diff options
Diffstat (limited to 'arch/mips/lib')
-rw-r--r-- | arch/mips/lib/Makefile | 2 | ||||
-rw-r--r-- | arch/mips/lib/csum_partial.S | 735 | ||||
-rw-r--r-- | arch/mips/lib/csum_partial_copy.c | 49 | ||||
-rw-r--r-- | arch/mips/lib/memset.S | 166 | ||||
-rw-r--r-- | arch/mips/lib/uncached.c | 4 |
5 files changed, 767 insertions, 189 deletions
diff --git a/arch/mips/lib/Makefile b/arch/mips/lib/Makefile index 888b61ea12fe..5ad501b30b43 100644 --- a/arch/mips/lib/Makefile +++ b/arch/mips/lib/Makefile @@ -2,7 +2,7 @@ # Makefile for MIPS-specific library files.. # -lib-y += csum_partial.o csum_partial_copy.o memcpy.o promlib.o \ +lib-y += csum_partial.o memcpy.o memset.o promlib.o \ strlen_user.o strncpy_user.o strnlen_user.o uncached.o obj-y += iomap.o diff --git a/arch/mips/lib/csum_partial.S b/arch/mips/lib/csum_partial.S index 15611d9df7ac..c0a77fe038be 100644 --- a/arch/mips/lib/csum_partial.S +++ b/arch/mips/lib/csum_partial.S @@ -8,47 +8,72 @@ * Copyright (C) 1998, 1999 Ralf Baechle * Copyright (C) 1999 Silicon Graphics, Inc. */ +#include <linux/errno.h> #include <asm/asm.h> +#include <asm/asm-offsets.h> #include <asm/regdef.h> #ifdef CONFIG_64BIT -#define T0 ta0 -#define T1 ta1 -#define T2 ta2 -#define T3 ta3 -#define T4 t0 -#define T7 t3 -#else -#define T0 t0 -#define T1 t1 -#define T2 t2 -#define T3 t3 -#define T4 t4 -#define T7 t7 +/* + * As we are sharing code base with the mips32 tree (which use the o32 ABI + * register definitions). We need to redefine the register definitions from + * the n64 ABI register naming to the o32 ABI register naming. + */ +#undef t0 +#undef t1 +#undef t2 +#undef t3 +#define t0 $8 +#define t1 $9 +#define t2 $10 +#define t3 $11 +#define t4 $12 +#define t5 $13 +#define t6 $14 +#define t7 $15 + +#define USE_DOUBLE #endif +#ifdef USE_DOUBLE + +#define LOAD ld +#define ADD daddu +#define NBYTES 8 + +#else + +#define LOAD lw +#define ADD addu +#define NBYTES 4 + +#endif /* USE_DOUBLE */ + +#define UNIT(unit) ((unit)*NBYTES) + #define ADDC(sum,reg) \ - addu sum, reg; \ + ADD sum, reg; \ sltu v1, sum, reg; \ - addu sum, v1 + ADD sum, v1 -#define CSUM_BIGCHUNK(src, offset, sum, _t0, _t1, _t2, _t3) \ - lw _t0, (offset + 0x00)(src); \ - lw _t1, (offset + 0x04)(src); \ - lw _t2, (offset + 0x08)(src); \ - lw _t3, (offset + 0x0c)(src); \ +#define CSUM_BIGCHUNK1(src, offset, sum, _t0, _t1, _t2, _t3) \ + LOAD _t0, (offset + UNIT(0))(src); \ + LOAD _t1, (offset + UNIT(1))(src); \ + LOAD _t2, (offset + UNIT(2))(src); \ + LOAD _t3, (offset + UNIT(3))(src); \ ADDC(sum, _t0); \ ADDC(sum, _t1); \ ADDC(sum, _t2); \ - ADDC(sum, _t3); \ - lw _t0, (offset + 0x10)(src); \ - lw _t1, (offset + 0x14)(src); \ - lw _t2, (offset + 0x18)(src); \ - lw _t3, (offset + 0x1c)(src); \ - ADDC(sum, _t0); \ - ADDC(sum, _t1); \ - ADDC(sum, _t2); \ - ADDC(sum, _t3); \ + ADDC(sum, _t3) + +#ifdef USE_DOUBLE +#define CSUM_BIGCHUNK(src, offset, sum, _t0, _t1, _t2, _t3) \ + CSUM_BIGCHUNK1(src, offset, sum, _t0, _t1, _t2, _t3) +#else +#define