+ added 32-bit and 64-bit unsigned asm optimized multiplication routines for

i8086, contributed by Max Nazhalov

git-svn-id: trunk@26306 -

.gitattributes

@@ -7945,6 +7945,7 @@ rtl/i386/strings.inc svneol=native#text/plain
 rtl/i386/stringss.inc svneol=native#text/plain
 rtl/i386/strpas.inc svneol=native#text/plain
 rtl/i8086/i8086.inc svneol=native#text/plain
+rtl/i8086/int32p.inc svneol=native#text/plain
 rtl/i8086/int64p.inc svneol=native#text/plain
 rtl/i8086/makefile.cpu svneol=native#text/plain
 rtl/i8086/math.inc svneol=native#text/plain

rtl/i8086/int32p.inc

@@ -0,0 +1,78 @@
+{
+    This file is part of the Free Pascal run time library.
+    Copyright (c) 2013 by the Free Pascal development team
+
+    This file contains some helper routines for longint and dword
+
+    See the file COPYING.FPC, included in this distribution,
+    for details about the copyright.
+
+    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.
+
+ **********************************************************************}
+
+{$define FPC_SYSTEM_HAS_MUL_DWORD}
+function fpc_mul_dword( f1, f2: dword; checkoverflow: boolean ): dword; [public,alias: 'FPC_MUL_DWORD']; compilerproc;
+begin
+{ routine contributed by Max Nazhalov
+
+//////// 16-bit multiplications summary:
+(A1:A0*B1:B0) = (A1*B1)<<32 + (A1*B0)<<16 + (A0*B1)<<16 + (A0*B0)
+
+  A1*B1 [only needed for overflow checking; overflow if <>0]
+  A1*B0
+  A0*B1
+  A0:B0
+  A3*B0 [only lo_word is needed; overflow if hi_word<>0]
+  A2*B1 [only lo_word is needed; overflow if hi_word<>0]
+  A2*B0
+  A1*B2 [only lo_word is needed; overflow if hi_word<>0]
+  A0*B3 [only lo_word is needed; overflow if hi_word<>0]
+  A0*B2
+}
+  asm
+    mov     cx,word[f1]
+    mov     ax,word[f1+2]
+    mov     di,word[f2]
+    mov     si,word[f2+2]
+    cmp     checkoverflow,0
+    jne     @@checked
+    mul     di
+    xchg    ax,si
+    mul     cx
+    add     si,ax
+    mov     ax,di
+    mul     cx
+    add     dx,si
+    jmp     @@done
+@@checked:
+    test    ax,ax
+    jz      @@skip
+    test    si,si
+    jnz     @@done
+    mul     di
+    test    dx,dx
+    jnz     @@done
+@@skip:
+    xchg    ax,si
+    mul     cx
+    test    dx,dx
+    jnz     @@done
+    add     si,ax
+    jc      @@done
+    mov     ax,di
+    mul     cx
+    add     dx,si
+    jc      @@done
+    // checked and succeed
+    mov     checkoverflow,0
+@@done:
+    mov     word[result],ax
+    mov     word[result+2],dx
+  end [ 'ax','cx','dx','si','di' ];
+  if checkoverflow then
+    HandleErrorAddrFrameInd(215,get_pc_addr,get_frame);
+end;
+

rtl/i8086/int64p.inc

@@ -1,6 +1,6 @@
 {
     This file is part of the Free Pascal run time library.
-    Copyright (c) 1999-2000 by the Free Pascal development team
+    Copyright (c) 2013 by the Free Pascal development team
 
