From: Alexandre Oliva Date: Fri, 21 Mar 2003 19:44:24 +0000 (+0000) Subject: * sysdeps/mips/mips64/memcpy.S, sysdeps/mips/mips64/memset.S: New. * sysdeps/mips... X-Git-Tag: upstream/2.30~19658 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=ee7b703635bf9c2df1c064a303eceda80d579b2d;p=external%2Fglibc.git * sysdeps/mips/mips64/memcpy.S, sysdeps/mips/mips64/memset.S: New. * sysdeps/mips/memcpy.S, sysdeps/mips/memset.S: Update comments. 2003-03-21 Alexandre Oliva * sysdeps/mips/mips64/memcpy.S, sysdeps/mips/mips64/memset.S: New. * sysdeps/mips/memcpy.S, sysdeps/mips/memset.S: Update comments. --- diff --git a/ChangeLog b/ChangeLog index 6bcdb22..aadb507 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,8 @@ +2003-03-21 Alexandre Oliva + + * sysdeps/mips/mips64/memcpy.S, sysdeps/mips/mips64/memset.S: New. + * sysdeps/mips/memcpy.S, sysdeps/mips/memset.S: Update comments. + 2003-03-21 Roland McGrath * sysdeps/i386/i486/bits/atomic.h diff --git a/sysdeps/mips/memcpy.S b/sysdeps/mips/memcpy.S index c77f1b8..1e9130f 100644 --- a/sysdeps/mips/memcpy.S +++ b/sysdeps/mips/memcpy.S @@ -21,12 +21,7 @@ #include -/* void *memcpy(void *s1, const void *s2, size_t n); - - This routine could be optimized further for MIPS64, but this is left - as an exercise for the future. When it is done, the file should be kept - as a sisterfile to this one, and placed in the sysdeps/mips/mips64 - directory. */ +/* 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 */ @@ -40,19 +35,6 @@ # define SWLO swl /* low part is left in little-endian */ #endif -#ifndef t0 -# define t0 a4 -#endif -#ifndef t1 -# define t1 a5 -#endif -#ifndef t2 -# define t2 a6 -#endif -#ifndef t3 -# define t3 a7 -#endif - ENTRY (memcpy) .set noreorder diff --git a/sysdeps/mips/memset.S b/sysdeps/mips/memset.S index 4681134..b372d29 100644 --- a/sysdeps/mips/memset.S +++ b/sysdeps/mips/memset.S @@ -21,12 +21,7 @@ #include -/* void *memset(void *s, int c, size_t n). - - This routine could be optimized further for MIPS64, but this is left - as an exercise for the future. When it is done, the file should be kept - as a sisterfile to this one, and placed in the sysdeps/mips/mips64 - directory. */ +/* void *memset(void *s, int c, size_t n). */ #if __BYTE_ORDER == __BIG_ENDIAN # define SWHI swl /* high part is left in big-endian */ @@ -34,13 +29,6 @@ # define SWHI swr /* high part is right in little-endian */ #endif -#ifndef t0 -# define t0 a4 -#endif -#ifndef t1 -# define t1 a5 -#endif - ENTRY (memset) .set noreorder diff --git a/sysdeps/mips/mips64/memcpy.S b/sysdeps/mips/mips64/memcpy.S new file mode 100644 index 0000000..3dbb31f --- /dev/null +++ b/sysdeps/mips/mips64/memcpy.S @@ -0,0 +1,139 @@ +/* Copyright (C) 2002, 2003 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Hartvig Ekner , 2002. + Ported to mips3 n32/n64 by Alexandre Oliva + + 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 +#include +#include + + +/* void *memcpy(void *s1, const void *s2, size_t n); + + This could probably be optimized further. */ + +#if __BYTE_ORDER == __BIG_ENDIAN +# define LDHI ldl /* high part is left in big-endian */ +# define SDHI sdl /* high part is left in big-endian */ +# define LDLO ldr /* low part is right in big-endian */ +# define SDLO sdr /* low part is right in big-endian */ +#else +# define LDHI ldr /* high part is right in little-endian */ +# define SDHI sdr /* high part is right in little-endian */ +# define LDLO ldl /* low part is left in little-endian */ +# define SDLO sdl /* low part is left in little-endian */ +#endif + +ENTRY (memcpy) + .set noreorder + + slti a4, a2, 8 # Less than 8? + bne a4, zero, L(last8) + move v0, a0 # Setup exit value before too late + + xor a4, a1, a0 # Find a0/a1 displacement + andi a4, 0x7 + bne a4, zero, L(shift) # Go handle the unaligned case + PTR_SUBU a5, zero, a1 + andi a5, 0x7 # a0/a1 are aligned, but are we + beq a5, zero, L(chk8w) # starting