original files of m241

components/ip_net2-6.9/ipcom/arch/powerpc/src/ipcom_powerpc_gcc.s

@@ -0,0 +1,500 @@
+###############################################################################
+#                     INTERPEAK SOURCE FILE
+#
+#   Document no: @(#) $Name: VXWORKS_ITER40_2019102207 $ $RCSfile: ipcom_powerpc_gcc.s,v $ $Revision: 1.5 $
+#   $Source: /home/interpeak/CVSRoot/ipcom/arch/powerpc/src/ipcom_powerpc_gcc.s,v $
+#   $Author: lob $
+#   $State: Exp $ $Locker:  $
+#
+#       Copyright (C) 1995-98 by ENEA DATA AB. All rights reserved.          
+#                 Written by Anders Hagstrom for INET.
+#   Copyright Interpeak AB 2000-2004 <www.interpeak.se>. All rights reserved.
+###############################################################################
+#
+
+
+#
+#############################################################################
+#                       DESCRIPTION
+#############################################################################
+# 
+
+#
+#############################################################################
+#                       CONFIGURATION
+#############################################################################
+# 
+
+#
+#############################################################################
+#                       SECTION
+#############################################################################
+# 
+
+	.text
+
+#
+#############################################################################
+#                       DEFINES
+#############################################################################
+# 
+
+#
+#############################################################################
+#                       TYPES
+#############################################################################
+# 
+
+#
+#############################################################################
+#                       EXTERN PROTOTYPES
+#############################################################################
+# 
+
+#
+#############################################################################
+#                       LOCAL PROTOTYPES
+#############################################################################
+# 
+
+#
+#############################################################################
+#                       DATA
+#############################################################################
+# 
+
+#
+#############################################################################
+#                       STATIC FUNCTIONS
+#############################################################################
+# 
+
+#
+#############################################################################
+#                       GLOBAL FUNCTIONS
+#############################################################################
+# 
+
+#
+#############################################################################
+#                       PUBLIC FUNCTIONS
+#############################################################################
+# 
+
+#
+#############################################################################
+#                    ipcom_in_checksum_update2_asm
+#############################################################################
+# Prototype:	IP_PUBLIC Ip_u32
+#               ipcom_in_checksum_update2_asm(void *pbuf, Ip_size_t bytes);
+# Description:  Computes the internet checksum (RFC 1071). NOTE
+#               it does not compute the 1 complement checksum.
+# Parameters:   16bit aligned pbuf and bytes > 20
+# Returns:      A standard internet checksum..
+#
+        .globl ipcom_in_checksum_update2_asm
+ipcom_in_checksum_update2_asm:
+	andi.	r6,r3,0x2	#Check if halfword or word alignment.
+        li      r5,0            #Init checksum.
+	beq+	1f
+	lhz	r5,0(r3)	#Get first halword to get word align.
+	addi	r3,r3,2		#Move pointer.
+	subi	r4,r4,2		#Compensate numBytes.
+1:		
+        andi.   r6,r4,0xc       #Extract extra word accesses.
+        srawi   r7,r4,4         #numBytes / 16.
+
+        cmpi    cr1,r6,0x8
+        addic   r7,r7,1         #Compensate loop counter, clear carry.
+        subi    r3,r3,4         #Compensate for lwzu.
+        mtspr   ctr,r7          #Setup loop counter.
+        beq     chk0            #Branch if no extra word.
+        bgt     cr1,chk3        #Branch if 3 extra word.
+        blt     cr1,chk1        #Branch if 1 extra word.
+        b       chk2            #2 extra word required.
+chk_loop:
+        lwzu    r6,4(r3)
+        adde    r5,r5,r6        #Add to checksum.
+chk3:
+        lwzu    r7,4(r3)
+        adde    r5,r5,r7        #Add to checksum.
+chk2:
+        lwzu    r6,4(r3)
+        adde    r5,r5,r6        #Add to checksum.
+chk1:
+        lwzu    r7,4(r3)
+        adde    r5,r5,r7        #Add to checksum.
