4b64912793
Begin doing some data flow analysis.
...
In te beginning, live variables.
2012-11-15 13:44:08 +09:00
c00e666668
Expose find_operands.
...
It has proven to be a generally useful function, not just for dags.
2012-11-15 13:44:08 +09:00
0a2ce6d381
Fix some doxygen comments.
2012-11-15 13:44:08 +09:00
0b49bd343c
Collect information on statements that set or use vars.
...
First, it turns out using daglabels wasn't such a workable plan (due to
labels being flushed every sblock). Instead, flowvars are used. Each actual
variable (whether normal or temp) has a pointer to the flowvar attached to
that variable.
For each variable, the statements that use or define the variable are
recorded in the appropriate set attached to each (flow)variable.
2012-11-15 13:44:08 +09:00
acedc65de1
Make an array of statements for the function.
...
With this, sets can be used to represent statements.
2012-11-15 13:44:07 +09:00
98245f6d00
Bury a dead field.
2012-11-15 13:44:07 +09:00
8844ac61a2
Rename dump_flow to dump_sblock and move to dot_sblock.c
2012-11-15 13:44:07 +09:00
acde6ca0a0
Handle jumpb statements in dot_sblock and flow graphs.
2012-11-15 13:44:07 +09:00
ffa08a99c2
Store the initializer expression in the def.
...
Initially, this will be used for collecting the targets of a jumpb
instruction.
2012-11-15 13:44:07 +09:00
d7177a78e9
Rewrite the flow graph code.
...
The flow graph nodes are now properly separated from the graph, and edge
information is stored in the graph struct. This actually made for much
cleaner code (partly thanks to the use of sets and set iterators).
Flow graph reduction has been (temporarily) ripped out as the entire
approach was wrong. There was also a bug in that I didn't really understand
the dragon book about selecting nodes and thus messed things up. The
depth-first search tree "fixed" the problem, but was really the wrong
solution (sledge hammer :P).
Also, now that I understand that dot's directed graphs must be acyclic, I
now have much better control over the graphs (back edges need to be
flipped).
2012-11-15 13:44:07 +09:00
b9599a7119
Better set memory management.
...
At the cost of one pointer per set, sets can now be allocated and freed
more efficiently (especially since malloc might round things up).
2012-11-15 13:44:07 +09:00
2385cf65c2
Rename setstate_t to set_iter_t.
...
After all, it is an iterator :).
2012-11-15 13:44:07 +09:00
8c5a2b3d11
Implement depth-first searching for flow graphs.
...
This fixes the bogusly irreducible graphs.
2012-11-15 13:44:07 +09:00
53622dbb99
Perform graph reduction on the function's flow graph.
...
The reduction is performed itteratively until the graph is irreducible, but
such that each reduction wraps the previous graph. Unfortunately, due
depth-first searching not being implemented, graphs that should be reduced
(ie, those with natural loops).
2012-11-15 13:44:07 +09:00
388b6e6e8f
Add set_size ().
...
Return the number of members in the set.
2012-11-15 13:44:07 +09:00
8567508963
Rewrite set_first() and add set_next().
...
set_first() now returns a pointer to a setstate_t struct that holds the
state necessary for scanning a set. set_next() will automatically delete
the state block when the end of the set is reached. set_delstate() is also
provided to allow early termination of the scan.
2012-11-15 13:44:07 +09:00
2da038ce0f
Bring back dot_flow.c.
...
However, it produces simplified flow graphs (ie, just nodes and edges, no
statements, though later on data flow information will be added).
2012-11-15 13:44:07 +09:00
8b374305d2
Rename dot_flow.c and print_flow.
...
They're now dot_sblock.c and print_sblock. The new names both better
reflect their purpose and free up "flow" for outputting the real flow
analysis graphs.
2012-11-15 13:44:07 +09:00
c483770121
Rework flow.c with the aim of graph reduction.
...
Much of the data recently added to sblock_t has been moved to flownode_t.
No graph reduction is carried out yet, but the initial (innermost level)
graph has been built.
2012-11-15 13:44:07 +09:00
da446fd343
Detect natural loops.
...
Non-nested loops that have the same head-node are merged into one loop.
However, loop nesting is not currently detected (next step).
2012-11-15 13:44:07 +09:00
8fb958603c
Add set_first.
...
It returns the first element of a set. If the set is empty, -1 (unsigned)
is returned.
2012-11-15 13:44:07 +09:00
1fead50f4e
Calculate the dominators of each node in the flow graph.
...
The dominators are represented by sets using the node numbers.
