qf/quakeforge
0
0
Fork 0
mirror of https://git.code.sf.net/p/quake/quakeforge synced 2024-11-10 23:32:09 +00:00
Code Issues Projects Releases Packages Wiki Activity
quakeforge/libs/gamecode/pr_exec.c

1737 lines
46 KiB
C
Raw Normal View History

initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
/*
pr_exec.c
(description)
Copyright (C) 1996-1997 Id Software, Inc.
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to:
Free Software Foundation, Inc.
59 Temple Place - Suite 330
Boston, MA 02111-1307, USA
*/
#ifdef HAVE_CONFIG_H
# include "config.h"
#endif
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
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
#ifdef HAVE_STRING_H
# include <string.h>
#endif
#ifdef HAVE_STRINGS_H
# include <strings.h>
#endif
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
#include <signal.h>
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
#include <stdarg.h>
move the api headers into include/QF and clean up (most of) the resulting mess. target specific files that I don't build won't compile yet. just put QF/ infront of the offending headers. Also move ver_check into libqfutils
2001-03-27 20:33:07 +00:00
#include "QF/cvar.h"
fix missed header commit
2002-09-14 07:51:53 +00:00
#include "QF/dstring.h"
LordHavoc's VectorIsNull changes from 0.3 but as VectorIsZero (no such thing as a `null' vector) plus a couple of other bits from his patch.
2001-09-11 05:18:15 +00:00
#include "QF/mathlib.h"
move the api headers into include/QF and clean up (most of) the resulting mess. target specific files that I don't build won't compile yet. just put QF/ infront of the offending headers. Also move ver_check into libqfutils
2001-03-27 20:33:07 +00:00
#include "QF/progs.h"
#include "QF/sys.h"
pr_comp.h: add OP_ADD_S. WARNING!!! this /will/ move. progs.h: add prototype for PR_PrintStatement pr_edict.c: add OP_ADD_S support in the progs checker pr_exec.c: implement OP_ADD_S tools/qfcc/include/.gitignore: add config.h.in qfcc.h: nuke PR_NameImmediate and change PR_ParseImmediate's prototype (see pr_imm.c) pr_comp.c: add ADD_S, adjust for PR_ParseImmediate's prototype, make PR_ParseExpression work with non-sequential opcodes (slow, will work on that next). Fix up initialised global parsing. pr_imm.c: nuke PR_NameImmediate. didn't work well and wasn't such a good idea anyway. PR_ParseImmediate now accepts a def_t * arg. if null, will allocate a new global def, otherwise it will initialize the def passed in. qwaq/main.c: sports some debugging code (dumps info about the progs it's running) qwaq/main.qc: better ADD_S testing
2001-06-03 17:36:49 +00:00
#include "QF/zone.h"
#include "compat.h"
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
rhamph's deadbeaf locals patch as-is :)
2001-08-13 20:29:33 +00:00
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
/*
PR_RunError
Aborts the currently executing function
*/
the big dso visibility patch :). Sure, we have to have unique names for static builds, but with controlled visibitly we should get faster program loads (although this isn't C++, so it's not as bad) and complex plugins are cleaner.
2007-03-10 12:00:59 +00:00
VISIBLE void
fear the mighty const correct patch
2001-07-15 07:04:17 +00:00
PR_RunError (progs_t * pr, const char *error, ...)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[gamecode] Add a param pointer to debug_handler I decided I want events for VM enter/exit but enter needs to somehow pass the function which will be executed (even if a builtin). A generic void * param seemed the best idea, which meant the error string could be passed via the param instead of a "global" string in the progs struct.
2020-03-26 02:44:02 +00:00
dstring_t *string = dstring_new ();//FIXME leaks when debugging
move the debug related stuff from pr_exec.c to pr_debug.c
2002-10-23 20:42:02 +00:00
va_list argptr;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
va_start (argptr, error);
move the debug related stuff from pr_exec.c to pr_debug.c
2002-10-23 20:42:02 +00:00
dvsprintf (string, error, argptr);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
va_end (argptr);
[gamecode] Add debug hooks to the VM engine While there was a breakpoint hook, it was for only breakpoints and more was needed. Now there's a generic hook that is called for tracing, breakpoints, watch points, runtime errors and VM errors, with the "event" type passed as the first parameter and a data pointer in the second.
2020-03-24 06:35:42 +00:00
if (pr->debug_handler) {
[gamecode] Add a param pointer to debug_handler I decided I want events for VM enter/exit but enter needs to somehow pass the function which will be executed (even if a builtin). A generic void * param seemed the best idea, which meant the error string could be passed via the param instead of a "global" string in the progs struct.
2020-03-26 02:44:02 +00:00
pr->debug_handler (prd_runerror, string->str, pr->debug_data);
[gamecode] Add debug hooks to the VM engine While there was a breakpoint hook, it was for only breakpoints and more was needed. Now there's a generic hook that is called for tracing, breakpoints, watch points, runtime errors and VM errors, with the "event" type passed as the first parameter and a data pointer in the second.
2020-03-24 06:35:42 +00:00
// not expected to return, but if so, behave as if there was no handler
}
move the debug related stuff from pr_exec.c to pr_debug.c
2002-10-23 20:42:02 +00:00
Sys_Printf ("%s\n", string->str);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Print the error message before the state dump. Normally, the order doesn't matter, but when tracing code, it becomes very difficult to tell where the trace ends and the dump begins. Printing the message first puts the message between the trace and the dump: much easier :)
2012-12-16 03:27:15 +00:00
PR_DumpState (pr);
break out the backtrace code from PR_RunError into PR_DumpState and use PR_DumpState instead of PR_RunError in seg_fault_handler so we get both a progs trace /and/ a core file (ulimit allowing)
2001-10-08 03:46:44 +00:00
// dump the stack so PR_Error can shutdown functions
Massive whitespace cleanup. Lots of trailing whitespace and otherwise blank lines.
2012-05-21 23:23:22 +00:00
pr->pr_depth = 0;
Dump locals stack when dumping main stack Any shutdown functions in progs will need locals too.
2020-02-25 06:18:15 +00:00
pr->localstack_used = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
move the debug related stuff from pr_exec.c to pr_debug.c
2002-10-23 20:42:02 +00:00
PR_Error (pr, "Program error: %s", string->str);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Allow nested use of PR_SaveParams() PR_SaveParams() is required for implementing the +initialize diversion used by Objective-QuakeC because builtins do not have local def spaces (of course, a normal stack calling convention would help). However, it is entirely possible for a call to +initialize to trigger another call to +initialize, thus the need for stacking parameter stashes. As a bonus, this implementation cleans up some fields in progs_t.
2020-02-25 05:30:26 +00:00
VISIBLE pr_stashed_params_t *
_PR_SaveParams (progs_t *pr, pr_stashed_params_t *params)
correctly save and restore the parameters when calling +initialize. fixes the segfault in the qwaq test.
2007-06-09 13:44:06 +00:00
{
int i;
int size = pr->pr_param_size * sizeof (pr_type_t);
Allow nested use of PR_SaveParams() PR_SaveParams() is required for implementing the +initialize diversion used by Objective-QuakeC because builtins do not have local def spaces (of course, a normal stack calling convention would help). However, it is entirely possible for a call to +initialize to trigger another call to +initialize, thus the need for stacking parameter stashes. As a bonus, this implementation cleans up some fields in progs_t.
2020-02-25 05:30:26 +00:00
params->param_ptrs[0] = pr->pr_params[0];
params->param_ptrs[1] = pr->pr_params[1];
correctly save and restore the parameters when calling +initialize. fixes the segfault in the qwaq test.
2007-06-09 13:44:06 +00:00
pr->pr_params[0] = pr->pr_real_params[0];
pr->pr_params[1] = pr->pr_real_params[1];
for (i = 0; i < pr->pr_argc; i++) {
Allow nested use of PR_SaveParams() PR_SaveParams() is required for implementing the +initialize diversion used by Objective-QuakeC because builtins do not have local def spaces (of course, a normal stack calling convention would help). However, it is entirely possible for a call to +initialize to trigger another call to +initialize, thus the need for stacking parameter stashes. As a bonus, this implementation cleans up some fields in progs_t.
2020-02-25 05:30:26 +00:00
memcpy (params->params + i * pr->pr_param_size,
correctly save and restore the parameters when calling +initialize. fixes the segfault in the qwaq test.
2007-06-09 13:44:06 +00:00
pr->pr_real_params[i], size);
Allow nested use of PR_SaveParams() PR_SaveParams() is required for implementing the +initialize diversion used by Objective-QuakeC because builtins do not have local def spaces (of course, a normal stack calling convention would help). However, it is entirely possible for a call to +initialize to trigger another call to +initialize, thus the need for stacking parameter stashes. As a bonus, this implementation cleans up some fields in progs_t.
2020-02-25 05:30:26 +00:00
if (i < 2) { //XXX FIXME what the what?!?
memcpy (pr->pr_real_params[i], params->param_ptrs[0], size);
}
correctly save and restore the parameters when calling +initialize. fixes the segfault in the qwaq test.
2007-06-09 13:44:06 +00:00
}
Allow nested use of PR_SaveParams() PR_SaveParams() is required for implementing the +initialize diversion used by Objective-QuakeC because builtins do not have local def spaces (of course, a normal stack calling convention would help). However, it is entirely possible for a call to +initialize to trigger another call to +initialize, thus the need for stacking parameter stashes. As a bonus, this implementation cleans up some fields in progs_t.
2020-02-25 05:30:26 +00:00
params->argc = pr->pr_argc;
return params;
correctly save and restore the parameters when calling +initialize. fixes the segfault in the qwaq test.
2007-06-09 13:44:06 +00:00
}
VISIBLE void
Allow nested use of PR_SaveParams() PR_SaveParams() is required for implementing the +initialize diversion used by Objective-QuakeC because builtins do not have local def spaces (of course, a normal stack calling convention would help). However, it is entirely possible for a call to +initialize to trigger another call to +initialize, thus the need for stacking parameter stashes. As a bonus, this implementation cleans up some fields in progs_t.
2020-02-25 05:30:26 +00:00
PR_RestoreParams (progs_t *pr, pr_stashed_params_t *params)
correctly save and restore the parameters when calling +initialize. fixes the segfault in the qwaq test.
2007-06-09 13:44:06 +00:00
{
int i;
int size = pr->pr_param_size * sizeof (pr_type_t);
Allow nested use of PR_SaveParams() PR_SaveParams() is required for implementing the +initialize diversion used by Objective-QuakeC because builtins do not have local def spaces (of course, a normal stack calling convention would help). However, it is entirely possible for a call to +initialize to trigger another call to +initialize, thus the need for stacking parameter stashes. As a bonus, this implementation cleans up some fields in progs_t.
2020-02-25 05:30:26 +00:00
pr->pr_params[0] = params->param_ptrs[0];
pr->pr_params[1] = params->param_ptrs[1];
pr->pr_argc = params->argc;
for (i = 0; i < pr->pr_argc; i++) {
correctly save and restore the parameters when calling +initialize. fixes the segfault in the qwaq test.
2007-06-09 13:44:06 +00:00
memcpy (pr->pr_real_params[i],
Allow nested use of PR_SaveParams() PR_SaveParams() is required for implementing the +initialize diversion used by Objective-QuakeC because builtins do not have local def spaces (of course, a normal stack calling convention would help). However, it is entirely possible for a call to +initialize to trigger another call to +initialize, thus the need for stacking parameter stashes. As a bonus, this implementation cleans up some fields in progs_t.
2020-02-25 05:30:26 +00:00
params->params + i * pr->pr_param_size, size);
}
correctly save and restore the parameters when calling +initialize. fixes the segfault in the qwaq test.
2007-06-09 13:44:06 +00:00
}
the big dso visibility patch :). Sure, we have to have unique names for static builds, but with controlled visibitly we should get faster program loads (although this isn't C++, so it's not as bad) and complex plugins are cleaner.
2007-03-10 12:00:59 +00:00
VISIBLE inline void
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
PR_PushFrame (progs_t *pr)
{
prstack_t *frame;
if (pr->pr_depth == MAX_STACK_DEPTH)
PR_RunError (pr, "stack overflow");
frame = pr->pr_stack + pr->pr_depth++;
frame->s = pr->pr_xstatement;
frame->f = pr->pr_xfunction;
frame->tstr = pr->pr_xtstr;
[gamecode] Add function PR_PushTempString This "pushes" a temp string onto the callee's stack frame after removing it from the caller's stack frame. This is so builtins can pass auto-freed memory to called progs code. No checking is done, but mayhem is likely to ensue if a string is pushed that was allocated in an earlier frame.
2020-03-09 14:36:09 +00:00
pr->pr_xtstr = pr->pr_pushtstr;
pr->pr_pushtstr = 0;
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
pr->pr_xfunction = 0;
}
the big dso visibility patch :). Sure, we have to have unique names for static builds, but with controlled visibitly we should get faster program loads (although this isn't C++, so it's not as bad) and complex plugins are cleaner.
2007-03-10 12:00:59 +00:00
VISIBLE inline void
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
PR_PopFrame (progs_t *pr)
{
prstack_t *frame;
if (pr->pr_depth <= 0)
PR_Error (pr, "prog stack underflow");
if (pr->pr_xtstr)
PR_FreeTempStrings (pr);
[gamecode] Add function PR_PushTempString This "pushes" a temp string onto the callee's stack frame after removing it from the caller's stack frame. This is so builtins can pass auto-freed memory to called progs code. No checking is done, but mayhem is likely to ensue if a string is pushed that was allocated in an earlier frame.