+chk0:
+        bc      16,0,chk_loop
+
+	addi	r3,r3,4		#Update pointer.
+
+#       Handle trailing halfword.
+        andi.   r7,r4,2
+        beq+    no_halfword
+        lhz	r6,0(r3)        #Load last halfword.
+	addi	r3,r3,2		#Update pointer.
+        adde    r5,r5,r6        #Add to checksum.
+no_halfword:
+
+#       Handle trailing byte.
+        andi.   r7,r4,1
+        beq+    no_byte
+        lbz	r6,0(r3)	#Load last byte.
+        rlwinm  r6,r6,8,16,23   #Shift left 8 times.
+        adde    r5,r5,r6        #Add to checksum.
+no_byte:
+
+#       Convert checksum from 32 bit to 16 bit.
+
+        rlwinm  r6,r5,16,16,31  #Extract high part of checksum.
+        rlwinm  r5,r5,0,16,31   #Truncate high part of checksum.
+        adde    r5,r5,r6        #High part + low part + carry.
+        rlwinm  r6,r5,16,16,31  #Extract carry from checksum.
+        rlwinm  r5,r5,0,16,31   #Truncate high part of checksum.
+        add     r3,r5,r6        #Add carry to low part.
+        blr
+
+
+#
+#############################################################################
+#                    ipcom_memcpy_asm
+#############################################################################
+# Prototype:	IP_PUBLIC void *
+#               ipcom_memcpy_asm(void *s1, const void *s2, Ip_size_t n);
+# Description:  Copy data.
+# Parameters:   
+# Returns:      
+#
+        .globl ipcom_memcpy_asm
+ipcom_memcpy_asm:
+        cmpli   cr0,r5,17       #Threshold for byte copy.
+        xor     r8,r3,r4	#Compare dest and src.
+	mr	r6,r3		#Leave r3 (dest) as return value.
+        blt-    byte_copy
+        andi.   r7,r8,1         #Check for same byte alignment.
+        bne-    byte_copy       #Byte copy if even/odd alignment.
+
+        andi.   r7,r6,1         #Check for leading odd address.
+        beq+    check_hword_align
+        lbz     r7,0(r4)        #Copy one byte to get halfword align.
+        subi    r5,r5,1         #Decrement numBytes.
+        stb     r7,0(r6)
+        addi    r4,r4,1         #Move pointers.
+        addi    r6,r6,1
+check_hword_align:
+	
+        andi.   r7,r8,2         #Check for same halfword alignment.
+        beq+    word_align      #Word copy if possible.
+
+#	Do halfword copy.
+        subi    r4,r4,2         #Compensate pointers for lhzu/sthu.
+        subi    r6,r6,2
+        srawi   r8,r5,4         #numBytes / 16.
+	rlwinm.	r5,r5,0,28,31	#Calculate trailing bytes.
+        mtspr   ctr,r8
+hword_loop:
+        lhzu    r7,2(r4)
+        sthu    r7,2(r6)
+        lhzu    r8,2(r4)
+        sthu    r8,2(r6)
+        lhzu    r7,2(r4)
+        sthu    r7,2(r6)
+        lhzu    r8,2(r4)
+        sthu    r8,2(r6)
+        lhzu    r7,2(r4)
+        sthu    r7,2(r6)
+        lhzu    r8,2(r4)
+        sthu    r8,2(r6)
+        lhzu    r7,2(r4)
+        sthu    r7,2(r6)
+        lhzu    r8,2(r4)
+        sthu    r8,2(r6)
+        bc      16,0,hword_loop
+
+        beqlr                   #Done if numBytes = 0.
+	addi	r4,r4,1		#Move pointers.
+	addi	r6,r6,1
+        b       byte_copy_cont  #Continue with trailing bytes.
+
+#	Do word copy.
+
+word_align:
+        andi.   r7,r6,2         #Check for leading halfword address.
+        beq-    word_copy
+        lhz     r7,0(r4)        #Copy halfword to get word align.
