mirror of
https://gitlab.com/freepascal.org/fpc/source.git
synced 2025-04-19 03:19:35 +02:00

* tregisteralloctor renamed to trgobj * removed rgobj from a lot of units * moved location_* and reference_* to cgobj * first things for mmx register allocation
526 lines
18 KiB
ObjectPascal
526 lines
18 KiB
ObjectPascal
{
|
|
$Id$
|
|
Copyright (c) 1998-2002 by Florian Klaempfl
|
|
|
|
Generate i386 inline nodes
|
|
|
|
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 the Free Software
|
|
Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|
|
|
****************************************************************************
|
|
}
|
|
unit n386inl;
|
|
|
|
{$i fpcdefs.inc}
|
|
|
|
interface
|
|
|
|
uses
|
|
node,ninl,ncginl;
|
|
|
|
type
|
|
ti386inlinenode = class(tcginlinenode)
|
|
{ first pass override
|
|
so that the code generator will actually generate
|
|
these nodes.
|
|
}
|
|
function first_pi: tnode ; override;
|
|
function first_arctan_real: tnode; override;
|
|
function first_abs_real: tnode; override;
|
|
function first_sqr_real: tnode; override;
|
|
function first_sqrt_real: tnode; override;
|
|
function first_ln_real: tnode; override;
|
|
function first_cos_real: tnode; override;
|
|
function first_sin_real: tnode; override;
|
|
{ second pass override to generate these nodes }
|
|
procedure second_IncludeExclude;override;
|
|
procedure second_pi; override;
|
|
procedure second_arctan_real; override;
|
|
procedure second_abs_real; override;
|
|
procedure second_sqr_real; override;
|
|
procedure second_sqrt_real; override;
|
|
procedure second_ln_real; override;
|
|
procedure second_cos_real; override;
|
|
procedure second_sin_real; override;
|
|
private
|
|
procedure load_fpu_location;
|
|
end;
|
|
|
|
implementation
|
|
|
|
uses
|
|
systems,
|
|
cutils,verbose,
|
|
defutil,
|
|
aasmtai,
|
|
cgbase,pass_2,
|
|
cpubase,paramgr,
|
|
nbas,ncon,ncal,ncnv,nld,
|
|
cga,cgx86,cgobj;
|
|
|
|
|
|
{*****************************************************************************
|
|
TI386INLINENODE
|
|
*****************************************************************************}
|
|
|
|
function ti386inlinenode.first_pi : tnode;
|
|
begin
|
|
expectloc:=LOC_FPUREGISTER;
|
|
registersfpu:=1;
|
|
first_pi := nil;
|
|
end;
|
|
|
|
|
|
function ti386inlinenode.first_arctan_real : tnode;
|
|
begin
|
|
expectloc:=LOC_FPUREGISTER;
|
|
registers32:=left.registers32;
|
|
registersfpu:=max(left.registersfpu,2);
|
|
{$ifdef SUPPORT_MMX}
|
|
registersmmx:=left.registersmmx;
|
|
{$endif SUPPORT_MMX}
|
|
first_arctan_real := nil;
|
|
end;
|
|
|
|
function ti386inlinenode.first_abs_real : tnode;
|
|
begin
|
|
expectloc:=LOC_FPUREGISTER;
|
|
registers32:=left.registers32;
|
|
registersfpu:=max(left.registersfpu,1);
|
|
{$ifdef SUPPORT_MMX}
|
|
registersmmx:=left.registersmmx;
|
|
{$endif SUPPORT_MMX}
|
|
first_abs_real := nil;
|
|
end;
|
|
|
|
function ti386inlinenode.first_sqr_real : tnode;
|
|
begin
|
|
expectloc:=LOC_FPUREGISTER;
|
|
registers32:=left.registers32;
|
|
registersfpu:=max(left.registersfpu,1);
|
|
{$ifdef SUPPORT_MMX}
