Torsten Ruger
f798173a09
change to_risc and builtin code according to last commit
...
Wherever space was loaded to get to the next_message
we now load the Message factory.
Otherwise much the same, only the attribute is next_object, not next_message
The binary is growing with 1k objects per factory, so i had to fix (hack) arm to handle bigger constants
close #14
2018-09-01 11:28:53 +03:00
Torsten Ruger
d964e9ea9d
let spce keep the messages in a factory #14
...
Like Integers and addresses before, messages are now in a factory
Factories keep allocated (uninitialised) objects, had to make init public to call it
2018-09-01 11:24:16 +03:00
Torsten Ruger
0a390cc5a9
shaves an instruction off resolve_method
...
by loading nil directly, not space first
2018-09-01 11:20:59 +03:00
Torsten Ruger
b2339dc330
fix address nil values
2018-08-30 16:38:00 +03:00
Torsten Ruger
dc12c1d70b
add the addresses from labels as constants
2018-08-29 21:06:29 +03:00
Torsten Ruger
ea7f3c9653
remove the last_object from chain
...
also tried to keep the first around as was done in space
partially to not have to add the consrtants seperately
didn't work, as chain gets broken
also this has to be redone anyway
2018-08-29 21:05:54 +03:00
Torsten Ruger
c983dcf0eb
move return address generation to factory
...
removes the list from space
adds a ReturnAddress factory instead
and uses these throughout
2018-08-29 21:02:49 +03:00
Torsten Ruger
f993ccefe3
litte bit of docs
2018-08-24 18:49:44 +03:00
Torsten Ruger
71ab369c71
use factory to generte intergers in space
...
start with just integer factory in space
change all the hand-out code
still #14
2018-08-24 18:49:21 +03:00
Torsten Ruger
d396da16e3
start with #14 by implementing factory
...
page was maybe a too low level name
pages may be the unit of the syscall, but after that objects desolve (maybe later to be added on from different pages)
Factory has the job of handing out a new instance of a type
it keeps a freelist for that and a reserve
2018-08-23 19:55:06 +03:00
Torsten Ruger
9687d6611f
avoid adding risc instructions twice
...
that causes loops in the chain
infinite loops in the code that are hard to debug
closes #11
2018-08-19 17:29:04 +03:00
Torsten Ruger
f85fe8a2cb
fix bug in slot_load and definition
...
move parfait helper for reuse
2018-08-19 15:36:51 +03:00
Torsten Ruger
57dc6c45bb
remove the code_builder
...
this is core of #11
rename compiler_builder to just builder
and change all builder uses to use that
some test change as code is not returned anymore
2018-08-19 13:16:07 +03:00
Torsten Ruger
da9dc30c20
move dynamic_call to builder
...
no change of test, great
also remove some duplicate code that was hanging around in method_compiler
2018-08-16 10:43:41 +03:00
Torsten Ruger
ce157ffa94
change block_yield to_risc to use builder
...
only changes the order of two instructions
2018-08-16 08:58:49 +03:00
Torsten Ruger
1dabe0fda1
finish the idea behind #8 , conditionally creating variables
...
by using space? , the ? makes for conditional creation of the variable
This is especially useful for constants (ie space)
2018-08-15 19:59:17 +03:00
Torsten Ruger
01752aab05
remove the last use of an add_xx function, closing #9
2018-08-15 19:37:03 +03:00
Torsten Ruger
43fa7ccbcc
redid div10 with builder
...
div10 is right at the edge of what can be understood
no matter the (assmbler) syntax
2018-08-15 19:30:40 +03:00
Torsten Ruger
252ae6de72
finally get rid of the fixme in div10
...
create (load/reduce) the int once and transfer.
Save a cruicial 2 instructions
Also expanded the variable name possibilities with _self, __const , _1 and _2
2018-08-15 18:18:21 +03:00
Torsten Ruger
15337e10be
move putstring to use builder
2018-08-15 17:52:21 +03:00
Torsten Ruger
e953cc90d5
remove some more add_xx functions
...
also load labels by dsl now
2018-08-14 20:24:48 +03:00
Torsten Ruger
520dc7b41f
remove ad_transfer in save_message
...
towards removing all add_
change to set_builder and chaining in reg values
2018-08-14 20:08:58 +03:00
Torsten Ruger
37461a1727
add the ! syntax to create variables in builder, fix #8
...
now a variable has to be created before being used
thus it is save to develop contracts where a certain name
must exist in the scope
Maybe the syntax starts getting a bit weird, but at least the ! is a common symbol in ruby
2018-08-14 19:39:46 +03:00
Torsten Ruger
0bf008b351
bit of documentation
2018-08-14 11:23:19 +03:00
Torsten Ruger
fb54d68020
redo restore after syscall with builder
2018-08-13 18:48:54 +03:00
Torsten Ruger
55bc9c5273
rewrite new_int with builder
2018-08-13 18:02:34 +03:00
Torsten Ruger
fee9e261a5
still some names that needed changing
2018-08-12 14:48:20 +03:00
Torsten Ruger
3d113b4d83
fix more of the changed names
...
add a few self.