CSUM_BIGCHUNK(src, offset, sum, _t0, _t1, _t2, _t3) \ + CSUM_BIGCHUNK1(src, offset, sum, _t0, _t1, _t2, _t3); \ + CSUM_BIGCHUNK1(src, offset + 0x10, sum, _t0, _t1, _t2, _t3) +#endif /* * a0: source address @@ -61,86 +86,27 @@ .text .set noreorder - -/* unknown src alignment and < 8 bytes to go */ -small_csumcpy: - move a1, T2 - - andi T0, a1, 4 - beqz T0, 1f - andi T0, a1, 2 - - /* Still a full word to go */ - ulw T1, (src) - PTR_ADDIU src, 4 - ADDC(sum, T1) - -1: move T1, zero - beqz T0, 1f - andi T0, a1, 1 - - /* Still a halfword to go */ - ulhu T1, (src) - PTR_ADDIU src, 2 - -1: beqz T0, 1f - sll T1, T1, 16 - - lbu T2, (src) - nop - -#ifdef __MIPSEB__ - sll T2, T2, 8 -#endif - or T1, T2 - -1: ADDC(sum, T1) - - /* fold checksum */ - sll v1, sum, 16 - addu sum, v1 - sltu v1, sum, v1 - srl sum, sum, 16 - addu sum, v1 - - /* odd buffer alignment? */ - beqz T7, 1f - nop - sll v1, sum, 8 - srl sum, sum, 8 - or sum, v1 - andi sum, 0xffff -1: - .set reorder - /* Add the passed partial csum. */ - ADDC(sum, a2) - jr ra - .set noreorder - -/* ------------------------------------------------------------------------- */ - .align 5 LEAF(csum_partial) move sum, zero - move T7, zero + move t7, zero sltiu t8, a1, 0x8 bnez t8, small_csumcpy /* < 8 bytes to copy */ - move T2, a1 + move t2, a1 - beqz a1, out - andi T7, src, 0x1 /* odd buffer? */ + andi t7, src, 0x1 /* odd buffer? */ hword_align: - beqz T7, word_align + beqz t7, word_align andi t8, src, 0x2 - lbu T0, (src) + lbu t0, (src) LONG_SUBU a1, a1, 0x1 #ifdef __MIPSEL__ - sll T0, T0, 8 + sll t0, t0, 8 #endif - ADDC(sum, T0) + ADDC(sum, t0) PTR_ADDU src, src, 0x1 andi t8, src, 0x2 @@ -148,9 +114,9 @@ word_align: beqz t8, dword_align sltiu t8, a1, 56 - lhu T0, (src) + lhu t0, (src) LONG_SUBU a1, a1, 0x2 - ADDC(sum, T0) + ADDC(sum, t0) sltiu t8, a1, 56 PTR_ADDU src, src, 0x2 @@ -162,9 +128,9 @@ dword_align: beqz t8, qword_align andi t8, src, 0x8 - lw T0, 0x00(src) + lw t0, 0x00(src) LONG_SUBU a1, a1, 0x4 - ADDC(sum, T0) + ADDC(sum, t0) PTR_ADDU src, src, 0x4 andi t8, src, 0x8 @@ -172,11 +138,17 @@ qword_align: beqz t8, oword_align andi t8, src, 0x10 - lw T0, 0x00(src) - lw T1, 0x04(src) +#ifdef USE_DOUBLE + ld t0, 0x00(src) LONG_SUBU a1, a1, 0x8 - ADDC(sum, T0) - ADDC(sum, T1) + ADDC(sum, t0) +#else + lw t0, 0x00(src) + lw t1, 0x04(src) + LONG_SUBU a1, a1, 0x8 + ADDC(sum, t0) + ADDC(sum, t1) +#endif PTR_ADDU src, src, 0x8 andi t8, src, 0x10 @@ -184,75 +156,560 @@ oword_align: beqz t8, begin_movement LONG_SRL t8, a1, 0x7 - lw T3, 0x08(src) - lw T4, 0x0c(src) - lw T0, 0x00(src) - lw T1, 0x04(src) - ADDC(sum, T3) - ADDC(sum, T4) - ADDC(sum, T0) - ADDC(sum, T1) +#ifdef USE_DOUBLE + ld t0, 0x00(src) + ld t1, 0x08(src) + ADDC(sum, t0) + ADDC(sum, t1) +#else + CSUM_BIGCHUNK1(src, 0x00, sum, t0, t1, t3, t4) +#endif LONG_SUBU a1, a1, 0x10 PTR_ADDU src, src, 0x10 LONG_SRL t8, a1, 0x7 begin_movement: beqz t8, 1f - andi T2, a1, 0x40 + andi t2, a1, 0x40 move_128bytes: - CSUM_BIGCHUNK(src, 0x00, sum, T0, T1, T3, T4) - CSUM_BIGCHUNK(src, 0x20, sum, T0, T1, T3, T4) - CSUM_BIGCHUNK(src, 