33 commits

Author	SHA1	Message	Date
Alex Shinn	5266d15004	off by one error in computing tail jump offset	2011-07-05 21:38:17 +09:00
Alex Shinn	0de463a7c4	dropping needless drops	2011-07-04 23:07:19 +09:00
Alex Shinn	463400fc7e	skipping over calls to reserve on tail jumps	2011-07-04 22:27:48 +09:00
Alex Shinn	7ddfc9f25a	trying out tail jumps instead of using the tail-call opcode where possible	2011-07-04 02:22:18 +09:00
Alex Shinn	bacc7d9399	adding simple VM-level profiler	2011-07-03 21:00:11 +09:00
Alex Shinn	d7147bf67e	adding reserve opcode to avoid pushing dummy values. ... also adding register-optimization! primitive.	2011-06-13 20:52:42 +09:00
Alex Shinn	2961f23d1d	don't bother consing up rest arguments if they're never used	2011-06-11 14:52:30 +09:00
Alex Shinn	f15885a1b6	better simplification - old apply no longer needed	2011-05-22 03:31:39 -07:00
Alex Shinn	1fea7687e2	simplifying apply and fixing variadic applications	2011-05-22 03:21:44 -07:00
Alex Shinn	1e02e95590	srfi-99 fixes from rotty	2011-05-21 23:02:44 -07:00
Alex Shinn	c5126fb2b0	adding optional strict top-level matching for syntax-rules literals	2011-04-25 23:29:30 +09:00
Alex Shinn	226d98d49f	preserving port line number under all input operations	2011-04-09 15:55:26 +09:00
Alex Shinn	0f9a23f94f	-e/-p options in main had a long-standing stupid gc bug	2011-04-03 22:15:48 +09:00
Alex Shinn	017548cc46	adding string cursors, optimizing in-string(-reverse) loops	2011-03-21 14:57:52 +09:00
Alex Shinn	a7c346806c	Adjusting context parents stacks when growing stack.	2011-02-27 20:51:52 +09:00
Alex Shinn	01dd712a50	resume/cc can grow the current stack if the continuation is larger ... (as could happen resuming a continuation from a separate thread)	2011-02-27 18:14:22 +09:00
Alex Shinn	810e693a8c	begin more careful about syncing the stack top	2011-02-19 11:56:58 +00:00
Alex Shinn	52d0322910	auto-growing stack to counter for the extremely small default size	2011-02-12 21:20:12 +09:00
Alex Shinn	8d1db07541	need to write top whenever we might cons ... (consider always writing once at the start of the loop or using a simple vector for the stack and just undefining values when we pop)	2011-02-12 20:44:34 +09:00
Alex Shinn	14f586453f	debugging result when DEBUG_VM is set	2011-02-12 16:59:51 +09:00
Alex Shinn	2c036c7e7a	adding experimental auto-force capability	2011-01-11 22:54:23 +09:00
Alex Shinn	23382e9540	allowing per-heap maximum size limit as a new argument to sexp_make_eval_context ... can specify on the command line with -h 1M/2M (set to 0 for no limit)	2010-12-30 11:26:15 +09:00
Alex Shinn	c163599685	making stack trace check for null procedures	2010-12-29 16:13:30 +09:00
Alex Shinn	96161ffcf6	fixing non-threaded build	2010-12-26 14:24:19 +09:00
Alex Shinn	2affd2c677	creating initial API for C functions to block on input ports	2010-12-11 22:15:17 -08:00
Alex Shinn	6a239130af	don't preserve the same object multiple times in bytecode-literals	2010-12-06 21:37:54 -08:00
Alex Shinn	2eaf4068e5	several gc bugfixes (fixes #52)	2010-12-05 16:50:57 -08:00
Alex Shinn	c01721e110	Replacing eval opcode with compile + normal application.	2010-11-18 21:37:46 -08:00
Alex Shinn	5ec136adcb	more parameter updates ... exception handling in (chibi repl) still broken	2010-09-26 01:04:55 +09:00
Alex Shinn	83e91a20c7	supporting parameter converters on mutation, not just parameterize. ... also fixing bug in interaction-environment parameter from default repl.	2010-09-24 22:46:01 +09:00
Alex Shinn	276db59353	fixed parameter mutation (not triggered in current test suite, need to add a case for this)	2010-09-24 21:48:08 +09:00
Alex Shinn	b95a7cac42	parameters now check for thread-specific overrides (still need to set these in parameterize)	2010-09-24 07:23:43 +09:00
Alex Shinn	acf3d1c120	fixes for SEXP_USE_SAFE_ACCESSORS on macosx	2010-09-18 14:50:24 +09:00