365b39e0cc
VM: change lot of visitation objects to use cool lambda functions instead
2015-07-06 22:49:38 +02:00
fce5d4353a
VM: replace binary_payload_start() with slot_count()
...
This should simplify a little the address calculations when visiting
slots of an object
2015-07-06 22:49:38 +02:00
66eb9f1fe5
VM: split the size() method into base_size() and aligned_size()
2015-07-06 22:49:38 +02:00
b7bd0668b3
VM: macro FACTOR_FOR_EACH to make stl container iteration easier to express
2015-06-05 02:14:48 +02:00
8e1f3a0e51
VM: after reset_datastack and retainstack clear the stack segment. makes
...
it easier to find bad accesses to shadow data
2015-05-22 12:22:28 -07:00
9bfc43144e
VM: debug macros FACTOR_PRINT and FACTOR_PRINT_MARK to make better debug
...
printing messages than just using std::cout
2015-05-09 20:04:21 -07:00
fa0ff77c45
VM: reset the unused parts of the data and retain stack segments with a bit pattern (good for debugging)
2015-04-29 09:31:54 -07:00
2cd2aa498a
VM: the compiler doesn't emit any live overinitialized locations anymore
...
Any overinitialized location immediately becomes garbage and is not
reused again. This is very nice because then the gc maps can be
simpler again.
2015-04-29 09:31:53 -07:00
1d251f7b6a
VM: new function visit_object to replace trace_object
2015-02-28 14:14:30 -08:00
b5a206c587
VM: new function visit_mark_stack() to factor out the mark stack iteration
2015-02-28 14:14:29 -08:00
5490c50217
VM: visit_roots and visit_contexts where always called in tandem, make a new method visit_all_roots that calls them both
2015-02-28 14:14:29 -08:00
4d7bc4c169
VM: merge of slot_visitor and code_block_visitor
...
Looks like there is no reason for them to be separate classes and by
merging them you can simplify lots of code.
2015-01-15 09:29:14 +00:00
af270cb4d8
VM converts the methods callstack::frame_top_at and code_block::offset to use cells instead of void*'s
2015-01-15 09:29:13 +00:00
eb7a5ab1ba
VM: some more fixing to make the types match better, converting void* to cell
2015-01-15 09:29:13 +00:00
1cec8a7814
VM: modifying the uninitialized_blocks map in place instead of creating a new one is more efficient
2014-12-04 10:23:34 -08:00
8741dbefce
VM: bignum_roots traversal code deleted!
2014-11-17 17:27:35 -08:00
85dade6914
VM: use a free_list_allocator for the callbacks, that way they can
...
individually be freed
2014-09-29 07:30:20 -07:00
4015565cf6
VM: "formalize" the callback_heaps object allocation using a
...
bump_allocator<code_block>
The advantage of doing it this way is that you can reuse the existing
each_object method to iterate all callbacks
2014-09-29 07:30:20 -07:00
237772daae
VM: dont scrub and check when visiting a callstack object
...
callstack objects doesn't have stacks so it doesn't make sense to scrub
and check. also added the methods check_stack and scrub_stack to deal
with code duplication
2014-09-08 14:54:18 -07:00
9836d6a1d0
VM: move the stack scrubbing logic to call_frame_slot_visitor to avoid visiting the callstack twice
...
primitive_minor_gc() iterates the stack twice, first to scrub stack
locations, then to trace overinitialized ones and gc roots. By running
visit_callstack() before visit_stack_elements() you only need to do it
once.
2014-09-08 14:54:18 -07:00
8fb8313251
VM: use the check_d and check_r to trace the overinitialized stack locations
2014-09-08 14:54:18 -07:00
a6e0867b22
VM: change type of bignum_roots and remove unnecessary not-null checks
...
Declaring bignum_roots to contain bignum** instead of cell avoids some
superfluous casts. Casting it to cell is wrong because the items in it
are never tagged. And due to a earlier commit, bignum_roots will never
contain NULL:s so checking for them is not needed.
2014-06-12 13:50:10 -07:00
0c9e61ca99
VM: refactoring that removes data_root_ranges
...
instead of storing data_root_ranges in data_roots, you can just store
cell pointers directly. the advantage with doing it that way is that
registration and traversal code becomes simpler (and slightly faster).