|
|
registersmmx:=left.registersmmx;
|
|
{$endif SUPPORT_MMX}
|
|
first_sqr_real := nil;
|
|
end;
|
|
|
|
function ti386inlinenode.first_sqrt_real : tnode;
|
|
begin
|
|
expectloc:=LOC_FPUREGISTER;
|
|
registers32:=left.registers32;
|
|
registersfpu:=max(left.registersfpu,1);
|
|
{$ifdef SUPPORT_MMX}
|
|
registersmmx:=left.registersmmx;
|
|
{$endif SUPPORT_MMX}
|
|
first_sqrt_real := nil;
|
|
end;
|
|
|
|
function ti386inlinenode.first_ln_real : tnode;
|
|
begin
|
|
expectloc:=LOC_FPUREGISTER;
|
|
registers32:=left.registers32;
|
|
registersfpu:=max(left.registersfpu,2);
|
|
{$ifdef SUPPORT_MMX}
|
|
registersmmx:=left.registersmmx;
|
|
{$endif SUPPORT_MMX}
|
|
first_ln_real := nil;
|
|
end;
|
|
|
|
function ti386inlinenode.first_cos_real : tnode;
|
|
begin
|
|
expectloc:=LOC_FPUREGISTER;
|
|
registers32:=left.registers32;
|
|
registersfpu:=max(left.registersfpu,1);
|
|
{$ifdef SUPPORT_MMX}
|
|
registersmmx:=left.registersmmx;
|
|
{$endif SUPPORT_MMX}
|
|
first_cos_real := nil;
|
|
end;
|
|
|
|
function ti386inlinenode.first_sin_real : tnode;
|
|
begin
|
|
expectloc:=LOC_FPUREGISTER;
|
|
registers32:=left.registers32;
|
|
registersfpu:=max(left.registersfpu,1);
|
|
{$ifdef SUPPORT_MMX}
|
|
registersmmx:=left.registersmmx;
|
|
{$endif SUPPORT_MMX}
|
|
first_sin_real := nil;
|
|
end;
|
|
|
|
|
|
procedure ti386inlinenode.second_Pi;
|
|
begin
|
|
location_reset(location,LOC_FPUREGISTER,def_cgsize(resulttype.def));
|
|
emit_none(A_FLDPI,S_NO);
|
|
tcgx86(cg).inc_fpu_stack;
|
|
location.register:=NR_FPU_RESULT_REG;
|
|
end;
|
|
|
|
{ load the FPU into the an fpu register }
|
|
procedure ti386inlinenode.load_fpu_location;
|
|
begin
|
|
location_reset(location,LOC_FPUREGISTER,def_cgsize(resulttype.def));
|
|
location.register:=NR_FPU_RESULT_REG;
|
|
secondpass(left);
|
|
case left.location.loc of
|
|
LOC_FPUREGISTER:
|
|
;
|
|
LOC_CFPUREGISTER:
|
|
begin
|
|
cg.a_loadfpu_reg_reg(exprasmlist,left.location.size,
|
|
left.location.register,location.register);
|
|
end;
|
|
LOC_REFERENCE,LOC_CREFERENCE:
|
|
begin
|
|
cg.a_loadfpu_ref_reg(exprasmlist,
|
|
def_cgsize(left.resulttype.def),
|
|
left.location.reference,location.register);
|
|
location_release(exprasmlist,left.location);
|
|
end
|
|
else
|
|
internalerror(309991);
|
|
end;
|
|
end;
|
|
|
|
|
|
procedure ti386inlinenode.second_arctan_real;
|
|
begin
|
|
load_fpu_location;
|
|
emit_none(A_FLD1,S_NO);
|
|
emit_none(A_FPATAN,S_NO);
|
|
end;
|
|
|
|
procedure ti386inlinenode.second_abs_real;
|
|
begin
|
|
load_fpu_location;
|
|
emit_none(A_FABS,S_NO);
|
|
end;
|
|
|
|
|
|
procedure ti386inlinenode.second_sqr_real;
|
|
|
|
begin
|
|
load_fpu_location;
|
|
emit_reg_reg(A_FMUL,S_NO,NR_ST0,NR_ST0);
|
|
end;
|
|
|
|
procedure ti386inlinenode.second_sqrt_real;
|
|
begin
|
|
load_fpu_location;
|
|
emit_none(A_FSQRT,S_NO);
|
|
end;
|
|
|
|
procedure ti386inlinenode.second_ln_real;
|
|
begin
|
|
load_fpu_location;
|
|
emit_none(A_FLDLN2,S_NO);
|
|
emit_none(A_FXCH,S_NO);
|
|
emit_none(A_FYL2X,S_NO);
|
|
end;
|
|
|
|
procedure ti386inlinenode.second_cos_real;
|
|
begin
|
|