2020-03-09 14:36:09 +00:00
// normally, this won't happen, but if a builtin pushed a temp string
// when calling a function and the callee was another builtin that
// did not call a progs function, then the push strings will still be
// valid because PR_EnterFunction was never called
[gamecode] Be more careful with temp strings If a temp string is found in the return slot, PR_FreeTempStrings won't delete the string. However, PR_PopFrame was blindly stomping on the possibly surviving temp string with the push strings, which would cause a leak.
2020-03-09 18:24:31 +00:00
// however, not if a temp string survived: better to hold on to the push
// strings a little longer than lose one erroneously
if (!pr->pr_xtstr && pr->pr_pushtstr) {
[gamecode] Add function PR_PushTempString This "pushes" a temp string onto the callee's stack frame after removing it from the caller's stack frame. This is so builtins can pass auto-freed memory to called progs code. No checking is done, but mayhem is likely to ensue if a string is pushed that was allocated in an earlier frame.
2020-03-09 14:36:09 +00:00
pr->pr_xtstr = pr->pr_pushtstr;
pr->pr_pushtstr = 0;
PR_FreeTempStrings (pr);
}
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
// up stack
frame = pr->pr_stack + --pr->pr_depth;
pr->pr_xfunction = frame->f;
pr->pr_xstatement = frame->s;
pr->pr_xtstr = frame->tstr;
}
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
static __attribute__((pure)) long
align_offset (long offset, dparmsize_t parmsize)
{
int mask = (1 << parmsize.alignment) - 1;
return (offset + mask) & ~mask;
}
static void
copy_param (pr_type_t *dst, pr_type_t *src, size_t size)
{
while (size--) {
memcpy (dst++, src++, sizeof (pr_type_t));
}
}
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
/** Setup the stackframe prior to calling a progs function. Saves all local
data the called function will trample on and copies the parameters used
by the function into the function's local data space.
\param pr pointer to progs_t VM struct
\param f pointer to the descriptor for the called function
\note Passing a descriptor for a builtin function will result in
undefined behavior.
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
*/
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
static void
Fix the access alignment bug on 64 bit archs.
2007-04-07 01:41:23 +00:00
PR_EnterFunction (progs_t *pr, bfunction_t *f)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
pr_int_t i;
pr_type_t *dstParams[MAX_PARMS];
Copy only the necessary number of vararg parameters
2020-02-16 09:00:29 +00:00
pointer_t paramofs = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[gamecode] Be more informative with function calls PR_EnterFunction and PR_LeaveFunction now print the entered/left/returned function when tracing. Makes debugging obj code much easier.
2020-03-24 04:24:55 +00:00
if (pr->pr_trace) {
Sys_Printf ("Entering function %s\n",
PR_GetString (pr, f->descriptor->s_name));
}
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
PR_PushFrame (pr);
heh, slight oversight when implementing rcall using the param pointer trick. ran into the exact same problem as passing addresses of locals to QC functions. well, actually, that's exactly what it was, but done in the engine. anyway, wound up potentially overwriting the params passed to the function. fortunatly, due to how rcall works, only the first two params are an issue.
2005-06-14 13:40:34 +00:00
if (f->numparms > 0) {
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
paramofs = f->parm_start;
for (i = 0; i < f->numparms; i++) {
paramofs = align_offset (paramofs, f->parm_size[i]);
dstParams[i] = pr->pr_globals + paramofs;
paramofs += f->parm_size[i].size;
if (pr->pr_params[i] != pr->pr_real_params[i]) {
copy_param (pr->pr_real_params[i], pr->pr_params[i],
f->parm_size[i].size);
[gamecode] Replace lost parameter auto-reset rcallN messes with the progs parameter pointers and not resetting them can cause incorrect data to be copied into the called function.
2020-03-11 10:38:50 +00:00
pr->pr_params[i] = pr->pr_real_params[i];
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
}
heh, slight oversight when implementing rcall using the param pointer trick. ran into the exact same problem as passing addresses of locals to QC functions. well, actually, that's exactly what it was, but done in the engine. anyway, wound up potentially overwriting the params passed to the function. fortunatly, due to how rcall works, only the first two params are an issue.
2005-06-14 13:40:34 +00:00
}
} else if (f->numparms < 0) {
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
paramofs = f->parm_start + 2; // argc and argv
for (i = 0; i < -f->numparms - 1; i++) {
paramofs = align_offset (paramofs, f->parm_size[i]);
dstParams[i] = pr->pr_globals + paramofs;
paramofs += f->parm_size[i].size;
if (pr->pr_params[i] != pr->pr_real_params[i]) {
copy_param (pr->pr_real_params[i], pr->pr_params[i],
f->parm_size[i].size);
[gamecode] Replace lost parameter auto-reset rcallN messes with the progs parameter pointers and not resetting them can cause incorrect data to be copied into the called function.
2020-03-11 10:38:50 +00:00
pr->pr_params[i] = pr->pr_real_params[i];
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
}
heh, slight oversight when implementing rcall using the param pointer trick. ran into the exact same problem as passing addresses of locals to QC functions. well, actually, that's exactly what it was, but done in the engine. anyway, wound up potentially overwriting the params passed to the function. fortunatly, due to how rcall works, only the first two params are an issue.
2005-06-14 13:40:34 +00:00
}
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
dparmsize_t parmsize = { pr->pr_param_size, pr->pr_param_alignment };
paramofs = align_offset (paramofs, parmsize );
if (i < MAX_PARMS) {
dstParams[i] = pr->pr_globals + paramofs;
heh, slight oversight when implementing rcall using the param pointer trick. ran into the exact same problem as passing addresses of locals to QC functions. well, actually, that's exactly what it was, but done in the engine. anyway, wound up potentially overwriting the params passed to the function. fortunatly, due to how rcall works, only the first two params are an issue.
2005-06-14 13:40:34 +00:00
}
Copy only the necessary number of vararg parameters
2020-02-16 09:00:29 +00:00
for (; i < pr->pr_argc; i++) {
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
if (pr->pr_params[i] != pr->pr_real_params[i]) {
copy_param (pr->pr_real_params[i], pr->pr_params[i],
Copy the correct number of words for rcall
2020-02-16 08:58:36 +00:00
parmsize.size);
[gamecode] Replace lost parameter auto-reset rcallN messes with the progs parameter pointers and not resetting them can cause incorrect data to be copied into the called function.
2020-03-11 10:38:50 +00:00
pr->pr_params[i] = pr->pr_real_params[i];
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
}
heh, slight oversight when implementing rcall using the param pointer trick. ran into the exact same problem as passing addresses of locals to QC functions. well, actually, that's exactly what it was, but done in the engine. anyway, wound up potentially overwriting the params passed to the function. fortunatly, due to how rcall works, only the first two params are an issue.
2005-06-14 13:40:34 +00:00
}
}
elimination of bare printf (with some exceptions) from libs: Sys_Printf is used instead
2001-10-18 16:42:14 +00:00
//Sys_Printf("%s:\n", PR_GetString(pr,f->s_name));
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
pr->pr_xfunction = f;
pr->pr_xstatement = f->first_statement - 1; // offset the st++
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
use memcpy instead of ugly casts to transfer locals
2001-06-05 23:57:51 +00:00
// save off any locals that the new function steps on
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
if (pr->localstack_used + f->locals > LOCALSTACK_SIZE)
correct the function names in PR_EnterFunction and PR_LeaveFunction
2002-05-21 21:14:32 +00:00
PR_RunError (pr, "PR_EnterFunction: locals stack overflow");
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
use memcpy instead of ugly casts to transfer locals
2001-06-05 23:57:51 +00:00
memcpy (&pr->localstack[pr->localstack_used],
&pr->pr_globals[f->parm_start],
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
sizeof (pr_type_t) * f->locals);
pr->localstack_used += f->locals;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
- add some comments to libs/console/inputline.c - rename pr_deadbeef to pr_deadbeef_ents - add pr_deadbeef_locals, which does uninited locals. (crashes from it are undeniably buggy code, unlike pr_deadeef_ents) - add a missing break to packetlog printing's switch - add sv_kickfake, which kicks people for attempting to fake messages, or replaces the ^Ms with # if disabled.
2001-09-23 00:36:21 +00:00
if (pr_deadbeef_locals->int_val)
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
for (i = f->parm_start; i < f->parm_start + f->locals; i++)
pr->pr_globals[i].integer_var = 0xdeadbeef;
rhamph's deadbeaf locals patch as-is :)
2001-08-13 20:29:33 +00:00
Whitespace.
2001-09-10 12:56:23 +00:00
// copy parameters
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
if (f->numparms >= 0) {
for (i = 0; i < f->numparms; i++) {
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
copy_param (dstParams[i], pr->pr_params[i], f->parm_size[i].size);
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
}
} else {
Copy only the necessary number of vararg parameters
2020-02-16 09:00:29 +00:00
int copy_args;
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
pr_type_t *argc = &pr->pr_globals[f->parm_start + 0];
pr_type_t *argv = &pr->pr_globals[f->parm_start + 1];
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
for (i = 0; i < -f->numparms - 1; i++) {
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
copy_param (dstParams[i], pr->pr_params[i], f->parm_size[i].size);
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
}
Copy only the necessary number of vararg parameters
2020-02-16 09:00:29 +00:00
copy_args = pr->pr_argc - i;
argc->integer_var = copy_args;
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
argv->integer_var = dstParams[i] - pr->pr_globals;
can now handle > vector sized parameters
2003-04-22 20:11:16 +00:00
if (i < MAX_PARMS) {
Copy only the necessary number of vararg parameters
2020-02-16 09:00:29 +00:00
memcpy (dstParams[i], pr->pr_params[i],
(copy_args * pr->pr_param_size) * sizeof (pr_type_t));
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
}
}
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
static void
[gamecode] Be more informative with function calls PR_EnterFunction and PR_LeaveFunction now print the entered/left/returned function when tracing. Makes debugging obj code much easier.
2020-03-24 04:24:55 +00:00
PR_LeaveFunction (progs_t *pr, int to_engine)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
Fix the access alignment bug on 64 bit archs.
2007-04-07 01:41:23 +00:00
bfunction_t *f = pr->pr_xfunction;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
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
PR_PopFrame (pr);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[gamecode] Be more informative with function calls PR_EnterFunction and PR_LeaveFunction now print the entered/left/returned function when tracing. Makes debugging obj code much easier.
2020-03-24 04:24:55 +00:00
if (pr->pr_trace) {
Sys_Printf ("Leaving function %s\n",
PR_GetString (pr, f->descriptor->s_name));
if (to_engine) {
Sys_Printf ("Returning to engine\n");
} else {
bfunction_t *rf = pr->pr_xfunction;
[gamecode] Fix some silly errors I forgot to compile-test :(
2020-03-24 14:16:25 +00:00
if (rf) {
Sys_Printf ("Returning to function %s\n",
PR_GetString (pr, rf->descriptor->s_name));
}
[gamecode] Be more informative with function calls PR_EnterFunction and PR_LeaveFunction now print the entered/left/returned function when tracing. Makes debugging obj code much easier.
2020-03-24 04:24:55 +00:00
}
}
Whitespace.
2001-09-10 12:56:23 +00:00
// restore locals from the stack
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
pr->localstack_used -= f->locals;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
if (pr->localstack_used < 0)
correct the function names in PR_EnterFunction and PR_LeaveFunction
2002-05-21 21:14:32 +00:00
PR_RunError (pr, "PR_LeaveFunction: locals stack underflow");
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
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
memcpy (&pr->pr_globals[f->parm_start],
Implement parameter alignment in the engine The engine now requires non-v6 progs to store the log2 alignment for the param struct in .param_alignment. PR_EnterFunction is clearer and possibly more efficient.
2020-02-16 08:13:45 +00:00
&pr->localstack[pr->localstack_used],
sizeof (pr_type_t) * f->locals);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Really nail down the args param in rua_obj_msg_sendv().
2010-01-13 06:36:16 +00:00
VISIBLE void
Change pointer_t to unsigned and clean up the mess. It doesn't make sense to have negative pointers. The size of the commit is from enabling gcc's -Wtype-limits warning and cleaning up that mess too.
2012-12-21 12:53:13 +00:00
PR_BoundsCheckSize (progs_t *pr, pointer_t addr, unsigned size)
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
{
Fix some signed/unsigned comparison warnings. It seems mingw's gcc treats pointer subtraction differently to how linux's gcc does.
2013-01-05 08:40:59 +00:00
if (addr < (pointer_t) (pr->pr_return - pr->pr_globals))
Check for null VM pointer accesses. Currently only four words of "null pointer protection" are available, but it's a start.
2012-12-22 05:24:11 +00:00
PR_RunError (pr, "null pointer access");
Change pointer_t to unsigned and clean up the mess. It doesn't make sense to have negative pointers. The size of the commit is from enabling gcc's -Wtype-limits warning and cleaning up that mess too.
2012-12-21 12:53:13 +00:00
if (addr >= pr->globals_size
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
|| size > (unsigned) (pr->globals_size - addr))
PR_RunError (pr, "invalid memory access: %d (0 to %d-%d)", addr,
pr->globals_size, size);
Add improved memory checking to progs. With pr_boundscheck >= 2, pointer access will be checked against allocated blocks (qfvalgrind?:). Currently extremely basic, but it seems to work.
2012-11-19 11:03:21 +00:00
if (pr_boundscheck->int_val >= 2
&& PR_GetPointer (pr, addr + size) > (pr_type_t *) pr->zone) {
void *mem = (void *) PR_GetPointer (pr, addr);
Z_CheckPointer (pr->zone, mem, size * sizeof (pr_type_t));
}
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
}
Really nail down the args param in rua_obj_msg_sendv().
