diff options
| author | André Fabian Silva Delgado <emulatorman@parabola.nu> | 2015-08-05 17:04:01 -0300 |
|---|---|---|
| committer | André Fabian Silva Delgado <emulatorman@parabola.nu> | 2015-08-05 17:04:01 -0300 |
| commit | 57f0f512b273f60d52568b8c6b77e17f5636edc0 (patch) | |
| tree | 5e910f0e82173f4ef4f51111366a3f1299037a7b /arch/arm/lib/copy_page.S | |
Initial import
Diffstat (limited to 'arch/arm/lib/copy_page.S')
| -rw-r--r-- | arch/arm/lib/copy_page.S | 47 |
1 files changed, 47 insertions, 0 deletions
diff --git a/arch/arm/lib/copy_page.S b/arch/arm/lib/copy_page.S new file mode 100644 index 000000000..6ee2f6706 --- /dev/null +++ b/arch/arm/lib/copy_page.S @@ -0,0 +1,47 @@ +/* + * linux/arch/arm/lib/copypage.S + * + * Copyright (C) 1995-1999 Russell King + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * ASM optimised string functions + */ +#include <linux/linkage.h> +#include <asm/assembler.h> +#include <asm/asm-offsets.h> +#include <asm/cache.h> + +#define COPY_COUNT (PAGE_SZ / (2 * L1_CACHE_BYTES) PLD( -1 )) + + .text + .align 5 +/* + * StrongARM optimised copy_page routine + * now 1.78bytes/cycle, was 1.60 bytes/cycle (50MHz bus -> 89MB/s) + * Note that we probably achieve closer to the 100MB/s target with + * the core clock switching. + */ +ENTRY(copy_page) + stmfd sp!, {r4, lr} @ 2 + PLD( pld [r1, #0] ) + PLD( pld [r1, #L1_CACHE_BYTES] ) + mov r2, #COPY_COUNT @ 1 + ldmia r1!, {r3, r4, ip, lr} @ 4+1 +1: PLD( pld [r1, #2 * L1_CACHE_BYTES]) + PLD( pld [r1, #3 * L1_CACHE_BYTES]) +2: + .rept (2 * L1_CACHE_BYTES / 16 - 1) + stmia r0!, {r3, r4, ip, lr} @ 4 + ldmia r1!, {r3, r4, ip, lr} @ 4 + .endr + subs r2, r2, #1 @ 1 + stmia r0!, {r3, r4, ip, lr} @ 4 + ldmgtia r1!, {r3, r4, ip, lr} @ 4 + bgt 1b @ 1 + PLD( ldmeqia r1!, {r3, r4, ip, lr} ) + PLD( beq 2b ) + ldmfd sp!, {r4, pc} @ 3 +ENDPROC(copy_page) |