in the middle of a word? + PTR_SUBU a2, a5 + LDHI a4, 0(a1) # Yes we are... take care of that + PTR_ADDU a1, a5 + SDHI a4, 0(a0) + PTR_ADDU a0, a5 + +L(chk8w): + andi a4, a2, 0x3f # 64 or more bytes left? + beq a4, a2, L(chk1w) + PTR_SUBU a3, a2, a4 # Yes + PTR_ADDU a3, a1 # a3 = end address of loop + move a2, a4 # a2 = what will be left after loop +L(lop8w): + ld a4, 0(a1) # Loop taking 8 words at a time + ld a5, 8(a1) + ld a6, 16(a1) + ld a7, 24(a1) + ld t4, 32(a1) + ld t5, 40(a1) + ld t6, 48(a1) + ld t7, 56(a1) + PTR_ADDIU a0, 64 + PTR_ADDIU a1, 64 + sd a4, -64(a0) + sd a5, -56(a0) + sd a6, -48(a0) + sd a7, -40(a0) + sd t4, -32(a0) + sd t5, -24(a0) + sd t6, -16(a0) + bne a1, a3, L(lop8w) + sd t7, -8(a0) + +L(chk1w): + andi a4, a2, 0x7 # 4 or more bytes left? + beq a4, a2, L(last8) + PTR_SUBU a3, a2, a4 # Yes, handle them one word at a time + PTR_ADDU a3, a1 # a3 again end address + move a2, a4 +L(lop1w): + ld a4, 0(a1) + PTR_ADDIU a0, 8 + PTR_ADDIU a1, 8 + bne a1, a3, L(lop1w) + sd a4, -8(a0) + +L(last8): + blez a2, L(lst8e) # Handle last 8 bytes, one at a time + PTR_ADDU a3, a2, a1 +L(lst8l): + lb a4, 0(a1) + PTR_ADDIU a0, 1 + PTR_ADDIU a1, 1 + bne a1, a3, L(lst8l) + sb a4, -1(a0) +L(lst8e): + jr ra # Bye, bye + nop + +L(shift): + PTR_SUBU a3, zero, a0 # Src and Dest unaligned + andi a3, 0x7 # (unoptimized case...) + beq a3, zero, L(shfa5) + PTR_SUBU a2, a3 # a2 = bytes left + LDHI a4, 0(a1) # Take care of first odd part + LDLO a4, 7(a1) + PTR_ADDU a1, a3 + SDHI a4, 0(a0) + PTR_ADDU a0, a3 +L(shfa5): + andi a4, a2, 0x7 + PTR_SUBU a3, a2, a4 + PTR_ADDU a3, a1 +L(shfth): + LDHI a5, 0(a1) # Limp through, word by word + LDLO a5, 7(a1) + PTR_ADDIU a0, 8 + PTR_ADDIU a1, 8 + bne a1, a3, L(shfth) + sd a5, -8(a0) + b L(last8) # Handle anything which may be left + move a2, a4 + + .set reorder +END (memcpy) diff --git a/sysdeps/mips/mips64/memset.S b/sysdeps/mips/mips64/memset.S new file mode 100644 index 0000000..6a3b154 --- /dev/null +++ b/sysdeps/mips/mips64/memset.S @@ -0,0 +1,91 @@ +/* Copyright (C) 2002, 2003 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Hartvig Ekner , 2002. + Ported to mips3 n32/n64 by Alexandre Oliva + + 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 +#include +#include + + +/* void *memset(void *s, int c, size_t n); + + This could probably be optimized further. */ + +#if __BYTE_ORDER == __BIG_ENDIAN +# define SDHI sdl /* high part is left in big-endian */ +#else +# define SDHI sdr /* high part is right in little-endian */ +#endif + +ENTRY (memset) + .set noreorder + + slti t5, a2, 8 # Less than 8? + bne t5, zero, L(last8) + move v0, a0 # Setup exit value before too late + + beq a1, zero, L(ueven) # If zero pattern, no need to extend + andi a1, 0xff # Avoid problems with bogus arguments + dsll t4, a1, 8 + or a1, t4 + dsll t4, a1, 16 + or a1, t4 # a1 is now pattern in full word + dsll t4, a1, 32 + or a1, t4 # a1 is now pattern in double word + +L(ueven): + PTR_SUBU t4, zero, a0 # Unaligned address? + andi t4, 0x7 + beq t4, zero, L(chkw) + PTR_SUBU a2, t4 + SDHI a1, 0(a0) # Yes, handle first unaligned part + PTR_ADDU a0, t4 # Now both a0 and a2 are updated + +L(chkw): + andi t4, a2, 0xf # Enough left for one loop iteration? + beq t4, a2, L(chkl) + PTR_SUBU a3, a2, t4 + PTR_ADDU a3, a0 # a3 is last loop address +1 + move a2, t4 # a2 is now # of bytes left after loop +L(loopw): + PTR_ADDIU a0, 16 # Handle 2 words pr. iteration + sd a1, -16(a0) + bne a0, a3, L(loopw) + sd a1, -8(a0) + +L(chkl): + andi t4, a2, 0x8 # Check if there is at least a double + beq t4, zero, L(last8) # word remaining after the loop + PTR_SUBU a2, t4 + sd a1, 0(a0) # Yes... + PTR_ADDIU a0, 8 + +L(last8): + blez a2, L(exit) # Handle last 8 bytes (if cnt>0) + PTR_ADDU a3, a2, a0 # a3 is last address +1 +L(lst8l): + PTR_ADDIU a0, 1 + bne a0, a3, L(lst8l) + sb a1, -1(a0) +L(exit): + j ra # Bye, bye + nop + + .set reorder +END (memset)