patch-2.3.40 linux/arch/arm/lib/csumpartialcopy.S

Next file: linux/arch/arm/lib/csumpartialcopyuser.S
Previous file: linux/arch/arm/lib/csumpartial.S
Back to the patch index
Back to the overall index

diff -u --recursive --new-file v2.3.39/linux/arch/arm/lib/csumpartialcopy.S linux/arch/arm/lib/csumpartialcopy.S
@@ -0,0 +1,257 @@
+/*
+ * linux/arch/arm/lib/csumpartialcopy.S
+ *
+ * Copyright (C) 1995-1998 Russell King
+ */
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+
+		.text
+
+/* Function: __u32 csum_partial_copy_nocheck(const char *src, char *dst, int len, __u32 sum)
+ * Params  : r0 = src, r1 = dst, r2 = len, r3 = checksum
+ * Returns : r0 = new checksum
+ */
+ENTRY(csum_partial_copy_nocheck)
+		mov	ip, sp
+		stmfd	sp!, {r4 - r8, fp, ip, lr, pc}
+		sub	fp, ip, #4
+		cmp	r2, #4
+		blt	Ltoo_small
+		tst	r1, #2			@ Test destination alignment
+		beq	Ldst_aligned
+		ldrb	ip, [r0], #1
+		ldrb	r8, [r0], #1
+		subs	r2, r2, #2		@ We do not know if SRC is aligned...
+		orr	ip, ip, r8, lsl #8
+		adds	r3, r3, ip
+		adcs	r3, r3, #0
+		strb	ip, [r1], #1
+		mov	ip, ip, lsr #8
+		strb	ip, [r1], #1		@ Destination now aligned
+Ldst_aligned:	tst	r0, #3
+		bne	Lsrc_not_aligned
+		adds	r3, r3, #0
+		bics	ip, r2, #15		@ Routine for src & dst aligned
+		beq	3f
+1:		ldmia	r0!, {r4, r5, r6, r7}
+		stmia	r1!, {r4, r5, r6, r7}
+		adcs	r3, r3, r4
+		adcs	r3, r3, r5
+		adcs	r3, r3, r6
+		adcs	r3, r3, r7
+		sub	ip, ip, #16
+		teq	ip, #0
+		bne	1b
+3:		ands	ip, r2, #12
+		beq	5f
+		tst	ip, #8
+		beq	4f
+		ldmia	r0!, {r4, r5}
+		stmia	r1!, {r4, r5}
+		adcs	r3, r3, r4
+		adcs	r3, r3, r5
+		tst	ip, #4
+		beq	5f
+4:		ldr	r4, [r0], #4
+		str	r4, [r1], #4
+		adcs	r3, r3, r4
+5:		ands	r2, r2, #3
+		adceq	r0, r3, #0
+		LOADREGS(eqea,fp,{r4 - r8, fp, sp, pc})
+		ldr	r4, [r0], #4
+		tst	r2, #2
+		beq	Lexit_r4
+		adcs	r3, r3, r4, lsl #16
+		strb	r4, [r1], #1
+		mov	r4, r4, lsr #8
+		strb	r4, [r1], #1
+		mov	r4, r4, lsr #8
+		b	Lexit_r4
+
+Ltoo_small:	teq	r2, #0
+		LOADREGS(eqea,fp,{r4 - r8, fp, sp, pc})
+		cmp	r2, #2
+		blt	Ltoo_small1
+		ldrb	ip, [r0], #1
+		ldrb	r8, [r0], #1
+		orr	ip, ip, r8, lsl #8
+		adds	r3, r3, ip
+		strb	ip, [r1], #1
+		strb	r8, [r1], #1
+		tst	r2, #1
+Ltoo_small1:	ldrneb	r4, [r0], #1
+Lexit_r4:	tst	r2, #1
+		strneb	r4, [r1], #1
+		andne	r4, r4, #255
+		adcnes	r3, r3, r4
+		adcs	r0, r3, #0
+		LOADREGS(ea,fp,{r4 - r8, fp, sp, pc})
+
+Lsrc_not_aligned:
+		cmp	r2, #4
+		blt	Ltoo_small
+		and	ip, r0, #3
+		bic	r0, r0, #3
+		ldr	r4, [r0], #4
+		cmp	ip, #2
+		beq	Lsrc2_aligned
+		bhi	Lsrc3_aligned
+		mov	r4, r4, lsr #8
+		adds	r3, r3, #0
+		bics	ip, r2, #15
+		beq	2f
+1:		ldmia	r0!, {r5, r6, r7, r8}
+		orr	r4, r4, r5, lsl #24
+		mov	r5, r5, lsr #8
+		orr	r5, r5, r6, lsl #24
+		mov	r6, r6, lsr #8
+		orr	r6, r6, r7, lsl #24
+		mov	r7, r7, lsr #8
+		orr	r7, r7, r8, lsl #24
+		stmia	r1!, {r4, r5, r6, r7}
+		adcs	r3, r3, r4
+		adcs	r3, r3, r5
+		adcs	r3, r3, r6
+		adcs	r3, r3, r7
+		mov	r4, r8, lsr #8