2012-11-15 13:44:07 +09:00
964ea7f9fe
Add set_as_string for debug output.
2012-11-15 13:44:07 +09:00
7853bf1859
Calculate a node's successors and predecessors.
...
The dot graphs are a little odd (arrow heads on the wrong end of the
predecessor edges), but things seem to be correct.
2012-11-15 13:44:06 +09:00
739ebc522c
Begin working on flow analysis.
...
For now, variable and flow graph node maps are built.
2012-11-15 13:44:06 +09:00
f30010b367
Add some docs to function.h
2012-11-15 13:44:06 +09:00
cdc790d0c8
Add a set type and support code.
2012-11-15 13:44:06 +09:00
dc9d2a982e
Make the returned dag a linked list of root-nodes.
...
A dag can have mulitple "root" nodes (ie, nodes with no parents).
2012-11-15 13:44:06 +09:00
a22260030a
Use value for pointer operands.
...
This allows daglabels on pointer operands without a redundant daglabel
pointer.
2012-11-15 13:44:06 +09:00
0c58cd9067
Generate only one daglabel for value operands.
...
As each value is now unique, their labels can also be unique, which will
improve CSE.
2012-11-15 13:44:06 +09:00
0ccf4093e4
Handle attached identifier labels correctly.
...
I'd forgotten I had the prev field. Makes for much cleaner code.
2012-11-15 13:44:06 +09:00
28ce35f1c1
Make values independent objects.
...
values are now uniquely allocated (for the current object file). With
this, constants in dags will work.
2012-11-15 13:44:06 +09:00
dbf1aa8f72
Attach dags to statement blocks.
...
Now, if a dag is attached to a statement block, it will be included in the
flowgraph with that statement block.
2012-11-15 13:44:06 +09:00
d9354255a3
Avoid double-printing dag nodes.
2012-11-15 13:44:05 +09:00
7b2e426545
Support temp operands in dags.
2012-11-15 13:44:05 +09:00
8015e1b1e9
Put the dags stuff into a qfcc sub-section.
...
Not much point writting docs if they can't be found.
2012-11-15 13:44:05 +09:00
aa2943709e
Start work on building dags from basic blocks.
...
Doesn't compile and the design is in flux anyway, but I want to get back to
iqm.
2012-11-15 13:44:05 +09:00
60108c688c
Change dags from binary to trinary.
...
I'd forgotten some instructions take three inputs that can benefit from
CSE.
2012-11-15 13:44:05 +09:00
0fefeb73fe
Initial support for basic block dags.
...
Currently, only dumping to dot, but that seems to be a very sensible place
to start: debug support.
2012-11-15 13:44:05 +09:00
d5c831b6da
Add a little test case for accessing types.
...
Using ruamoko to access ruamoko type encodings is actually quite pleasant
:)
2012-11-15 13:40:11 +09:00
552142497e
Support progs.dat when dumping type encodings.
...
Knock up a quick and dirty progs-to-qfo "converter" (strings and types
only) to re-use the qfo types dumper.
2012-11-15 11:58:13 +09:00
9691b50f0f
Make .type_encodings more useful.
...
Rather than just a pointer, it is now a struct with a pointer and size
indicating the extent of the type encodings block.
2012-11-15 11:57:16 +09:00
453ac11bc3
Correct some comments.
2012-11-14 17:13:14 +09:00
e05c12e16d
Make linker_add_def take a pointer for the value.
...
While it looks like I might not use it for the type encodings, it should be
useful in the future.
2012-11-14 13:12:58 +09:00
5b7b49b100
Add a QFO_TYPE macro to get a qfot_type_t pointer.
...
In the end, it wasn't used, but it could be handy in the future.
2012-11-13 21:45:54 +09:00
93b007b1f8
Take care of relocs when defining external defs.
...
For most of the cases, some relocs may have been lost, but they were
certainly lost for type encodings. Now that all seems to be fixed.
2012-11-12 16:14:09 +09:00
c96718493a
A bit of a cleanup before fixing class type defs.
...
All internal structs now have "proper" names, and fit the naming convention
(eg, obj_module (like objective-c's types, but obj instead of objc). Some
redundant types got removed (holdovers from before proper struct tag
handling).
Also, it has proven to be unnecessary to build internal classes, so
make_class and make_class_struct are gone, too.
2012-11-12 11:12:33 +09:00
d13ce81b4f
Ensure the type's encoding string is valid.
...
When encoding a type to a qfo file, the type's encoding string is written
and thus needs to be valid prior to actually doing the encoding. The
problem occurs mostly in self-referential structs (particularly, obj_class)
because the struct is being encoded prior to the pointer to the struct.
