qf/quakeforge
0
0
Fork 0
mirror of https://git.code.sf.net/p/quake/quakeforge synced 2024-11-13 08:27:39 +00:00
Code Issues Projects Releases Packages Wiki Activity
7377 commits 17 branches 20 tags 29 MiB
Commit graph

288 commits

Author SHA1 Message Date
Bill Currie
acd54afff7 move to using a hash table for builtin number -> builtin lookup so sparse 
ranges can be used efficiently. move the auto-allocated builtins to
0x10000000-0x7fffffff. should be more than enough :)

use static builtin tables ("nul" terminated) instead of a series of
function calls to add builtins to a vm. should be more memory efficient.
 2004-01-06 05:51:09 +00:00
Bill Currie
fbe6c2f631 make the bad builtin number a warning and provide a suitable function in 
case the function is called: allows progs with bad (but unused) builtins
to run anyway
 2004-01-06 03:18:44 +00:00
Bill Currie
0ddb5eb822 report all bad builtins rather than just the first 2004-01-05 08:52:20 +00:00
Bill Currie
595280decb do the builtin function lookup at load time rather than runtime and store 
the function address in the progs function descriptor. this will speed up
calls to builtin functions, especially when ranges get introduced.
 2004-01-05 08:08:46 +00:00
Bill Currie
dfc83c1f5f add PR_PushFrame and PR_PopFrame so temp strings are easy to use for 
parameters to progs functions.

double the progs stack sizes (call depth of 32? eek)

allow one extra call on the progs stack :)

misc minor cleanups
 2004-01-05 07:10:32 +00:00
Bill Currie
211908e741 implement a couple more obj functions 2004-01-05 01:41:20 +00:00
Bill Currie
dfe7f263c7 new internal printf engine for progs, mostly thanks to Deek. 
also a few cleanup bits
 2004-01-04 07:42:43 +00:00
Bill Currie
a533761770 Nuke PR_GarbageCollect as it's both redundant with the new temp strings 
and wrong anyway (could free strings it wasn't supposed to).

Don't free the string pointed to by a strref since it's already been
implicity freed (whole memory space nuked). Fixes new map crash.
 2004-01-04 02:03:30 +00:00
Bill Currie
bd561fafc3 o all progs strings are now stored in progs memory 
o  temporary strings automaticly get freed when the function they're
    created in (results of str + str and most strings returned from
	builtins). a way to keep temp strings will be provided later
 o	fix up qwaq to test the temp strings
 2004-01-03 08:43:57 +00:00
Bill Currie
0307618b15 use the passed "size" parameter rather than qfs_filesize: no telling 
/where/ the progs really came from :)
 2003-12-22 19:58:46 +00:00
Bill Currie
cd8035319d PR_GetString should have been returning const char *. now does 2003-11-21 21:34:53 +00:00
Bill Currie
e74cd75a0d d'oh, forgot to relocate the builtins :P 2003-11-20 08:01:27 +00:00
Bill Currie
5f5662a063 PF_VarString return a plain char * and make it easier to add "at load" init 
functions
 2003-11-20 07:46:56 +00:00
Bill Currie
aa37db997c check for null function 2003-09-18 22:50:42 +00:00
Bill Currie
fea382d1d9 fix an accounting error 2003-09-15 21:13:13 +00:00
Bill Currie
30e0cba25d better formats for load* and store[bp]* 2003-09-04 05:30:55 +00:00
Bill Currie
a1fa0b3feb make sure the pointed to def has a name before attempting to print it 2003-09-02 04:52:54 +00:00
Bill Currie
cf93f5f65e oops, who let C++ into C? :) 2003-08-25 01:47:05 +00:00
Bill Currie
96202a5984 bah, forgot to -1 the instruction pointer for jumpb 2003-08-24 07:23:12 +00:00
Bill Currie
afb42f5ed0 fix the unsigned comparison opcodes 2003-08-24 05:53:15 +00:00
Bill Currie
253d19876c no code to dump for a null function 2003-08-24 05:08:47 +00:00
Bill Currie
5d8bebfff8 implement class_get_class_method and (partially) class_pose_as 2003-08-24 03:44:12 +00:00
Bill Currie
b50d283b28 make sel_get_uid and class_get_instance_method work 2003-08-23 22:16:42 +00:00
Bill Currie
edb716ada0 oops, forgot to remove developer forcing 2003-08-23 06:21:35 +00:00
Bill Currie
385a90e8fb correct pr_symtab_t to include the pointer to the array of selectors used 
in the module. Unfortunatly, this requires a version bump on both qfo and
progs files due to the changes structure layout.