2010-01-13 06:36:16 +00:00
VISIBLE void
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
PR_BoundsCheck (progs_t *pr, int addr, etype_t type)
{
PR_BoundsCheckSize (pr, addr, pr_type_size[type]);
}
this should speed up progs a little bit
2002-06-07 19:41:13 +00:00
#define OPA (*op_a)
#define OPB (*op_b)
#define OPC (*op_c)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
#define OPA_double_var (*((double *) (op_a)))
#define OPB_double_var (*((double *) (op_b)))
#define OPC_double_var (*((double *) (op_c)))
make float != 0 tests work on alpha for when the value is actually an integer (works fine on intel, but on alpha the denormals either produce an exception or true zero: not quite desirable:)
2001-11-19 17:51:31 +00:00
/*
This gets around the problem of needing to test for -0.0 but denormals
causing exceptions (or wrong results for what we need) on the alpha.
*/
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
#define FNZ(x) ((x).uinteger_var & ~0x80000000u)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
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
static int
signal_hook (int sig, void *data)
{
progs_t *pr = (progs_t *) data;
Massive whitespace cleanup. Lots of trailing whitespace and otherwise blank lines.
2012-05-21 23:23:22 +00:00
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
if (sig == SIGFPE && pr_faultchecks->int_val) {
dstatement_t *st;
pr_type_t *op_a, *op_b, *op_c;
st = pr->pr_statements + pr->pr_xstatement;
op_a = pr->pr_globals + st->a;
op_b = pr->pr_globals + st->b;
op_c = pr->pr_globals + st->c;
switch (st->op) {
case OP_DIV_F:
if ((OPA.integer_var & 0x80000000)
^ (OPB.integer_var & 0x80000000))
OPC.integer_var = 0xff7fffff;
else
OPC.integer_var = 0x7f7fffff;
return 1;
case OP_DIV_I:
if (OPA.integer_var & 0x80000000)
OPC.integer_var = -0x80000000;
else
OPC.integer_var = 0x7fffffff;
return 1;
Implement true modulo in the engine
2020-02-16 02:53:56 +00:00
case OP_MOD_I:
case OP_MOD_F:
Rename the mod instruction to rem Because % really implements remainder rather than true modulo, and I plan on adding %% to implement true modulo.
2020-02-16 02:04:30 +00:00
case OP_REM_I:
case OP_REM_F:
handle x % 0 by giving a 0 result
2004-02-20 00:25:08 +00:00
OPC.integer_var = 0x00000000;
return 1;
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
default:
break;
}
}
PR_DumpState (pr);
Ensure stdout is flushed in the progs signal_hook()
2011-03-20 04:34:28 +00:00
fflush (stdout);
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
return 0;
}
Allow Sys_Error to be hooked. This makes debugging builtins that wrap normal functions a little easier by giving a progs dump when such an error occurs.
2016-01-03 14:04:00 +00:00
static void
error_handler (void *data)
{
progs_t *pr = (progs_t *) data;
PR_DumpState (pr);
fflush (stdout);
}
the big dso visibility patch :). Sure, we have to have unique names for static builds, but with controlled visibitly we should get faster program loads (although this isn't C++, so it's not as bad) and complex plugins are cleaner.
2007-03-10 12:00:59 +00:00
VISIBLE int
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
PR_CallFunction (progs_t *pr, func_t fnum)
{
Fix the access alignment bug on 64 bit archs.
2007-04-07 01:41:23 +00:00
bfunction_t *f;
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
if (!fnum)
PR_RunError (pr, "NULL function");
Fix the access alignment bug on 64 bit archs.
2007-04-07 01:41:23 +00:00
f = pr->function_table + fnum;
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
if (f->first_statement < 0) {
// negative statements are built in functions
Indicate calls to builtins when tracing. This makes following progs traces a little easier.
2016-01-03 07:13:59 +00:00
if (pr->pr_trace) {
Sys_Printf ("Calling builtin %s @ %p\n",
PR_GetString (pr, f->descriptor->s_name), f->func);
}
Fix the access alignment bug on 64 bit archs.
2007-04-07 01:41:23 +00:00
f->func (pr);
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
return 0;
} else {
PR_EnterFunction (pr, f);
return 1;
}
}
Add a stack and push and pop instructions. The stack is for data, not return addresses.
2018-10-11 04:24:03 +00:00
static void
check_stack_pointer (progs_t *pr, pointer_t stack, int size)
{
if (stack < pr->stack_bottom) {
PR_RunError (pr, "Progs stack overflow");
}
if (stack > pr->globals_size - size) {
PR_RunError (pr, "Progs stack underflow");
}
}
[gamecode] Add memset instructions
2020-03-11 13:48:55 +00:00
static inline void
pr_memset (pr_type_t *dst, int val, int count)
{
while (count-- > 0) {
(*dst++).integer_var = val;
}
}
Whitespace.
2001-09-10 12:56:23 +00:00
/*
PR_ExecuteProgram
The interpretation main loop
*/
the big dso visibility patch :). Sure, we have to have unique names for static builds, but with controlled visibitly we should get faster program loads (although this isn't C++, so it's not as bad) and complex plugins are cleaner.
2007-03-10 12:00:59 +00:00
VISIBLE void
Add a stack and push and pop instructions. The stack is for data, not return addresses.
2018-10-11 04:24:03 +00:00
PR_ExecuteProgram (progs_t *pr, func_t fnum)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
make progs related code a little more consistent with its int type usage
2007-04-06 00:47:41 +00:00
int exitdepth, profile, startprofile;
pr_uint_t pointer;
dstatement_t *st;
edict_t *ed;
pr_type_t *ptr;
[gamecode] Allow watch-point to be updated With the watch-point address cached in a local var, it was not possible to update the watch-point while the progs were executing.
2020-03-24 10:45:39 +00:00
pr_type_t old_val = {0};
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Whitespace.
2001-09-10 12:56:23 +00:00
// make a stack frame
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
exitdepth = pr->pr_depth;
startprofile = profile = 0;
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
Sys_PushSignalHook (signal_hook, pr);
Allow Sys_Error to be hooked. This makes debugging builtins that wrap normal functions a little easier by giving a progs dump when such an error occurs.
2016-01-03 14:04:00 +00:00
Sys_PushErrorHandler (error_handler, pr);
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
[gamecode] Add VM enter/exit events And rename prd_exit to prd_terminate (the idea is the host will terminate the VM). This makes it possible for the debugger to pause the VM before any code, even a builtin function, is executed. Breaks the debugger source window, but only because it's not updating on file change (I think).
2020-03-26 03:30:32 +00:00
if (pr->debug_handler) {
pr->debug_handler (prd_subenter, &fnum, pr->debug_data);
}
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
if (!PR_CallFunction (pr, fnum)) {
// called a builtin instead of progs code
Allow Sys_Error to be hooked. This makes debugging builtins that wrap normal functions a little easier by giving a progs dump when such an error occurs.
2016-01-03 14:04:00 +00:00
goto exit_program;
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
}
st = pr->pr_statements + pr->pr_xstatement;
"hardware" (haha) watch points (one!) in progs. only catches changes done by progs, and the expression parser is as flaky as anything, but it's better than nothing :)
2007-05-08 02:04:47 +00:00
if (pr->watch) {
[gamecode] Fix some silly errors I forgot to compile-test :(
2020-03-24 14:16:25 +00:00
old_val = *pr->watch;
"hardware" (haha) watch points (one!) in progs. only catches changes done by progs, and the expression parser is as flaky as anything, but it's better than nothing :)
2007-05-08 02:04:47 +00:00
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
while (1) {
this should speed up progs a little bit
2002-06-07 19:41:13 +00:00
pr_type_t *op_a, *op_b, *op_c;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
st++;
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
++pr->pr_xstatement;
if (pr->pr_xstatement != st - pr->pr_statements)
PR_RunError (pr, "internal error");
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
if (++profile > 1000000 && !pr->no_exec_limit) {
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
PR_RunError (pr, "runaway loop error");
}
this should speed up progs a little bit
2002-06-07 19:41:13 +00:00
op_a = pr->pr_globals + st->a;
op_b = pr->pr_globals + st->b;
op_c = pr->pr_globals + st->c;
[gamecode] Add debug hooks to the VM engine While there was a breakpoint hook, it was for only breakpoints and more was needed. Now there's a generic hook that is called for tracing, breakpoints, watch points, runtime errors and VM errors, with the "event" type passed as the first parameter and a data pointer in the second.
2020-03-24 06:35:42 +00:00
if (pr->pr_trace) {
if (pr->debug_handler) {
[gamecode] Add a param pointer to debug_handler I decided I want events for VM enter/exit but enter needs to somehow pass the function which will be executed (even if a builtin). A generic void * param seemed the best idea, which meant the error string could be passed via the param instead of a "global" string in the progs struct.
2020-03-26 02:44:02 +00:00
pr->debug_handler (prd_trace, 0, pr->debug_data);
[gamecode] Add debug hooks to the VM engine While there was a breakpoint hook, it was for only breakpoints and more was needed. Now there's a generic hook that is called for tracing, breakpoints, watch points, runtime errors and VM errors, with the "event" type passed as the first parameter and a data pointer in the second.
2020-03-24 06:35:42 +00:00
} else {
PR_PrintStatement (pr, st, 1);
}
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Add a breakpoint flag to opcodes The progs execution code will call a breakpoint handler just before executing an instruction with the flag set. This means there's no need for the breakpoint handler to mess with execution state or even the instruction in order to continue past the breakpoint. The flag being set in a progs file is invalid.
2020-02-26 04:40:26 +00:00
if (st->op & OP_BREAK) {
[gamecode] Add debug hooks to the VM engine While there was a breakpoint hook, it was for only breakpoints and more was needed. Now there's a generic hook that is called for tracing, breakpoints, watch points, runtime errors and VM errors, with the "event" type passed as the first parameter and a data pointer in the second.
2020-03-24 06:35:42 +00:00
if (pr->debug_handler) {
[gamecode] Add a param pointer to debug_handler I decided I want events for VM enter/exit but enter needs to somehow pass the function which will be executed (even if a builtin). A generic void * param seemed the best idea, which meant the error string could be passed via the param instead of a "global" string in the progs struct.
2020-03-26 02:44:02 +00:00
pr->debug_handler (prd_breakpoint, 0, pr->debug_data);
Add a breakpoint flag to opcodes The progs execution code will call a breakpoint handler just before executing an instruction with the flag set. This means there's no need for the breakpoint handler to mess with execution state or even the instruction in order to continue past the breakpoint. The flag being set in a progs file is invalid.
2020-02-26 04:40:26 +00:00
} else {
PR_RunError (pr, "breakpoint hit");
}
}
pr_opcode_e op = st->op & ~OP_BREAK;
switch (op) {
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_ADD_D:
OPC_double_var = OPA_double_var + OPB_double_var;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_ADD_F:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = OPA.float_var + OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_ADD_V:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
VectorAdd (OPA.vector_var, OPB.vector_var, OPC.vector_var);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_ADD_Q:
QuatAdd (OPA.quat_var, OPB.quat_var, OPC.quat_var);
break;
pr_comp.h: add OP_ADD_S. WARNING!!! this /will/ move. progs.h: add prototype for PR_PrintStatement pr_edict.c: add OP_ADD_S support in the progs checker pr_exec.c: implement OP_ADD_S tools/qfcc/include/.gitignore: add config.h.in qfcc.h: nuke PR_NameImmediate and change PR_ParseImmediate's prototype (see pr_imm.c) pr_comp.c: add ADD_S, adjust for PR_ParseImmediate's prototype, make PR_ParseExpression work with non-sequential opcodes (slow, will work on that next). Fix up initialised global parsing. pr_imm.c: nuke PR_NameImmediate. didn't work well and wasn't such a good idea anyway. PR_ParseImmediate now accepts a def_t * arg. if null, will allocate a new global def, otherwise it will initialize the def passed in. qwaq/main.c: sports some debugging code (dumps info about the progs it's running) qwaq/main.qc: better ADD_S testing
2001-06-03 17:36:49 +00:00
case OP_ADD_S:
avoid using the hunk in the progs engine (except for the override-able load and alloc functions) so tools (qfprogs and qwaq) don't need to create it. register only the first instance of a progs provided string This is an imperfect revision of history.