2012-11-11 20:22:48 +09:00
86091856be
Revamp the static class creation.
...
Type names are cleaned up, as is the creation. Also, the class pointer in
the type encoding now gets emitted. However, Still need to actually create
_OBJ_CLASS_Class and fix the type encoding reloc handling in the linker.
2012-11-10 12:56:19 +09:00
9095e1eabc
Rework build_switch to use initialize_def.
...
This required support for label reference expressions, whose purpose is to
represent the address of a label.
2012-11-09 12:22:34 +09:00
afbab60c25
Create linker_find_def.
...
It turns out I didn't need it, but it could be useful in the future, so
I'll leave it in.
2012-11-09 11:57:06 +09:00
eb86a66ce5
Do some class object type renaming.
...
type_object -> type_obj_object
type_Class -> type_obj_class
type_ClassPtr -> type_Class
In the process, one more FIXME is gone :)
2012-11-08 20:58:48 +09:00
377f8f9d56
Use an explicit flag for class initialization.
...
Much cleaner than messing with class_Class, which is about to go away.
2012-11-08 17:14:33 +09:00
7e293a38e8
Check for double-chaining of types.
...
Chaining a type twice will form a loop in the type list, causing find_type
to go into an infinite loop.
2012-11-08 17:14:33 +09:00
a3a22f2a1b
Add the ability to dump qfo type encoding.
2012-11-08 17:06:38 +09:00
ed4018fd8d
Clear out the function overloading tables between compiles.
...
Fixes some issues with multiple compiles (especially mixed languages!).
2012-11-01 20:53:33 +09:00
f8bdd909cc
Separate out C and html quoting into separate functions.
...
Dot sometimes wants normal C quotes, othertimes html. Quite annoying,
really, but this is cleaner anyway.
2012-10-30 12:31:22 +09:00
034139b806
Move some code around so it's more accessible.
2012-10-27 11:43:29 +09:00
144387c3e3
Add shared.h to the dist files.
2012-10-26 16:16:46 +09:00
f14d8060e0
Merge qfpc into qfcc.
...
Since gnu bison and flex are required anyway, no harm in using their api
prefix options. Now, qfcc can compile both QC/Ruamoko and Pascal files
(Pascal is (currently?) NOT supported in progs.src mode), selecting the
language based on the extension: .r, .qc and .c select QC/Ruamoko, .pas and
.p select Pascal, while anything else is treated as an object file (as
before).
2012-10-26 16:01:41 +09:00
1364bff91b
Add an extended mode to qfcc.
...
Extended mode allows extra keywords (switch, for, etc) that are compatible
with v6 progs.
2012-07-14 17:16:33 +09:00
23a38738fc
Massive whitespace cleanup.
...
Lots of trailing whitespace and otherwise blank lines.
2012-05-22 08:23:22 +09:00
e866619de6
Output basic block flow diagrams to files.
...
The output can be controlled via --block-dot (not yet documented). The
files a named <sourcefile>.<function>.<stage>.dot. Currently, stage will be
one of "initial" (after expression to statement conversion), "thread"
(after jump threading), "dead" (after dead block removal), "final" (final
state before actual code emission).
2012-05-04 09:45:51 +09:00
5f676c367f
Add a non-error diagnostic that can't be silenced.
...
It is inteded for flagging buggy conditions in the compiler, particularly
after having fixed the original bug (in case something comes back from the
dead).
2012-05-03 13:24:24 +09:00
cb8fe4eb38
Correct a "spelling" error.
2012-05-03 11:54:36 +09:00
e267e0a664
Allow control of progdefs.h CRC writing.
...
CRC writing defaults to off for non-v6 progs, but on for v6 progs. The
--progdefs option forces CRC writing.
2012-05-03 00:22:24 +09:00
bc1b483525
Nuke the rcsid stuff.
...
It's pretty useless in git.
2012-04-22 10:56:32 +09:00
c8163fc0de
Add an option to prevent the use of default paths.
...
Despair has things locked down such that running qfcc during a build fails
due to lack of read access to /usr/local/lib. This is actually a good
thing as accidentally hitting old includes/libs (when a file gets deleted
in the tree) hides bugs. Thus, --no-default-paths to turn off default
search paths.
2012-01-04 13:53:04 +09:00
d910c14935
Clean up some doxygen warnings.
2011-07-10 19:12:07 +09:00
55cc0f9206
Bring back the unsigned type (PROGS version bump)
...
This is only low-level support (the unsigned keyword still does not work),
but sufficient to make switch statements using jump tables work.