 2003-08-23 06:15:19 +00:00
Bill Currie
98beb38540 implement jumpb correctly 2003-08-22 19:48:36 +00:00
Bill Currie
cb6eba87ca print the value even for unknown defs if the type is known 2003-08-22 17:32:55 +00:00
Bill Currie
fe3b544528 handle "return 0" a bit better. assumes location is is always 0, but it 
always has been (and always /should/ be anyway:)
 2003-08-13 17:27:34 +00:00
Bill Currie
6f519c7e0f don't bother doing the copy if the value to be returned is already in the 
return slot
 2003-08-13 17:07:05 +00:00
Bill Currie
7ead5a91f8 don't let gcc automaticly inline functions but make gcc inline bigger 
functions when told to. also make gcc warn if it can't inline a function.

Explicitly inline several functions (including moving VectorNormalize to
mathlib.h so it /can/ be) resulting in a 5.5% speedup for spam2 (88 to 92
fps)
 2003-08-11 06:05:07 +00:00
Bill Currie
cbc6c40d33 get some single source instruction formats right and handle empty paths 
properly
 2003-08-05 17:27:47 +00:00
Bill Currie
59c26b5ce0 quick fix to handle .ent files starting with comments 2003-08-04 05:21:18 +00:00
Bill Currie
dd9ab1e050 get unsigned mostly working 2003-08-01 21:20:04 +00:00
Jeff Teunissen
f04d67f319 Nuke the old-style Object's dead built-in methods. 2003-08-01 16:42:02 +00:00
Bill Currie
cbd5c70320 oops, %O should be in hex 2003-07-31 01:57:29 +00:00
Bill Currie
72eca6a956 get the default right :P 2003-07-30 22:24:44 +00:00
Bill Currie
0806fc33dd debugging improvements. opcodes now sport an optional format specifier for 
printing their args (default is "%Ga, %Gb, %gc") and PR_ValueString will
try to print what a pointer points to
 2003-07-30 22:24:16 +00:00
Bill Currie
dcd143f197 missing/incorrect symbols fixed 2003-07-30 17:44:38 +00:00
Bill Currie
3886723467 fix some builtin method names due to the selector name fixes in qfcc 2003-07-27 19:57:40 +00:00
Jeff Teunissen
c9858f4f65 Preliminary "unsigned" support for qfcc and the gamecode engine. Should still 
be compatible.
 2003-07-26 21:07:51 +00:00
Bill Currie
5481273639 start working on plist entity files 2003-07-17 18:27:00 +00:00
Bill Currie
cb6bf2e2ae minor house-keeping 2003-07-15 16:34:43 +00:00
Bill Currie
dedb2fa6bf get [super ...] working correctly 2003-05-15 05:58:31 +00:00
Bill Currie
5f3573143a the beginnings of a jihad against buffer overflows 2003-05-08 23:24:02 +00:00
Bill Currie
f7d806e8d5 can now handle > vector sized parameters 2003-04-22 20:11:16 +00:00
Bill Currie
e732cf591d remove all references to OFS_* from the C code. only the macros in progs.h 
use them now
 2003-04-22 18:20:15 +00:00
Bill Currie
eb099ae0d3 make qf gcc 3.3 clean 2003-04-17 00:01:48 +00:00
Bill Currie
8fc721a2bb plug another overflow 2003-04-08 04:01:20 +00:00
Bill Currie
58541b6855 support printing a couple more types (especially for edicts) 2003-03-26 18:45:44 +00:00
Bill Currie
fafc578440 correct a comment and ws 2003-03-25 17:40:02 +00:00
Bill Currie
548d3b68f8 pr_load.c: 
PR_RelocateBuiltins must be called before PR_InitRuntime
pr_obj.c:
	don't segfault on broken classes
 2003-03-13 18:45:45 +00:00
Bill Currie
2858bc4922 d'oh 2003-03-12 22:38:02 +00:00
Bill Currie
71a60717d9 make pr_source_path a ; separated list of dirs and make it controllable in 
qfprogs
 2003-03-12 22:31:44 +00:00
Bill Currie
bfaef7747b the first fielddef isn't always null (esp with qfcc), so check to see if it 
needs printing too
 2003-03-05 18:18:42 +00:00
Bill Currie
93834f2694 better source line tracking 2003-02-24 16:01:52 +00:00
Bill Currie