2018-08-12 13:10:44 +03:00
Torsten Ruger
66e0d4ae26
cache index resolution
...
resolve once and reuse
also use non checking version of get/set
as index per definition must be ok
This creates some speedup, but mostly avoids some weird hang in Thread::Queue.pop for many seconds
2018-08-12 13:09:34 +03:00
Torsten Ruger
84de400529
some test change because of renaming
...
next is a keyword, can't be used as instance writer
fake_memory gets the pobject for debug
2018-08-11 19:17:20 +03:00
Torsten Ruger
e6df473647
changing the adapter to have fake memory for every object
...
Before instance variables were used to store the actual data
While this worked it was a double mental loop and (more importantly) did not allow the same memory access
Ie, in the interpreter the types for args had to be set correctly
But they don't need to be until we walk the list, when we can get the types from the method.
In short, now the fake memory may be used as memory, by indexing into it
We now have to use attr reader/writers that map to the fake memory
And while this is possible, it means a lot more self. than i would like
Also the compiler whill have to undo those attr to instance acess at some point
2018-08-11 19:13:09 +03:00
Torsten Ruger
d74e9c2c40
removed some obsolete builder helpers
...
cleaner code with dsl,
just div10 undone
2018-08-09 21:10:05 +03:00
Torsten Ruger
99a95d40a6
also redid the get/set internal byte with builder
...
had to pimp register value to use a semblance of the dsl
using <= for bytes and << for words
2018-08-09 20:16:55 +03:00
Torsten Ruger
7002956e81
redid get_set internal byte with builder
...
trying to change style, get rid of some of builders functions
2018-08-09 20:15:52 +03:00
Torsten Ruger
0d52b620ed
redid integer comparison with builder
...
had to add some helpers to builder
2018-08-08 20:53:06 +03:00
Torsten Ruger
fc6aa4e28b
rework operator with builder
2018-08-08 15:49:47 +03:00
Torsten Ruger
78466090b3
add reduce_int and fix bug
...
hardcoded operator, buh
reduce_int to avoid that long constant
2018-08-08 15:49:07 +03:00
Torsten Ruger
6200a35562
add shortcut for class_name
...
for when type is not set
2018-08-08 12:02:59 +03:00
Torsten Ruger
71a7161200
redid div4 with builder
2018-08-08 12:02:24 +03:00
Torsten Ruger
c63e55c2bc
add an operator function
...
to be able to use the builder more
also add ability to load fixnums (for the int functions)
2018-08-07 20:48:36 +03:00
Torsten Ruger
595e032edf
add function_return helper to risc_value
...
create a function_return from within builder
2018-08-06 13:03:33 +03:00
Torsten Ruger
c9d8f750e5
extend type inference for dsl
...
and finally test it
2018-08-06 10:11:12 +03:00
Torsten Ruger
5346077a72
use the return jump to jump to the return sequence
...
thus every method only has one exit
should make multi return messages smaller
especially when we have escape analisis
(maybe will help with inlining too)
2018-08-02 17:36:39 +03:00
Torsten Ruger
4b4528abb2
propagate constants from block_compiler up
...
up to method_compiler, where they are collected by mom_compiler
(and included in binary)
2018-08-01 16:27:34 +03:00
Torsten Ruger
04bcfea8ce
fix scoping of blockcompiler
...
Was accessing caller scope, but must use caller's caller as the yield itself is a call.