+        subi    r5,r5,2         #Decrement numBytes.
+        sth     r7,0(r6)
+        subi    r4,r4,2         #Compensate pointers for lwzu/stwu.
+        subi    r6,r6,2
+        b       word_copy_cont
+
+word_copy:
+        subi    r4,r4,4         #Compensate pointers for lwzu/stwu.
+        subi    r6,r6,4
+word_copy_cont:
+	andi.	r7,r5,0xc	#Extract extra word accesses.
+        srawi   r8,r5,4         #numBytes / 16.
+	cmpi	cr1,r7,0x8
+	addi	r8,r8,1		#Compensate loop counter.
+        mtspr   ctr,r8
+        beq     wcopy0		#Branch if no extra word.
+        bgt     cr1,wcopy3	#Branch if 3 extra word.
+        blt     cr1,wcopy1	#Branch if 1 extra word.
+        b       wcopy2		#2 extra word required.
+
+word_loop:
+        lwzu    r7,4(r4)
+        stwu    r7,4(r6)
+wcopy3:
+        lwzu    r8,4(r4)
+        stwu    r8,4(r6)
+wcopy2:
+        lwzu    r7,4(r4)
+        stwu    r7,4(r6)
+wcopy1:
+        lwzu    r8,4(r4)
+        stwu    r8,4(r6)
+wcopy0:
+        bc      16,0,word_loop
+
+        rlwinm. r5,r5,0,30,31   #Calculate number of trailing bytes.
+	addi	r4,r4,3		#Move pointers.
+	addi	r6,r6,3
+        beqlr                   #Done if numBytes = 0.
+        b       byte_copy_cont  #Continue with trailing bytes.
+byte_copy:
+	subi	r4,r4,1		#Move pointers.
+	subi	r6,r6,1
+byte_copy_cont:
+	andi.	r7,r5,0x3	#Extract extra byte accesses.
+        srawi   r8,r5,2         #numBytes / 4.
+	cmpi	cr1,r7,2
+	addi	r8,r8,1		#Compensate loop counter.
+        mtspr   ctr,r8
+        beq     bcopy0		#Branch if no extra byte.
+        bgt     cr1,bcopy3	#Branch if 3 extra byte.
+        blt     cr1,bcopy1	#Branch if 1 extra byte.
+        b       bcopy2		#2 extra byte required.
+byte_loop:
+	lbzu	r7,1(r4)
+	stbu	r7,1(r6)
+bcopy3:
+	lbzu	r8,1(r4)
+	stbu	r8,1(r6)
+bcopy2:
+	lbzu	r7,1(r4)
+	stbu	r7,1(r6)
+bcopy1:
+	lbzu	r8,1(r4)
+	stbu	r8,1(r6)
+bcopy0:
+        bc      16,0,byte_loop
+        blr
+
+	
+#
+#############################################################################
+#                    ipcom_in_checksum_memcpy_asm
+#############################################################################
+# Prototype:	IP_PUBLIC Ip_u32
+#               ipcom_in_checksum_memcpy(void *dst, void *src, Ip_size_t len);
+# Description:  Copy data and calculate internet checksum.
+# Parameters:   
+# Returns:      
+#
+        .globl ipcom_in_checksum_memcpy_asm
+ipcom_in_checksum_memcpy_asm:
+	xor     r6,r6,r6        # Interpeak always use 0 as initial checksum
+        cmpli   cr0,r5,17       #Threshold for byte copy.
+	li	r12,0		#Clear byte swap flag.
+        xor     r8,r3,r4	#Compare dest and src.
+	addic	r5,r5,0		#Make dummy add to clear carry.
+        blt-    scpy_byte_copy
+        andi.   r7,r8,1         #Check for same byte alignment.
+        bne-    scpy_byte_copy	#Byte copy if even/odd alignment.
+
+        andi.   r7,r3,1         #Check for leading odd address.
+        beq+    scpy_check_hword_align
+        lbz     r7,0(r4)        #Copy one byte to get halfword align.