2004-11-02 04:59:00 +00:00
OPC.string_var = PR_CatStrings (pr,
PR_GetString (pr,
OPA.string_var),
PR_GetString (pr,
OPB.string_var));
pr_comp.h: add OP_ADD_S. WARNING!!! this /will/ move. progs.h: add prototype for PR_PrintStatement pr_edict.c: add OP_ADD_S support in the progs checker pr_exec.c: implement OP_ADD_S tools/qfcc/include/.gitignore: add config.h.in qfcc.h: nuke PR_NameImmediate and change PR_ParseImmediate's prototype (see pr_imm.c) pr_comp.c: add ADD_S, adjust for PR_ParseImmediate's prototype, make PR_ParseExpression work with non-sequential opcodes (slow, will work on that next). Fix up initialised global parsing. pr_imm.c: nuke PR_NameImmediate. didn't work well and wasn't such a good idea anyway. PR_ParseImmediate now accepts a def_t * arg. if null, will allocate a new global def, otherwise it will initialize the def passed in. qwaq/main.c: sports some debugging code (dumps info about the progs it's running) qwaq/main.qc: better ADD_S testing
2001-06-03 17:36:49 +00:00
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_SUB_D:
OPC_double_var = OPA_double_var - OPB_double_var;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_SUB_F:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = OPA.float_var - OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_SUB_V:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
VectorSubtract (OPA.vector_var, OPB.vector_var, OPC.vector_var);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_SUB_Q:
QuatSubtract (OPA.quat_var, OPB.quat_var, OPC.quat_var);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_MUL_D:
OPC_double_var = OPA_double_var * OPB_double_var;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_MUL_F:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = OPA.float_var * OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_MUL_V:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
OPC.float_var = DotProduct (OPA.vector_var, OPB.vector_var);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_MUL_DV:
{
// avoid issues with the likes of x = x.x * x;
// makes for faster code, too
double scale = OPA_double_var;
VectorScale (OPB.vector_var, scale, OPC.vector_var);
}
break;
case OP_MUL_VD:
{
// avoid issues with the likes of x = x * x.x;
// makes for faster code, too
double scale = OPB_double_var;
VectorScale (OPA.vector_var, scale, OPC.vector_var);
}
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_MUL_FV:
Fix the vector/quaternion scaling instructions. It was pointed out by Blub\w (gmqcc) that OP_MUL_FV and friends were buggy when the operands overlapped (eg, x = x.x * x) as the result would become 'x.x*x.x x.y*x.x*x.x x.z*x.x*x.x' (note the x.x squared for y and z). On testing, sure enough the bug was present (and is a nice demonstration that QF's VM does NOT have strict-aliasing bugs). As a very nice benefit: the code produced by the fixes is actually faster than the broken version :). The ruamoko code used for testing: void (string fmt, ...) printf = #0; vector foo (vector x) { x = x * x.x; return x; } vector bar (vector x) { x = x.x * x; return x; } int main () { vector x = '2 3 4'; vector y = foo (x); vector z = bar (x); printf ("x=%v y=%v z=%v 2*x=%v\n", x, y, z, 2*x); return 0; }
2013-01-17 01:23:02 +00:00
{
// avoid issues with the likes of x = x.x * x;
// makes for faster code, too
float scale = OPA.float_var;
VectorScale (OPB.vector_var, scale, OPC.vector_var);
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_MUL_VF:
Fix the vector/quaternion scaling instructions. It was pointed out by Blub\w (gmqcc) that OP_MUL_FV and friends were buggy when the operands overlapped (eg, x = x.x * x) as the result would become 'x.x*x.x x.y*x.x*x.x x.z*x.x*x.x' (note the x.x squared for y and z). On testing, sure enough the bug was present (and is a nice demonstration that QF's VM does NOT have strict-aliasing bugs). As a very nice benefit: the code produced by the fixes is actually faster than the broken version :). The ruamoko code used for testing: void (string fmt, ...) printf = #0; vector foo (vector x) { x = x * x.x; return x; } vector bar (vector x) { x = x.x * x; return x; } int main () { vector x = '2 3 4'; vector y = foo (x); vector z = bar (x); printf ("x=%v y=%v z=%v 2*x=%v\n", x, y, z, 2*x); return 0; }
2013-01-17 01:23:02 +00:00
{
// avoid issues with the likes of x = x * x.x;
// makes for faster code, too
float scale = OPB.float_var;
VectorScale (OPA.vector_var, scale, OPC.vector_var);
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_MUL_Q:
QuatMult (OPA.quat_var, OPB.quat_var, OPC.quat_var);
break;
Add vec = quat * vec to the progs engine.
2012-04-26 01:29:21 +00:00
case OP_MUL_QV:
QuatMultVec (OPA.quat_var, OPB.vector_var, OPC.vector_var);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_MUL_DQ:
{
// avoid issues with the likes of x = x.s * x;
// makes for faster code, too
double scale = OPA_double_var;
QuatScale (OPB.quat_var, scale, OPC.quat_var);
}
break;
case OP_MUL_QD:
{
// avoid issues with the likes of x = x * x.s;
// makes for faster code, too
double scale = OPB_double_var;
QuatScale (OPA.quat_var, scale, OPC.quat_var);
}
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_MUL_FQ:
Fix the vector/quaternion scaling instructions. It was pointed out by Blub\w (gmqcc) that OP_MUL_FV and friends were buggy when the operands overlapped (eg, x = x.x * x) as the result would become 'x.x*x.x x.y*x.x*x.x x.z*x.x*x.x' (note the x.x squared for y and z). On testing, sure enough the bug was present (and is a nice demonstration that QF's VM does NOT have strict-aliasing bugs). As a very nice benefit: the code produced by the fixes is actually faster than the broken version :). The ruamoko code used for testing: void (string fmt, ...) printf = #0; vector foo (vector x) { x = x * x.x; return x; } vector bar (vector x) { x = x.x * x; return x; } int main () { vector x = '2 3 4'; vector y = foo (x); vector z = bar (x); printf ("x=%v y=%v z=%v 2*x=%v\n", x, y, z, 2*x); return 0; }
2013-01-17 01:23:02 +00:00
{
// avoid issues with the likes of x = x.s * x;
// makes for faster code, too
float scale = OPA.float_var;
QuatScale (OPB.quat_var, scale, OPC.quat_var);
}
engine side quaternion support
2004-04-08 00:56:30 +00:00
break;
case OP_MUL_QF:
Fix the vector/quaternion scaling instructions. It was pointed out by Blub\w (gmqcc) that OP_MUL_FV and friends were buggy when the operands overlapped (eg, x = x.x * x) as the result would become 'x.x*x.x x.y*x.x*x.x x.z*x.x*x.x' (note the x.x squared for y and z). On testing, sure enough the bug was present (and is a nice demonstration that QF's VM does NOT have strict-aliasing bugs). As a very nice benefit: the code produced by the fixes is actually faster than the broken version :). The ruamoko code used for testing: void (string fmt, ...) printf = #0; vector foo (vector x) { x = x * x.x; return x; } vector bar (vector x) { x = x.x * x; return x; } int main () { vector x = '2 3 4'; vector y = foo (x); vector z = bar (x); printf ("x=%v y=%v z=%v 2*x=%v\n", x, y, z, 2*x); return 0; }
2013-01-17 01:23:02 +00:00
{
// avoid issues with the likes of x = x * x.s;
// makes for faster code, too
float scale = OPB.float_var;
QuatScale (OPA.quat_var, scale, OPC.quat_var);
}
engine side quaternion support
2004-04-08 00:56:30 +00:00
break;
quaternion conjugate support (engine side)
2004-04-08 04:57:17 +00:00
case OP_CONJ_Q:
QuatConj (OPA.quat_var, OPC.quat_var);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_DIV_D:
OPC_double_var = OPA_double_var / OPB_double_var;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_DIV_F:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = OPA.float_var / OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_BITAND:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = (int) OPA.float_var & (int) OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_BITOR:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = (int) OPA.float_var | (int) OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
^ and ~ operators for qc (^ from Rhamphoryncus) and clean up the punctuation line in qc-lex.l (why didn't I think to do that in the first place?!?)
2001-08-09 16:34:46 +00:00
case OP_BITXOR_F:
OPC.float_var = (int) OPA.float_var ^ (int) OPB.float_var;
break;
case OP_BITNOT_F:
OPC.float_var = ~ (int) OPA.float_var;
break;
got a few new operators for qc :) << >> % which all act like their C counterparts += -= *= /= &= ^= |= <<= >>= %= which just expand into the obvious, also like the C versions.
2001-08-10 16:17:00 +00:00
case OP_SHL_F:
OPC.float_var = (int) OPA.float_var << (int) OPB.float_var;
break;
case OP_SHR_F:
OPC.float_var = (int) OPA.float_var >> (int) OPB.float_var;
break;
case OP_SHL_I:
OPC.integer_var = OPA.integer_var << OPB.integer_var;
break;
case OP_SHR_I:
OPC.integer_var = OPA.integer_var >> OPB.integer_var;
break;
get unsigned mostly working
2003-08-01 21:20:04 +00:00
case OP_SHR_U:
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 01:07:47 +00:00
OPC.uinteger_var = OPA.uinteger_var >> OPB.integer_var;
get unsigned mostly working
2003-08-01 21:20:04 +00:00
break;
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_GE_F:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = OPA.float_var >= OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_LE_F:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = OPA.float_var <= OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_GT_F:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = OPA.float_var > OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_LT_F:
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
OPC.float_var = OPA.float_var < OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
change the results of the test instructions from float to int.
2001-07-22 19:58:17 +00:00
case OP_AND: // OPA and OPB have to be float for -0.0
make float != 0 tests work on alpha for when the value is actually an integer (works fine on intel, but on alpha the denormals either produce an exception or true zero: not quite desirable:)
2001-11-19 17:51:31 +00:00
OPC.integer_var = FNZ (OPA) && FNZ (OPB);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
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
case OP_OR: // OPA and OPB have to be float for -0.0
make float != 0 tests work on alpha for when the value is actually an integer (works fine on intel, but on alpha the denormals either produce an exception or true zero: not quite desirable:)
2001-11-19 17:51:31 +00:00
OPC.integer_var = FNZ (OPA) || FNZ (OPB);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_NOT_F:
make float != 0 tests work on alpha for when the value is actually an integer (works fine on intel, but on alpha the denormals either produce an exception or true zero: not quite desirable:)
2001-11-19 17:51:31 +00:00
OPC.integer_var = !FNZ (OPA);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_NOT_V:
LordHavoc's VectorIsNull changes from 0.3 but as VectorIsZero (no such thing as a `null' vector) plus a couple of other bits from his patch.
2001-09-11 05:18:15 +00:00
OPC.integer_var = VectorIsZero (OPA.vector_var);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_NOT_Q:
OPC.integer_var = QuatIsZero (OPA.quat_var);
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_NOT_S:
Whitespace.
2001-09-10 12:56:23 +00:00
OPC.integer_var = !OPA.string_var ||
!*PR_GetString (pr, OPA.string_var);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_NOT_FN:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = !OPA.func_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_NOT_ENT:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = !OPA.entity_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_EQ_F:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.float_var == OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_EQ_V:
more mathlib usage
2001-11-06 20:36:20 +00:00
OPC.integer_var = VectorCompare (OPA.vector_var,
OPB.vector_var);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_EQ_Q:
OPC.integer_var = QuatCompare (OPA.quat_var, OPB.quat_var);
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_EQ_E:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.integer_var == OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_EQ_FN:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.func_var == OPB.func_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_NE_F:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.float_var != OPB.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_NE_V:
more mathlib usage
2001-11-06 20:36:20 +00:00
OPC.integer_var = !VectorCompare (OPA.vector_var,
OPB.vector_var);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_NE_Q:
OPC.integer_var = !QuatCompare (OPA.quat_var, OPB.quat_var);
break;
string comparison operators
2001-06-04 03:36:35 +00:00
case OP_LE_S:
case OP_GE_S:
case OP_LT_S:
case OP_GT_S:
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_NE_S:
string comparison operators
2001-06-04 03:36:35 +00:00
case OP_EQ_S:
{
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
int cmp = strcmp (PR_GetString (pr, OPA.string_var),
PR_GetString (pr, OPB.string_var));
string comparison operators
2001-06-04 03:36:35 +00:00
switch (st->op) {
case OP_LE_S: cmp = (cmp <= 0); break;
case OP_GE_S: cmp = (cmp >= 0); break;
case OP_LT_S: cmp = (cmp < 0); break;
case OP_GT_S: cmp = (cmp > 0); break;
case OP_NE_S: break;
case OP_EQ_S: cmp = !cmp; break;
blah, enums and switch... (at least gcc 3.1 is warning about that again:)
2002-06-09 16:31:08 +00:00
default: break;
string comparison operators
2001-06-04 03:36:35 +00:00
}
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = cmp;
string comparison operators
2001-06-04 03:36:35 +00:00
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_NE_E:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.integer_var != OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_NE_FN:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.func_var != OPB.func_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
new operand handling from nq
2001-02-27 08:21:40 +00:00
// ==================
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_STORE_F:
case OP_STORE_ENT:
case OP_STORE_FLD: // integers
case OP_STORE_S:
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_STORE_FN: // pointers
redundant code removal
2001-11-02 03:04:01 +00:00
case OP_STORE_I:
forgot to implement store.p
2002-01-23 22:33:22 +00:00
case OP_STORE_P:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPB.integer_var = OPA.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_STORE_V:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
VectorCopy (OPA.vector_var, OPB.vector_var);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_STORE_Q:
QuatCopy (OPA.quat_var, OPB.quat_var);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_STORE_D:
OPB_double_var = OPA_double_var;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_STOREP_F:
case OP_STOREP_ENT:
case OP_STOREP_FLD: // integers
case OP_STOREP_S:
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_STOREP_FN: // pointers
redundant code removal
2001-11-02 03:04:01 +00:00
case OP_STOREP_I:
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
case OP_STOREP_P:
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
pointer = OPB.integer_var;
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_integer);
}
ptr = pr->pr_globals + pointer;
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
ptr->integer_var = OPA.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_STOREP_V:
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
pointer = OPB.integer_var;
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_vector);
}
ptr = pr->pr_globals + pointer;
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
VectorCopy (OPA.vector_var, ptr->vector_var);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_STOREP_Q:
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
pointer = OPB.integer_var;
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_quat);
}
ptr = pr->pr_globals + pointer;
engine side quaternion support
2004-04-08 00:56:30 +00:00
QuatCopy (OPA.quat_var, ptr->quat_var);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_STOREP_D:
pointer = OPB.integer_var;
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_double);
}
ptr = pr->pr_globals + pointer;
*(double *) ptr = OPA_double_var;
break;
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_ADDRESS:
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
if (pr_boundscheck->int_val) {
if (OPA.entity_var < 0
|| OPA.entity_var >= pr->pr_edictareasize)
PR_RunError (pr, "Progs attempted to address an out "
"of bounds edict");
if (OPA.entity_var == 0 && pr->null_bad)
PR_RunError (pr, "assignment to world entity");
make qf gcc 3.3 clean
2003-04-17 00:01:48 +00:00
if (OPB.uinteger_var >= pr->progs->entityfields)
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
PR_RunError (pr, "Progs attempted to address an "
"invalid field in an edict");
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
ed = PROG_TO_EDICT (pr, OPA.entity_var);
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
OPC.integer_var = &ed->v[OPB.integer_var] - pr->pr_globals;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
Add a void address instruction.
