pa-risc2.s

mediastreamer2是开源的网络传输媒体流的库

	.PROCEND	

;-----------------------------------------------------------------------------
;
;void bn_sqr_comba4(BN_ULONG *r, BN_ULONG *a)
; arg0 = r_ptr
; arg1 = a_ptr
;

bn_sqr_comba4
	.proc
	.callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE
	.EXPORT	bn_sqr_comba4,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
    .entry
	.align 64
    STD     %r3,0(%sp)          ; save r3
    STD     %r4,8(%sp)          ; save r4
    STD     %r5,16(%sp)         ; save r5
    STD     %r6,24(%sp)         ; save r6

	;
	; Zero out carries
	;
	COPY     %r0,c1
	COPY     %r0,c2
	COPY     %r0,c3

	LDO      128(%sp),%sp       ; bump stack
    DEPDI,Z -1,32,33,high_mask   ; Create Mask 0xffffffff80000000L
    DEPDI,Z  1,31,1,high_one     ; Create Value  1 << 32

	;
	; Load up all of the values we are going to use
	;
    FLDD     0(a_ptr),a0       
    FLDD     8(a_ptr),a1       
    FLDD    16(a_ptr),a2       
    FLDD    24(a_ptr),a3       
    FLDD    32(a_ptr),a4       
    FLDD    40(a_ptr),a5       
    FLDD    48(a_ptr),a6       
    FLDD    56(a_ptr),a7       

	SQR_ADD_C a0L,a0R,c1,c2,c3

	STD     c1,0(r_ptr)          ; r[0] = c1;
	COPY    %r0,c1

	SQR_ADD_C2 a1L,a1R,a0L,a0R,c2,c3,c1

	STD     c2,8(r_ptr)          ; r[1] = c2;
	COPY    %r0,c2

	SQR_ADD_C a1L,a1R,c3,c1,c2
	SQR_ADD_C2 a2L,a2R,a0L,a0R,c3,c1,c2

	STD     c3,16(r_ptr)            ; r[2] = c3;
	COPY    %r0,c3

	SQR_ADD_C2 a3L,a3R,a0L,a0R,c1,c2,c3
	SQR_ADD_C2 a2L,a2R,a1L,a1R,c1,c2,c3

	STD     c1,24(r_ptr)           ; r[3] = c1;
	COPY    %r0,c1

	SQR_ADD_C a2L,a2R,c2,c3,c1
	SQR_ADD_C2 a3L,a3R,a1L,a1R,c2,c3,c1

	STD     c2,32(r_ptr)           ; r[4] = c2;
	COPY    %r0,c2

	SQR_ADD_C2 a3L,a3R,a2L,a2R,c3,c1,c2
	STD     c3,40(r_ptr)           ; r[5] = c3;
	COPY    %r0,c3

	SQR_ADD_C a3L,a3R,c1,c2,c3
	STD     c1,48(r_ptr)           ; r[6] = c1;
	STD     c2,56(r_ptr)           ; r[7] = c2;

    .EXIT
    LDD     -104(%sp),%r6        ; restore r6
    LDD     -112(%sp),%r5        ; restore r5
    LDD     -120(%sp),%r4        ; restore r4
    BVE     (%rp)
    LDD,MB  -128(%sp),%r3

	.PROCEND	


;---------------------------------------------------------------------------

MUL_ADD_C  .macro  A0L,A0R,B0L,B0R,C1,C2,C3
    XMPYU   A0L,B0R,ftemp1        ; m1 = bl*ht
    FSTD    ftemp1,-16(%sp)       ;
    XMPYU   A0R,B0L,ftemp2        ; m = bh*lt
    FSTD    ftemp2,-8(%sp)        ;
    XMPYU   A0R,B0R,ftemp3        ; lt = bl*lt
    FSTD    ftemp3,-32(%sp)
    XMPYU   A0L,B0L,ftemp4        ; ht = bh*ht
    FSTD    ftemp4,-24(%sp)       ;

    LDD     -8(%sp),m             ; r21 = m
    LDD     -16(%sp),m1           ; r19 = m1
    ADD,L   m,m1,m                ; m+m1

    DEPD,Z  m,31,32,temp3         ; (m+m1<<32)
    LDD     -24(%sp),ht           ; r24 = ht

    CMPCLR,*>>= m,m1,%r0          ; if (m < m1)
    ADD,L   ht,high_one,ht        ; ht+=high_one