+        subi    r5,r5,1         #Decrement numBytes.
+        li	r12,1		#Set flag.
+	rlwinm	r8,r6,8,16,23	#Swap bytes for intitial checksum.
+        rlwinm	r6,r6,24,24,31
+	or	r6,r6,r8
+        stb     r7,0(r3)
+        adde    r6,r6,r7        #Add to checksum.
+        addi    r4,r4,1         #Move pointers.
+        addi    r3,r3,1
+scpy_check_hword_align:
+	
+        andi.   r7,r8,2         #Check for same halfword alignment.
+        beq+    scpy_word_align	#Word copy if possible.
+
+#	Do halfword copy.
+
+	andi.	r7,r5,0x6	#Extract extra word accesses.
+        subi    r4,r4,2         #Compensate pointers for lhzu/sthu.
+        subi    r3,r3,2
+        srawi   r8,r5,3         #numBytes / 8.
+	cmpi	cr1,r7,0x4
+	addi	r8,r8,1		#Compensate loop counter.
+        mtspr   ctr,r8
+        beq     scpy_hwcopy0	#Branch if no extra hword.
+        bgt     cr1,scpy_hwcopy3 #Branch if 3 extra hword.
+        blt     cr1,scpy_hwcopy1 #Branch if 1 extra hword.
+        b       scpy_hwcopy2	#2 extra hword required.
+	
+scpy_hword_loop:
+        lhzu    r7,2(r4)
+        sthu    r7,2(r3)
+        adde    r6,r6,r7        #Add to checksum.
+scpy_hwcopy3:
+        lhzu    r8,2(r4)
+        sthu    r8,2(r3)
+        adde    r6,r6,r8        #Add to checksum.
+scpy_hwcopy2:
+        lhzu    r7,2(r4)
+        sthu    r7,2(r3)
+        adde    r6,r6,r7        #Add to checksum.
+scpy_hwcopy1:
+        lhzu    r8,2(r4)
+        sthu    r8,2(r3)
+        adde    r6,r6,r8        #Add to checksum.
+scpy_hwcopy0:
+        bc      16,0,scpy_hword_loop
+
+	rlwinm.	r5,r5,0,31,31	#Check for trailing byte.
+        beq+	scpy_end	#Done if numBytes = 0.
+	addi	r4,r4,1		#Move pointers.
+	addi	r3,r3,1
+        b       scpy_trailing_bytes #Continue with trailing bytes.
+
+#	Do word copy.
+
+scpy_word_align:
+        andi.   r7,r3,2         #Check for leading halfword address.
+        beq-    scpy_word_copy
+        lhz     r7,0(r4)        #Copy halfword to get word align.
+        subi    r5,r5,2         #Decrement numBytes.
+        sth     r7,0(r3)
+        adde    r6,r6,r7        #Add to checksum.
+        subi    r4,r4,2         #Compensate pointers for lwzu/stwu.
+        subi    r3,r3,2
+        b       scpy_word_copy_cont
+
+scpy_word_copy:
+        subi    r4,r4,4         #Compensate pointers for lwzu/stwu.
+        subi    r3,r3,4
+scpy_word_copy_cont:
+	andi.	r7,r5,0xc	#Extract extra word accesses.
+        srawi   r8,r5,4         #numBytes / 16.
+	cmpi	cr1,r7,0x8
+	addi	r8,r8,1		#Compensate loop counter.
+        mtspr   ctr,r8
+        beq     scpy_wcopy0	#Branch if no extra word.
+        bgt     cr1,scpy_wcopy3	#Branch if 3 extra word.
+        blt     cr1,scpy_wcopy1	#Branch if 1 extra word.
+        b       scpy_wcopy2	#2 extra word required.
+
+scpy_word_loop:
+        lwzu    r7,4(r4)
+        stwu    r7,4(r3)
+        adde    r6,r6,r7        #Add to checksum.
+scpy_wcopy3:
+        lwzu    r8,4(r4)
+        stwu    r8,4(r3)
+        adde    r6,r6,r8        #Add to checksum.