2011-03-10 10:28:03 +00:00
case OP_ADDRESS_VOID:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
case OP_ADDRESS_F:
case OP_ADDRESS_V:
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_ADDRESS_Q:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
case OP_ADDRESS_S:
case OP_ADDRESS_ENT:
case OP_ADDRESS_FLD:
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_ADDRESS_FN:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
case OP_ADDRESS_I:
case OP_ADDRESS_P:
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_ADDRESS_D:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
OPC.integer_var = st->a;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_LOAD_F:
case OP_LOAD_FLD:
case OP_LOAD_ENT:
case OP_LOAD_S:
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_LOAD_FN:
redundant code removal
2001-11-02 03:04:01 +00:00
case OP_LOAD_I:
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
case OP_LOAD_P:
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
if (pr_boundscheck->int_val) {
if (OPA.entity_var < 0
|| OPA.entity_var >= pr->pr_edictareasize)
PR_RunError (pr, "Progs attempted to read an out of "
"bounds edict number");
make qf gcc 3.3 clean
2003-04-17 00:01:48 +00:00
if (OPB.uinteger_var >= pr->progs->entityfields)
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
PR_RunError (pr, "Progs attempted to read an invalid "
"field in an edict");
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
ed = PROG_TO_EDICT (pr, OPA.entity_var);
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = ed->v[OPB.integer_var].integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_LOAD_V:
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
if (pr_boundscheck->int_val) {
if (OPA.entity_var < 0
|| OPA.entity_var >= pr->pr_edictareasize)
engine side quaternion support
2004-04-08 00:56:30 +00:00
PR_RunError (pr, "Progs attempted to read an out of "
"bounds edict number");
make qf gcc 3.3 clean
2003-04-17 00:01:48 +00:00
if (OPB.uinteger_var + 2 >= pr->progs->entityfields)
engine side quaternion support
2004-04-08 00:56:30 +00:00
PR_RunError (pr, "Progs attempted to read an invalid "
"field in an edict");
}
ed = PROG_TO_EDICT (pr, OPA.entity_var);
memcpy (&OPC, &ed->v[OPB.integer_var], 3 * sizeof (OPC));
break;
case OP_LOAD_Q:
if (pr_boundscheck->int_val) {
if (OPA.entity_var < 0
|| OPA.entity_var >= pr->pr_edictareasize)
PR_RunError (pr, "Progs attempted to read an out of "
"bounds edict number");
if (OPB.uinteger_var + 3 >= pr->progs->entityfields)
PR_RunError (pr, "Progs attempted to read an invalid "
"field in an edict");
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
ed = PROG_TO_EDICT (pr, OPA.entity_var);
Fix a size error in OP_LOAD_Q. C&P from OP_LOAD_V :/
2018-10-11 04:25:32 +00:00
memcpy (&OPC, &ed->v[OPB.integer_var], 4 * sizeof (OPC));
new operand handling from nq
2001-02-27 08:21:40 +00:00
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_LOAD_D:
if (pr_boundscheck->int_val) {
if (OPA.entity_var < 0
|| OPA.entity_var >= pr->pr_edictareasize)
PR_RunError (pr, "Progs attempted to read an out of "
"bounds edict number");
if (OPB.uinteger_var + 1 >= pr->progs->entityfields)
PR_RunError (pr, "Progs attempted to read an invalid "
"field in an edict");
}
ed = PROG_TO_EDICT (pr, OPA.entity_var);
OPC_double_var = *(double *) (ed->v + OPB.integer_var);
break;
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
case OP_LOADB_F:
case OP_LOADB_S:
case OP_LOADB_ENT:
case OP_LOADB_FLD:
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_LOADB_FN:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
case OP_LOADB_I:
case OP_LOADB_P:
pointer = OPA.integer_var + OPB.integer_var;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_integer);
}
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
ptr = pr->pr_globals + pointer;
OPC.integer_var = ptr->integer_var;
break;
case OP_LOADB_V:
pointer = OPA.integer_var + OPB.integer_var;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_vector);
}
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
ptr = pr->pr_globals + pointer;
VectorCopy (ptr->vector_var, OPC.vector_var);
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_LOADB_Q:
pointer = OPA.integer_var + OPB.integer_var;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_quat);
}
engine side quaternion support
2004-04-08 00:56:30 +00:00
ptr = pr->pr_globals + pointer;
QuatCopy (ptr->quat_var, OPC.quat_var);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_LOADB_D:
pointer = OPA.integer_var + OPB.integer_var;
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_double);
}
ptr = pr->pr_globals + pointer;
OPC_double_var = *(double *) ptr;
break;
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
case OP_LOADBI_F:
case OP_LOADBI_S:
case OP_LOADBI_ENT:
case OP_LOADBI_FLD:
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_LOADBI_FN:
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
case OP_LOADBI_I:
case OP_LOADBI_P:
pointer = OPA.integer_var + (short) st->b;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_integer);
}
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
ptr = pr->pr_globals + pointer;
OPC.integer_var = ptr->integer_var;
break;
case OP_LOADBI_V:
pointer = OPA.integer_var + (short) st->b;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_vector);
}
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
ptr = pr->pr_globals + pointer;
VectorCopy (ptr->vector_var, OPC.vector_var);
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_LOADBI_Q:
pointer = OPA.integer_var + (short) st->b;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_quat);
}
engine side quaternion support
2004-04-08 00:56:30 +00:00
ptr = pr->pr_globals + pointer;
QuatCopy (ptr->quat_var, OPC.quat_var);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_LOADBI_D:
pointer = OPA.integer_var + (short) st->b;
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_quat);
}
ptr = pr->pr_globals + pointer;
OPC_double_var = *(double *) ptr;
break;
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
case OP_LEA:
pointer = OPA.integer_var + OPB.integer_var;
OPC.integer_var = pointer;
break;
ok, structures seem to generate correct code now
2001-12-08 08:19:48 +00:00
case OP_LEAI:
pointer = OPA.integer_var + (short) st->b;
OPC.integer_var = pointer;
break;
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
case OP_STOREB_F:
case OP_STOREB_S:
case OP_STOREB_ENT:
case OP_STOREB_FLD:
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_STOREB_FN:
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
case OP_STOREB_I:
case OP_STOREB_P:
pointer = OPB.integer_var + OPC.integer_var;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_integer);
}
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
ptr = pr->pr_globals + pointer;
ptr->integer_var = OPA.integer_var;
break;
case OP_STOREB_V:
pointer = OPB.integer_var + OPC.integer_var;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_vector);
}
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
ptr = pr->pr_globals + pointer;
VectorCopy (OPA.vector_var, ptr->vector_var);
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_STOREB_Q:
pointer = OPB.integer_var + OPC.integer_var;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_quat);
}
engine side quaternion support
2004-04-08 00:56:30 +00:00
ptr = pr->pr_globals + pointer;
QuatCopy (OPA.quat_var, ptr->quat_var);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_STOREB_D:
pointer = OPB.integer_var + OPC.integer_var;
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_quat);
}
ptr = pr->pr_globals + pointer;
*(double *) ptr = OPA_double_var;
break;
pr_comp.h: - new opcodes for pointers progs.h: - prototype PR_Check_Opcodes pr_edict.c: - move the static bounds checking to PR_Check_Opcodes in pr_opcode.c pr_exec.c: - vector instruction cleanup - temporarily nuke bounds checking from storep.* - move base of address results from pr->edicts to pr->pr_globals - implement new pointer related instructions - nuke proposed global pointer instructions (not flexible enough) pr_opcode.c: - add new pointer instuctions - re-implement static bounds/invalid opcode checking. no more multiple case statements to keep track of.
2001-11-02 22:41:11 +00:00
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
case OP_STOREBI_F:
case OP_STOREBI_S:
case OP_STOREBI_ENT:
case OP_STOREBI_FLD:
change FNC to FN and fnc to fn for function based opcodes
2004-02-11 06:52:19 +00:00
case OP_STOREBI_FN:
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
case OP_STOREBI_I:
case OP_STOREBI_P:
pointer = OPB.integer_var + (short) st->c;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_integer);
}
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
ptr = pr->pr_globals + pointer;
ptr->integer_var = OPA.integer_var;
break;
case OP_STOREBI_V:
pointer = OPB.integer_var + (short) st->c;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_vector);
}
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
ptr = pr->pr_globals + pointer;
VectorCopy (OPA.vector_var, ptr->vector_var);
break;
engine side quaternion support
2004-04-08 00:56:30 +00:00
case OP_STOREBI_Q:
pointer = OPB.integer_var + (short) st->c;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_quat);
}
engine side quaternion support
2004-04-08 00:56:30 +00:00
ptr = pr->pr_globals + pointer;
QuatCopy (OPA.quat_var, ptr->quat_var);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_STOREBI_D:
pointer = OPB.integer_var + (short) st->c;
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_quat);
}
ptr = pr->pr_globals + pointer;
*(double *) ptr = OPA_double_var;
break;
immidiate indexed pointer access and some whitespace + reorg
2001-12-07 20:07:38 +00:00
Add a stack and push and pop instructions. The stack is for data, not return addresses.