2018-07-31 18:00:42 +03:00
Torsten Ruger
9c6a099cde
block test
...
working on return semanitcs
(still avoiding the implicit return)
2018-07-30 14:10:24 +03:00
Torsten Ruger
165036ea39
misc
2018-07-30 10:26:47 +03:00
Torsten Ruger
2dc03f8d1b
push the callable into the callable compiler
...
thus generalizing for .callable access
keep block and method aliases for destinction in derived classes
2018-07-30 10:26:11 +03:00
Torsten Ruger
285a88b59f
generalize assemblers to use callables
...
not just methods, they are almost the same anyway
2018-07-30 10:23:42 +03:00
Torsten Ruger
4055709529
push the name down into callable
...
blocks need a name too
if just for debug, and stacks
2018-07-30 10:21:43 +03:00
Torsten Ruger
1cb07a4164
block arg access was off by one
2018-07-27 12:16:06 +03:00
Torsten Ruger
e1d5592c55
fix boot to include blocks correctly
2018-07-27 10:48:45 +03:00
Torsten Ruger
f5c284b3a0
bring the blocks down to mom level
...
reusing message_setup, but adding yield specific instructions
2018-07-24 11:35:49 +03:00
Torsten Ruger
d80ef4bf4e
store method, not name, in message
...
changes the calling and thus too many test
2018-07-22 22:29:32 +03:00
Torsten Ruger
3c1137066b
more block tests reveal a compiler bug
...
copy/paste, args twice in switch
2018-07-18 10:13:19 +03:00
Torsten Ruger
ff8b95f21a
get block resolution to use the extra
...
block_compiler puts in the extra, that we get out when resolving the type
Thus block args work, though only by assuming direct call
2018-07-17 10:37:33 +03:00
Torsten Ruger
280ea8a8c4
remove resolve_to_register
...
which was, quite simply, from another era when more than message was pinned
2018-07-16 19:19:49 +03:00
Torsten Ruger
73645e1507
rename resolve_type to infer_type
...
as we go from a symbol to a type, more or less by convention
also remove the compiler from there and other not used cases
No tests change, good refactoring
2018-07-16 19:17:57 +03:00
Torsten Ruger
4cc04787e9
remove Risc.resolve_to_index
...
mostly by using better typed registers,
which cleans up the code where it was used
2018-07-16 19:00:04 +03:00
Torsten Ruger
29363e7f72
split compilers resolve_type into the three possibilities
...
cleaner code, though temporary it shifts some dirt
into the index method. up next
2018-07-16 12:03:40 +03:00
Torsten Ruger
3343017dba
move type resolution to compiler
...
from risc_value.
also unite mock compilers
2018-07-16 11:46:18 +03:00
Torsten Ruger
e3673e579c
pass extra info into register init, not just value
...
extra info may be hash, maybe just type info
2018-07-16 11:23:09 +03:00
Torsten Ruger
3bc85805a4
must pass registers to slot_to_reg and reg_to_slot
...
as they are typed, those functions don't resolve on Risc, but the register type
miscother changes from previous commits
2018-07-15 16:30:50 +03:00
Torsten Ruger
f31d22d901
use real type in register, not symbol
...
this has some more consequences, upcoming
2018-07-15 15:16:12 +03:00
Torsten Ruger
a71a6d34fb
use message reg not message
...
for upcoming change
2018-07-15 15:14:38 +03:00
Torsten Ruger
58c7294abd
finish the loop in slot_load
...
thus rewrite of that old nested if thing is done
2018-07-14 11:04:21 +03:00
Torsten Ruger
61c840c023
start to move slot_load code to register_value
...
the iea is to iterate through register_values while reducing the slot_load to a number of Slot_to_regs
wip
2018-07-13 21:56:55 +03:00
Torsten Ruger
21009b0e9b
delegate resolve to compiler when needed
...
and forgotten space test
2018-07-13 21:55:27 +03:00
Torsten Ruger
a095515b0e
new get_type_by_class_name helper for space
...
sorely needed, with 36 occurences replaced
2018-07-13 21:50:40 +03:00
Torsten Ruger
27a142f2a3
use compiler base class for blocks too
...
can now go to mom level and add
test harness still looks overly complicated, but works
first block tests at mom level
2018-07-10 22:03:32 +03:00
Torsten Ruger
a3e758357c
extract base class from method_compiler
...
to be used for blocks too
2018-07-10 22:02:11 +03:00
Torsten Ruger
edea9ac080
makeing the method of the compiler more or less private
...
in preparation for using the same code for bocks
2018-07-09 19:32:17 +03:00
Torsten Ruger
63b55f2aa4
refactor a bit for upcoming delegations
...
must get method to be private
more tell, not ask
2018-07-09 18:16:51 +03:00
Torsten Ruger
4ac89ece66
start delegating scope matters to the compiler(s)
...
slot_type_for to return the slot way to access variable
this is off course version 0.0.1 alpha, no types are checked or errors handled
2018-07-09 17:53:56 +03:00
Torsten Ruger
dd544214b3
start with block_compiler
...
