Commit Graph

297 Commits

Author SHA1 Message Date
Roberto Ierusalimschy 722bdbe17d no more 'getBMode'-'getCMode' (imprecise + we will need more space
for op mode) + better control of op modes
2017-09-28 13:53:29 -03:00
Roberto Ierusalimschy 00e728af88 binary operators use R instead of RK
faster + nobody uses RK(B), so B can be smaller (freeing one bit
for more opcodes, soon)
2017-09-26 15:14:45 -03:00
Roberto Ierusalimschy 13256a4173 detail
put explicit all cases for a switch of an enumeration
2017-09-26 14:10:49 -03:00
Roberto Ierusalimschy abb17cf19b new opcode OP_LOADF (load immediate float) 2017-09-19 15:38:14 -03:00
Roberto Ierusalimschy 80d9b09f35 jumps do not close upvalues (to be faster and simpler);
explicit instruction to close upvalues; command 'break' not
handled like a 'goto' (to optimize removal of uneeded 'close'
instructions)
2017-09-13 16:50:08 -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 b42430fd3a 'lineinfo' in prototypes saved as differences instead of absolute
values, so that the array can use bytes instead of ints, reducing
its size. (A new array 'abslineinfo' is used when line differences
do not fit in a byte.)
2017-06-27 08:35:31 -03:00
Roberto Ierusalimschy de74289049 table field names for dedicated opcodes can be restricted to
small strings for slightly faster access
2017-05-18 16:44:19 -03:00
Roberto Ierusalimschy 502a1d1108 new opcodes for table access with constant keys (strings and integers) 2017-04-28 17:57:45 -03:00
Roberto Ierusalimschy 173e41b2eb new opcode OP_ADDI (for immediate integer operand) (Experimental) 2017-04-26 14:46:52 -03:00
Roberto Ierusalimschy a3f9c1a77a detail (using unsigned comparison in range check for LOADI) 2017-04-25 17:01:14 -03:00
Roberto Ierusalimschy 6dbae1b5d9 registers in a VINDEXED expression must be freed in order 2017-04-25 15:28:25 -03:00
Roberto Ierusalimschy cb3d5dce30 opcodes 'OP_GETTABUP'/'OP_SETTABUP' operate only with string keys,
so they can use fast-track table access
2017-04-24 17:26:39 -03:00
Roberto Ierusalimschy 6a98aa0bb0 new opcode LOADI (for loading immediate integers) 2017-04-20 16:53:55 -03:00
Roberto Ierusalimschy 2a235312f0 detail (removing spaces at end of lines) 2016-12-22 11:08:50 -02:00
Roberto Ierusalimschy fc6b32bcaa bug: Lua can generate wrong code in functions with too many constants 2016-07-19 14:12:07 -03:00
Roberto Ierusalimschy 6487fb11fc all 'static' variables should be 'const' 2016-06-20 16:12:46 -03:00
Roberto Ierusalimschy fbd8614bdb comment (FALLTHROUGH) 2016-05-13 16:09:46 -03:00
Roberto Ierusalimschy 1f259be52a 'getcode' -> 'getinstruction' 2016-01-05 14:22:37 -02:00
Roberto Ierusalimschy b12b635a90 more refactoring 2016-01-04 11:40:57 -02:00
Roberto Ierusalimschy 384d1b47b0 comments (*lots* of them) + asserts 2015-12-18 11:53:36 -02:00
Roberto Ierusalimschy a01eba657e reordering of some functions + 'code_label' renamed to 'code_loadbool' 2015-12-17 13:02:44 -02:00
Roberto Ierusalimschy b7446ea88d detail ('while' -> 'for') in 'luaK_patchclose' 2015-12-17 12:52:53 -02:00
Roberto Ierusalimschy d103312661 details (typos in comments) 2015-11-19 17:16:22 -02:00
Roberto Ierusalimschy 82129b9266 use macro 'hasjumps' to test whether expression has jumps... 2015-10-26 12:27:47 -02:00
Roberto Ierusalimschy a1935b9cba error message ("too complex" -> "too many registers") + MAXREGS
changed to 255 (no reason not to use maximum allowed)
2015-04-29 15:24:11 -03:00
Roberto Ierusalimschy e723c75c02 details (avoid 'lint' warnings) 2015-03-28 16:14:47 -03:00
Roberto Ierusalimschy 8e5290d81e all "divisions" (div,idiv,mod) by zero are not folded, to avoid
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)
2014-12-29 14:49:25 -02:00
Roberto Ierusalimschy 2b83711fba new macro 'nvalue' (to convert an object to a float when we know
object is a number)
2014-12-19 11:36:32 -02:00
Roberto Ierusalimschy bf163ea7f0 bug ('#3' causes seg. fault in 5.3-beta) + comments + 'codearith' ->
'codeexpval' (confusion about what operations function accept was
one of the reasons for the bug)
2014-11-24 12:59:22 -02:00
Roberto Ierusalimschy 5fbd40dbe5 'x//y' extended to floats 2014-11-21 10:15:57 -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 4542fb9f78 comments (a few extra quotes around identifiers) 2014-10-30 16:53:28 -02:00
Roberto Ierusalimschy 4b526045fa comment 2014-10-28 15:27:41 -02:00
Roberto Ierusalimschy 491103708e 'MAXSTACK' -> 'MAXREGS' and moved to 'lcode.c' (there is no reason
to change that constant...)