0x40, sum, T0, T1, T3, T4) - CSUM_BIGCHUNK(src, 0x60, sum, T0, T1, T3, T4) + CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4) + CSUM_BIGCHUNK(src, 0x20, sum, t0, t1, t3, t4) + CSUM_BIGCHUNK(src, 0x40, sum, t0, t1, t3, t4) + CSUM_BIGCHUNK(src, 0x60, sum, t0, t1, t3, t4) LONG_SUBU t8, t8, 0x01 bnez t8, move_128bytes PTR_ADDU src, src, 0x80 1: - beqz T2, 1f - andi T2, a1, 0x20 + beqz t2, 1f + andi t2, a1, 0x20 move_64bytes: - CSUM_BIGCHUNK(src, 0x00, sum, T0, T1, T3, T4) - CSUM_BIGCHUNK(src, 0x20, sum, T0, T1, T3, T4) + CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4) + CSUM_BIGCHUNK(src, 0x20, sum, t0, t1, t3, t4) PTR_ADDU src, src, 0x40 1: - beqz T2, do_end_words + beqz t2, do_end_words andi t8, a1, 0x1c move_32bytes: - CSUM_BIGCHUNK(src, 0x00, sum, T0, T1, T3, T4) + CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4) andi t8, a1, 0x1c PTR_ADDU src, src, 0x20 do_end_words: - beqz t8, maybe_end_cruft - LONG_SRL t8, t8, 0x2 + beqz t8, small_csumcpy + andi t2, a1, 0x3 + LONG_SRL t8, t8, 0x2 end_words: - lw T0, (src) + lw t0, (src) LONG_SUBU t8, t8, 0x1 - ADDC(sum, T0) + ADDC(sum, t0) bnez t8, end_words PTR_ADDU src, src, 0x4 -maybe_end_cruft: - andi T2, a1, 0x3 +/* unknown src alignment and < 8 bytes to go */ +small_csumcpy: + move a1, t2 -small_memcpy: - j small_csumcpy; move a1, T2 /* XXX ??? */ - beqz t2, out - move a1, T2 + andi t0, a1, 4 + beqz t0, 1f + andi t0, a1, 2 -end_bytes: - lb T0, (src) - LONG_SUBU a1, a1, 0x1 - bnez a2, end_bytes - PTR_ADDU src, src, 0x1 + /* Still a full word to go */ + ulw t1, (src) + PTR_ADDIU src, 4 + ADDC(sum, t1) + +1: move t1, zero + beqz t0, 1f + andi t0, a1, 1 + + /* Still a halfword to go */ + ulhu t1, (src) + PTR_ADDIU src, 2 -out: +1: beqz t0, 1f + sll t1, t1, 16 + + lbu t2, (src) + nop + +#ifdef __MIPSEB__ + sll t2, t2, 8 +#endif + or t1, t2 + +1: ADDC(sum, t1) + + /* fold checksum */ +#ifdef USE_DOUBLE + dsll32 v1, sum, 0 + daddu sum, v1 + sltu v1, sum, v1 + dsra32 sum, sum, 0 + addu sum, v1 +#endif + sll v1, sum, 16 + addu sum, v1 + sltu v1, sum, v1 + srl sum, sum, 16 + addu sum, v1 + + /* odd buffer alignment? */ + beqz t7, 1f + nop + sll v1, sum, 8 + srl sum, sum, 8 + or sum, v1 + andi sum, 0xffff +1: + .set reorder + /* Add the passed partial csum. */ + ADDC(sum, a2) jr ra - move v0, sum + .set noreorder END(csum_partial) + + +/* + * checksum and copy routines based on memcpy.S + * + * csum_partial_copy_nocheck(src, dst, len, sum) + * __csum_partial_copy_user(src, dst, len, sum, errp) + * + * See "Spec" in memcpy.S for details. Unlike __copy_user, all + * function in this file use the standard calling convention. + */ + +#define src a0 +#define dst a1 +#define len a2 +#define psum a3 +#define sum v0 +#define odd t8 +#define errptr t9 + +/* + * The exception handler for loads requires that: + * 1- AT contain the address of the byte just past the end of the source + * of the copy, + * 2- src_entry <= src < AT, and + * 3- (dst - src) == (dst_entry - src_entry), + * The _entry