2018-10-11 04:24:03 +00:00
case OP_PUSH_F:
case OP_PUSH_FLD:
case OP_PUSH_ENT:
case OP_PUSH_S:
case OP_PUSH_FN:
case OP_PUSH_I:
case OP_PUSH_P:
{
pointer_t stack = *pr->globals.stack - 1;
pr_type_t *stk = pr->pr_globals + stack;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 1);
}
stk->integer_var = OPA.integer_var;
*pr->globals.stack = stack;
}
break;
case OP_PUSH_V:
{
pointer_t stack = *pr->globals.stack - 3;
pr_type_t *stk = pr->pr_globals + stack;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 3);
}
memcpy (stk, &OPA, 3 * sizeof (OPC));
*pr->globals.stack = stack;
}
break;
case OP_PUSH_Q:
{
pointer_t stack = *pr->globals.stack - 4;
pr_type_t *stk = pr->pr_globals + stack;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 4);
}
memcpy (stk, &OPA, 4 * sizeof (OPC));
*pr->globals.stack = stack;
}
break;
case OP_PUSHB_F:
case OP_PUSHB_S:
case OP_PUSHB_ENT:
case OP_PUSHB_FLD:
case OP_PUSHB_FN:
case OP_PUSHB_I:
case OP_PUSHB_P:
{
pointer_t stack = *pr->globals.stack - 1;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + OPB.integer_var;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 1);
PR_BoundsCheck (pr, pointer, ev_integer);
}
stk->integer_var = ptr->integer_var;
*pr->globals.stack = stack;
}
break;
case OP_PUSHB_V:
{
pointer_t stack = *pr->globals.stack - 3;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + OPB.integer_var;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 3);
PR_BoundsCheck (pr, pointer, ev_integer);
}
VectorCopy (ptr->vector_var, stk->vector_var);
*pr->globals.stack = stack;
}
break;
case OP_PUSHB_Q:
{
pointer_t stack = *pr->globals.stack - 4;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + OPB.integer_var;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 4);
PR_BoundsCheck (pr, pointer, ev_quat);
}
QuatCopy (ptr->quat_var, stk->quat_var);
*pr->globals.stack = stack;
}
break;
case OP_PUSHBI_F:
case OP_PUSHBI_S:
case OP_PUSHBI_ENT:
case OP_PUSHBI_FLD:
case OP_PUSHBI_FN:
case OP_PUSHBI_I:
case OP_PUSHBI_P:
{
pointer_t stack = *pr->globals.stack - 1;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + st->b;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 1);
PR_BoundsCheck (pr, pointer, ev_integer);
}
stk->integer_var = ptr->integer_var;
*pr->globals.stack = stack;
}
break;
case OP_PUSHBI_V:
{
pointer_t stack = *pr->globals.stack - 3;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + st->b;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 3);
PR_BoundsCheck (pr, pointer, ev_integer);
}
VectorCopy (ptr->vector_var, stk->vector_var);
*pr->globals.stack = stack;
}
break;
case OP_PUSHBI_Q:
{
pointer_t stack = *pr->globals.stack - 4;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + st->b;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 4);
PR_BoundsCheck (pr, pointer, ev_quat);
}
QuatCopy (ptr->quat_var, stk->quat_var);
*pr->globals.stack = stack;
}
break;
case OP_POP_F:
case OP_POP_FLD:
case OP_POP_ENT:
case OP_POP_S:
case OP_POP_FN:
case OP_POP_I:
case OP_POP_P:
{
pointer_t stack = *pr->globals.stack;
pr_type_t *stk = pr->pr_globals + stack;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 1);
}
stk->integer_var = OPA.integer_var;
*pr->globals.stack = stack + 1;
}
break;
case OP_POP_V:
{
pointer_t stack = *pr->globals.stack;
pr_type_t *stk = pr->pr_globals + stack;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 3);
}
memcpy (stk, &OPA, 3 * sizeof (OPC));
*pr->globals.stack = stack + 3;
}
break;
case OP_POP_Q:
{
pointer_t stack = *pr->globals.stack;
pr_type_t *stk = pr->pr_globals + stack;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 4);
}
memcpy (stk, &OPA, 4 * sizeof (OPC));
*pr->globals.stack = stack + 4;
}
break;
case OP_POPB_F:
case OP_POPB_S:
case OP_POPB_ENT:
case OP_POPB_FLD:
case OP_POPB_FN:
case OP_POPB_I:
case OP_POPB_P:
{
pointer_t stack = *pr->globals.stack;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + OPB.integer_var;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 1);
PR_BoundsCheck (pr, pointer, ev_integer);
}
stk->integer_var = ptr->integer_var;
*pr->globals.stack = stack + 1;
}
break;
case OP_POPB_V:
{
pointer_t stack = *pr->globals.stack;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + OPB.integer_var;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 3);
PR_BoundsCheck (pr, pointer, ev_integer);
}
VectorCopy (ptr->vector_var, stk->vector_var);
*pr->globals.stack = stack + 3;
}
break;
case OP_POPB_Q:
{
pointer_t stack = *pr->globals.stack;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + OPB.integer_var;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 4);
PR_BoundsCheck (pr, pointer, ev_quat);
}
QuatCopy (ptr->quat_var, stk->quat_var);
*pr->globals.stack = stack + 4;
}
break;
case OP_POPBI_F:
case OP_POPBI_S:
case OP_POPBI_ENT:
case OP_POPBI_FLD:
case OP_POPBI_FN:
case OP_POPBI_I:
case OP_POPBI_P:
{
pointer_t stack = *pr->globals.stack;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + st->b;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 1);
PR_BoundsCheck (pr, pointer, ev_integer);
}
stk->integer_var = ptr->integer_var;
*pr->globals.stack = stack + 1;
}
break;
case OP_POPBI_V:
{
pointer_t stack = *pr->globals.stack;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + st->b;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 3);
PR_BoundsCheck (pr, pointer, ev_integer);
}
VectorCopy (ptr->vector_var, stk->vector_var);
*pr->globals.stack = stack + 3;
}
break;
case OP_POPBI_Q:
{
pointer_t stack = *pr->globals.stack;
pr_type_t *stk = pr->pr_globals + stack;
pointer = OPA.integer_var + st->b;
ptr = pr->pr_globals + pointer;
if (pr_boundscheck->int_val) {
check_stack_pointer (pr, stack, 4);
PR_BoundsCheck (pr, pointer, ev_quat);
}
QuatCopy (ptr->quat_var, stk->quat_var);
*pr->globals.stack = stack + 4;
}
break;
new operand handling from nq
2001-02-27 08:21:40 +00:00
// ==================
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_IFNOT:
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
if (!OPA.integer_var) {
bah, forgot to -1 the instruction pointer for jumpb
2003-08-24 07:23:12 +00:00
pr->pr_xstatement += (short)st->b - 1; // offset the st++
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
st = pr->pr_statements + pr->pr_xstatement;
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_IF:
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
if (OPA.integer_var) {
bah, forgot to -1 the instruction pointer for jumpb
2003-08-24 07:23:12 +00:00
pr->pr_xstatement += (short)st->b - 1; // offset the st++
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
st = pr->pr_statements + pr->pr_xstatement;
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
case OP_IFBE:
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
if (OPA.integer_var <= 0) {
bah, forgot to -1 the instruction pointer for jumpb
2003-08-24 07:23:12 +00:00
pr->pr_xstatement += (short)st->b - 1; // offset the st++
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
st = pr->pr_statements + pr->pr_xstatement;
}
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
break;
case OP_IFB:
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
if (OPA.integer_var < 0) {
bah, forgot to -1 the instruction pointer for jumpb
2003-08-24 07:23:12 +00:00
pr->pr_xstatement += (short)st->b - 1; // offset the st++
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
st = pr->pr_statements + pr->pr_xstatement;
}
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
break;
case OP_IFAE:
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
if (OPA.integer_var >= 0) {
bah, forgot to -1 the instruction pointer for jumpb
2003-08-24 07:23:12 +00:00
pr->pr_xstatement += (short)st->b - 1; // offset the st++
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
st = pr->pr_statements + pr->pr_xstatement;
}
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
break;
case OP_IFA:
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
if (OPA.integer_var > 0) {
bah, forgot to -1 the instruction pointer for jumpb
2003-08-24 07:23:12 +00:00
pr->pr_xstatement += (short)st->b - 1; // offset the st++
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
st = pr->pr_statements + pr->pr_xstatement;
}
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_GOTO:
bah, forgot to -1 the instruction pointer for jumpb
2003-08-24 07:23:12 +00:00
pr->pr_xstatement += (short)st->a - 1; // offset the st++
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
st = pr->pr_statements + pr->pr_xstatement;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
case OP_JUMP:
if (pr_boundscheck->int_val
&& (OPA.uinteger_var >= pr->progs->numstatements)) {
search for `.self' before `self'
2002-05-18 00:49:16 +00:00
PR_RunError (pr, "Invalid jump destination");
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
}
Fix the jump instruction. The offset to compensate for st++ was missing. Obviously, the code has never been tested. Found while looking at the jump code and thinking about using 32-bit addresses for the jump tables.
2013-06-26 00:59:19 +00:00
pr->pr_xstatement = OPA.uinteger_var - 1; // offset the st++
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
st = pr->pr_statements + pr->pr_xstatement;
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
break;
pr_comp.h: o add OP_JUMPB o OP_JUMPB renumberd some opcodes, so up PROG_VERSION pr_edict.c: o make the version error reporting more informative pr_exec.c: o implement OP_JUMPB (goto *(ptr + index)) pr_opcode.c: (libs/gamecode/engine) o add OP_JUMPB to the table expr.h: o ex_uinteger support o some const correctness o prototype new_label_expr qfcc.h: o uinteger ussport o add pointers for op_jump and op_jumpb o prototype PR_GetArray emit.c: o general uinteger support o new reference/reloc type 3: absolute statement address o jumpb support (binary goto) expr.c: o uinteger support o break the label name creation out of new_label_expr into new_label_name o some const correctness pr_def.c: o add PR_GetArray to allocate an array in global space o factor out some code common to PR_GetDef and PR_GetArray that would otherwise be duplicated pr_imm.c: o some const correctness o uinteger support pr_lex.c: o uinteger support pr_opcode.c: (tools/qfcc/source) o support jump and jumpb switch.c: o rewrite the binary search code to support ranges.
2001-11-13 08:58:54 +00:00
case OP_JUMPB:
implement jumpb correctly
2003-08-22 19:48:36 +00:00
pointer = st->a + OPB.integer_var;
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheck (pr, pointer, ev_integer);
}
pr_comp.h: o add OP_JUMPB o OP_JUMPB renumberd some opcodes, so up PROG_VERSION pr_edict.c: o make the version error reporting more informative pr_exec.c: o implement OP_JUMPB (goto *(ptr + index)) pr_opcode.c: (libs/gamecode/engine) o add OP_JUMPB to the table expr.h: o ex_uinteger support o some const correctness o prototype new_label_expr qfcc.h: o uinteger ussport o add pointers for op_jump and op_jumpb o prototype PR_GetArray emit.c: o general uinteger support o new reference/reloc type 3: absolute statement address o jumpb support (binary goto) expr.c: o uinteger support o break the label name creation out of new_label_expr into new_label_name o some const correctness pr_def.c: o add PR_GetArray to allocate an array in global space o factor out some code common to PR_GetDef and PR_GetArray that would otherwise be duplicated pr_imm.c: o some const correctness o uinteger support pr_lex.c: o uinteger support pr_opcode.c: (tools/qfcc/source) o support jump and jumpb switch.c: o rewrite the binary search code to support ranges.
2001-11-13 08:58:54 +00:00
ptr = pr->pr_globals + pointer;
pointer = ptr->integer_var;
if (pr_boundscheck->int_val
&& (pointer >= pr->progs->numstatements)) {
search for `.self' before `self'
2002-05-18 00:49:16 +00:00
PR_RunError (pr, "Invalid jump destination");
pr_comp.h: o add OP_JUMPB o OP_JUMPB renumberd some opcodes, so up PROG_VERSION pr_edict.c: o make the version error reporting more informative pr_exec.c: o implement OP_JUMPB (goto *(ptr + index)) pr_opcode.c: (libs/gamecode/engine) o add OP_JUMPB to the table expr.h: o ex_uinteger support o some const correctness o prototype new_label_expr qfcc.h: o uinteger ussport o add pointers for op_jump and op_jumpb o prototype PR_GetArray emit.c: o general uinteger support o new reference/reloc type 3: absolute statement address o jumpb support (binary goto) expr.c: o uinteger support o break the label name creation out of new_label_expr into new_label_name o some const correctness pr_def.c: o add PR_GetArray to allocate an array in global space o factor out some code common to PR_GetDef and PR_GetArray that would otherwise be duplicated pr_imm.c: o some const correctness o uinteger support pr_lex.c: o uinteger support pr_opcode.c: (tools/qfcc/source) o support jump and jumpb switch.c: o rewrite the binary search code to support ranges.
2001-11-13 08:58:54 +00:00
}
bah, forgot to -1 the instruction pointer for jumpb
2003-08-24 07:23:12 +00:00
pr->pr_xstatement = pointer - 1; // offset the st++
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
st = pr->pr_statements + pr->pr_xstatement;
pr_comp.h: o add OP_JUMPB o OP_JUMPB renumberd some opcodes, so up PROG_VERSION pr_edict.c: o make the version error reporting more informative pr_exec.c: o implement OP_JUMPB (goto *(ptr + index)) pr_opcode.c: (libs/gamecode/engine) o add OP_JUMPB to the table expr.h: o ex_uinteger support o some const correctness o prototype new_label_expr qfcc.h: o uinteger ussport o add pointers for op_jump and op_jumpb o prototype PR_GetArray emit.c: o general uinteger support o new reference/reloc type 3: absolute statement address o jumpb support (binary goto) expr.c: o uinteger support o break the label name creation out of new_label_expr into new_label_name o some const correctness pr_def.c: o add PR_GetArray to allocate an array in global space o factor out some code common to PR_GetDef and PR_GetArray that would otherwise be duplicated pr_imm.c: o some const correctness o uinteger support pr_lex.c: o uinteger support pr_opcode.c: (tools/qfcc/source) o support jump and jumpb switch.c: o rewrite the binary search code to support ranges.
2001-11-13 08:58:54 +00:00
break;
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
implement Raven's calling convention (rcall: first 2 params in opb and opc). should give a general speedup to most progs.
2005-06-12 09:54:01 +00:00
case OP_RCALL2:
case OP_RCALL3:
case OP_RCALL4:
case OP_RCALL5:
case OP_RCALL6:
case OP_RCALL7:
case OP_RCALL8:
pr->pr_params[1] = &OPC;
goto op_rcall;
case OP_RCALL1:
pr->pr_params[1] = pr->pr_real_params[1];
op_rcall:
pr->pr_params[0] = &OPB;
pr->pr_argc = st->op - OP_RCALL1 + 1;
goto op_call;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_CALL0:
case OP_CALL1:
case OP_CALL2:
case OP_CALL3:
case OP_CALL4:
case OP_CALL5:
case OP_CALL6:
case OP_CALL7:
case OP_CALL8:
implement Raven's calling convention (rcall: first 2 params in opb and opc). should give a general speedup to most progs.
2005-06-12 09:54:01 +00:00
PR_RESET_PARAMS (pr);
pr->pr_argc = st->op - OP_CALL0;
op_call:
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
pr->pr_xfunction->profile += profile - startprofile;
startprofile = profile;
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
PR_CallFunction (pr, OPA.func_var);
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
st = pr->pr_statements + pr->pr_xstatement;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_DONE:
case OP_RETURN:
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
if (!st->a)
memset (&R_INT (pr), 0,
pr->pr_param_size * sizeof (OPA));
else if (&R_INT (pr) != &OPA.integer_var)
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
memcpy (&R_INT (pr), &OPA,
pr->pr_param_size * sizeof (OPA));
Add a "void return" instruction. I got fed up with always having to explicty return something.
2011-01-12 15:29:56 +00:00
// fallthrough
case OP_RETURN_V:
fix an accounting error
2003-09-15 21:13:13 +00:00
pr->pr_xfunction->profile += profile - startprofile;
startprofile = profile;
[gamecode] Be more informative with function calls PR_EnterFunction and PR_LeaveFunction now print the entered/left/returned function when tracing. Makes debugging obj code much easier.
2020-03-24 04:24:55 +00:00
PR_LeaveFunction (pr, pr->pr_depth == exitdepth);
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
st = pr->pr_statements + pr->pr_xstatement;
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
if (pr->pr_depth == exitdepth) {
make a common call function and fix pr_trace such that it autoclears only when the execution session it's set in is exited This is an imperfect revision of history.
2004-11-07 03:00:00 +00:00
if (pr->pr_trace && pr->pr_depth <= pr->pr_trace_depth)
pr->pr_trace = false;
Allow Sys_Error to be hooked. This makes debugging builtins that wrap normal functions a little easier by giving a progs dump when such an error occurs.
2016-01-03 14:04:00 +00:00
// all done
goto exit_program;
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
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_STATE:
really clean up global and function access. Unfortunatly, frames are still borked, at least for the view entity.
2001-03-01 08:34:30 +00:00
ed = PROG_TO_EDICT (pr, *pr->globals.self);
Whitespace.
2001-09-10 12:56:23 +00:00
ed->v[pr->fields.nextthink].float_var = *pr->globals.time +
0.1;
Massive clean up of progs accessors. Just the cleanup of the casts fixed the CustomTF doors issue.