as a copy of method_compiler
re-merge later, when we know what's needed
2018-07-09 16:48:23 +03:00
Torsten Ruger
0d900de695
block creation and insertion with correct types
2018-07-07 22:42:00 +03:00
Torsten Ruger
2f07cc34f3
add parfait block
2018-07-07 15:50:43 +03:00
Torsten Ruger
9005513368
rename typed_method to callable_method
...
seems to make the essence clearer
also extracted base class
2018-07-07 09:11:09 +03:00
Torsten Ruger
acd5cd8f30
rename for_type to self_type
...
and split a base class off TypedMethod
2018-07-06 20:01:17 +03:00
Torsten Ruger
d8b58d8da6
fix inital main (again)
...
stop even compiling a fake main if a real exists (in builtin)
previous version was still buggy: using builting methods types (especially locals) even comiled version had different
2018-07-04 09:17:30 +03:00
Torsten Ruger
6f936f190d
misc to_s and small fixes
2018-07-04 08:28:29 +03:00
Torsten Ruger
e099014d63
fix dunamic jump in interpreter and misc
2018-07-03 19:15:36 +03:00
Torsten Ruger
bb1d1495db
fix constant propagation through the layers
...
so they can end up in the binary
2018-07-03 10:12:22 +03:00
Torsten Ruger
56f2742e89
fix code listener logic bug
...
next's position was only updated initially
previous logic relied on pushing everything long rightly
but it wasn't right, and no harm in setting position twice,
as events only fire if there is real change
2018-07-02 23:03:00 +03:00
Torsten Ruger
a1197fb70c
interpreter tests working again
2018-07-02 17:29:26 +03:00
Torsten Ruger
6af651a886
text writer working
...
adjusted ordering as psoitioning, ie
methods in order of the assemblers of the linker
(not the object space)
2018-07-02 17:05:02 +03:00
Torsten Ruger
474cd4b348
linker and collector test work again
...
had to give space a fixed number of addresses
2018-07-02 16:19:01 +03:00
Torsten Ruger
6a4c90e2bd
fix the cpu_init for linker
...
cpu_init still translates from risc_init
but since interpreter is a platform risc_init is a private thing
and the cpu_init does not need to be lazily created, so it's done in the init
2018-07-02 15:51:50 +03:00
Torsten Ruger
1132309f6a
unify space collection attribute naming
...
currently space is still acting as a sort of memory manager.
For proper linking, all objects must be reachable from space, hence the plural versions like messages and addresses (even they are instances, it is the list that is important)
To dish out instance to use, the head must be kept, ie next_XXX for intergers, return addresses and messages
2018-07-02 15:49:51 +03:00
Torsten Ruger
07a154be70
code listener now get platform explicitly
...
used to grab it from global machine.
Now passed in from linker
tests fixed accordingly
2018-07-02 09:39:07 +03:00
Torsten Ruger
8952b39446
linker goes through assemblers, not space
...
used to traverse all methods in space
now "just" use the passed assemblers, which usually should be compiled and builtin (but for testing . . .)
tests fixed accordingly
2018-07-02 09:37:58 +03:00
Torsten Ruger
c6f0dc636d
platform helper
2018-07-02 09:36:29 +03:00
Torsten Ruger
f4b1290e58
position tests working
2018-07-02 09:35:57 +03:00
Torsten Ruger
22408b86c6
return the linker from mom_compiler
...
linker holds assemblers
assemblers come from method compilation
and so the layers come into focus
2018-07-01 21:51:48 +03:00
Torsten Ruger
5f2a256608
fix platform derivation and some tests
2018-07-01 21:27:27 +03:00
Torsten Ruger
c947c27a14
clean up booting
...
many machine boot became obsolete
or just neede parfait to boot
actual linker functionality pending
2018-07-01 14:12:42 +03:00
Torsten Ruger
e75aef933d
have to pass constants around now
...
as they are not global anymore
2018-07-01 14:11:29 +03:00
Torsten Ruger
e10092a4be
rename machine to linker
...
even work is not quite finished, but helps root out offenders
(needless to say this is wip, deep red wip)
2018-07-01 13:45:14 +03:00
Torsten Ruger
57fc67ddf6
move address list to object_space
...
where i guess it should have been from the start
2018-07-01 13:43:53 +03:00
Torsten Ruger
35b10c46a3
adds the risc instructions to the compiler
...
since they are not in the method anymore
2018-07-01 11:59:52 +03:00