load_fpu_location;
|
|
emit_none(A_FCOS,S_NO);
|
|
end;
|
|
|
|
procedure ti386inlinenode.second_sin_real;
|
|
begin
|
|
load_fpu_location;
|
|
emit_none(A_FSIN,S_NO)
|
|
end;
|
|
|
|
{*****************************************************************************
|
|
INCLUDE/EXCLUDE GENERIC HANDLING
|
|
*****************************************************************************}
|
|
|
|
procedure ti386inlinenode.second_IncludeExclude;
|
|
var
|
|
hregister : tregister;
|
|
asmop : tasmop;
|
|
L : cardinal;
|
|
cgop : topcg;
|
|
begin
|
|
secondpass(tcallparanode(left).left);
|
|
if tcallparanode(tcallparanode(left).right).left.nodetype=ordconstn then
|
|
begin
|
|
{ calculate bit position }
|
|
l:=cardinal(1 shl (tordconstnode(tcallparanode(tcallparanode(left).right).left).value mod 32));
|
|
|
|
{ determine operator }
|
|
if inlinenumber=in_include_x_y then
|
|
cgop:=OP_OR
|
|
else
|
|
begin
|
|
cgop:=OP_AND;
|
|
l:=not(l);
|
|
end;
|
|
if (tcallparanode(left).left.location.loc=LOC_REFERENCE) then
|
|
begin
|
|
inc(tcallparanode(left).left.location.reference.offset,
|
|
(tordconstnode(tcallparanode(tcallparanode(left).right).left).value div 32)*4);
|
|
cg.a_op_const_ref(exprasmlist,cgop,OS_INT,l,tcallparanode(left).left.location.reference);
|
|
location_release(exprasmlist,tcallparanode(left).left.location);
|
|
end
|
|
else
|
|
{ LOC_CREGISTER }
|
|
begin
|
|
cg.a_op_const_reg(exprasmlist,cgop,tcallparanode(left).left.location.size,
|
|
l,tcallparanode(left).left.location.register);
|
|
end;
|
|
end
|
|
else
|
|
begin
|
|
{ generate code for the element to set }
|
|
secondpass(tcallparanode(tcallparanode(left).right).left);
|
|
{ determine asm operator }
|
|
if inlinenumber=in_include_x_y then
|
|
asmop:=A_BTS
|
|
else
|
|
asmop:=A_BTR;
|
|
|
|
if tcallparanode(tcallparanode(left).right).left.location.loc in [LOC_CREGISTER,LOC_REGISTER] then
|
|
{ we don't need a mod 32 because this is done automatically }
|
|
{ by the bts instruction. For proper checking we would }
|
|
|
|
{ note: bts doesn't do any mod'ing, that's why we can also use }
|
|
{ it for normalsets! (JM) }
|
|
|
|
{ need a cmp and jmp, but this should be done by the }
|
|
{ type cast code which does range checking if necessary (FK) }
|
|
begin
|
|
hregister:=cg.makeregsize(Tcallparanode(Tcallparanode(left).right).left.location.register,OS_INT);
|
|
end
|
|
else
|
|
begin
|
|
hregister:=cg.getintregister(exprasmlist,OS_INT);
|
|
end;
|
|
location_release(exprasmlist,tcallparanode(tcallparanode(left).right).left.location);
|
|
cg.a_load_loc_reg(exprasmlist,OS_INT,tcallparanode(tcallparanode(left).right).left.location,hregister);
|
|
if (tcallparanode(left).left.location.loc=LOC_REFERENCE) then
|
|
emit_reg_ref(asmop,S_L,hregister,tcallparanode(left).left.location.reference)
|
|
else
|
|
emit_reg_reg(asmop,S_L,hregister,tcallparanode(left).left.location.register);
|
|
cg.ungetregister(exprasmlist,hregister);
|
|
location_release(exprasmlist,Tcallparanode(left).left.location);
|
|
end;
|
|
end;
|
|
|
|
|
|
begin
|
|
cinlinenode:=ti386inlinenode;
|
|
end.
