nor 'cpcall' functions. (With light C functions they 
are obsolete.)
 2010-04-14 12:14:21 -03:00
Roberto Ierusalimschy
d20ff60615 new macro LUA_NUMTAGS 2010-04-12 13:07:29 -03:00