Commit Graph

775 Commits

Author SHA1 Message Date
Torsten Ruger
047c77e186 fixed that darn method address bug 2015-07-01 09:47:10 +03:00
Torsten Ruger
9541712af8 fixing all the indexes
with the type word and layout
but the list starting at 1, indexes still need 1 added
and for arm x 4.
Tried to get all that into one function, resolve_index
2015-06-30 18:38:56 +03:00
Torsten Ruger
88f45cbf15 weird 4 to add for call works 2015-06-30 18:36:57 +03:00
Torsten Ruger
c0f11d70f1 rephrased the retry, didnt find the bug though 2015-06-30 18:35:37 +03:00
Torsten Ruger
b9e2d4dd21 remove some debug 2015-06-30 10:22:54 +03:00
Torsten Ruger
7e9b940890 add init_message to space
so kernel::__init can use that and the first next_message does not need
to be reused
2015-06-30 09:52:17 +03:00
Torsten Ruger
df2a080bd0 move register_machine to register 2015-06-30 09:46:11 +03:00
Torsten Ruger
ee2294ded7 remove unused reference 2015-06-30 09:45:45 +03:00
Torsten Ruger
60c8f8ef73 move generator functions to the instructions they generate 2015-06-30 09:43:50 +03:00
Torsten Ruger
bd77db656a remove class_eval on slots to map registers
Use their class names (or for now called object names)
these are the same that Register helper functions accept already anyway
2015-06-30 09:39:45 +03:00
Torsten Ruger
c4502c5030 get rid of the last constants
return them by method, nicer code
2015-06-30 09:38:32 +03:00
Torsten Ruger
b8090e0ecd fix init to load into new message 2015-06-30 09:37:25 +03:00
Torsten Ruger
3a6d3143c5 change method debug format for gdb autocomplete 2015-06-30 09:37:01 +03:00
Torsten Ruger
0f2c8e4201 finally scoping builtin to register
had put this off because it breaks history
but now the references to register stuff which
builtin is off course full of, become much shorter
2015-06-29 21:03:58 +03:00
Torsten Ruger
a03dcecbbd fix namespacing and remove indexes 2015-06-29 20:58:52 +03:00
Torsten Ruger
553f30c874 more ripples from removing index constants 2015-06-29 20:58:06 +03:00
Torsten Ruger
1e18db00c9 add set_slot and save_return
as functions to generate instructions
2015-06-29 20:57:16 +03:00
Torsten Ruger
a24e96efbc asserts to debug 2015-06-29 20:56:11 +03:00
Torsten Ruger
218fafca05 renames 2015-06-29 20:55:45 +03:00
Torsten Ruger
5ce7b6c7c9 removing those ugly slot index constants
The constants were bad enough,
but they were also at the wrong level

Now register level is defining mappings from
symbol names to indexes, by using the layout
2015-06-29 10:55:22 +03:00
Torsten Ruger
06b5d24729 fixing require order
so Register stuff is defined when builtin is loaded
2015-06-29 10:53:27 +03:00
Torsten Ruger
a16abeb3e6 giving the method to return and save
so they can make decisions
like wether to create a frame or not
2015-06-28 22:03:21 +03:00
Torsten Ruger
02615db508 minor 2015-06-28 22:02:27 +03:00
Torsten Ruger
8674c322c4 correctly linking frames and messages 2015-06-28 22:02:07 +03:00
Torsten Ruger
31635d9747 fixing index issues 2015-06-28 22:01:13 +03:00
Torsten Ruger
99b59d920c constant fixes 2015-06-28 10:52:02 +03:00
Torsten Ruger
a00a49ecdb fix frame indexes 2015-06-28 10:50:47 +03:00
Torsten Ruger
ef42abe611 unifying register comstants
were in several files with different names
many files touched, but just renames
2015-06-27 21:16:46 +03:00
Torsten Ruger
97b4c469f8 fixing register order in some instructions
and their use
Arm is confusing as it has result as first arg
we use forward logic, i.e. from -> to
2015-06-27 20:09:21 +03:00
Torsten Ruger
add79e5157 use sod names for debug info 2015-06-27 20:08:07 +03:00
Torsten Ruger
294ea8de4f fix the move, correct for funny pipeline 2015-06-27 20:07:42 +03:00
Torsten Ruger
485d6566ad write object refs at offset
that is (only) the linux/arm offset off course
2015-06-27 15:17:15 +03:00
Torsten Ruger
d0d857bbe0 fix object loading
by (ahem) copying the code from MoveInst.
2015-06-26 20:36:00 +03:00
Torsten Ruger
5539d8fe32 just a little more precise 2015-06-26 20:35:16 +03:00
Torsten Ruger
e20fba61d9 allowing for relinking to add more instructions 2015-06-26 20:34:40 +03:00
Torsten Ruger
55b7ca83e1 fixing load constant
which needs to add to pc to get an object reference
2015-06-26 20:01:52 +03:00
Torsten Ruger
108cc8af45 minor cleanup 2015-06-26 20:00:50 +03:00
Torsten Ruger
37403f1139 fix misunderstood mov usage
When moving a reference (pointer) to a register
one needs to ADD to the pc
i.e. one needs an add, not mov instruction
2015-06-26 20:00:33 +03:00
Torsten Ruger
32e1903884 finished init routine
and better implemented sys calls
2015-06-25 16:31:09 +03:00
Torsten Ruger
cedc6e1b61 Fix parfait bug for word_length 2015-06-24 16:08:06 +03:00
Torsten Ruger
59188105ba finish syscall
works, even had to fix parfait bug to see
2015-06-24 16:07:27 +03:00
Torsten Ruger
1a82ebcd69 mor on sys call implementation 2015-06-23 19:55:54 +03:00
Torsten Ruger
2aba926f1f start on reworking sys calls 2015-06-22 22:48:42 +03:00
Torsten Ruger
9c21f4274d better names for index constants 2015-06-21 21:09:15 +03:00
Torsten Ruger
836089a249 clean up Get/SetSlot
document and make arguments consistent
2015-06-21 21:00:16 +03:00
Torsten Ruger
f3ee11fca5 simple set optimisation
the price of having simple code (generation) is that
it is sometimes stupid code that gets generated
Filter some of the really daft stuff out
2015-06-21 17:25:27 +03:00
Torsten Ruger
aafb179c61 move frame_impl to register 2015-06-21 17:23:15 +03:00
Torsten Ruger
3fe35e34ec fix slot constant access 2015-06-21 17:22:51 +03:00
Torsten Ruger
cdf17a73b2 changed Set order
from to seems more logical than to,from
2015-06-21 13:29:27 +03:00
Torsten Ruger
6867175bd1 slot docs and rename 2015-06-21 00:21:42 +03:00