2001-06-07 21:59:24 +00:00
ed->v[pr->fields.frame].float_var = OPA.float_var;
ed->v[pr->fields.think].func_var = OPB.func_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
new opcode: state.f. same as state, but takes a 3rd float operand to specify the step for calculating nextthink. accessed using [frame, think, step] (state is [frame, think])
2004-02-11 01:43:33 +00:00
case OP_STATE_F:
ed = PROG_TO_EDICT (pr, *pr->globals.self);
ed->v[pr->fields.nextthink].float_var = *pr->globals.time +
OPC.float_var;
ed->v[pr->fields.frame].float_var = OPA.float_var;
ed->v[pr->fields.think].func_var = OPB.func_var;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_ADD_I:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.integer_var + OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_SUB_I:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.integer_var - OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_MUL_I:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.integer_var * OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_DIV_I:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.integer_var / OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
Implement true modulo in the engine
2020-02-16 02:53:56 +00:00
case OP_MOD_I:
{
// implement true modulo for integers:
// 5 mod 3 = 2
// -5 mod 3 = 1
// 5 mod -3 = -1
// -5 mod -3 = -2
int a = OPA.integer_var;
int b = OPB.integer_var;
int c = a % b;
// % is really remainder and so has the same sign rules
// as division: -5 % 3 = -2, so need to add b (3 here)
// if c's sign is incorrect, but only if c is non-zero
int mask = (a ^ b) >> 31;
Make gcc accept the bitwise-boolean magic Forgot to compile check...
2020-02-16 03:08:08 +00:00
mask &= ~(!!c + 0) + 1; // +0 to convert bool to int (gcc)
Implement true modulo in the engine
2020-02-16 02:53:56 +00:00
OPC.integer_var = c + (mask & b);
}
break;
Rename the mod instruction to rem Because % really implements remainder rather than true modulo, and I plan on adding %% to implement true modulo.
2020-02-16 02:04:30 +00:00
case OP_REM_I:
got a few new operators for qc :) << >> % which all act like their C counterparts += -= *= /= &= ^= |= <<= >>= %= which just expand into the obvious, also like the C versions.
2001-08-10 16:17:00 +00:00
OPC.integer_var = OPA.integer_var % OPB.integer_var;
break;
Implement true modulo in the engine
2020-02-16 02:53:56 +00:00
case OP_MOD_D:
{
double a = OPA_double_var;
double b = OPB_double_var;
// floating point modulo is so much easier :P
OPC_double_var = a - b * floor (a / b);
}
break;
Rename the mod instruction to rem Because % really implements remainder rather than true modulo, and I plan on adding %% to implement true modulo.
2020-02-16 02:04:30 +00:00
case OP_REM_D:
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
{
double a = OPA_double_var;
double b = OPB_double_var;
OPC_double_var = a - b * trunc (a / b);
}
break;
Implement true modulo in the engine
2020-02-16 02:53:56 +00:00
case OP_MOD_F:
{
float a = OPA.float_var;
float b = OPB.float_var;
OPC.float_var = a - b * floorf (a / b);
}
break;
Rename the mod instruction to rem Because % really implements remainder rather than true modulo, and I plan on adding %% to implement true modulo.
2020-02-16 02:04:30 +00:00
case OP_REM_F:
Don't truncat float % float This allows full usage, eg, x % pi, but otherwise maintains compatibility with integer %
2020-02-14 04:54:26 +00:00
{
float a = OPA.float_var;
float b = OPB.float_var;
OPC.float_var = a - b * truncf (a / b);
}
got a few new operators for qc :) << >> % which all act like their C counterparts += -= *= /= &= ^= |= <<= >>= %= which just expand into the obvious, also like the C versions.
2001-08-10 16:17:00 +00:00
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_CONV_IF:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.float_var = OPA.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_CONV_FI:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.float_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_BITAND_I:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.integer_var & OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_BITOR_I:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
OPC.integer_var = OPA.integer_var | OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
^ and ~ operators for qc (^ from Rhamphoryncus) and clean up the punctuation line in qc-lex.l (why didn't I think to do that in the first place?!?)
2001-08-09 16:34:46 +00:00
case OP_BITXOR_I:
OPC.integer_var = OPA.integer_var ^ OPB.integer_var;
break;
case OP_BITNOT_I:
OPC.integer_var = ~OPA.integer_var;
break;
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_GE_I:
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_GE_P:
Initial integer type support. qfcc /is/ partially broken when it comes to integer constants and float function args/return values. pr_comp.h: o add the integer opcodes to pr_opcode_e pr_edict.c: o add "quaternion" and "integer" to type_name[] o support quatnernion and integers types when printing values o support the integer opcodes when bounds checking pr_exec.c o enable the integer opcodes pr_opcode: o add the integer opcodes to the opcode table o logical operators all result in an integer rather than a value expr.h: o rename int_val to integer_val qfcc.h: o kill another magic number expr.c: o move the opcode to string conversion out of type_mismatch and into get_op_string o rename int_val to integer_val o general integer type support. o generate an internal comipiler error for null opcodes rather than segging. pr_imm.c: o rename int_val to integer_val o support integer constants, converting to float when needed. pr_lex.c: o magic number death and support quaternions and integers in type_size[] qc-lex.l o rename int_val to integer_val o support quaternion and integer type keywords qc-parse.y: o rename int_val to integer_val o use binary_expr instead of new_binary_expr for local initialized variables builtins.c: o rename int_val to integer_val o fix most (all?) of the INT related FIXMEs defs.qc: o use integer instead of float where it makes sense main.c: o read_result is now integer rather than float main.qc: o float -> integer where appropriate o new test for int const to float arg
2001-07-23 01:31:22 +00:00
OPC.integer_var = OPA.integer_var >= OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
fix the unsigned comparison opcodes
2003-08-24 05:53:15 +00:00
case OP_GE_U:
OPC.integer_var = OPA.uinteger_var >= OPB.uinteger_var;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_LE_I:
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_LE_P:
Initial integer type support. qfcc /is/ partially broken when it comes to integer constants and float function args/return values. pr_comp.h: o add the integer opcodes to pr_opcode_e pr_edict.c: o add "quaternion" and "integer" to type_name[] o support quatnernion and integers types when printing values o support the integer opcodes when bounds checking pr_exec.c o enable the integer opcodes pr_opcode: o add the integer opcodes to the opcode table o logical operators all result in an integer rather than a value expr.h: o rename int_val to integer_val qfcc.h: o kill another magic number expr.c: o move the opcode to string conversion out of type_mismatch and into get_op_string o rename int_val to integer_val o general integer type support. o generate an internal comipiler error for null opcodes rather than segging. pr_imm.c: o rename int_val to integer_val o support integer constants, converting to float when needed. pr_lex.c: o magic number death and support quaternions and integers in type_size[] qc-lex.l o rename int_val to integer_val o support quaternion and integer type keywords qc-parse.y: o rename int_val to integer_val o use binary_expr instead of new_binary_expr for local initialized variables builtins.c: o rename int_val to integer_val o fix most (all?) of the INT related FIXMEs defs.qc: o use integer instead of float where it makes sense main.c: o read_result is now integer rather than float main.qc: o float -> integer where appropriate o new test for int const to float arg
2001-07-23 01:31:22 +00:00
OPC.integer_var = OPA.integer_var <= OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
fix the unsigned comparison opcodes
2003-08-24 05:53:15 +00:00
case OP_LE_U:
OPC.integer_var = OPA.uinteger_var <= OPB.uinteger_var;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_GT_I:
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_GT_P:
Initial integer type support. qfcc /is/ partially broken when it comes to integer constants and float function args/return values. pr_comp.h: o add the integer opcodes to pr_opcode_e pr_edict.c: o add "quaternion" and "integer" to type_name[] o support quatnernion and integers types when printing values o support the integer opcodes when bounds checking pr_exec.c o enable the integer opcodes pr_opcode: o add the integer opcodes to the opcode table o logical operators all result in an integer rather than a value expr.h: o rename int_val to integer_val qfcc.h: o kill another magic number expr.c: o move the opcode to string conversion out of type_mismatch and into get_op_string o rename int_val to integer_val o general integer type support. o generate an internal comipiler error for null opcodes rather than segging. pr_imm.c: o rename int_val to integer_val o support integer constants, converting to float when needed. pr_lex.c: o magic number death and support quaternions and integers in type_size[] qc-lex.l o rename int_val to integer_val o support quaternion and integer type keywords qc-parse.y: o rename int_val to integer_val o use binary_expr instead of new_binary_expr for local initialized variables builtins.c: o rename int_val to integer_val o fix most (all?) of the INT related FIXMEs defs.qc: o use integer instead of float where it makes sense main.c: o read_result is now integer rather than float main.qc: o float -> integer where appropriate o new test for int const to float arg
2001-07-23 01:31:22 +00:00
OPC.integer_var = OPA.integer_var > OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
fix the unsigned comparison opcodes
2003-08-24 05:53:15 +00:00
case OP_GT_U:
OPC.integer_var = OPA.uinteger_var > OPB.uinteger_var;
break;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_LT_I:
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_LT_P:
oops, signage error
2005-05-01 08:30:30 +00:00
OPC.integer_var = OPA.integer_var < OPB.integer_var;
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
break;
fix the unsigned comparison opcodes
2003-08-24 05:53:15 +00:00
case OP_LT_U:
OPC.integer_var = OPA.uinteger_var < OPB.uinteger_var;
break;
pr_comp.h: o add ev_uniteger to the types enum o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui progs.h: o add uinteger accessors pr_exec.c: o implement ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui pr_opcode.c: o add opcodes for ifbe, ifb, ifae, ifa, jump, lt.ui, gt.ui, le.ui, ge.ui expr.h: o prototype inc_users qfcc.h: o add externs for op_ifbe, op_ifb, op_ifae and op_ifa emit.c: o don't bother emiting an assignment to a temp def that's only used once (ie, it's never read, only written to) o support the new if* instructions expr.c: o support the new if* insructions o dectect expression loops in append_expr o support unsigned integers o re-work temp def usage counting pr_def.c o debugging for temp def usage counts pr_opcode.c: o support the new if* instructions qc-parse.y: o provide defines for IFBE IFB IFAE IFA switch.c: o do binary searches for strings, floats and ints if there are more than 8 cases in a switch. Strings need more testing.