    EXTRD,U m,31,32,temp1         ; m >> 32
    LDD     -32(%sp),lt           ; lt
    ADD,L   ht,temp1,ht           ; ht+= m>>32
    ADD     lt,temp3,lt           ; lt = lt+m1
    ADD,DC  ht,%r0,ht             ; ht++

    ADD     C1,lt,C1              ; c1=c1+lt
    ADD,DC  ht,%r0,ht             ; bump c3 if overflow,nullify otherwise

    ADD     C2,ht,C2              ; c2 = c2 + ht
    ADD,DC  C3,%r0,C3             ; add in carry (c3++)
.endm


;
;void bn_mul_comba8(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
; arg0 = r_ptr
; arg1 = a_ptr
; arg2 = b_ptr
;

bn_mul_comba8
	.proc
	.callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE
	.EXPORT	bn_mul_comba8,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
    .entry
	.align 64

    STD     %r3,0(%sp)          ; save r3
    STD     %r4,8(%sp)          ; save r4
    STD     %r5,16(%sp)         ; save r5
    STD     %r6,24(%sp)         ; save r6
    FSTD    %fr12,32(%sp)       ; save r6
    FSTD    %fr13,40(%sp)       ; save r7

	;
	; Zero out carries
	;
	COPY     %r0,c1
	COPY     %r0,c2
	COPY     %r0,c3

	LDO      128(%sp),%sp       ; bump stack
    DEPDI,Z  1,31,1,high_one     ; Create Value  1 << 32

	;
	; Load up all of the values we are going to use
	;
    FLDD      0(a_ptr),a0       
    FLDD      8(a_ptr),a1       
    FLDD     16(a_ptr),a2       
    FLDD     24(a_ptr),a3       
    FLDD     32(a_ptr),a4       
    FLDD     40(a_ptr),a5       
    FLDD     48(a_ptr),a6       
    FLDD     56(a_ptr),a7       

    FLDD      0(b_ptr),b0       
    FLDD      8(b_ptr),b1       
    FLDD     16(b_ptr),b2       
    FLDD     24(b_ptr),b3       
    FLDD     32(b_ptr),b4       
    FLDD     40(b_ptr),b5       
    FLDD     48(b_ptr),b6       
    FLDD     56(b_ptr),b7       

	MUL_ADD_C a0L,a0R,b0L,b0R,c1,c2,c3
	STD       c1,0(r_ptr)
	COPY      %r0,c1

	MUL_ADD_C a0L,a0R,b1L,b1R,c2,c3,c1
	MUL_ADD_C a1L,a1R,b0L,b0R,c2,c3,c1
	STD       c2,8(r_ptr)
	COPY      %r0,c2

	MUL_ADD_C a2L,a2R,b0L,b0R,c3,c1,c2
	MUL_ADD_C a1L,a1R,b1L,b1R,c3,c1,c2
	MUL_ADD_C a0L,a0R,b2L,b2R,c3,c1,c2
	STD       c3,16(r_ptr)
	COPY      %r0,c3

	MUL_ADD_C a0L,a0R,b3L,b3R,c1,c2,c3
	MUL_ADD_C a1L,a1R,b2L,b2R,c1,c2,c3
	MUL_ADD_C a2L,a2R,b1L,b1R,c1,c2,c3
	MUL_ADD_C a3L,a3R,b0L,b0R,c1,c2,c3
	STD       c1,24(r_ptr)
	COPY      %r0,c1

	MUL_ADD_C a4L,a4R,b0L,b0R,c2,c3,c1
	MUL_ADD_C a3L,a3R,b1L,b1R,c2,c3,c1
	MUL_ADD_C a2L,a2R,b2L,b2R,c2,c3,c1
	MUL_ADD_C a1L,a1R,b3L,b3R,c2,c3,c1
	MUL_ADD_C a0L,a0R,b4L,b4R,c2,c3,c1
	STD       c2,32(r_ptr)
	COPY      %r0,c2

	MUL_ADD_C a0L,a0R,b5L,b5R,c3,c1,c2
	MUL_ADD_C a1L,a1R,b4L,b4R,c3,c1,c2
	MUL_ADD_C a2L,a2R,b3L,b3R,c3,c1,c2
	MUL_ADD_C a3L,a3R,b2L,b2R,c3,c1,c2
	MUL_ADD_C a4L,a4R,b1L,b1R,c3,c1,c2
	MUL_ADD_C a5L,a5R,b0L,b0R,c3,c1,c2
	STD       c3,40(r_ptr)
	COPY      %r0,c3