+		sub	ip, ip, #16
+		teq	ip, #0
+		bne	1b
+2:		ands	ip, r2, #12
+		beq	4f
+		tst	ip, #8
+		beq	3f
+		ldmia	r0!, {r5, r6}
+		orr	r4, r4, r5, lsl #24
+		mov	r5, r5, lsr #8
+		orr	r5, r5, r6, lsl #24
+		stmia	r1!, {r4, r5}
+		adcs	r3, r3, r4
+		adcs	r3, r3, r5
+		mov	r4, r6, lsr #8
+		tst	ip, #4
+		beq	4f
+3:		ldr	r5, [r0], #4
+		orr	r4, r4, r5, lsl #24
+		str	r4, [r1], #4
+		adcs	r3, r3, r4
+		mov	r4, r5, lsr #8
+4:		ands	r2, r2, #3
+		adceq	r0, r3, #0
+		LOADREGS(eqea,fp,{r4 - r8, fp, sp, pc})
+		tst	r2, #2
+		beq	Lexit_r4
+		adcs	r3, r3, r4, lsl #16
+		strb	r4, [r1], #1
+		mov	r4, r4, lsr #8
+		strb	r4, [r1], #1
+		mov	r4, r4, lsr #8
+		b	Lexit_r4
+
+Lsrc2_aligned:	mov	r4, r4, lsr #16
+		adds	r3, r3, #0
+		bics	ip, r2, #15
+		beq	2f
+1:		ldmia	r0!, {r5, r6, r7, r8}
+		orr	r4, r4, r5, lsl #16
+		mov	r5, r5, lsr #16
+		orr	r5, r5, r6, lsl #16
+		mov	r6, r6, lsr #16
+		orr	r6, r6, r7, lsl #16
+		mov	r7, r7, lsr #16
+		orr	r7, r7, r8, lsl #16
+		stmia	r1!, {r4, r5, r6, r7}
+		adcs	r3, r3, r4
+		adcs	r3, r3, r5
+		adcs	r3, r3, r6
+		adcs	r3, r3, r7
+		mov	r4, r8, lsr #16
+		sub	ip, ip, #16
+		teq	ip, #0
+		bne	1b
+2:		ands	ip, r2, #12
+		beq	4f
+		tst	ip, #8
+		beq	3f
+		ldmia	r0!, {r5, r6}
+		orr	r4, r4, r5, lsl #16
+		mov	r5, r5, lsr #16
+		orr	r5, r5, r6, lsl #16
+		stmia	r1!, {r4, r5}
+		adcs	r3, r3, r4
+		adcs	r3, r3, r5
+		mov	r4, r6, lsr #16
+		tst	ip, #4
+		beq	4f
+3:		ldr	r5, [r0], #4
+		orr	r4, r4, r5, lsl #16
+		str	r4, [r1], #4
+		adcs	r3, r3, r4
+		mov	r4, r5, lsr #16
+4:		ands	r2, r2, #3
+		adceq	r0, r3, #0
+		LOADREGS(eqea,fp,{r4 - r8, fp, sp, pc})
+		tst	r2, #2
+		beq	Lexit_r4
+		adcs	r3, r3, r4, lsl #16
+		strb	r4, [r1], #1
+		mov	r4, r4, lsr #8
+		strb	r4, [r1], #1
+		ldrb	r4, [r0], #1
+		b	Lexit_r4
+
+Lsrc3_aligned:	mov	r4, r4, lsr #24
+		adds	r3, r3, #0
+		bics	ip, r2, #15
+		beq	2f
+1:		ldmia	r0!, {r5, r6, r7, r8}
+		orr	r4, r4, r5, lsl #8
+		mov	r5, r5, lsr #24
+		orr	r5, r5, r6, lsl #8
+		mov	r6, r6, lsr #24
+		orr	r6, r6, r7, lsl #8
+		mov	r7, r7, lsr #24
+		orr	r7, r7, r8, lsl #8
+		stmia	r1!, {r4, r5, r6, r7}
+		adcs	r3, r3, r4
+		adcs	r3, r3, r5
+		adcs	r3, r3, r6
+		adcs	r3, r3, r7
+		mov	r4, r8, lsr #24
+		sub	ip, ip, #16
+		teq	ip, #0
+		bne	1b
+2:		ands	ip, r2, #12
+		beq	4f
+		tst	ip, #8
+		beq	3f
+		ldmia	r0!, {r5, r6}
+		orr	r4, r4, r5, lsl #8
+		mov	r5, r5, lsr #24
+		orr	r5, r5, r6, lsl #8
+		stmia	r1!, {r4, r5}
+		adcs	r3, r3, r4
+		adcs	r3, r3, r5
+		mov	r4, r6, lsr #24
+		tst	ip, #4
+		beq	4f
+3:		ldr	r5, [r0], #4
+		orr	r4, r4, r5, lsl #8
+		str	r4, [r1], #4
+		adcs	r3, r3, r4
+		mov	r4, r5, lsr #24
+4:		ands	r2, r2, #3
+		adceq	r0, r3, #0
+		LOADREGS(eqea,fp,{r4 - r8, fp, sp, pc})
+		tst	r2, #2
+		beq	Lexit_r4
+		adcs	r3, r3, r4, lsl #16
+		strb	r4, [r1], #1
+		ldr	r4, [r0], #4
+		strb	r4, [r1], #1
+		adcs	r3, r3, r4, lsl #24
+		mov	r4, r4, lsr #8
+		b	Lexit_r4
+
+

FUNET's LINUX-ADM group, linux-adm@nic.funet.fi
TCL-scripts by Sam Shen (who was at: slshen@lbl.gov)