     This file contains some helper routines for int64 and qword
 
@@ -12,3 +12,185 @@
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 
  **********************************************************************}
+
+{$I int32p.inc}
+
+{$define FPC_SYSTEM_HAS_MUL_QWORD}
+function fpc_mul_qword( f1, f2: qword; checkoverflow: longbool ): qword; [public,alias: 'FPC_MUL_QWORD']; compilerproc;
+begin
+{ routine contributed by Max Nazhalov
+
+64-bit multiplication via 16-bit digits: (A3:A2:A1:A0)*(B3:B2:B1:B0)
+
+//////// STEP 1; break-down to 32-bit multiplications, each of them generates 64-bit result:
+  (A3:A2*B3:B2)<<64 + (A3:A2*B1:B0)<<32 + (A1:A0*B3:B2)<<32 + (A1:A0*B1:B0)
+
+(A1:A0*B1:B0) = (A1*B1)<<32 + (A1*B0)<<16 + (A0*B1)<<16 + (A0:B0)
+ -- never overflows, forms the base of the final result, name it as "R64"
+
+(A3:A2*B3:B2) is not required for the 64-bit result if overflow is not checked, since it is completely beyond the resulting width.
+ -- always overflows if "<>0", so can be checked as "((A2|A3)<>0)&&(B2|B3)<>0)"
+
+(A3:A2*B1:B0) and (A1:A0*B3:B2) are partially required for the final result
+ -- to be calculated on steps 2 and 3 as a correction for the "R64"
+
+//////// STEP 2; calculate "R64+=(A3:A2*B1:B0)<<32" (16-bit multiplications, each of them generates 32-bit result):
+  (A3*B1)<<32 + (A3*B0)<<16 + (A2*B1)<<16 + (A2*B0)
+
+((A3*B1)<<32)<<32 is not required for the 64-bit result if overflow is not checked, since it is completely beyond the resulting width.
+ -- always overflows if "<>0", so can be checked as "(A3<>0)&&(B1<>0)"
+
+((A3*B0)<<16)<<32: only low word of "A3*B0" contributes to the final result if overflow is not checked.
+ -- overflows if the hi_word "<>0"
+ -- overflows if R64+(lo_word<<48) produces C-flag
+
+((A2*B1)<<16)<<32: only low word of "A2*B1" contributes to the final result if overflow is not checked.
+ -- overflows if the hi_word "<>0"
+ -- overflows if R64+(lo_word<<48) produces C-flag
+
+(A2*B0)<<32: the whole dword is significand, name it as "X"
+ -- overflows if R64+(X<<32) produces C-flag
+
+//////// STEP 3; calculate "R64+=(A1:A0*B3:B2)<<32" (16-bit multiplications, each of them generates 32-bit result):
+  (A1*B3)<<32 + (A1*B2)<<16 + (A0*B3)<<16 + (A0*B2)
+
+((A1*B3)<<32)<<32 is not required for the 64-bit result if overflow is not checked, since it is completely beyond the resulting width.
+ -- always overflows if "<>0", so can be checked as "(A1<>0)&&(B3<>0)"
+
+((A1*B2)<<16)<<32: only low word of "A1*B2" contributes to the final result if overflow is not checked.
+ -- overflows if the hi_word "<>0"
+ -- overflows if R64+(lo_word<<48) produces C-flag
+
+((A0*B3)<<16)<<32: only low word "A0*B3" contributes to the final result if overflow is not checked.
+ -- overflows if the hi_word "<>0"
+ -- overflows if R64+(lo_word<<48) produces C-flag
+
+(A0*B2)<<32: the whole dword is significand, name it as "Y"
+ -- overflows if R64+(Y<<32) produces C-flag
+}
+  asm
+    mov     di,word[f1]
+    mov     bx,word[f1+2]
+    mov     si,word[f2]
+    mov     ax,word[f2+2]
+    push    bp
+    mov     cx,ax
+    mul     bx
+    xchg    ax,bx
+    mov     bp,dx
+    mul     si
+    xchg    ax,cx
+    add     bx,dx
+    adc     bp,0
+    mul     di
+    add     cx,ax
+    adc     bx,dx
+    adc     bp,0
+    mov     ax,di
+    mul     si
+    add     cx,dx
+    adc     bx,0
+    adc     bp,0
+    mov     dx,bp
+    pop     bp
+    mov     word[result],ax
+    mov     word[result+2],cx
+    mov     word[result+4],bx
+    mov     word[result+6],dx
+    mov     si,word[f1+4]
+    mov     ax,word[f1+6]
+    mov     bx,word[checkoverflow]
+    or      bx,word[checkoverflow+2]
+    jnz     @@checked
+    mov     di,word[f2]
+    mul     di
+    mov     cx,ax
+    mov     ax,word[f2+2]
+    mul     si
+    add     cx,ax
+    mov     ax,di
+    mul     si
+    mov     bx,ax
+    add     cx,dx
+    mov     si,word[f2+4]
+    mov     ax,word[f2+6]
+    mov     di,word[f1]
+    mul     di
+    add     cx,ax
+    mov     ax,word[f1+2]
+    mul     si
+    add     cx,ax
+    mov     ax,di
+    mul     si
+    add     bx,ax
+    adc     cx,dx
+    add     word[result+4],bx
+    adc     word[result+6],cx
+    jmp     @@done
+@@checked:
+    mov     bx,word[f2+6]
+    mov     cx,ax
+    or      cx,si
+    jz      @@nover1
+    mov     cx,word[f2+4]
+    or      cx,bx
+    jnz     @@done
+@@nover1:
+    test    bx,bx
+    jz      @@nover2
+    mov     bx,word[f1+2]
+    test    bx,bx
+    jnz     @@done
+@@nover2:
+    test    ax,ax
+    jz      @@nover3
+    or      bx,word[f2+2]
+    jnz     @@done
+@@nover3:
+    mov     di,word[f2]
+    mul     di
+    test    dx,dx
+    jnz     @@done
+    mov     cx,ax
+    mov     ax,word[f2+2]
+    mul     si
+    test    dx,dx
+    jnz     @@done
+    add     cx,ax
+    jc      @@done
+    mov     ax,di
+    mul     si
+    mov     bx,ax
+    add     cx,dx
+    jc      @@done
+    mov     si,word[f2+4]
+    mov     ax,word[f2+6]
+    mov     di,word[f1]
+    mul     di
+    test    dx,dx
+    jnz     @@done
+    add     cx,ax
+    jc      @@done
+    mov     ax,word[f1+2]
+    mul     si
+    test    dx,dx
+    jnz     @@done
+    add     cx,ax
+    jc      @@done
+    mov     ax,di
+    mul     si
+    add     bx,ax
+    adc     cx,dx
+    jc      @@done
+    add     word[result+4],bx
+    adc     word[result+6],cx
+    jc      @@done
+    // checked and succeed
+    xor     ax,ax
+    mov     word[checkoverflow],ax
+    mov     word[checkoverflow+2],ax
+@@done:
+  end [ 'ax','bx','cx','dx','si','di' ];
+  if checkoverflow then
+    HandleErrorAddrFrameInd(215,get_pc_addr,get_frame);
+end;