+scpy_wcopy2:
+        lwzu    r7,4(r4)
+        stwu    r7,4(r3)
+        adde    r6,r6,r7        #Add to checksum.
+scpy_wcopy1:
+        lwzu    r8,4(r4)
+        stwu    r8,4(r3)
+        adde    r6,r6,r8        #Add to checksum.
+scpy_wcopy0:
+        bc      16,0,scpy_word_loop
+
+        rlwinm. r5,r5,0,30,31   #Calculate number of trailing bytes.
+	addi	r4,r4,3		#Move pointers.
+	addi	r3,r3,3
+        beq	scpy_end	#Done if numBytes = 0.
+        b       scpy_trailing_bytes #Continue with trailing bytes.
+
+#	Byte copy, bytes are not swaped.
+#	Four bytes for each loop, trailing bytes are copied later.
+scpy_byte_copy:
+	subi	r4,r4,1		#Move pointers.
+	subi	r3,r3,1
+        srawi   r8,r5,2         #numBytes / 4.
+	rlwinm	r5,r5,0,30,31	#Calculate number of trailing bytes.
+	addi	r8,r8,1		#Compensate loop counter.
+        mtspr   ctr,r8
+        b       scpy_bcopy0
+scpy_byte_loop:
+	lbzu	r7,1(r4)
+        rlwinm  r8,r7,8,16,23	#Shift left 8 times.
+	stbu	r7,1(r3)
+        adde    r6,r6,r8        #Add to checksum.
+scpy_bcopy3:
+	lbzu	r8,1(r4)
+	stbu	r8,1(r3)
+        adde    r6,r6,r8        #Add to checksum.
+scpy_bcopy2:
+	lbzu	r7,1(r4)
+        rlwinm  r8,r7,8,16,23	#Shift left 8 times.
+	stbu	r7,1(r3)
+        adde    r6,r6,r8        #Add to checksum.
+scpy_bcopy1:
+	lbzu	r8,1(r4)
+	stbu	r8,1(r3)
+        adde    r6,r6,r8        #Add to checksum.
+scpy_bcopy0:
+        bc      16,0,scpy_byte_loop
+
+#	Copy trailing bytes, byte swap is checked.
+scpy_trailing_bytes:
+	mr	r9,r12		#Copy swap flag.
+trailing_loop:
+	cmpli	cr0,r5,0	#Test loop counter.
+	subi	r5,r5,1
+	beq-	scpy_trailing_end
+	cmpli	cr0,r9,0
+	lbzu	r7,1(r4)
+	stbu	r7,1(r3)
+	bne	no_shift
+        rlwinm  r7,r7,8,16,23	#Shift left 8 times.
+no_shift:
+	xori	r9,r9,1		#Toggle flag.
+        adde    r6,r6,r7        #Add to checksum.
+	b	trailing_loop
+scpy_trailing_end:
+	
+scpy_end:
+	cmpli	cr0,r12,0	#Test swap flag.
+	
+#       Convert checksum from 32 bit to 16 bit.
+
+        rlwinm  r7,r6,16,16,31  #Extract high part of checksum.
+        rlwinm  r6,r6,0,16,31   #Truncate high part of checksum.
+        adde    r6,r6,r7        #High part + low part + carry.
+        rlwinm  r7,r6,16,16,31  #Extract carry from checksum.
+        rlwinm  r6,r6,0,16,31   #Truncate high part of checksum.
+        add     r3,r6,r7        #Add carry to low part.
+
+#	Check if bytes has been swaped, swap if needed.
+	beqlr			#Return if no swap.
+	rlwinm	r4,r3,24,24,31	#Swap bytes.