|
|
{
|
|
$Log$
|
|
Revision 1.70 2003-10-10 17:48:14 peter
|
|
* old trgobj moved to x86/rgcpu and renamed to trgx86fpu
|
|
* tregisteralloctor renamed to trgobj
|
|
* removed rgobj from a lot of units
|
|
* moved location_* and reference_* to cgobj
|
|
* first things for mmx register allocation
|
|
|
|
Revision 1.69 2003/10/09 21:31:37 daniel
|
|
* Register allocator splitted, ans abstract now
|
|
|
|
Revision 1.68 2003/10/01 20:34:49 peter
|
|
* procinfo unit contains tprocinfo
|
|
* cginfo renamed to cgbase
|
|
* moved cgmessage to verbose
|
|
* fixed ppc and sparc compiles
|
|
|
|
Revision 1.67 2003/09/28 21:48:20 peter
|
|
* fix register leaks
|
|
|
|
Revision 1.66 2003/09/07 22:09:35 peter
|
|
* preparations for different default calling conventions
|
|
* various RA fixes
|
|
|
|
Revision 1.65 2003/09/03 15:55:01 peter
|
|
* NEWRA branch merged
|
|
|
|
Revision 1.64.2.2 2003/08/31 15:46:26 peter
|
|
* more updates for tregister
|
|
|
|
Revision 1.64.2.1 2003/08/29 17:29:00 peter
|
|
* next batch of updates
|
|
|
|
Revision 1.64 2003/07/02 22:18:04 peter
|
|
* paraloc splitted in callerparaloc,calleeparaloc
|
|
* sparc calling convention updates
|
|
|
|
Revision 1.63 2003/06/03 13:01:59 daniel
|
|
* Register allocator finished
|
|
|
|
Revision 1.62 2003/06/01 21:38:06 peter
|
|
* getregisterfpu size parameter added
|
|
* op_const_reg size parameter added
|
|
* sparc updates
|
|
|
|
Revision 1.61 2003/05/30 23:49:18 jonas
|
|
* a_load_loc_reg now has an extra size parameter for the destination
|
|
register (properly fixes what I worked around in revision 1.106 of
|
|
ncgutil.pas)
|
|
|
|
Revision 1.60 2003/04/23 09:50:31 peter
|
|
* wrong location_copy for include/exclude
|
|
|
|
Revision 1.59 2003/04/22 23:50:23 peter
|
|
* firstpass uses expectloc
|
|
* checks if there are differences between the expectloc and
|
|
location.loc from secondpass in EXTDEBUG
|
|
|
|
Revision 1.58 2003/04/22 14:33:38 peter
|
|
* removed some notes/hints
|
|
|
|
Revision 1.57 2003/04/22 10:09:35 daniel
|
|
+ Implemented the actual register allocator
|
|
+ Scratch registers unavailable when new register allocator used
|
|
+ maybe_save/maybe_restore unavailable when new register allocator used
|
|
|
|
Revision 1.56 2003/02/19 22:00:15 daniel
|
|
* Code generator converted to new register notation
|
|
- Horribily outdated todo.txt removed
|
|
|
|
Revision 1.55 2003/01/08 18:43:57 daniel
|
|
* Tregister changed into a record
|
|
|
|
Revision 1.54 2002/11/25 17:43:26 peter
|
|
* splitted defbase in defutil,symutil,defcmp
|
|
* merged isconvertable and is_equal into compare_defs(_ext)
|
|
* made operator search faster by walking the list only once
|
|
|
|
Revision 1.53 2002/09/07 15:25:10 peter
|
|
* old logs removed and tabs fixed
|
|
|
|
Revision 1.52 2002/08/02 07:44:31 jonas
|
|
* made assigned() handling generic
|
|
* add nodes now can also evaluate constant expressions at compile time
|
|
that contain nil nodes
|
|
|
|
Revision 1.51 2002/07/26 11:16:35 jonas
|
|
* fixed (actual and potential) range errors
|
|
|
|
Revision 1.50 2002/07/25 18:02:33 carl
|
|
+ added generic inline nodes
|
|
|
|
Revision 1.49 2002/07/20 11:58:02 florian
|
|
* types.pas renamed to defbase.pas because D6 contains a types
|
|
unit so this would conflicts if D6 programms are compiled
|
|
+ Willamette/SSE2 instructions to assembler added
|
|
|
|
Revision 1.48 2002/07/11 14:41:33 florian
|
|
* start of the new generic parameter handling
|
|
|
|
Revision 1.47 2002/07/07 09:52:34 florian
|
|
* powerpc target fixed, very simple units can be compiled
|
|
* some basic stuff for better callparanode handling, far from being finished
|
|
|
|
Revision 1.46 2002/07/01 18:46:33 peter
|
|
* internal linker
|
|
* reorganized aasm layer
|
|
|
|
Revision 1.45 2002/07/01 16:23:56 peter
|
|
* cg64 patch
|
|
* basics for currency