38d5c8c41a better line search algorithm (not perfect yet, though) 2003-02-24 03:34:11 +00:00
Bill Currie
5c3892cbb2 make sure the string is valid before attempting to print it 2003-02-24 03:07:26 +00:00
Bill Currie
72e5221f06 minor debugging tweaks 2003-02-21 15:09:02 +00:00
Bill Currie
b6a3bc8d1c try to give line numbers in the traceback 2003-02-19 06:26:49 +00:00
Bill Currie
6a687f5e45 rename all the globals in quakefs.[ch] to be QFS_* or qfs_* 2003-02-14 19:46:07 +00:00
Bill Currie
8a9cd78072 fix up a couple of error cases (and catch them!) 2003-01-29 20:32:44 +00:00
Bill Currie
761a7546dd re-arrange things so __attribute__ and __builtin_expect are properly 
autoconfiscated so rcsid will continue to work with gcc 3.3
 2003-01-15 15:31:36 +00:00
Bill Currie
71196fedf6 make gcc more anal about prototypes, string constants and function 
visibility (ie, global functions must have a prototype)
 2003-01-06 18:28:13 +00:00
Bill Currie
d78f34e795 fix some spelling errors 2002-11-20 21:21:22 +00:00
Bill Currie
06df342655 fix a stray ; 2002-11-15 14:56:57 +00:00
Bill Currie
d1199f23da get categories working in the runtime :) 2002-11-14 20:42:48 +00:00
Bill Currie
836c469737 sys.[ch]: 
handle SIGFPE and allow the registered signal handler to do recovery
	rather than bail
progs.h, pr_exec.c, pr_load.c:
	if pr_faultchecks is 1 (0 is default), handle division by 0 gracefully
	by loading the maximum representable number into the answer
	Closes: #58
the rest:
	kill the SIGFPE stuff
 2002-11-13 19:26:44 +00:00
Bill Currie
ec0273e3dc do a pre and post pass for maps/mapname.cfg and use maps_default.cfg if 
maps/mapname.cfg doesn't exist.
 2002-11-08 02:43:04 +00:00
Bill Currie
ca4b3acd6c big protototype cleanup. Now, except for a few cases, all non-static 
prototypes are in headers files.
 2002-11-05 19:12:51 +00:00
Bill Currie
aa36b85ba6 add PR_StringValid and PR_EdictValid to check the validity of progs string 
and entity values without crashing.
 2002-10-29 05:07:10 +00:00
Bill Currie
012cc58eb4 move the debug related stuff from pr_exec.c to pr_debug.c 2002-10-23 20:42:02 +00:00
Bill Currie
3a3da6a1c4 nuke PR_InitEdicts as it's redundant and only used in the tools due to 
forgetting to remove it
 2002-10-23 15:55:38 +00:00
Bill Currie
a4a5bb2d72 split out the parsing code 2002-10-23 04:57:08 +00:00
Bill Currie
80e85226b9 pr_xstatement now always reflects the currently executing statement making 
debugging easier in the event of a segfault
 2002-10-22 15:07:54 +00:00
Bill Currie
6203d450eb resture pr_xstatement on return from PR_ExecuteProgram so recursive calls 
don't mangle things with the new semantics of builtin calls
 2002-10-22 02:51:57 +00:00
Bill Currie
aad04efe25 move the correct amount of bytes (I keep forgetting the * 4 :P) 2002-10-21 16:38:45 +00:00
Bill Currie
4810046f39 copy the correct number of bytes 2002-10-17 19:33:06 +00:00
Bill Currie
4da662eb47 implement pr_obj_msg_sendv 2002-10-16 22:17:26 +00:00
Bill Currie
be4b8096d4 two new instructions: move and movep: basicly memmove. qfcc support coming 
as soon as I get it working :)
 2002-10-16 06:44:41 +00:00
Bill Currie
1f0b6e84a5 change pr_obj_msgSend and pr_obj_msgSend_super so they either call builtin 
methods directly or setup the progs struct for a normal function call to
call a progs method rather than recursively calling PR_ExecuteProgram. This
will reduce method call overhead (both cpu and stack usage), fix the loss
of tracing when calling a method and makes it possible to break out of the
progs engine simply if threaded progs are ever implemented.
