|
|
@@ -14,9 +14,137 @@
|
|
|
|
|
|
**********************************************************************}
|
|
|
|
|
|
+{$define FPC_SYSTEM_HAS_DIV_QWORD}
|
|
|
+ function fpc_div_qword(n,z : qword) : qword;assembler;[public,alias: 'FPC_DIV_QWORD']; {$ifdef hascompilerproc} compilerproc; {$endif}
|
|
|
+ { from the ppc compiler writers guide }
|
|
|
+ assembler;
|
|
|
+ asm
|
|
|
+ // (R5:R6) = (R5:R6) / (R3:R4) (64b) = (64b / 64b)
|
|
|
+ // quo dvd dvs
|
|
|
+ //
|
|
|
+ // Remainder is returned in R3:R4.
|
|
|
+ //
|
|
|
+ // Code comment notation:
|
|
|
+ // msw = most-significant (high-order) word, i.e. bits 0..31
|
|
|
+ // lsw = least-significant (low-order) word, i.e. bits 32..63
|
|
|
+ // LZ = Leading Zeroes
|
|
|
+ // SD = Significant Digits
|
|
|
+ //
|
|
|
+ // R5:R6 = dvd (input dividend); quo (output quotient)
|
|
|
+ // R3:R4 = dvs (input divisor); rem (output remainder)
|
|
|
+ //
|
|
|
+ // R7:R8 = tmp
|
|
|
+ // count the number of leading 0s in the dividend
|
|
|
+ or. R0,R3,R4 // dvs = 0?
|
|
|
+ cmpwi cr1,R5,0 // dvd.msw == 0?
|
|
|
+ cntlzw R0,R5 // R0 = dvd.msw.LZ
|
|
|
+ cntlzw R9,R6 // R9 = dvd.lsw.LZ
|
|
|
+ bne+ .LNoDivByZero
|
|
|
+ b FPC_DIVBYZERO
|
|
|
+ .LNoDivByZero:
|
|
|
+ bne cr1,.Llab1 // if(dvd.msw == 0) dvd.LZ = dvd.msw.LZ
|
|
|
+ addi R0,R9,32 // dvd.LZ = dvd.lsw.LZ + 32
|
|
|
+ .Llab1:
|
|
|
+ // count the number of leading 0s in the divisor
|
|
|
+ cmpwi cr0,R3,0 // dvd.msw == 0?
|
|
|
+ cntlzw R9,R3 // R9 = dvs.msw.LZ
|
|
|
+ cntlzw R10,R4 // R10 = dvs.lsw.LZ
|
|
|
+ bne cr0,.Llab2 // if(dvs.msw == 0) dvs.LZ = dvs.msw.LZ
|
|
|
+ addi R9,R10,32 // dvs.LZ = dvs.lsw.LZ + 32
|
|
|
+ .Llab2:
|
|
|
+ // determine shift amounts to minimize the number of iterations
|
|
|
+ cmpw cr0,R0,R9 // compare dvd.LZ to dvs.LZ
|
|
|
+ subfic R10,R0,64 // R10 = dvd.SD
|
|
|
+ bgt cr0,.Llab9 // if(dvs > dvd) quotient = 0
|
|
|
+ addi R9,R9,1 // ++dvs.LZ (or --dvs.SD)
|
|
|
+ subfic R9,R9,64 // R9 = dvs.SD
|
|
|
+ add R0,R0,R9 // (dvd.LZ + dvs.SD) = left shift of dvd for
|
|
|
+ // initial dvd
|
|
|
+ subf R9,R9,R10 // (dvd.SD - dvs.SD) = right shift of dvd for
|
|
|
+ // initial tmp
|
|
|
+ mtctr R9 // number of iterations = dvd.SD - dvs.SD
|
|
|
+ // R7:R8 = R5:R6 >> R9
|
|
|
+ cmpwi cr0,R9,32 // compare R9 to 32
|
|
|
+ addi R7,R9,-32
|
|
|
+ blt cr0,.Llab3 // if(R9 < 32) jump to .Llab3
|
|
|
+ srw R8,R5,R7 // tmp.lsw = dvd.msw >> (R9 - 32)
|
|
|
+ li R7,0 // tmp.msw = 0
|
|
|
+ b .Llab4
|
|
|
+ .Llab3:
|
|
|
+ srw R8,R6,R9 // R8 = dvd.lsw >> R9
|
|
|
+ subfic R7,R9,32
|
|
|
+ slw R7,R5,R7 // R7 = dvd.msw << 32 - R9