2001-11-09 00:58:16 +00:00
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_AND_I:
Initial integer type support. qfcc /is/ partially broken when it comes to integer constants and float function args/return values. pr_comp.h: o add the integer opcodes to pr_opcode_e pr_edict.c: o add "quaternion" and "integer" to type_name[] o support quatnernion and integers types when printing values o support the integer opcodes when bounds checking pr_exec.c o enable the integer opcodes pr_opcode: o add the integer opcodes to the opcode table o logical operators all result in an integer rather than a value expr.h: o rename int_val to integer_val qfcc.h: o kill another magic number expr.c: o move the opcode to string conversion out of type_mismatch and into get_op_string o rename int_val to integer_val o general integer type support. o generate an internal comipiler error for null opcodes rather than segging. pr_imm.c: o rename int_val to integer_val o support integer constants, converting to float when needed. pr_lex.c: o magic number death and support quaternions and integers in type_size[] qc-lex.l o rename int_val to integer_val o support quaternion and integer type keywords qc-parse.y: o rename int_val to integer_val o use binary_expr instead of new_binary_expr for local initialized variables builtins.c: o rename int_val to integer_val o fix most (all?) of the INT related FIXMEs defs.qc: o use integer instead of float where it makes sense main.c: o read_result is now integer rather than float main.qc: o float -> integer where appropriate o new test for int const to float arg
2001-07-23 01:31:22 +00:00
OPC.integer_var = OPA.integer_var && OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_OR_I:
Initial integer type support. qfcc /is/ partially broken when it comes to integer constants and float function args/return values. pr_comp.h: o add the integer opcodes to pr_opcode_e pr_edict.c: o add "quaternion" and "integer" to type_name[] o support quatnernion and integers types when printing values o support the integer opcodes when bounds checking pr_exec.c o enable the integer opcodes pr_opcode: o add the integer opcodes to the opcode table o logical operators all result in an integer rather than a value expr.h: o rename int_val to integer_val qfcc.h: o kill another magic number expr.c: o move the opcode to string conversion out of type_mismatch and into get_op_string o rename int_val to integer_val o general integer type support. o generate an internal comipiler error for null opcodes rather than segging. pr_imm.c: o rename int_val to integer_val o support integer constants, converting to float when needed. pr_lex.c: o magic number death and support quaternions and integers in type_size[] qc-lex.l o rename int_val to integer_val o support quaternion and integer type keywords qc-parse.y: o rename int_val to integer_val o use binary_expr instead of new_binary_expr for local initialized variables builtins.c: o rename int_val to integer_val o fix most (all?) of the INT related FIXMEs defs.qc: o use integer instead of float where it makes sense main.c: o read_result is now integer rather than float main.qc: o float -> integer where appropriate o new test for int const to float arg
2001-07-23 01:31:22 +00:00
OPC.integer_var = OPA.integer_var || OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_NOT_I:
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_NOT_P:
Initial integer type support. qfcc /is/ partially broken when it comes to integer constants and float function args/return values. pr_comp.h: o add the integer opcodes to pr_opcode_e pr_edict.c: o add "quaternion" and "integer" to type_name[] o support quatnernion and integers types when printing values o support the integer opcodes when bounds checking pr_exec.c o enable the integer opcodes pr_opcode: o add the integer opcodes to the opcode table o logical operators all result in an integer rather than a value expr.h: o rename int_val to integer_val qfcc.h: o kill another magic number expr.c: o move the opcode to string conversion out of type_mismatch and into get_op_string o rename int_val to integer_val o general integer type support. o generate an internal comipiler error for null opcodes rather than segging. pr_imm.c: o rename int_val to integer_val o support integer constants, converting to float when needed. pr_lex.c: o magic number death and support quaternions and integers in type_size[] qc-lex.l o rename int_val to integer_val o support quaternion and integer type keywords qc-parse.y: o rename int_val to integer_val o use binary_expr instead of new_binary_expr for local initialized variables builtins.c: o rename int_val to integer_val o fix most (all?) of the INT related FIXMEs defs.qc: o use integer instead of float where it makes sense main.c: o read_result is now integer rather than float main.qc: o float -> integer where appropriate o new test for int const to float arg
2001-07-23 01:31:22 +00:00
OPC.integer_var = !OPA.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_EQ_I:
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_EQ_P:
Initial integer type support. qfcc /is/ partially broken when it comes to integer constants and float function args/return values. pr_comp.h: o add the integer opcodes to pr_opcode_e pr_edict.c: o add "quaternion" and "integer" to type_name[] o support quatnernion and integers types when printing values o support the integer opcodes when bounds checking pr_exec.c o enable the integer opcodes pr_opcode: o add the integer opcodes to the opcode table o logical operators all result in an integer rather than a value expr.h: o rename int_val to integer_val qfcc.h: o kill another magic number expr.c: o move the opcode to string conversion out of type_mismatch and into get_op_string o rename int_val to integer_val o general integer type support. o generate an internal comipiler error for null opcodes rather than segging. pr_imm.c: o rename int_val to integer_val o support integer constants, converting to float when needed. pr_lex.c: o magic number death and support quaternions and integers in type_size[] qc-lex.l o rename int_val to integer_val o support quaternion and integer type keywords qc-parse.y: o rename int_val to integer_val o use binary_expr instead of new_binary_expr for local initialized variables builtins.c: o rename int_val to integer_val o fix most (all?) of the INT related FIXMEs defs.qc: o use integer instead of float where it makes sense main.c: o read_result is now integer rather than float main.qc: o float -> integer where appropriate o new test for int const to float arg
2001-07-23 01:31:22 +00:00
OPC.integer_var = OPA.integer_var == OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case OP_NE_I:
implement comparisons and not for pointers
2002-01-30 21:20:12 +00:00
case OP_NE_P:
Initial integer type support. qfcc /is/ partially broken when it comes to integer constants and float function args/return values. pr_comp.h: o add the integer opcodes to pr_opcode_e pr_edict.c: o add "quaternion" and "integer" to type_name[] o support quatnernion and integers types when printing values o support the integer opcodes when bounds checking pr_exec.c o enable the integer opcodes pr_opcode: o add the integer opcodes to the opcode table o logical operators all result in an integer rather than a value expr.h: o rename int_val to integer_val qfcc.h: o kill another magic number expr.c: o move the opcode to string conversion out of type_mismatch and into get_op_string o rename int_val to integer_val o general integer type support. o generate an internal comipiler error for null opcodes rather than segging. pr_imm.c: o rename int_val to integer_val o support integer constants, converting to float when needed. pr_lex.c: o magic number death and support quaternions and integers in type_size[] qc-lex.l o rename int_val to integer_val o support quaternion and integer type keywords qc-parse.y: o rename int_val to integer_val o use binary_expr instead of new_binary_expr for local initialized variables builtins.c: o rename int_val to integer_val o fix most (all?) of the INT related FIXMEs defs.qc: o use integer instead of float where it makes sense main.c: o read_result is now integer rather than float main.qc: o float -> integer where appropriate o new test for int const to float arg
2001-07-23 01:31:22 +00:00
OPC.integer_var = OPA.integer_var != OPB.integer_var;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
Rename move to movei.
2011-03-09 01:28:40 +00:00
case OP_MOVEI:
move the correct amount of bytes (I keep forgetting the * 4 :P)
2002-10-21 16:38:45 +00:00
memmove (&OPC, &OPA, st->b * 4);
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
break;
case OP_MOVEP:
implement bounds checking on all opcodes
2010-01-13 06:19:50 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheckSize (pr, OPC.integer_var, OPB.uinteger_var);
PR_BoundsCheckSize (pr, OPA.integer_var, OPB.uinteger_var);
}
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
memmove (pr->pr_globals + OPC.integer_var,
move the correct amount of bytes (I keep forgetting the * 4 :P)
2002-10-21 16:38:45 +00:00
pr->pr_globals + OPA.integer_var,
OPB.uinteger_var * 4);
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
break;
Add a movepi instruction to support indirect moves of fixed size.
2011-03-09 01:29:24 +00:00
case OP_MOVEPI:
if (pr_boundscheck->int_val) {
PR_BoundsCheckSize (pr, OPC.integer_var, st->b);
PR_BoundsCheckSize (pr, OPA.integer_var, st->b);
}
memmove (pr->pr_globals + OPC.integer_var,
pr->pr_globals + OPA.integer_var,
st->b * 4);
break;
[gamecode] Rework implementation of memset* The memset instructions now match the move* instructions other than the first operand (always int). Probably breaks much, but fixed in next few commits.
2020-03-13 08:50:57 +00:00
case OP_MEMSETI:
pr_memset (&OPC, OPA.integer_var, st->b);
break;
case OP_MEMSETP:
[gamecode] Add memset instructions
2020-03-11 13:48:55 +00:00
if (pr_boundscheck->int_val) {
[gamecode] Rework implementation of memset* The memset instructions now match the move* instructions other than the first operand (always int). Probably breaks much, but fixed in next few commits.
2020-03-13 08:50:57 +00:00
PR_BoundsCheckSize (pr, OPC.pointer_var, OPB.integer_var);
[gamecode] Add memset instructions
2020-03-11 13:48:55 +00:00
}
pr_memset (pr->pr_globals + OPC.pointer_var, OPA.integer_var,
OPB.integer_var);
break;
[gamecode] Rework implementation of memset* The memset instructions now match the move* instructions other than the first operand (always int). Probably breaks much, but fixed in next few commits.
2020-03-13 08:50:57 +00:00
case OP_MEMSETPI:
[gamecode] Add memset instructions
2020-03-11 13:48:55 +00:00
if (pr_boundscheck->int_val) {
PR_BoundsCheckSize (pr, OPC.pointer_var, st->b);
}
pr_memset (pr->pr_globals + OPC.pointer_var, OPA.integer_var,
st->b);
break;
Add support for doubles to Ruamoko Only as scalars, I still need to think about what to do for vectors and quaternions due to param size issues. Also, doubles are not yet guaranteed to be correctly aligned.
2020-02-14 07:38:37 +00:00
case OP_GE_D:
OPC.float_var = OPA_double_var >= OPB_double_var;
break;
case OP_LE_D:
OPC.float_var = OPA_double_var <= OPB_double_var;
break;
case OP_GT_D:
OPC.float_var = OPA_double_var > OPB_double_var;
break;
case OP_LT_D:
OPC.float_var = OPA_double_var < OPB_double_var;
break;
case OP_NOT_D:
OPC.integer_var = (op_a[0].integer_var
|| (op_a[1].integer_var & ~0x80000000u));
break;
case OP_EQ_D:
OPC.integer_var = OPA_double_var == OPB_double_var;
break;
case OP_NE_D:
OPC.integer_var = OPA_double_var != OPB_double_var;
break;
case OP_CONV_ID:
OPC_double_var = OPA.integer_var;
break;
case OP_CONV_DI:
OPC.integer_var = OPA_double_var;
break;
case OP_CONV_FD:
OPC_double_var = OPA.float_var;
break;
case OP_CONV_DF:
OPC.float_var = OPA_double_var;
break;
Initial integer type support. qfcc /is/ partially broken when it comes to integer constants and float function args/return values. pr_comp.h: o add the integer opcodes to pr_opcode_e pr_edict.c: o add "quaternion" and "integer" to type_name[] o support quatnernion and integers types when printing values o support the integer opcodes when bounds checking pr_exec.c o enable the integer opcodes pr_opcode: o add the integer opcodes to the opcode table o logical operators all result in an integer rather than a value expr.h: o rename int_val to integer_val qfcc.h: o kill another magic number expr.c: o move the opcode to string conversion out of type_mismatch and into get_op_string o rename int_val to integer_val o general integer type support. o generate an internal comipiler error for null opcodes rather than segging. pr_imm.c: o rename int_val to integer_val o support integer constants, converting to float when needed. pr_lex.c: o magic number death and support quaternions and integers in type_size[] qc-lex.l o rename int_val to integer_val o support quaternion and integer type keywords qc-parse.y: o rename int_val to integer_val o use binary_expr instead of new_binary_expr for local initialized variables builtins.c: o rename int_val to integer_val o fix most (all?) of the INT related FIXMEs defs.qc: o use integer instead of float where it makes sense main.c: o read_result is now integer rather than float main.qc: o float -> integer where appropriate o new test for int const to float arg
2001-07-23 01:31:22 +00:00
// LordHavoc: to be enabled when Progs version 7 (or whatever it will be numbered) is finalized
/*
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
case OP_BOUNDCHECK:
rename int_var to integer_var
2001-07-22 20:20:46 +00:00
if (OPA.integer_var < 0 || OPA.integer_var >= st->b) {
Whitespace.
2001-09-10 12:56:23 +00:00
PR_RunError (pr, "Progs boundcheck failed at line number "
search for `.self' before `self'
2002-05-18 00:49:16 +00:00
"%d, value is < 0 or >= %d", st->b, st->c);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
break;
*/
default:
fix up some overzelous s///
2001-06-07 22:15:37 +00:00
PR_RunError (pr, "Bad opcode %i", st->op);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
[gamecode] Allow watch-point to be updated With the watch-point address cached in a local var, it was not possible to update the watch-point while the progs were executing.
2020-03-24 10:45:39 +00:00
if (pr->watch && pr->watch->integer_var != old_val.integer_var) {
[gamecode] Improve watchpoint handling a little Mostly just update the old value if the condition didn't trigger.
2020-03-24 04:26:35 +00:00
if (!pr->wp_conditional
[gamecode] Allow watch-point to be updated With the watch-point address cached in a local var, it was not possible to update the watch-point while the progs were executing.
2020-03-24 10:45:39 +00:00
|| pr->watch->integer_var == pr->wp_val.integer_var) {
[gamecode] Add debug hooks to the VM engine While there was a breakpoint hook, it was for only breakpoints and more was needed. Now there's a generic hook that is called for tracing, breakpoints, watch points, runtime errors and VM errors, with the "event" type passed as the first parameter and a data pointer in the second.
2020-03-24 06:35:42 +00:00
if (pr->debug_handler) {
[gamecode] Add a param pointer to debug_handler I decided I want events for VM enter/exit but enter needs to somehow pass the function which will be executed (even if a builtin). A generic void * param seemed the best idea, which meant the error string could be passed via the param instead of a "global" string in the progs struct.
2020-03-26 02:44:02 +00:00
pr->debug_handler (prd_watchpoint, 0, pr->debug_data);
[gamecode] Add debug hooks to the VM engine While there was a breakpoint hook, it was for only breakpoints and more was needed. Now there's a generic hook that is called for tracing, breakpoints, watch points, runtime errors and VM errors, with the "event" type passed as the first parameter and a data pointer in the second.
2020-03-24 06:35:42 +00:00
} else {
PR_RunError (pr, "watchpoint hit: %d -> %d",
[gamecode] Fix some silly errors I forgot to compile-test :(
2020-03-24 14:16:25 +00:00
old_val.integer_var, pr->watch->integer_var);
[gamecode] Add debug hooks to the VM engine While there was a breakpoint hook, it was for only breakpoints and more was needed. Now there's a generic hook that is called for tracing, breakpoints, watch points, runtime errors and VM errors, with the "event" type passed as the first parameter and a data pointer in the second.
2020-03-24 06:35:42 +00:00
}
[gamecode] Improve watchpoint handling a little Mostly just update the old value if the condition didn't trigger.
2020-03-24 04:26:35 +00:00
}
[gamecode] Fix some silly errors I forgot to compile-test :(
2020-03-24 14:16:25 +00:00
old_val.integer_var = pr->watch->integer_var;
[gamecode] Improve watchpoint handling a little Mostly just update the old value if the condition didn't trigger.
2020-03-24 04:26:35 +00:00
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Allow Sys_Error to be hooked. This makes debugging builtins that wrap normal functions a little easier by giving a progs dump when such an error occurs.
2016-01-03 14:04:00 +00:00
exit_program:
[gamecode] Add VM enter/exit events And rename prd_exit to prd_terminate (the idea is the host will terminate the VM). This makes it possible for the debugger to pause the VM before any code, even a builtin function, is executed. Breaks the debugger source window, but only because it's not updating on file change (I think).
2020-03-26 03:30:32 +00:00
if (pr->debug_handler) {
pr->debug_handler (prd_subexit, 0, pr->debug_data);
}
Improve handling of pr_argc It is now set to 0 when progs are loaded and every time PR_ExecuteProgram() returns. This takes care of the default case, but when setting parameters, pr_argc needs to be set correctly in case a vararg function is called.
2020-02-25 08:28:32 +00:00
pr->pr_argc = 0;
Allow Sys_Error to be hooked. This makes debugging builtins that wrap normal functions a little easier by giving a progs dump when such an error occurs.
2016-01-03 14:04:00 +00:00
Sys_PopErrorHandler ();
Sys_PopSignalHook ();
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Reference in a new issue Copy permalink