 2002-10-15 18:53:33 +00:00
Jeff Teunissen
575689296a Tidy up progs dump output, and do some whitespace. 2002-10-08 19:46:42 +00:00
Bill Currie
eec3b7d8bc fix missed header commit 2002-09-14 07:51:53 +00:00
Jeff Teunissen
e551c1f22b New dump format for progs. It looks like assembly code. setting pr_debug to 2 
or higher adds more debug information.
 2002-09-14 07:00:40 +00:00
Bill Currie
ec47e08e1c go back to using QFile, quakeio.h and quakefs.h 2002-08-27 07:16:28 +00:00
Bill Currie
5009213564 Move the signal handling code from video/targets/* to sys.c (I'm leaving 
the SDL target files for Despair since he's working on the currently). Also
make progs segfault handling much more robust (and general: sigill etc too:)
 2002-08-20 23:04:57 +00:00
Bill Currie
b99a72e876 pr_obj.c: 
call obj_find_message with super instead of class when doing a super
	lookup
client_menu.[ch]:
	use the InputLine object instead of directly using the api
InputLine.r:
	builint functions are = #0, not = 0 :P
function.[ch]:
	provide copy_params
method.[ch]
	provide copy_keywordargs
expr.c:
	call inc_users for the parm temp if it's a bind expression
	revers a copy of the selector when building the selector name so the
	selector doesn't get mangled
qc-parse.y:
	catch erronous func = const inintializers (should be func = #const)
 2002-08-20 02:09:34 +00:00
Bill Currie
c1c6c05cf5 pr_obj.c: 
delay class fixups so order doesn't matter
class.c:
	write out the class and category diffs properly
 2002-08-16 22:53:37 +00:00
Bill Currie
c1f5659977 bi_inputline.c: 
add InputLine_SetUserData
pr_resolve.c:
	report the correct missing symbol (self instead of time)
 2002-08-16 21:52:37 +00:00
Bill Currie
2ce0f2ded2 we now require modern autoconf (2.52+), automake (1.6+) and libtool (1.4+). 
This gives us a bunch more flexibility. plugins no longer have the "lib"
prefix or the version suffix, they're now installed in
$fs_sharepath/QFplugins,, builds should take much less time (in general,
only one of pic or non-pic versions are build), bins and libs can have
individual CFLAGS
 2002-08-01 19:49:10 +00:00
Brian Koropoff
439f9838d3 Whoops, missed some stuff that needed to be in idparse.h 2002-07-31 22:13:15 +00:00
Bill Currie
ad6e025a91 take cmd.[ch] back to basics. the gib code is temporarily gone pending 
renovations. cmd_source handling is a little hacky in nq, but it's working.
 2002-07-31 05:19:03 +00:00
Ragnvald Maartmann-Moe IV
227da6afba Change pr_boundscheck default to 0. *grmbl* 2002-07-21 23:35:36 +00:00
Bill Currie
bff6291bfb don't need to force developer now 2002-07-16 22:06:52 +00:00
Bill Currie
ac1392eb74 object.r: 
define __obj_exec_class, obj_msgSend and obj_msgSend_super for the way
	qfcc now handles these defs
pr_edict.c:
	PR_ValueString now checks that the function is valid before trying to
	print its name.
