diff options
author | Ulrich Drepper <drepper@redhat.com> | 2004-12-22 20:10:10 +0000 |
---|---|---|
committer | Ulrich Drepper <drepper@redhat.com> | 2004-12-22 20:10:10 +0000 |
commit | a334319f6530564d22e775935d9c91663623a1b4 (patch) | |
tree | b5877475619e4c938e98757d518bb1e9cbead751 /sysdeps/mips/memcpy.S | |
parent | 0ecb606cb6cf65de1d9fc8a919bceb4be476c602 (diff) | |
download | glibc-a334319f6530564d22e775935d9c91663623a1b4.zip glibc-a334319f6530564d22e775935d9c91663623a1b4.tar.gz glibc-a334319f6530564d22e775935d9c91663623a1b4.tar.bz2 |
(CFLAGS-tst-align.c): Add -mpreferred-stack-boundary=4.
Diffstat (limited to 'sysdeps/mips/memcpy.S')
-rw-r--r-- | sysdeps/mips/memcpy.S | 136 |
1 files changed, 136 insertions, 0 deletions
diff --git a/sysdeps/mips/memcpy.S b/sysdeps/mips/memcpy.S new file mode 100644 index 0000000..2049d05 --- /dev/null +++ b/sysdeps/mips/memcpy.S @@ -0,0 +1,136 @@ +/* Copyright (C) 2002, 2003 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Hartvig Ekner <hartvige@mips.com>, 2002. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. */ + +#include <sysdep.h> +#include <endian.h> + + +/* void *memcpy(void *s1, const void *s2, size_t n); */ + +#if __BYTE_ORDER == __BIG_ENDIAN +# define LWHI lwl /* high part is left in big-endian */ +# define SWHI swl /* high part is left in big-endian */ +# define LWLO lwr /* low part is right in big-endian */ +# define SWLO swr /* low part is right in big-endian */ +#else +# define LWHI lwr /* high part is right in little-endian */ +# define SWHI swr /* high part is right in little-endian */ +# define LWLO lwl /* low part is left in little-endian */ +# define SWLO swl /* low part is left in little-endian */ +#endif + +ENTRY (memcpy) + .set noreorder + + slti t0, a2, 8 # Less than 8? + bne t0, zero, L(last8) + move v0, a0 # Setup exit value before too late + + xor t0, a1, a0 # Find a0/a1 displacement + andi t0, 0x3 + bne t0, zero, L(shift) # Go handle the unaligned case + subu t1, zero, a1 + andi t1, 0x3 # a0/a1 are aligned, but are we + beq t1, zero, L(chk8w) # starting in the middle of a word? + subu a2, t1 + LWHI t0, 0(a1) # Yes we are... take care of that + addu a1, t1 + SWHI t0, 0(a0) + addu a0, t1 + +L(chk8w): + andi t0, a2, 0x1f # 32 or more bytes left? + beq t0, a2, L(chk1w) + subu a3, a2, t0 # Yes + addu a3, a1 # a3 = end address of loop + move a2, t0 # a2 = what will be left after loop +L(lop8w): + lw t0, 0(a1) # Loop taking 8 words at a time + lw t1, 4(a1) + lw t2, 8(a1) + lw t3, 12(a1) + lw t4, 16(a1) + lw t5, 20(a1) + lw t6, 24(a1) + lw t7, 28(a1) + addiu a0, 32 + addiu a1, 32 + sw t0, -32(a0) + sw t1, -28(a0) + sw t2, -24(a0) + sw t3, -20(a0) + sw t4, -16(a0) + sw t5, -12(a0) + sw t6, -8(a0) + bne a1, a3, L(lop8w) + sw t7, -4(a0) + +L(chk1w): + andi t0, a2, 0x3 # 4 or more bytes left? + beq t0, a2, L(last8) + subu a3, a2, t0 # Yes, handle them one word at a time + addu a3, a1 # a3 again end address + move a2, t0 +L(lop1w): + lw t0, 0(a1) + addiu a0, 4 + addiu a1, 4 + bne a1, a3, L(lop1w) + sw t0, -4(a0) + +L(last8): + blez a2, L(lst8e) # Handle last 8 bytes, one at a time + addu a3, a2, a1 +L(lst8l): + lb t0, 0(a1) + addiu a0, 1 + addiu a1, 1 + bne a1, a3, L(lst8l) + sb t0, -1(a0) +L(lst8e): + jr ra # Bye, bye + nop + +L(shift): + subu a3, zero, a0 # Src and Dest unaligned + andi a3, 0x3 # (unoptimized case...) + beq a3, zero, L(shft1) + subu a2, a3 # a2 = bytes left + LWHI t0, 0(a1) # Take care of first odd part + LWLO t0, 3(a1) + addu a1, a3 + SWHI t0, 0(a0) + addu a0, a3 +L(shft1): + andi t0, a2, 0x3 + subu a3, a2, t0 + addu a3, a1 +L(shfth): + LWHI t1, 0(a1) # Limp through, word by word + LWLO t1, 3(a1) + addiu a0, 4 + addiu a1, 4 + bne a1, a3, L(shfth) + sw t1, -4(a0) + b L(last8) # Handle anything which may be left + move a2, t0 + + .set reorder +END (memcpy) +libc_hidden_builtin_def (memcpy) |