2014-06-12 13:50:10 -07:00
ffe41b3d7f
VM: Fixup cast formatting after clang-format
...
clang-format doesn't recognize casts to non-pointer/non-template types
so it winds up adding a space between the right paren and the expression
and then failing to recognize prefix operators in the process
(e.g. foo = (cell) & bar; should be foo = (cell)&bar;). This commit
manually fixes up the major cases (fixnum, cell, all types ending in _t).
2013-05-13 14:59:24 -04:00
88d7c10d03
VM: Replace u8-64, s8-64, cell, fixnum with stdint.h equivalents
2013-05-13 00:28:25 -04:00
7b1b1eef23
VM: Remove unnecessary _ suffix in constructors
2013-05-12 23:20:43 -04:00
26ff071205
VM: Remove unnecessary explicit keywords
2013-05-12 21:48:38 -04:00
6d5197923c
VM: Refactor slot_visitor.hpp to Factor style
2013-05-12 13:24:47 -04:00
ca2ea548be
VM: Remove exec bit from VM source files
2013-05-11 17:04:23 -04:00
d74f194b07
vm: remove _reversed from callstack iterator names
...
Now that they're the only game in town we can give them the short names
2011-12-14 09:56:49 -08:00
9236c05e60
vm: stage code block map fixup properly for GC
...
Don't update the map until the very last thing, and pass untranslated addresses to the iterator functors. Somewhat complicated by the fact that, for startup_fixup, the map is initialized with fixed-up values, so the fixup functor needs a flag indicating whether it operates with a fixed or unfixed code heap map.
2011-12-14 09:56:48 -08:00
e6a80aa9aa
vm: callstack_reversed in code_block_visitor
2011-12-14 09:56:48 -08:00
d12b6e375b
vm: iterate_callstack_reversed in slot_visitor
2011-12-14 09:56:47 -08:00
e116bc497b
vm: reversed callstack iteration
2011-12-14 09:56:47 -08:00
e2fafaed9c
WIP verify_callstack function
...
dumps callstack in reverse order so i can visually inspect that it matches the old forward walking logic
2011-12-14 09:56:47 -08:00
0a1a252cdb
vm: replace assert with FACTOR_ASSERT
...
Factor is finally a real C++ project and has a custom assert macro. Assertion failures were still getting caught as exceptions and causing failure loops. Write our own macro that calls factor::abort on failure.
2011-11-23 11:11:26 -08:00
c31530caec
vm: strip out call-counting profiler
...
This makes the separate "code" and "entry_point" fields in word and quotation redundant, so also remove them to reclaim an additional cell per word and quotation object, which should help with #318 .
2011-11-10 16:01:07 -08:00
be8db5aed3
vm: sample thread object rather than raw context
2011-11-02 13:23:18 -07:00
7da8a9776f
vm: sample code block owners instead of blocks
2011-11-02 13:23:17 -07:00
813e13140a
vm: fix 64-bit regression
2010-10-09 15:14:14 -07:00
0fc8ee8250
Fix win64 compile error
2010-10-09 16:03:46 -05:00
8a0d0fca83
Revert previous patch; not simply a typo. Still broken on win64.
2010-09-29 20:49:37 -05:00
466e630dc9
Win64 Windows SDK hates int/cell type errors
2010-09-29 17:43:27 -05:00
6b5fffc026
compiler: preliminary implementation of tracking derived pointers in GC maps
2010-09-27 20:12:44 -07:00
174463e156
vm: fix another compile warning on 32-bit Windows
2010-06-17 01:04:03 -05:00
e86f434f26
Add GC maps to ##box, ##box-long-long, ##alien-invoke, ##alien-indirect and ##call-gc; remove ##gc-map instruction
2010-06-14 19:39:46 -04:00
70b8a74208
vm: have to be extra careful when messing with return addresses
2010-06-13 18:12:57 -04:00
3f28c0e7f5
vm: context switching needs to scrub the return address
2010-06-12 19:54:31 -04:00
806e54630a
GC maps for more compact inline GC checks
2010-06-11 20:06:00 -04:00
560c119cd2
vm: actually use context callstacks when running code
2010-03-26 22:44:43 -04:00
Björn Lindqvist
Erik Charlebois
Joe Groff
Slava Pestov
Slava Pestov
Doug Coleman