2011-04-09 10:07:47 +09:00
c4a398850a
Fix defs as constants for vectors and quaternions.
2011-04-04 21:31:59 +09:00
8d3508cf20
Allocate space for temp defs using size rather than type.
...
Statement operands throw away the high level type information, so store
type size in the operand and use this size for allocating space for temps
rather than using the low-level type.
2011-03-23 21:32:14 +09:00
9b62c8c126
Rename ty_type_e to ty_meta_e and type_t.ty to type_t.meta
2011-03-22 16:06:47 +09:00
f1d9570dc8
Fix constant value type handling.
...
Rename immediate.[ch] to value.[ch] and clean up the mess.
Add convert_value() to convert the type of a value (only scalar types).
2011-03-22 12:24:39 +09:00
8368f57fbf
Close a potential buffer overflow.
2011-03-09 11:09:46 +09:00
39278ba8cc
Explicitly select between direct and indirect moves.
2011-03-09 10:30:57 +09:00
02a70ebe1c
Remove the auxfunction manipulation from the compiler proper.
...
It was broken code in the first place (realloc and pointers), and is
redundant with the new linking process.
2011-03-07 13:33:02 +09:00
29cac0fe3a
Optionally create *_[xyz] symbols for accessing vector components.
...
"vector-components" in code options controls this feature. The default is
off for advanced code and on for traditional code. Disabling
vector-components prevents the comonent names polluting the namespace and
reduces the number of globals needed for vector fields if the components
of that field are never used.
2011-03-07 10:21:40 +09:00
fe3d228658
Maintain a symbol table for entity fields.
2011-03-06 15:38:19 +09:00
fe4df03896
Add is_float ().
...
For now, it checks just floats, but I might one day add doubles.
2011-03-06 11:35:15 +09:00
85bffbcad0
Add is_integral() to check for integral types (integer, short, enum).
2011-03-06 11:25:12 +09:00
8cfa80b5d4
Implement think expressions.
2011-03-05 18:01:37 +09:00
8eb5fd653a
Perform the relocations!!!
...
line numbers seem to be broken, but the code itself looks good.
2011-03-05 08:41:14 +09:00
7fb22bd1e4
Make D_STRING() more useful and add D_GETSTR()
2011-03-04 21:28:12 +09:00
d6d8b5b49d
Generate the debug symbol file.
2011-03-04 18:46:05 +09:00
4192ef835c
Delay initialization of obj_module_t as long as possible.
...
This keeps the type descriptor out of progs that never use any
Objective-QC features.
2011-03-04 08:25:50 +09:00
c99f3aba0e
Add the new header files for make dist.
2011-03-04 00:11:13 +09:00
c5ecc170b6
Give labels a usage count to detect unused labels.
2011-03-03 15:28:49 +09:00
d937172243
Create and use alias operands for alias expressions.
...
This avoids the alias expression modifying the operand used in other
expressions.
2011-03-03 13:46:07 +09:00
45de7327dc
Implement alias expressions (finally).
...
Alias expressions are like cast expressions, but never do any conversions.
2011-03-03 11:06:10 +09:00
c12211d8bb
Support enum results for boolean expressions.
...
eg enum BOOL x = a && b;
2011-03-02 22:46:55 +09:00
4a99f8dcbd
Write the linker output to progs.dat.
...
There is much breakage, but qfcc now produces a progs.dat from either
progs.src or object files. Better yet, the progs.src result is passed
through the linker, removing much duplicate code.
2011-03-02 20:51:00 +09:00
29985efcef
Expose some linker functions to the rest of the compiler.
...
They proved to be useful for cleaning up the final compilation stages.
2011-03-02 20:48:55 +09:00
ecb995a8f7
Rewrite qfo_to_progs() to go direct from qfo to progs.dat.
...
Relocs and debug info is currently broken.
2011-03-02 19:06:43 +09:00
824d8bfc3c
Remove some now unnecessary functions.
2011-03-02 14:21:20 +09:00
9f126ed805
Add some quick type access macros.
2011-03-01 15:18:12 +09:00
b544196c21
Change ptrfld to fldptr for consistency with type.h.
2011-02-28 23:12:32 +09:00
8b661bbb89
Rename qfo_reloc_t's def to target and correct its usage.
2011-02-27 11:37:24 +09:00
4f6d18a328
Encode the basic type for functions, fields and pointers.
...
Very important for distinguishing them from other types.
2011-02-27 00:33:16 +09:00
55043310fc
Fix the QFO_POINTER macro
2011-02-27 00:31:52 +09:00
Bill Currie