class.c:
	__obj_exec_class is now extern
linker.c:
	fix more mangled relocs
method.c:
	obj_msgSend and obj_msgSend_super are now extern
qfcc.c:
	fixes so obj_module_t doesn't get zapped
type.c:
	fixes so obj_module_t doesn't get zapped
 2002-07-16 19:03:56 +00:00
Bill Currie
35d224f1da a little more info and force developer on for now 2002-07-16 15:39:49 +00:00
Bill Currie
34fa421638 whitespace 2002-07-13 03:41:33 +00:00
Bill Currie
717127e0d8 don't print ! on uninitialized global defs that are the correct type for 
the instruction
 2002-06-17 05:31:08 +00:00
Bill Currie
2264c28310 pr_strings: 
remove a bogus 0x which caused confusion in an error message
def.c:
	vector fields still need an extra 3 globals for their component fields
 2002-06-16 05:24:58 +00:00
Bill Currie
88fc0be127 arrays are /almost/ working again 2002-06-12 22:37:18 +00:00
Bill Currie
0b1d1ade73 qfprogs can now read the debug data 2002-06-11 17:24:37 +00:00
Bill Currie
a4b06b1074 blah, enums and switch... (at least gcc 3.1 is warning about that again:) 2002-06-09 16:31:08 +00:00
Bill Currie
5c943d4ca9 this should speed up progs a little bit 2002-06-07 19:41:13 +00:00
Bill Currie
55fcead77c look for the .this field but don't complain if not found 2002-05-31 17:21:38 +00:00
Bill Currie
a8e177652a initialize the meta class super class pointer correctly 2002-05-31 14:58:42 +00:00
Bill Currie
2ee3aec19b reverse the order of linking 2002-05-31 06:52:48 +00:00
Bill Currie
ea3c6995fd actually /use/ method_list.method_next and chain the class methods onto the 
instance methods for root classes.
 2002-05-31 06:01:47 +00:00
Bill Currie
a7b5b56a2c implement Z_Realloc, PR_Zone_Realloc and pr_obj_realloc 2002-05-29 20:58:53 +00:00
Bill Currie
33e055209e Semi-major (more than minor:) progs convenience macro cleanup. There will 
be more to come.
 2002-05-22 22:35:37 +00:00
Bill Currie
ea36fceec5 fix a few missed initializations 2002-05-22 21:45:15 +00:00
Bill Currie
34ec521198 vararg functions can now be imlemented in qc using @argc (number of params 
passed through ...) and @argv (array of vectors, one for each ... param)
 2002-05-22 20:43:29 +00:00
Bill Currie
018b3ae536 more runtime implementation 2002-05-22 19:02:26 +00:00
Bill Currie
43e609241c initial stab at supporting most of the obj runtime 2002-05-22 17:32:44 +00:00
Bill Currie
33daf1e724 provide temporary stubs for most of Object's methods 2002-05-21 23:39:21 +00:00
Bill Currie
3815d4be7e blah, unitialized variable 2002-05-21 22:27:41 +00:00
Bill Currie
05c8cad2d1 grr, really should make compat.h include string[s].h 
also, partially implement the messaging functions
 2002-05-21 22:13:05 +00:00
Bill Currie
9820c5d45a start working in the obj runtime. any and all functions named ".ctor" will 
be run in the order found.
 2002-05-21 21:29:49 +00:00
Bill Currie
a8c3fa8ffc correct the function names in PR_EnterFunction and PR_LeaveFunction 2002-05-21 21:14:32 +00:00
Bill Currie
f78f921285 search for .self' before self' 2002-05-18 00:49:16 +00:00
Bill Currie
057aa61bd3 `document' the type sizes 2002-05-17 16:44:47 +00:00
Bill Currie
91779b04f7 new tool: qfprogs. currently only disassembles progs, but the idea is for it 
to become a progs oriented nm, objdump etc
 2002-05-13 19:37:36 +00:00
Bill Currie
73e6cf062c implement type encoding 2002-05-10 23:19:57 +00:00
Bill Currie
8588d2d1cd fix a silly braino in the resources code that was causing the resource 
structs to not get linked into the chain
 2002-04-09 16:06:05 +00:00
Brian Koropoff
388ff7b597 Some bug fixes and memory leak fixes. Local variables break randomly, 
probably due to my misuse of hash tables.  Anyone have any ideas?