	MUL_ADD_C a6L,a6R,b0L,b0R,c1,c2,c3
	MUL_ADD_C a5L,a5R,b1L,b1R,c1,c2,c3
	MUL_ADD_C a4L,a4R,b2L,b2R,c1,c2,c3
	MUL_ADD_C a3L,a3R,b3L,b3R,c1,c2,c3
	MUL_ADD_C a2L,a2R,b4L,b4R,c1,c2,c3
	MUL_ADD_C a1L,a1R,b5L,b5R,c1,c2,c3
	MUL_ADD_C a0L,a0R,b6L,b6R,c1,c2,c3
	STD       c1,48(r_ptr)
	COPY      %r0,c1
	
	MUL_ADD_C a0L,a0R,b7L,b7R,c2,c3,c1
	MUL_ADD_C a1L,a1R,b6L,b6R,c2,c3,c1
	MUL_ADD_C a2L,a2R,b5L,b5R,c2,c3,c1
	MUL_ADD_C a3L,a3R,b4L,b4R,c2,c3,c1
	MUL_ADD_C a4L,a4R,b3L,b3R,c2,c3,c1
	MUL_ADD_C a5L,a5R,b2L,b2R,c2,c3,c1
	MUL_ADD_C a6L,a6R,b1L,b1R,c2,c3,c1
	MUL_ADD_C a7L,a7R,b0L,b0R,c2,c3,c1
	STD       c2,56(r_ptr)
	COPY      %r0,c2

	MUL_ADD_C a7L,a7R,b1L,b1R,c3,c1,c2
	MUL_ADD_C a6L,a6R,b2L,b2R,c3,c1,c2
	MUL_ADD_C a5L,a5R,b3L,b3R,c3,c1,c2
	MUL_ADD_C a4L,a4R,b4L,b4R,c3,c1,c2
	MUL_ADD_C a3L,a3R,b5L,b5R,c3,c1,c2
	MUL_ADD_C a2L,a2R,b6L,b6R,c3,c1,c2
	MUL_ADD_C a1L,a1R,b7L,b7R,c3,c1,c2
	STD       c3,64(r_ptr)
	COPY      %r0,c3

	MUL_ADD_C a2L,a2R,b7L,b7R,c1,c2,c3
	MUL_ADD_C a3L,a3R,b6L,b6R,c1,c2,c3
	MUL_ADD_C a4L,a4R,b5L,b5R,c1,c2,c3
	MUL_ADD_C a5L,a5R,b4L,b4R,c1,c2,c3
	MUL_ADD_C a6L,a6R,b3L,b3R,c1,c2,c3
	MUL_ADD_C a7L,a7R,b2L,b2R,c1,c2,c3
	STD       c1,72(r_ptr)
	COPY      %r0,c1

	MUL_ADD_C a7L,a7R,b3L,b3R,c2,c3,c1
	MUL_ADD_C a6L,a6R,b4L,b4R,c2,c3,c1
	MUL_ADD_C a5L,a5R,b5L,b5R,c2,c3,c1
	MUL_ADD_C a4L,a4R,b6L,b6R,c2,c3,c1
	MUL_ADD_C a3L,a3R,b7L,b7R,c2,c3,c1
	STD       c2,80(r_ptr)
	COPY      %r0,c2

	MUL_ADD_C a4L,a4R,b7L,b7R,c3,c1,c2
	MUL_ADD_C a5L,a5R,b6L,b6R,c3,c1,c2
	MUL_ADD_C a6L,a6R,b5L,b5R,c3,c1,c2
	MUL_ADD_C a7L,a7R,b4L,b4R,c3,c1,c2
	STD       c3,88(r_ptr)
	COPY      %r0,c3

	MUL_ADD_C a7L,a7R,b5L,b5R,c1,c2,c3
	MUL_ADD_C a6L,a6R,b6L,b6R,c1,c2,c3
	MUL_ADD_C a5L,a5R,b7L,b7R,c1,c2,c3
	STD       c1,96(r_ptr)
	COPY      %r0,c1

	MUL_ADD_C a6L,a6R,b7L,b7R,c2,c3,c1
	MUL_ADD_C a7L,a7R,b6L,b6R,c2,c3,c1
	STD       c2,104(r_ptr)
	COPY      %r0,c2

	MUL_ADD_C a7L,a7R,b7L,b7R,c3,c1,c2
	STD       c3,112(r_ptr)
	STD       c1,120(r_ptr)