2014-10-27 14:29:58 -02:00
Roberto Ierusalimschy bdf566a8a3 `name' in comments changed to 'name' 2014-10-25 09:50:46 -02:00
Roberto Ierusalimschy 42ff788e53 comments 2014-05-08 15:58:46 -03:00
Roberto Ierusalimschy ffa43df3cd 'setnvalue' -> 'setfltvalue' (for consitency with 'fltvalue') 2014-04-29 15:14:16 -03:00
Roberto Ierusalimschy e98ba351ce n^-m gives float result (instead of error) 2014-04-27 11:41:11 -03:00
Roberto Ierusalimschy f61ceee708 LUAI_FUNC is being used only in header files 2014-04-02 13:44:42 -03:00
Roberto Ierusalimschy 607be77ec8 some details to avoid warnings 2014-04-01 11:39:55 -03:00
Roberto Ierusalimschy da4811238a details (typos in comments) 2014-03-21 10:52:33 -03:00
Roberto Ierusalimschy 2aff75f58e small bug (from revision 2.80): LUA_OP* constants are not from
enumeration 'Opcode' (they are integers)
2014-03-09 16:21:34 -03:00
Roberto Ierusalimschy cd12410195 new macro 'cast_void' 2014-03-07 13:19:00 -03:00
Roberto Ierusalimschy 5ff1c18a71 back with 'L' for macros 'luai_num*', but now with a new macro
'luai_numinvalidop' to protect constant folding
2014-03-06 13:15:18 -03:00
Roberto Ierusalimschy 99ac4a260f 'constfolding' passes a proper Lua state to 'luaO_arith' 2014-03-06 10:58:28 -03:00
Roberto Ierusalimschy fa7e77fd38 detail ('codearith' uses 'LUA_OP*' constants instead of 'OP_*') 2014-03-06 10:39:05 -03:00
Roberto Ierusalimschy 19ad78d713 C++ needs casts for arithmetic with enums 2014-02-06 17:55:55 -02:00
Roberto Ierusalimschy 86dd8bf3f5 no more 'L' in macros "luai_num*" (several places that use those macros
cannot throw errors anyway...)
2014-01-27 11:34:32 -02:00
Roberto Ierusalimschy 1ea2d20f74 first implementation of '<<', '>>', and '~' (bitwise not) 2013-12-30 18:47:58 -02:00
Roberto Ierusalimschy f5133aa1a5 small change in handling of unary operations 2013-12-18 16:44:42 -02:00
Roberto Ierusalimschy c0edab0f6d first implementation of bitwise operators '&' (band), '|' (bor),
and '~' (bxor)
2013-12-18 12:12:03 -02:00
Roberto Ierusalimschy a948054a19 new order for binary operations (grouping them by type of result) 2013-12-16 17:06:52 -02:00
Roberto Ierusalimschy a8f8c7fd80 integer exponentiation with negative exponent is invalid 2013-12-16 12:30:22 -02:00
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 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 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 1f2b82bf25 correct way to avoid compile-time errors in integer divisions 2013-05-06 14:22:16 -03:00
Roberto Ierusalimschy 67532d5a10 constant folding and API arithmetic with integers 2013-05-02 09:37:24 -03:00
Roberto Ierusalimschy 00744a2729 'luaO_arith' -> 'luaO_numarith' 2013-04-29 13:57:48 -03:00
Roberto Ierusalimschy a2f5c28a80 new operation '//' (integer division) 2013-04-26 10:08:29 -03:00
Roberto Ierusalimschy 5951c79ae1 default increment for 'for' loop is an integer (1, not 1.0) 2013-04-25 16:35:19 -03:00
Roberto Ierusalimschy 1294b09d8e first implementation of literal integers (no constant folding yet) 2013-04-16 15:46:28 -03:00
Roberto Ierusalimschy 7a543cfae6 2013-04-15 12:43:34 -03:00
Roberto Ierusalimschy a3e1c40d6d remove of unecessary luaD_checkstack. (In some cases, C should
ensure stack space; in others, Lua can use the extra slots for
temporary values.)