+	rlwinm	r3,r3,8,16,23
+	or	r3,r3,r4
+        blr
+
+        .end
+
+
+#
+#############################################################################
+#                       END OF FILE
+#############################################################################
+# 

components/ip_net2-6.9/ipcom/arch/powerpc/src/ipcom_powerpc_vxworks.s

@@ -0,0 +1,409 @@
+#define _ASMLANGUAGE
+#include "vxWorks.h"
+#include "asm.h"
+#include "regs.h"
+        .text
+        .align 2
+
+        .globl ipcom_in_checksum_update2_asm
+ipcom_in_checksum_update2_asm:
+        andi.	r6,r3,0x2
+        li      r5,0
+        beq+	1f
+        lhz	    r5,0(r3)
+        addi	r3,r3,2
+        subi	r4,r4,2
+1:
+        andi.   r6,r4,0xc
+        srawi   r7,r4,4
+        cmpi    cr1,r6,0x8
+        addic   r7,r7,1
+        subi    r3,r3,4
+        mtspr   9,r7
+        beq     chk0
+        bgt     cr1,chk3
+        blt     cr1,chk1
+        b       chk2
+
+chk_loop:
+        lwzu    r6,4(r3)
+        adde    r5,r5,r6
+
+chk3:
+        lwzu    r7,4(r3)
+        adde    r5,r5,r7
+
+chk2:
+        lwzu    r6,4(r3)
+        adde    r5,r5,r6
+
+chk1:
+        lwzu    r7,4(r3)
+        adde    r5,r5,r7
+
+chk0:
+        bc      16,0,chk_loop
+
+        addi	r3,r3,4
+        andi.   r7,r4,2
+        beq+    no_halfword
+
+        lhz	    r6,0(r3)
+        addi	r3,r3,2
+        adde    r5,r5,r6
+
+no_halfword:
+        andi.   r7,r4,1
+        beq+    no_byte
+
+        lbz	    r6,0(r3)
+        rlwinm  r6,r6,8,16,23
+        adde    r5,r5,r6
+
+no_byte:
+        rlwinm  r6,r5,16,16,31
+        rlwinm  r5,r5,0,16,31
+        adde    r5,r5,r6
+        rlwinm  r6,r5,16,16,31
+        rlwinm  r5,r5,0,16,31
+        add     r3,r5,r6
+        blr
+
+        .globl ipcom_memcpy_asm
+ipcom_memcpy_asm:
+        cmpli   cr0,r5,17
+        xor     r8,r3,r4
+        mr      r6,r3
+        blt-    byte_copy
+
+        andi.   r7,r8,1
+        bne-    byte_copy
+
+        andi.   r7,r6,1
+        beq+    check_hword_align
+
+        lbz     r7,0(r4)
+        subi    r5,r5,1
+        stb     r7,0(r6)
+        addi    r4,r4,1
+        addi    r6,r6,1
+
+check_hword_align:
+        andi.   r7,r8,2
+        beq+    word_align
+
+        subi    r4,r4,2
+        subi    r6,r6,2
+        srawi   r8,r5,4
+        rlwinm.	r5,r5,0,28,31
+        mtspr   9,r8
+
+hword_loop:
+        lhzu    r7,2(r4)
+        sthu    r7,2(r6)
+        lhzu    r8,2(r4)
+        sthu    r8,2(r6)
+        lhzu    r7,2(r4)
+        sthu    r7,2(r6)
+        lhzu    r8,2(r4)
+        sthu    r8,2(r6)
+        lhzu    r7,2(r4)
+        sthu    r7,2(r6)
+        lhzu    r8,2(r4)
+        sthu    r8,2(r6)
+        lhzu    r7,2(r4)
+        sthu    r7,2(r6)
+        lhzu    r8,2(r4)
+        sthu    r8,2(r6)
+        bc      16,0,hword_loop
+
+        beqlr
+