    .EXIT
    FLDD    -88(%sp),%fr13 
    FLDD    -96(%sp),%fr12 
    LDD     -104(%sp),%r6        ; restore r6
    LDD     -112(%sp),%r5        ; restore r5
    LDD     -120(%sp),%r4        ; restore r4
    BVE     (%rp)
    LDD,MB  -128(%sp),%r3

	.PROCEND	

;-----------------------------------------------------------------------------
;
;void bn_mul_comba4(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
; arg0 = r_ptr
; arg1 = a_ptr
; arg2 = b_ptr
;

bn_mul_comba4
	.proc
	.callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE
	.EXPORT	bn_mul_comba4,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
    .entry
	.align 64

    STD     %r3,0(%sp)          ; save r3
    STD     %r4,8(%sp)          ; save r4
    STD     %r5,16(%sp)         ; save r5
    STD     %r6,24(%sp)         ; save r6
    FSTD    %fr12,32(%sp)       ; save r6
    FSTD    %fr13,40(%sp)       ; save r7

	;
	; Zero out carries
	;
	COPY     %r0,c1
	COPY     %r0,c2
	COPY     %r0,c3

	LDO      128(%sp),%sp       ; bump stack
    DEPDI,Z  1,31,1,high_one     ; Create Value  1 << 32

	;
	; Load up all of the values we are going to use
	;
    FLDD      0(a_ptr),a0       
    FLDD      8(a_ptr),a1       
    FLDD     16(a_ptr),a2       
    FLDD     24(a_ptr),a3       

    FLDD      0(b_ptr),b0       
    FLDD      8(b_ptr),b1       
    FLDD     16(b_ptr),b2       
    FLDD     24(b_ptr),b3       

	MUL_ADD_C a0L,a0R,b0L,b0R,c1,c2,c3
	STD       c1,0(r_ptr)
	COPY      %r0,c1

	MUL_ADD_C a0L,a0R,b1L,b1R,c2,c3,c1
	MUL_ADD_C a1L,a1R,b0L,b0R,c2,c3,c1
	STD       c2,8(r_ptr)
	COPY      %r0,c2

	MUL_ADD_C a2L,a2R,b0L,b0R,c3,c1,c2
	MUL_ADD_C a1L,a1R,b1L,b1R,c3,c1,c2
	MUL_ADD_C a0L,a0R,b2L,b2R,c3,c1,c2
	STD       c3,16(r_ptr)
	COPY      %r0,c3

	MUL_ADD_C a0L,a0R,b3L,b3R,c1,c2,c3
	MUL_ADD_C a1L,a1R,b2L,b2R,c1,c2,c3
	MUL_ADD_C a2L,a2R,b1L,b1R,c1,c2,c3
	MUL_ADD_C a3L,a3R,b0L,b0R,c1,c2,c3
	STD       c1,24(r_ptr)
	COPY      %r0,c1

	MUL_ADD_C a3L,a3R,b1L,b1R,c2,c3,c1
	MUL_ADD_C a2L,a2R,b2L,b2R,c2,c3,c1
	MUL_ADD_C a1L,a1R,b3L,b3R,c2,c3,c1
	STD       c2,32(r_ptr)
	COPY      %r0,c2

	MUL_ADD_C a2L,a2R,b3L,b3R,c3,c1,c2
	MUL_ADD_C a3L,a3R,b2L,b2R,c3,c1,c2
	STD       c3,40(r_ptr)
	COPY      %r0,c3

	MUL_ADD_C a3L,a3R,b3L,b3R,c1,c2,c3
	STD       c1,48(r_ptr)
	STD       c2,56(r_ptr)

    .EXIT
    FLDD    -88(%sp),%fr13 
    FLDD    -96(%sp),%fr12 
    LDD     -104(%sp),%r6        ; restore r6
    LDD     -112(%sp),%r5        ; restore r5
    LDD     -120(%sp),%r4        ; restore r4
    BVE     (%rp)
    LDD,MB  -128(%sp),%r3

	.PROCEND	


;--- not PIC	.SPACE	$TEXT$
;--- not PIC	.SUBSPA	$CODE$
;--- not PIC	.SPACE	$PRIVATE$,SORT=16
;--- not PIC	.IMPORT	$global$,DATA
;--- not PIC	.SPACE	$TEXT$
;--- not PIC	.SUBSPA	$CODE$
;--- not PIC	.SUBSPA	$LIT$,ACCESS=0x2c
;--- not PIC	C$7
;--- not PIC	.ALIGN	8
;--- not PIC	.STRINGZ	"Division would overflow (%d)\n"
	.END