|
|
|
+ or R8,R8,R7 // tmp.lsw = R8 | R7
|
|
|
+ srw R7,R5,R9 // tmp.msw = dvd.msw >> R9
|
|
|
+ .Llab4:
|
|
|
+ // R5:R6 = R5:R6 << R0
|
|
|
+ cmpwi cr0,R0,32 // compare R0 to 32
|
|
|
+ addic R9,R0,-32
|
|
|
+ blt cr0,.Llab5 // if(R0 < 32) jump to .Llab5
|
|
|
+ slw R5,R6,R9 // dvd.msw = dvd.lsw << R9
|
|
|
+ li R6,0 // dvd.lsw = 0
|
|
|
+ b .Llab6
|
|
|
+ .Llab5:
|
|
|
+ slw R5,R5,R0 // R5 = dvd.msw << R0
|
|
|
+ subfic R9,R0,32
|
|
|
+ srw R9,R6,R9 // R9 = dvd.lsw >> 32 - R0
|
|
|
+ or R5,R5,R9 // dvd.msw = R5 | R9
|
|
|
+ slw R6,R6,R0 // dvd.lsw = dvd.lsw << R0
|
|
|
+ .Llab6:
|
|
|
+ // restoring division shift and subtract loop
|
|
|
+ li R10,-1 // R10 = -1
|
|
|
+ addic R7,R7,0 // clear carry bit before loop starts
|
|
|
+ .Llab7:
|
|
|
+ // tmp:dvd is considered one large register
|
|
|
+ // each portion is shifted left 1 bit by adding it to itself
|
|
|
+ // adde sums the carry from the previous and creates a new carry
|
|
|
+ adde R6,R6,R6 // shift dvd.lsw left 1 bit
|
|
|
+ adde R5,R5,R5 // shift dvd.msw to left 1 bit
|
|
|
+ adde R8,R8,R8 // shift tmp.lsw to left 1 bit
|
|
|
+ adde R7,R7,R7 // shift tmp.msw to left 1 bit
|
|
|
+ subfc R0,R4,R8 // tmp.lsw - dvs.lsw
|
|
|
+ subfe. R9,R3,R7 // tmp.msw - dvs.msw
|
|
|
+ blt cr0,.Llab8 // if(result < 0) clear carry bit
|
|
|
+ mr R8,R0 // move lsw
|
|
|
+ mr R7,R9 // move msw
|
|
|
+ addic R0,R10,1 // set carry bit
|
|
|
+ .Llab8:
|
|
|
+ bdnz .Llab7
|
|
|
+ // write quotient and remainder
|
|
|
+ adde R4,R6,R6 // quo.lsw (lsb = CA)
|
|
|
+ adde R3,R5,R5 // quo.msw (lsb from lsw)
|
|
|
+ mr R6,R8 // rem.lsw
|
|
|
+ mr R5,R7 // rem.msw
|
|
|
+ b .Lqworddivdone // return
|
|
|
+ .Llab9:
|
|
|
+ // Quotient is 0 (dvs > dvd)
|
|
|
+ li R4,0 // dvd.lsw = 0
|
|
|
+ li R3,0 // dvd.msw = 0
|
|
|
+ .Lqworddivdone:
|
|
|
+ end;
|
|
|
+
|
|
|
+
|
|
|
+{$define FPC_SYSTEM_HAS_MOD_QWORD}
|
|
|
+ function fpc_mod_qword(n,z : qword) : qword;assembler;[public,alias: 'FPC_MOD_QWORD']; {$ifdef hascompilerproc} compilerproc; {$endif}
|
|
|
+ assembler;
|
|
|
+ var
|
|
|
+ oldlr: pointer;
|
|
|
+ asm
|
|
|
+ mflr r0
|
|
|
+ stw r0,oldlr
|
|
|
+ bl FPC_DIV_QWORD
|
|
|
+ lwz r0,oldlr
|
|
|
+ mtlr r0
|
|
|
+ mr R3,R5
|
|
|
+ mr R4,R6
|
|
|
+ end;
|
|
|
+
|
|
|
{
|
|
|
$Log$
|
|
|
- Revision 1.1 2003-09-14 11:34:13 peter
|
|
|
+ Revision 1.2 2004-01-12 18:03:30 jonas
|
|
|
+ + ppc implemetnation of fpc_mod/div_qword (from ppc compiler writers guide)
|
|
|
+
|
|
|
+ Revision 1.1 2003/09/14 11:34:13 peter
|
|
|
* moved int64 asm code to int64p.inc
|
|
|
* save ebx,esi
|
|
|
|
Jonas Maebe