Messages in this thread Patch in this message | | | From | Christophe Leroy <> | Subject | [PATCH v2] powerpc/32: implement strlen() in assembly | Date | Mon, 28 May 2018 12:10:45 +0000 (UTC) |
| |
The generic implementation of strlen() reads strings byte per byte.
This patch implements strlen() in assembly for PPC32 based on a read of entire words, in the same spirit as what some other arches and glibc do.
For long strings, the time spent in strlen is reduced by 50-60%
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr> --- Applies on top of serie "[v5 0/3] powerpc/lib: Optimisation of memcmp() and __clear_user() for PPC32"
Changes in v2: - Moved handling of unaligned strings outside of the main path as it is very unlikely. - Removed the verification of the fourth byte in case none of the three first ones are NUL.
arch/powerpc/include/asm/string.h | 3 +++ arch/powerpc/lib/string_32.S | 41 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 44 insertions(+)
diff --git a/arch/powerpc/include/asm/string.h b/arch/powerpc/include/asm/string.h index 9b8cedf618f4..5ecfdb776f87 100644 --- a/arch/powerpc/include/asm/string.h +++ b/arch/powerpc/include/asm/string.h @@ -13,6 +13,9 @@ #define __HAVE_ARCH_MEMCHR #define __HAVE_ARCH_MEMSET16 #define __HAVE_ARCH_MEMCPY_FLUSHCACHE +#ifdef CONFIG_PPC32 +#define __HAVE_ARCH_STRLEN +#endif extern char * strcpy(char *,const char *); extern char * strncpy(char *,const char *, __kernel_size_t); diff --git a/arch/powerpc/lib/string_32.S b/arch/powerpc/lib/string_32.S index 4fbaa046aa84..69593c917b35 100644 --- a/arch/powerpc/lib/string_32.S +++ b/arch/powerpc/lib/string_32.S @@ -47,6 +47,47 @@ _GLOBAL(memcmp) blr EXPORT_SYMBOL(memcmp) +_GLOBAL(strlen) + andi. r9, r3, 3 + addi r10, r3, -4 + bne- 1f +2: lis r6, 0x8080 + ori r6, r6, 0x8080 /* r6 = 0x80808080 (himagic) */ + rlwinm r7, r6, 1, 0xffffffff /* r7 = 0x01010101 (lomagic) */ +3: lwzu r9, 4(r10) + /* ((x - lomagic) & ~x & himagic) == 0 means no byte in x is NUL */ + subf r8, r7, r9 + andc r11, r6, r9 + and. r8, r8, r11 + beq+ 3b + rlwinm. r8, r9, 0, 0xff000000 + beq 20f + rlwinm. r8, r9, 0, 0x00ff0000 + beq 21f + rlwinm. r8, r9, 0, 0x0000ff00 + beq 22f + subf r3, r3, r10 + addi r3, r3, 3 + blr +22: subf r3, r3, r10 + addi r3, r3, 2 + blr +21: subf r3, r3, r10 + addi r3, r3, 1 + blr +19: addi r10, r10, 3 +20: subf r3, r3, r10 + blr + +1: lbz r9, 4(r10) + addi r10, r10, 1 + cmpwi cr1, r9, 0 + andi. r9, r10, 3 + beq cr1, 19b + bne 1b + b 2b +EXPORT_SYMBOL(strlen) + CACHELINE_BYTES = L1_CACHE_BYTES LG_CACHELINE_BYTES = L1_CACHE_SHIFT CACHELINE_MASK = (L1_CACHE_BYTES-1) -- 2.13.3
| |