|
|
* asnode updates for class and interface (not finished)
|
|
|
|
Revision 1.44 2002/05/18 13:34:25 peter
|
|
* readded missing revisions
|
|
|
|
Revision 1.43 2002/05/16 19:46:51 carl
|
|
+ defines.inc -> fpcdefs.inc to avoid conflicts if compiling by hand
|
|
+ try to fix temp allocation (still in ifdef)
|
|
+ generic constructor calls
|
|
+ start of tassembler / tmodulebase class cleanup
|
|
|
|
Revision 1.41 2002/05/13 19:54:38 peter
|
|
* removed n386ld and n386util units
|
|
* maybe_save/maybe_restore added instead of the old maybe_push
|
|
|
|
Revision 1.40 2002/05/12 16:53:17 peter
|
|
* moved entry and exitcode to ncgutil and cgobj
|
|
* foreach gets extra argument for passing local data to the
|
|
iterator function
|
|
* -CR checks also class typecasts at runtime by changing them
|
|
into as
|
|
* fixed compiler to cycle with the -CR option
|
|
* fixed stabs with elf writer, finally the global variables can
|
|
be watched
|
|
* removed a lot of routines from cga unit and replaced them by
|
|
calls to cgobj
|
|
* u32bit-s32bit updates for and,or,xor nodes. When one element is
|
|
u32bit then the other is typecasted also to u32bit without giving
|
|
a rangecheck warning/error.
|
|
* fixed pascal calling method with reversing also the high tree in
|
|
the parast, detected by tcalcst3 test
|
|
|
|
Revision 1.39 2002/04/23 19:16:35 peter
|
|
* add pinline unit that inserts compiler supported functions using
|
|
one or more statements
|
|
* moved finalize and setlength from ninl to pinline
|
|
|
|
Revision 1.38 2002/04/21 15:35:54 carl
|
|
* changeregsize -> rg.makeregsize
|
|
|
|
Revision 1.37 2002/04/19 15:39:35 peter
|
|
* removed some more routines from cga
|
|
* moved location_force_reg/mem to ncgutil
|
|
* moved arrayconstructnode secondpass to ncgld
|
|
|
|
Revision 1.36 2002/04/15 19:44:21 peter
|
|
* fixed stackcheck that would be called recursively when a stack
|
|
error was found
|
|
* generic changeregsize(reg,size) for i386 register resizing
|
|
* removed some more routines from cga unit
|
|
* fixed returnvalue handling
|
|
* fixed default stacksize of linux and go32v2, 8kb was a bit small :-)
|
|
|
|
Revision 1.35 2002/04/04 19:06:11 peter
|
|
* removed unused units
|
|
* use tlocation.size in cg.a_*loc*() routines
|
|
|
|
Revision 1.34 2002/04/02 17:11:36 peter
|
|
* tlocation,treference update
|
|
* LOC_CONSTANT added for better constant handling
|
|
* secondadd splitted in multiple routines
|
|
* location_force_reg added for loading a location to a register
|
|
of a specified size
|
|
* secondassignment parses now first the right and then the left node
|
|
(this is compatible with Kylix). This saves a lot of push/pop especially
|
|
with string operations
|
|
* adapted some routines to use the new cg methods
|
|
|
|
Revision 1.33 2002/03/31 20:26:39 jonas
|
|
+ a_loadfpu_* and a_loadmm_* methods in tcg
|
|
* register allocation is now second_d by a class and is mostly processor
|
|
independent (+rgobj.pas and i386/rgcpu.pas)
|
|
* temp allocation is now second_d by a class (+tgobj.pas, -i386\tgcpu.pas)
|
|
* some small improvements and fixes to the optimizer
|
|
* some register allocation fixes
|
|
* some fpuvaroffset fixes in the unary minus node
|
|
* push/popusedregisters is now called rg.save/restoreusedregisters and
|
|
(for i386) uses temps instead of push/pop's when using -Op3 (that code is
|
|
also better optimizable)
|
|
* fixed and optimized register saving/restoring for new/dispose nodes
|
|
* LOC_FPU locations now also require their "register" field to be set to
|
|
R_ST, not R_ST0 (the latter is used for LOC_CFPUREGISTER locations only)
|
|
- list field removed of the tnode class because it's not used currently
|
|
and can cause hard-to-find bugs
|
|
|
|
Revision 1.32 2002/03/04 19:10:14 peter
|
|
* removed compiler warnings
|
|
|
|
}
|