diff options
| author | Anton Blanchard via Libc-alpha <libc-alpha@sourceware.org> | 2020-05-14 09:08:35 +1000 |
|---|---|---|
| committer | Paul E. Murphy <murphyp@linux.vnet.ibm.com> | 2020-05-18 08:26:22 -0500 |
| commit | aa70d0563256b8ea053203177f756bca33b5cf37 (patch) | |
| tree | 152631b5f088b02e3a42f3e1f7defe63cefe33dd /sysdeps/powerpc/powerpc64/le/power9/strcpy.S | |
| parent | 39037048502d52ab6422c18f2d178d6228d2c7b9 (diff) | |
| download | glibc-aa70d0563256b8ea053203177f756bca33b5cf37.tar.gz glibc-aa70d0563256b8ea053203177f756bca33b5cf37.tar.xz glibc-aa70d0563256b8ea053203177f756bca33b5cf37.zip | |
powerpc: Optimized stpcpy for POWER9
Add stpcpy support to the POWER9 strcpy. This is up to 40% faster on
small strings and up to 90% faster on long relatively unaligned strings,
compared to the POWER8 version. A few examples:
__stpcpy_power9 __stpcpy_power8
Length 20, alignments in bytes 4/ 4: 2.58246 4.8788
Length 1024, alignments in bytes 1/ 6: 24.8186 47.8528
Diffstat (limited to 'sysdeps/powerpc/powerpc64/le/power9/strcpy.S')
| -rw-r--r-- | sysdeps/powerpc/powerpc64/le/power9/strcpy.S | 73 |
1 files changed, 58 insertions, 15 deletions
diff --git a/sysdeps/powerpc/powerpc64/le/power9/strcpy.S b/sysdeps/powerpc/powerpc64/le/power9/strcpy.S index 5749228054..ce8f503291 100644 --- a/sysdeps/powerpc/powerpc64/le/power9/strcpy.S +++ b/sysdeps/powerpc/powerpc64/le/power9/strcpy.S @@ -18,19 +18,35 @@ #include <sysdep.h> -#ifndef STRCPY -# define STRCPY strcpy -#endif +#ifdef USE_AS_STPCPY +# ifndef STPCPY +# define FUNC_NAME __stpcpy +# else +# define FUNC_NAME STPCPY +# endif +#else +# ifndef STRCPY +# define FUNC_NAME strcpy +# else +# define FUNC_NAME STRCPY +# endif +#endif /* !USE_AS_STPCPY */ /* Implements the function char * [r3] strcpy (char *dest [r3], const char *src [r4]) + or + + char * [r3] stpcpy (char *dest [r3], const char *src [r4]) + + if USE_AS_STPCPY is defined. + The implementation can load bytes past a null terminator, but only up to the next 16B boundary, so it never crosses a page. */ .machine power9 -ENTRY_TOCLESS (STRCPY, 4) +ENTRY_TOCLESS (FUNC_NAME, 4) CALL_MCOUNT 2 /* NULL string optimisation */ @@ -53,8 +69,8 @@ ENTRY_TOCLESS (STRCPY, 4) vperm v0,v18,v0,v1 vcmpequb v6,v0,v18 /* 0xff if byte is NULL, 0x00 otherwise */ - vctzlsbb r8,v6 /* Number of trailing zeroes */ - addi r8,r8,1 /* Add null terminator */ + vctzlsbb r7,v6 /* Number of trailing zeroes */ + addi r8,r7,1 /* Add null terminator */ /* r8 = bytes including null r9 = bytes to get source 16B aligned @@ -68,6 +84,11 @@ ENTRY_TOCLESS (STRCPY, 4) sldi r10,r8,56 /* stxvl wants size in top 8 bits */ stxvl 32+v0,r11,r10 /* Partial store */ +#ifdef USE_AS_STPCPY + /* stpcpy returns the dest address plus the size not counting the + final '\0'. */ + add r3,r11,r7 +#endif blr L(no_null): @@ -106,28 +127,43 @@ L(loop): L(tail1): vctzlsbb r8,v6 - addi r8,r8,1 - sldi r9,r8,56 /* stxvl wants size in top 8 bits */ + addi r9,r8,1 + sldi r9,r9,56 /* stxvl wants size in top 8 bits */ stxvl 32+v0,r11,r9 +#ifdef USE_AS_STPCPY + /* stpcpy returns the dest address plus the size not counting the + final '\0'. */ + add r3,r11,r8 +#endif blr L(tail2): stxv 32+v0,0(r11) vctzlsbb r8,v6 /* Number of trailing zeroes */ - addi r8,r8,1 /* Add null terminator */ - sldi r10,r8,56 /* stxvl wants size in top 8 bits */ + addi r9,r8,1 /* Add null terminator */ + sldi r10,r9,56 /* stxvl wants size in top 8 bits */ addi r11,r11,16 stxvl 32+v1,r11,r10 /* Partial store */ +#ifdef USE_AS_STPCPY + /* stpcpy returns the dest address plus the size not counting the + final '\0'. */ + add r3,r11,r8 +#endif blr L(tail3): stxv 32+v0,0(r11) stxv 32+v1,16(r11) vctzlsbb r8,v6 /* Number of trailing zeroes */ - addi r8,r8,1 /* Add null terminator */ - sldi r10,r8,56 /* stxvl wants size in top 8 bits */ + addi r9,r8,1 /* Add null terminator */ + sldi r10,r9,56 /* stxvl wants size in top 8 bits */ addi r11,r11,32 stxvl 32+v2,r11,r10 /* Partial store */ +#ifdef USE_AS_STPCPY + /* stpcpy returns the dest address plus the size not counting the + final '\0'. */ + add r3,r11,r8 +#endif blr L(tail4): @@ -135,10 +171,17 @@ L(tail4): stxv 32+v1,16(r11) stxv 32+v2,32(r11) vctzlsbb r8,v6 /* Number of trailing zeroes */ - addi r8,r8,1 /* Add null terminator */ - sldi r10,r8,56 /* stxvl wants size in top 8 bits */ + addi r9,r8,1 /* Add null terminator */ + sldi r10,r9,56 /* stxvl wants size in top 8 bits */ addi r11,r11,48 stxvl 32+v3,r11,r10 /* Partial store */ +#ifdef USE_AS_STPCPY + /* stpcpy returns the dest address plus the size not counting the + final '\0'. */ + add r3,r11,r8 +#endif blr -END (STRCPY) +END (FUNC_NAME) +#ifndef USE_AS_STPCPY libc_hidden_builtin_def (strcpy) +#endif |