Commit Graph

2905 Commits

Author SHA1 Message Date
Roberto Ierusalimschy a3af42de8e patch for "blow stack" bug 2008-05-08 13:55:08 -03:00
Roberto Ierusalimschy 198be23f36 added structure for local-variable information to allow extra
checkings if needed
2008-05-08 12:44:51 -03:00
Roberto Ierusalimschy 24359ae294 'string.find' cannot find things after subject's end 2008-04-14 12:54:59 -03:00
Roberto Ierusalimschy 6c84722afa make default GC behavior a little more agressive
(to avoid falling behind memory consumption)
2008-04-07 15:58:42 -03:00
Roberto Ierusalimschy facb0519a0 two bugs: invalid boolean values in constant table + too deep recursion
when reading nested functions
2008-04-07 15:44:23 -03:00
Roberto Ierusalimschy ffdca3522e 'table.sort' detects invalid order function before calling it
for nil elements
2008-04-07 15:43:00 -03:00
Roberto Ierusalimschy 4d7469b610 avoid constant folding for -0 (to avoid it colapsing to 0) 2008-04-07 15:41:47 -03:00
Roberto Ierusalimschy 2b84e36b93 patches for some bugs 2008-04-04 22:31:59 -03:00
Roberto Ierusalimschy 85bda9eef5 bugs: precheck must use check (instead of assert) and ensures that
code size is at least 1
2008-04-02 16:14:16 -03:00
Roberto Ierusalimschy a4d3080fe3 SETLIST extra argument now is an "instruction" (OP_EXTRAARG) 2008-04-02 14:38:54 -03:00
Roberto Ierusalimschy 5627d51e05 corrected error message ("too many constants" -> "too many functions") 2008-04-02 14:19:22 -03:00
Roberto Ierusalimschy e64b0cada8 several bugs related to precompiled code 2008-04-01 16:17:48 -03:00
Roberto Ierusalimschy 7316d61a66 'coroutine.running' should work for the main thread too 2008-02-25 11:33:57 -03:00
Roberto Ierusalimschy e2b366c760 userdata with finalizers are kept in a separated list 2008-02-19 15:55:09 -03:00
Roberto Ierusalimschy fa19baab7f proxies must be created with a __gc field in their metatables to work
properly with new semantics for finalizers
2008-02-19 15:54:09 -03:00
Roberto Ierusalimschy 2f8351fe48 turn off optimizations when testing 2008-02-15 15:44:34 -02:00
Roberto Ierusalimschy 32d42bdcdc bugs: lua_checkstack may have arithmetic overflow for large 'size' +
unpack with maximum indices may crash due to arithmetic overflow
2008-02-14 14:51:57 -02:00
Roberto Ierusalimschy 843d53aabb 'table.concat' may get confused with too large limits 2008-02-14 14:03:27 -02:00
Roberto Ierusalimschy 5ac3386888 bug: unpack with maximum indices may crash due to arithmetic overflow 2008-02-14 14:03:09 -02:00
Roberto Ierusalimschy 6d182faab6 bug: lua_checkstack may have arithmetic overflow for large 'size' 2008-02-14 14:02:58 -02:00
Roberto Ierusalimschy f86f4116c8 micro-optimization: avoid one API call in 'read_chars' 2008-02-12 15:05:36 -02:00
Roberto Ierusalimschy 7f69f0efb0 remove of useless test in 'read_chars' (if n == 0 then the resulting
string cannot be empty).
2008-02-12 14:51:03 -02:00
Roberto Ierusalimschy f6cab9666c corrected name of 'Patrick Donnelly' 2008-02-12 14:42:46 -02:00
Roberto Ierusalimschy 3deb38814d added assertions checking maximum number of upvalues 2008-02-12 11:34:12 -02:00
Roberto Ierusalimschy 3b5b14a085 LUAI_MAXCSTACK must be smaller than -LUA_REGISTRYINDEX +
coroutine.resume pushes element without ensuring stack size
2008-02-11 17:18:21 -02:00
Roberto Ierusalimschy aec671c126 maximum C stack should reserve some values for pseudo-indices 2008-02-11 17:17:19 -02:00
Roberto Ierusalimschy dd92af69db bug: pseudo-indices should be linked to maximum C stack size 2008-02-11 17:16:02 -02:00
Roberto Ierusalimschy 015e973899 bug: auxresume should reserve stack space for boolean result 2008-02-11 17:14:52 -02:00
Roberto Ierusalimschy 163f58dac5 in 'testC', booleans should be represented by booean values 2008-02-11 17:04:16 -02:00
Roberto Ierusalimschy dcee7d2fe0 new instructions 'newtable' and 'checkstack' in 'testC' 2008-02-11 16:04:26 -02:00
Roberto Ierusalimschy 5f0a2f6248 more options for 'kinds' of Garbage Collection + more agressive
shrinking of stacks and string hash
2008-02-11 13:46:03 -02:00
Roberto Ierusalimschy a73930228a more options for 'kinds' of Garbage Collection 2008-02-11 13:45:30 -02:00
Roberto Ierusalimschy a76fa25199 'ceillog2' now is exported (other modules may need it) 2008-01-30 16:05:23 -02:00
Roberto Ierusalimschy dd3519ab89 comment (LUA_GLOBALSINDEX does not need write barrier in lua_replace) 2008-01-25 11:42:12 -02:00
Roberto Ierusalimschy 3a515df086 debug.sethook/gethook may overflow the thread's stack 2008-01-21 12:41:19 -02:00
Roberto Ierusalimschy 74897bc52f bug: debug.sethook/gethook may overflow thread's stack 2008-01-21 11:37:08 -02:00
Roberto Ierusalimschy e7c989baf1 avoid calling "tail return" hooks if the hook itself turns off the event 2008-01-18 20:36:50 -02:00
Roberto Ierusalimschy 7a3c8314ac small bug (see http://lua-users.org/lists/lua-l/2007-08/msg00350.html) 2008-01-18 15:14:47 -02:00
Roberto Ierusalimschy 6f8257ec5f it seems useless trying to avoid this warning... 2008-01-18 13:37:10 -02:00
Roberto Ierusalimschy 87b0191753 better definition for LUAI_MAXCSTACK 2008-01-17 14:24:38 -02:00
Roberto Ierusalimschy 56c17ba4c7 detail 2008-01-17 14:24:30 -02:00
Roberto Ierusalimschy daddc57abd luaL_tostring -> luaL_tolstring (more generic) 2008-01-03 15:07:59 -02:00
Roberto Ierusalimschy f9cdd09191 larger limit for MAXCSTACK (plus removing of some useless definitions) 2008-01-02 14:36:19 -02:00
Roberto Ierusalimschy 47b4858357 detail ('name' of baselib is "_G") 2008-01-02 14:16:28 -02:00
Roberto Ierusalimschy b8d23d6e07 option to use Sun compiler (on lhf account) 2007-12-27 10:59:43 -02:00
Roberto Ierusalimschy fdecdd0cec stand-alone interpreter shows incorrect error message when the
'message' is a coroutine (already fixed in 5.2)
2007-12-27 10:58:29 -02:00
Roberto Ierusalimschy fabdaa37b6 'lua_pushfstring' raises an error if called with invalid format 2007-12-19 15:24:38 -02:00
Roberto Ierusalimschy bc82b4d78a new function 'package.searchpath' 2007-12-12 12:36:12 -02:00
Roberto Ierusalimschy dae850e0ee details 2007-12-08 09:54:32 -02:00
Roberto Ierusalimschy abcc124df0 BUG: lua_setfenv may crash if called over an invalid object 2007-11-28 16:27:38 -02:00