224 Commits

Author	SHA1	Message	Date
Roberto Ierusalimschy	cbc59592ff	new definition for luaD_call' and luaD_adjusttop'	2001-06-08 16:01:38 -03:00
Roberto Ierusalimschy	762d059a13	new implementation for the Virtual Machine	2001-06-05 15:17:01 -03:00
Roberto Ierusalimschy	a68635a919	list constructors do not adjust last expression	2001-04-06 15:25:00 -03:00
Roberto Ierusalimschy	5e870f86a2	optimization for tailcall does not seem to pay itself	2001-03-07 10:22:55 -03:00
Roberto Ierusalimschy	a04de4f0ad	no more END opcode	2001-01-15 14:13:24 -02:00
Roberto Ierusalimschy	4894c27962	lua_Number defined in lua.h (1st version)	2000-12-04 16:33:40 -02:00
Roberto Ierusalimschy	e833bd47c9	comments	2000-10-24 14:05:59 -02:00
Roberto Ierusalimschy	9d60598260	better definitions for MULTRET	2000-08-29 11:48:16 -03:00
Roberto Ierusalimschy	44eb7d91ac	comments	2000-08-15 15:28:48 -03:00
Roberto Ierusalimschy	afef009fce	new version of debug system	2000-06-26 16:28:31 -03:00
Roberto Ierusalimschy	f517759507	details	2000-06-21 14:05:49 -03:00
Roberto Ierusalimschy	dbfb810267	cleansing of lparser.c	2000-06-05 11:56:18 -03:00
Roberto Ierusalimschy	5c2dd7a9e0	array `luaK_opproperties' keeps delta stack and mode for each opcode	2000-05-22 15:44:46 -03:00
Roberto Ierusalimschy	93d93a0bfb	first implementation of `for' over tables	2000-05-15 16:48:04 -03:00
Roberto Ierusalimschy	c9f91c6fe3	[C++ Warning]: W8084 Suggest parentheses to clarify precedence.	2000-04-27 14:39:15 -03:00
Roberto Ierusalimschy	634c3d57e9	optimization for SETLOCAL was too specific.	2000-04-14 14:46:15 -03:00
Roberto Ierusalimschy	e7c8393682	optimization INCLOCAL is not necessary, with `for'	2000-04-13 13:51:01 -03:00
Roberto Ierusalimschy	f9cf402fbd	first implementation of FOR	2000-04-12 15:57:19 -03:00
Roberto Ierusalimschy	1eabd0549e	comments	2000-04-07 16:35:31 -03:00
Roberto Ierusalimschy	14251c5a56	name changes + new operation - old operation	2000-04-07 10:12:50 -03:00
Roberto Ierusalimschy	4e7e9e8de5	new opcode INCLOCAL.	2000-04-04 17:48:44 -03:00
Roberto Ierusalimschy	63e230a16b	comments	2000-03-27 17:08:33 -03:00
Roberto Ierusalimschy	54633d7c7d	new name for llimits.h	2000-03-24 16:49:23 -03:00
Roberto Ierusalimschy	213e9febc8	limits now are in `llims.n'	2000-03-24 14:26:08 -03:00
Roberto Ierusalimschy	06f08f5634	small problems with limits and jumps	2000-03-16 15:03:09 -03:00
Roberto Ierusalimschy	f0b697e01c	details	2000-03-13 17:37:16 -03:00
Roberto Ierusalimschy	73aa465a8e	some name changes	2000-03-10 15:37:44 -03:00
Roberto Ierusalimschy	e3cf93ddb6	first implementation of "threaded code" for boolean operations	2000-03-09 10:57:37 -03:00
Roberto Ierusalimschy	88b306f495	some optimizations	2000-03-08 21:19:22 -03:00
Roberto Ierusalimschy	563de491be	a better way to control optimizations.	2000-03-04 17:18:15 -03:00
Roberto Ierusalimschy	d1ea38580a	change of code generation design (independent functions for each opcode)	2000-03-03 15:53:17 -03:00
Roberto Ierusalimschy	3c9d999424	many details (most by lhf).	2000-03-03 11:58:26 -03:00
Roberto Ierusalimschy	fa14cbf8e8	comments	2000-03-02 09:32:53 -03:00
Roberto Ierusalimschy	3bc925138e	first version of code optimizer	2000-02-22 11:31:43 -02:00
Roberto Ierusalimschy	3afe85b2ce	new version for INSTRUCTION formats	2000-02-14 14:51:08 -02:00
Roberto Ierusalimschy	52aad0ab59	removing implementation for better error messages (must rethink it...)	2000-02-11 14:52:54 -02:00
Roberto Ierusalimschy	1f691a4fcd	renaming of some opcodes and fields	2000-01-28 14:53:00 -02:00
Roberto Ierusalimschy	d11e5adf55	`const' array in protos breaked in 3 arrays (for strings, numbers, and ... prototypes).	2000-01-25 11:57:18 -02:00