 2002-03-20 22:57:26 +00:00
Bill Currie
9ea013888b eliminate GetEdictFieldValue, mostly merge SV_AddGravity, and clean up 
some whitespace in sv_phys.c
 2002-02-25 07:44:17 +00:00
Bill Currie
b5c2c0e14e clean up a fixme. this doesn't fix serplord's out-of-bounds problem, but 
that's because his qc compiler is doing weird stuff.
 2002-02-11 19:36:36 +00:00
Bill Currie
681950cf26 print [val] for pointers 2002-01-30 22:14:13 +00:00
Bill Currie
6d63d1d55b implement comparisons and not for pointers 2002-01-30 21:20:12 +00:00
Bill Currie
e1abe4c44b finalize the crc of the progs file 2002-01-30 21:13:57 +00:00
Bill Currie
c76f7d88ce fix an API snafu (oops:) 2002-01-30 08:09:53 +00:00
Bill Currie
8561dfa93a basic resource management system for the VM 2002-01-30 06:21:20 +00:00
Bill Currie
0ea713eda4 redo the progs loader to allocate all the memory needed for progs in one 
shot, allowing for caller specified alloc and edicts and zone memory sizes.
(progs + edicts + zone).
 2002-01-29 20:53:44 +00:00
Bill Currie
21b6e68a8c break out the non-edict code into sensible locations 2002-01-28 17:15:36 +00:00
Bill Currie
08210bc93f fix an accidental double byteswap that broke progs for big-endian machines 2002-01-24 04:43:47 +00:00
Bill Currie
68637eea9e start implementing dynamic memory for progs 2002-01-23 22:35:48 +00:00
Bill Currie
16cc0c22c9 forgot to implement store.p 2002-01-23 22:33:22 +00:00
Bill Currie
6af9a5d8e6 make the builtins an array of pointers so the hash table doesn't get royally 
screwed up
 2002-01-19 03:31:51 +00:00
Bill Currie
7def88668a fix a goof with the progs loader for the server interface 2002-01-18 23:53:51 +00:00
Bill Currie
ac71d101d2 menu code now gets initialized (menu_init in the progs). 2002-01-18 22:02:59 +00:00
Bill Currie
bb96a5b736 first use of client side QuakeC :) (and having more than one progs instance 
seems to work, too). The menu code gets loaded, but is currently otherwise
unused as of yet.
 2002-01-18 19:19:33 +00:00
Bill Currie
d69762facd integer() and float() casts seem to work 2002-01-17 19:32:04 +00:00
Bill Currie
3dafbebf10 fix up #0 builtin functions at runtime. not done automaticly (need to call 
PR_RelocateBuiltins) and fix a bug with profiling and unlimited execution
counts
 2001-12-14 08:15:04 +00:00
Bill Currie
8ddce20949 pr_*: add storep.p so pointers can be stored into entities 
emit.c: convert & (address) to . in unary .
        rewrite emit_assign_expr
expr.c: remove debug code and make unary . less agressive on address calcs
        also convert & (address) to . for lvals
 2001-12-12 20:35:58 +00:00
Bill Currie
f9149a0d6c finally get around to renaming type_name to pr_type_name 2001-12-12 17:15:09 +00:00
Bill Currie
cc34e5954f lots of work on better array/pointer/struct handline. doesn't quite work yet 2001-12-12 08:39:47 +00:00
Bill Currie
d344aea51f store.p so pointers can be moved around 2001-12-08 21:43:44 +00:00
Bill Currie
00b2a45ecd change the operator for "address" from "." to "&" so load.p can be implement. 
this makes it so pointers in entities work.
 2001-12-08 20:17:20 +00:00
Bill Currie
b03a1bee42 ok, structures seem to generate correct code now 2001-12-08 08:19:48 +00:00
Bill Currie
028f8a116d immidiate indexed pointer access and some whitespace + reorg 2001-12-07 20:07:38 +00:00
Bill Currie
12a0a10d33 more alpha portability fixes 2001-11-29 19:35:25 +00:00
Bill Currie
67fde0c492 remove some debug code I accidently committed 2001-11-26 17:25:16 +00:00