2012-08-16 14:34:28 -03:00
Roberto Ierusalimschy 53480f1163 details (remove of some extra spaces) 2012-08-14 15:12:34 -03:00
Roberto Ierusalimschy ca458f1f52 'luaK_goiffalse' exported 2011-08-30 13:26:41 -03:00
Roberto Ierusalimschy 51471ba748 comments about use of 'luaH_set' 2011-08-15 16:41:58 -03:00
Roberto Ierusalimschy 5ab6d36d99 no more field 'L' in 'FuncState' 2011-07-27 15:07:37 -03:00
Roberto Ierusalimschy 32a12e2f3f detail (cleaned whitespaces at end of lines) 2011-07-15 09:50:29 -03:00
Roberto Ierusalimschy 821bd7025e more uniform names for 'equalobj'-related functions 2011-05-31 15:27:56 -03:00
Roberto Ierusalimschy 9b7dddad7d no need for two different implementations for equality (one raw and
one with metamethods)
2011-05-31 15:24:36 -03:00
Roberto Ierusalimschy d120ec29ca bug in OP_SELF when method name goes to a register 2011-04-28 11:00:11 -03:00
Roberto Ierusalimschy 4758113043 change in opcode OP_LOADNIL: B is used as a counter instead of a
register. (Avoids an assignment to R(B), not present in any other
instruction.)
2011-04-19 13:22:13 -03:00
Roberto Ierusalimschy 427ee519db new instruction OP_LOADKX (to replace OP_LOADK with extra argument) 2011-04-07 15:14:12 -03:00
Roberto Ierusalimschy dd547c55c8 new scheme to close upvalues in 'break'; jump instructions may
do the close, avoiding the need for a OP_CLOSE instruction
2011-02-01 16:03:10 -02:00
Roberto Ierusalimschy 026a7e92be bug in some short-circuit optimizations 2011-01-31 12:28:41 -02:00
Roberto Ierusalimschy 81dd13f4c6 new way to distinguish between indexing tables in registers and
tables in upvalues (+ fixed small bug when checking conflicts in
multiple assignments)
2010-07-07 13:27:29 -03:00
Roberto Ierusalimschy 6a02bbe1e2 better organization for fields in struct 'expdesc' 2010-07-02 17:42:40 -03:00
Roberto Ierusalimschy a71c5f6f53 typo in comments 2010-06-30 11:11:17 -03:00
Roberto Ierusalimschy c066ac10a9 better line numbers for function calls and unary/binary operators +
null statement + no more "ambiguous syntax" restriction
2010-04-17 09:46:32 -03:00
Roberto Ierusalimschy 25c557ec63 first version of _ENV; no more global variables 2010-03-12 16:14:06 -03:00
Roberto Ierusalimschy 0fe2576a39 new instructions to optimize indexing on upvalues 2010-02-26 17:40:29 -03:00
Roberto Ierusalimschy cef96b73e1 added casts from int to enumerations to follow C++ rules 2010-01-11 15:38:30 -02:00
Roberto Ierusalimschy fcc46467fa limit of constants per function changed to 2^26 using extra arguments
to opcodes LOADK, GETGLOBAL, and SETGLOBAL
2009-09-23 17:33:05 -03:00
Roberto Ierusalimschy 1b4480003b "stange numbers" (-0 and NaN) also go to the constant table (as
strings with their binary representation). Therefore, constant
folding may produce these results.
2009-08-10 12:31:44 -03:00
Roberto Ierusalimschy d7872dcf91 small optimization (reorder of BinOpr enum to unify some cases
in switches)
2009-06-18 13:35:05 -03:00
Roberto Ierusalimschy 96fa34ec32 constant folding uses 'luaO_arith' to perform basic operations 2009-06-17 14:49:09 -03:00
Roberto Ierusalimschy 3db5f60547 BUG: "(((1 or false) and true) or false)" gives wrong result 2009-06-15 10:52:08 -03:00
Roberto Ierusalimschy 2598138ece new function 'luaK_codek' (detail) 2009-06-10 13:52:03 -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 a4d3080fe3 SETLIST extra argument now is an "instruction" (OP_EXTRAARG) 2008-04-02 14:38:54 -03:00
Roberto Ierusalimschy 81fc3c4f45 bug: code generated for "-nil", "-true", and "-false" is wrong 2007-05-04 15:41:49 -03:00
Roberto Ierusalimschy 5f4f0d4a63 no more optimization to avoid LOADNIL at function start 2007-03-27 11:11:38 -03:00
Roberto Ierusalimschy 3b19bd4d57 bug: wrong code when constant is coded after it should be 2007-03-09 15:50:56 -03:00
Roberto Ierusalimschy 742b8be0c1 detail 2006-10-10 14:39:00 -03:00
Roberto Ierusalimschy 3c8865cf66 detail 2006-09-22 17:24:38 -03:00
Roberto Ierusalimschy bd869c7b31 details 2006-09-14 15:42:28 -03:00
Roberto Ierusalimschy d5a23dde90 little better error messages for internal arrays overflows 2006-09-14 09:59:06 -03:00
Roberto Ierusalimschy dfe2f1eeff macros luai_num* take a state L (when available) as argument, to allow
them to generate errors (and other facilities)
2006-08-07 16:14:30 -03:00