+        addi	r4,r4,1
+        addi	r6,r6,1
+        b       byte_copy_cont
+
+word_align:
+        andi.   r7,r6,2
+        beq-    word_copy
+
+        lhz     r7,0(r4)
+        subi    r5,r5,2
+        sth     r7,0(r6)
+        subi    r4,r4,2
+        subi    r6,r6,2
+        b       word_copy_cont
+
+word_copy:
+        subi    r4,r4,4
+        subi    r6,r6,4
+
+word_copy_cont:
+        andi.	r7,r5,0xc
+        srawi   r8,r5,4
+        cmpi	cr1,r7,0x8
+        addi	r8,r8,1
+        mtspr   9,r8
+        beq     wcopy0
+        bgt     cr1,wcopy3
+        blt     cr1,wcopy1
+        b       wcopy2
+
+word_loop:
+        lwzu    r7,4(r4)
+        stwu    r7,4(r6)
+
+wcopy3:
+        lwzu    r8,4(r4)
+        stwu    r8,4(r6)
+
+wcopy2:
+        lwzu    r7,4(r4)
+        stwu    r7,4(r6)
+wcopy1:
+        lwzu    r8,4(r4)
+        stwu    r8,4(r6)
+
+wcopy0:
+        bc      16,0,word_loop
+
+        rlwinm. r5,r5,0,30,31
+        addi	r4,r4,3
+        addi	r6,r6,3
+        beqlr
+        b       byte_copy_cont
+
+byte_copy:
+        subi	r4,r4,1
+        subi	r6,r6,1
+byte_copy_cont:
+        andi.	r7,r5,0x3
+        srawi   r8,r5,2
+        cmpi	cr1,r7,2
+        addi	r8,r8,1
+        mtspr   9,r8
+        beq     bcopy0
+        bgt     cr1,bcopy3
+        blt     cr1,bcopy1
+        b       bcopy2
+
+byte_loop:
+        lbzu	r7,1(r4)
+        stbu	r7,1(r6)
+
+bcopy3:
+        lbzu	r8,1(r4)
+        stbu	r8,1(r6)
+
+bcopy2:
+        lbzu	r7,1(r4)
+        stbu	r7,1(r6)
+
+bcopy1:
+        lbzu	r8,1(r4)
+        stbu	r8,1(r6)
+
+bcopy0:
+        bc      16,0,byte_loop
+
+        blr
+
+        .globl ipcom_in_checksum_memcpy_asm
+ipcom_in_checksum_memcpy_asm:
+        xor     r6,r6,r6
+        cmpli   cr0,r5,17
+        li      r12,0
+        xor     r8,r3,r4
+        addic	r5,r5,0
+        blt-    scpy_byte_copy
+
+        andi.   r7,r8,1
+        bne-    scpy_byte_copy
+
+        andi.   r7,r3,1
+        beq+    scpy_check_hword_align
+
+        lbz     r7,0(r4)
+        subi    r5,r5,1
+        li      r12,1
+        stb     r7,0(r3)
+        adde    r6,r6,r7
+        addi    r4,r4,1
+        addi    r3,r3,1
+
+scpy_check_hword_align:
+        andi.   r7,r8,2
+        beq+    scpy_word_align
+
+        andi.	r7,r5,0x6
+        subi    r4,r4,2
+        subi    r3,r3,2
+        srawi   r8,r5,3
+        cmpi	cr1,r7,0x4
+        addi	r8,r8,1
+        mtspr   9,r8
+        beq     scpy_hwcopy0
+        bgt     cr1,scpy_hwcopy3
+        blt     cr1,scpy_hwcopy1
+        b       scpy_hwcopy2
+
+scpy_hword_loop:
+        lhzu    r7,2(r4)
+        sthu    r7,2(r3)
+        adde    r6,r6,r7
+
+scpy_hwcopy3:
+        lhzu    r8,2(r4)
+        sthu    r8,2(r3)
+        adde    r6,r6,r8
+
+scpy_hwcopy2:
+        lhzu    r7,2(r4)
+        sthu    r7,2(r3)
+        adde    r6,r6,r7
+
+scpy_hwcopy1:
+        lhzu    r8,2(r4)
+        sthu    r8,2(r3)
+        adde    r6,r6,r8
+
+scpy_hwcopy0:
+        bc      16,0,scpy_hword_loop
+
+        rlwinm.	r5,r5,0,31,31
+        beq+	scpy_end
+
+        addi	r4,r4,1
+        addi	r3,r3,1
+        b       scpy_trailing_bytes
+
+scpy_word_align:
+        andi.   r7,r3,2
+        beq-    scpy_word_copy
+
+        lhz     r7,0(r4)
+        subi    r5,r5,2
+        sth     r7,0(r3)
+        adde    r6,r6,r7
+        subi    r4,r4,2
+        subi    r3,r3,2
+        b       scpy_word_copy_cont
+
+scpy_word_copy:
+        subi    r4,r4,4
+        subi    r3,r3,4
+
+scpy_word_copy_cont:
+        andi.	r7,r5,0xc
+        srawi   r8,r5,4
+        cmpi	cr1,r7,0x8
+        addi	r8,r8,1
+        mtspr   9,r8
+        beq     scpy_wcopy0
+        bgt     cr1,scpy_wcopy3
+        blt     cr1,scpy_wcopy1
+        b       scpy_wcopy2
+
+scpy_word_loop:
+        lwzu    r7,4(r4)
+        stwu    r7,4(r3)
+        adde    r6,r6,r7
+
+scpy_wcopy3:
+        lwzu    r8,4(r4)
+        stwu    r8,4(r3)
+        adde    r6,r6,r8
+
+scpy_wcopy2:
+        lwzu    r7,4(r4)
+        stwu    r7,4(r3)
+        adde    r6,r6,r7
+
+scpy_wcopy1:
+        lwzu    r8,4(r4)
+        stwu    r8,4(r3)
+        adde    r6,r6,r8
+
+scpy_wcopy0:
+        bc      16,0,scpy_word_loop
+
+        rlwinm. r5,r5,0,30,31
+        addi	r4,r4,3
+        addi	r3,r3,3
+        beq	    scpy_end
+        b       scpy_trailing_bytes
+
+scpy_byte_copy:
+        subi	r4,r4,1
+        subi	r3,r3,1
+        srawi   r8,r5,2
+        rlwinm	r5,r5,0,30,31
+        addi	r8,r8,1
+        mtspr   9,r8
+        b       scpy_bcopy0
+
+scpy_byte_loop:
+        lbzu	r7,1(r4)
+        rlwinm  r8,r7,8,16,23
+        stbu	r7,1(r3)
+        adde    r6,r6,r8
+
+scpy_bcopy3:
+        lbzu	r8,1(r4)
+        stbu	r8,1(r3)
+        adde    r6,r6,r8
+
+scpy_bcopy2:
+        lbzu	r7,1(r4)
+        rlwinm  r8,r7,8,16,23
+        stbu	r7,1(r3)
+        adde    r6,r6,r8
+
+scpy_bcopy1:
+        lbzu	r8,1(r4)
+        stbu	r8,1(r3)
+        adde    r6,r6,r8
+
+scpy_bcopy0:
+        bc      16,0,scpy_byte_loop
+
+scpy_trailing_bytes:
+        li	    r9,0
+
+trailing_loop:
+        cmpli	cr0,r5,0
+        subi	r5,r5,1
+        beq-	scpy_trailing_end
+
+        cmpli	cr0,r9,0
+        lbzu	r7,1(r4)
+        stbu	r7,1(r3)
+        bne	    no_shift
+
+        rlwinm  r7,r7,8,16,23
+
+no_shift:
+        xori	r9,r9,1
+        adde    r6,r6,r7
+        b	    trailing_loop
+
+scpy_trailing_end:
+
+scpy_end:
+	    cmpli	cr0,r12,0
+        rlwinm  r7,r6,16,16,31
+        rlwinm  r6,r6,0,16,31
+        adde    r6,r6,r7
+        rlwinm  r7,r6,16,16,31
+        rlwinm  r6,r6,0,16,31
+        add     r3,r6,r7
+
+        beqlr
+        rlwinm	r4,r3,24,24,31
+        rlwinm	r3,r3,8,16,23
+        or	    r3,r3,r4
+        blr
+
+        .end
+
+