Roberto Ierusalimschy	298d0abff7	first version of extra debug information (NAME)	1999-12-29 14:31:15 -02:00
Roberto Ierusalimschy	acdb0b741e	comments.	1999-12-27 15:33:22 -02:00
Roberto Ierusalimschy	33b8a01032	new way to code CALLs + passing multiple arguments between function calls	1999-11-25 16:59:43 -02:00
Roberto Ierusalimschy	36b6fe8d17	better treatment for arbitrary limits	1999-06-17 14:04:03 -03:00
Roberto Ierusalimschy	2679461637	no more assignment expressions (they don't fit in Lua...)	1999-03-10 11:09:45 -03:00
Roberto Ierusalimschy	0870a2d1d8	new opcode TAILCALL	1999-03-05 18:16:07 -03:00
Roberto Ierusalimschy	677188de8a	details.	1999-03-04 18:15:50 -03:00
Roberto Ierusalimschy	e63b542c9b	comments	1999-02-26 12:19:54 -03:00
Roberto Ierusalimschy	6a853fcb8b	details (from lhf)	1999-02-25 18:07:26 -03:00
Roberto Ierusalimschy	9d7bae0b6a	better instrumentation for internal debugging	1999-02-24 14:55:51 -03:00
Roberto Ierusalimschy	c364e9f97e	LONGARGW (for completeness).	1999-02-23 10:38:38 -03:00
Roberto Ierusalimschy	b82ff713e3	new name for opcode	1999-02-09 16:01:55 -02:00
Roberto Ierusalimschy	ad6c7b0dd4	small corrections in opcodes.	1999-02-09 13:59:10 -02:00
Roberto Ierusalimschy	8b2d97d187	assignment expression may be multiple	1999-02-08 16:54:19 -02:00
Roberto Ierusalimschy	fb1cf6ab2d	clearer way to set tables.	1999-02-08 15:07:59 -02:00
Roberto Ierusalimschy	cbc58af260	new opcode for "long" arguments (3 bytes)	1999-02-04 14:36:16 -02:00
Roberto Ierusalimschy	27407fc1f5	new syntax: assignment expressions + better order for opcodes	1999-02-02 17:41:17 -02:00
Roberto Ierusalimschy	1a17da2ff9	opcodes with builtin parameters are too complicated for very little extra ... performance.	1999-02-02 15:57:49 -02:00
Roberto Ierusalimschy	8f31eda649	details in comments	1998-06-25 11:37:00 -03:00
Roberto Ierusalimschy	2a2b64d6ac	opcode "CLOSURE" gets the prototipe (instead of a previous pushconstant)	1998-03-25 15:52:29 -03:00
Roberto Ierusalimschy	99cc4b20f2	details	1998-03-11 10:59:50 -03:00
Roberto Ierusalimschy	f0dffaa209	new way to hanlde arg information	1998-01-12 11:35:37 -02:00
Roberto Ierusalimschy	f76bca23ef	variants for "ARGS".	1997-12-30 17:08:23 -02:00
Roberto Ierusalimschy	4e0bf95622	variant opcodes for PUSHSELF	1997-12-29 15:35:46 -02:00
Roberto Ierusalimschy	18cd7adac6	optimization to handle <a.x> (new opcode).	1997-10-24 16:40:29 -02:00
Roberto Ierusalimschy	e78cf96c97	first version of Cclosures.	1997-10-24 15:17:24 -02:00
Roberto Ierusalimschy	3a89c973ff	details	1997-10-16 19:14:47 -02:00
Roberto Ierusalimschy	45ccb0e881	"nupvalues" is kept in Closure, not in prototype (as a preparation ... for C closures...)	1997-10-16 08:59:34 -02:00
Roberto Ierusalimschy	7c261a13b5	more uniform treatment to opcode variants.	1997-10-13 20:12:04 -02:00
Roberto Ierusalimschy	f6a9cc9a67	jumps are relative to next instruction	1997-10-06 12:51:11 -02:00
Roberto Ierusalimschy	28d47a0aaa	all jumps have byte variants; WHILE optimization	1997-10-01 17:05:34 -03:00
Roberto Ierusalimschy	0dd6d1080e	new opcode variants.	1997-09-24 16:43:11 -03:00
Roberto Ierusalimschy	d6c867ea50	better way to cope with opcode variants	1997-09-22 17:53:20 -03:00
Roberto Ierusalimschy	2079cfe8fa	new way to code globals, using const table instead of putting global ... index inside the opcode.	1997-09-19 18:17:52 -03:00
Roberto Ierusalimschy	dfe03c7abe	small optimizations (bit scrubbing)	1997-09-19 15:40:32 -03:00
Roberto Ierusalimschy	d985dc0629	Opcodes for Lua virtual machine	1997-09-16 16:25:59 -03:00