suffix denotes values when __copy_user was called. + * + * (1) is set up up by __csum_partial_copy_from_user and maintained by + * not writing AT in __csum_partial_copy + * (2) is met by incrementing src by the number of bytes copied + * (3) is met by not doing loads between a pair of increments of dst and src + * + * The exception handlers for stores stores -EFAULT to errptr and return. + * These handlers do not need to overwrite any data. + */ + +#define EXC(inst_reg,addr,handler) \ +9: inst_reg, addr; \ + .section __ex_table,"a"; \ + PTR 9b, handler; \ + .previous + +#ifdef USE_DOUBLE + +#define LOAD ld +#define LOADL ldl +#define LOADR ldr +#define STOREL sdl +#define STORER sdr +#define STORE sd +#define ADD daddu +#define SUB dsubu +#define SRL dsrl +#define SLL dsll +#define SLLV dsllv +#define SRLV dsrlv +#define NBYTES 8 +#define LOG_NBYTES 3 + +#else + +#define LOAD lw +#define LOADL lwl +#define LOADR lwr +#define STOREL swl +#define STORER swr +#define STORE sw +#define ADD addu +#define SUB subu +#define SRL srl +#define SLL sll +#define SLLV sllv +#define SRLV srlv +#define NBYTES 4 +#define LOG_NBYTES 2 + +#endif /* USE_DOUBLE */ + +#ifdef CONFIG_CPU_LITTLE_ENDIAN +#define LDFIRST LOADR +#define LDREST LOADL +#define STFIRST STORER +#define STREST STOREL +#define SHIFT_DISCARD SLLV +#define SHIFT_DISCARD_REVERT SRLV +#else +#define LDFIRST LOADL +#define LDREST LOADR +#define STFIRST STOREL +#define STREST STORER +#define SHIFT_DISCARD SRLV +#define SHIFT_DISCARD_REVERT SLLV +#endif + +#define FIRST(unit) ((unit)*NBYTES) +#define REST(unit) (FIRST(unit)+NBYTES-1) + +#define ADDRMASK (NBYTES-1) + + .set noat + +LEAF(__csum_partial_copy_user) + PTR_ADDU AT, src, len /* See (1) above. */ +#ifdef CONFIG_64BIT + move errptr, a4 +#else + lw errptr, 16(sp) +#endif +FEXPORT(csum_partial_copy_nocheck) + move sum, zero + move odd, zero + /* + * Note: dst & src may be unaligned, len may be 0 + * Temps + */ + /* + * The "issue break"s below are very approximate. + * Issue delays for dcache fills will perturb the schedule, as will + * load queue full replay traps, etc. + * + * If len < NBYTES use byte operations. + */ + sltu t2, len, NBYTES + and t1, dst, ADDRMASK + bnez t2, copy_bytes_checklen + and t0, src, ADDRMASK + andi odd, dst, 0x1 /* odd buffer? */ + bnez t1, dst_unaligned + nop + bnez t0, src_unaligned_dst_aligned + /* + * use delay slot for fall-through + * src and dst are aligned; need to compute rem + */ +both_aligned: + SRL t0, len, LOG_NBYTES+3 # +3 for 8 units/iter + beqz t0, cleanup_both_aligned # len < 8*NBYTES + nop + SUB len, 8*NBYTES # subtract here for bgez loop + .align 4 +1: +EXC( LOAD t0, UNIT(0)(src), l_exc) +EXC( LOAD t1, UNIT(1)(src), l_exc_copy) +EXC( LOAD t2, UNIT(2)(src), l_exc_copy) +EXC( LOAD t3, UNIT(3)(src), l_exc_copy) +EXC( LOAD t4, UNIT(4)(src), l_exc_copy) +EXC( LOAD t5, UNIT(5)(src), l_exc_copy) +EXC( LOAD t6, UNIT(6)(src), l_exc_copy) +EXC( LOAD t7, UNIT(7)(src), l_exc_copy) + SUB len, len, 8*NBYTES + ADD src, src, 8*NBYTES +EXC( STORE t0, UNIT(0)(dst), s_exc) + ADDC(sum, t0) +EXC( STORE t1, UNIT(1)(dst), s_exc) + ADDC(sum, t1) +EXC( STORE t2, UNIT(2)(dst), s_exc) + ADDC(sum, t2) +EXC( STORE t3, UNIT(3)(dst), s_exc) + ADDC(sum, t3) +EXC( STORE t4, UNIT(4)(dst), s_exc) + ADDC(sum, t4) +EXC( STORE t5, UNIT(5)(dst), s_exc) + ADDC(sum, t5) +EXC( STORE t6, UNIT(6)(dst), s_exc) + ADDC(sum, t6) +EXC( STORE t7, UNIT(7)(dst), s_exc) + ADDC(sum, t7) + bgez len, 1b + ADD dst, dst, 8*NBYTES + ADD len, 8*NBYTES # revert len (see above) + + /* + * len == the number of bytes left to copy < 8*NBYTES + */ +cleanup_both_aligned: +#define rem t7 + beqz len, done + sltu t0, len, 4*NBYTES + bnez t0, less_than_4units + and rem, len, (NBYTES-1) # rem = len % NBYTES + /* + * len >= 4*NBYTES + */ +EXC( LOAD t0, UNIT(0)(src), l_exc) +EXC( LOAD t1, UNIT(1)(src), l_exc_copy) +EXC( LOAD t2, UNIT(2)(src), l_exc_copy) +EXC( LOAD t3, UNIT(3)(src), l_exc_copy) + SUB len, len, 4*NBYTES + ADD src, src, 4*NBYTES +EXC( STORE t0, UNIT(0)(dst), s_exc) + ADDC(sum, t0) +EXC( STORE t1, UNIT(1)(dst), s_exc) + ADDC(sum, t1) +EXC( STORE t2, UNIT(2)(dst), s_exc) + ADDC(sum, t2) +EXC( STORE t3, UNIT(3)(dst), s_exc) + ADDC(sum, t3) + beqz len, done + ADD dst, dst, 4*NBYTES +less_than_4units: + /* + * rem = len % NBYTES + */ + beq rem, len, copy_bytes + nop +1: +EXC( LOAD t0, 0(src), l_exc) + ADD src, src, NBYTES + SUB len, len, NBYTES +EXC( STORE t0, 0(dst), s_exc) + ADDC(sum, t0) + bne rem, len, 1b + ADD dst, dst, NBYTES + + /* + * src and dst are aligned, need to copy rem bytes (rem < NBYTES) + * A loop would do only a byte at a time with possible branch + * mispredicts. Can't do an explicit LOAD dst,mask,or,STORE + * because can't assume read-access to dst. Instead, use + * STREST dst, which doesn't require read access to dst. + * + * This code should perform better than a simple loop on modern, + * wide-issue mips processors because the code has fewer branches and + * more instruction-level parallelism. + */ +#define bits t2 + beqz len, done + ADD t1, dst, len # t1 is just past last byte of dst + li bits, 8*NBYTES + SLL rem, len, 3 # rem = number of bits to keep +EXC( LOAD t0, 0(src), l_exc) + SUB bits, bits, rem # bits = number of bits to discard + SHIFT_DISCARD t0, t0, bits +EXC( STREST t0, -1(t1), s_exc) + SHIFT_DISCARD_REVERT t0, t0, bits + .set reorder + ADDC(sum, t0) + b done + .set noreorder +dst_unaligned: + /* + * dst is unaligned + * t0 = src & ADDRMASK + * t1 = dst & ADDRMASK; T1 > 0 + * len >= NBYTES + * + * Copy enough bytes to align dst + * Set match = (src and dst have same alignment) + */ +#define match rem +EXC( LDFIRST t3, FIRST(0)(src), l_exc) + ADD t2, zero, NBYTES +EXC( LDREST t3, REST(0)(src), l_exc_copy) + SUB t2, t2, t1 # t2 = number of bytes copied + xor match, t0, t1 +EXC( STFIRST t3, FIRST(0)(dst), s_exc) + SLL t4, t1, 3 # t4 = number of bits to discard + SHIFT_DISCARD t3, t3, t4 + /* no SHIFT_DISCARD_REVERT to handle odd buffer properly */ + ADDC(sum, t3) + beq len, t2, done + SUB len, len, t2 + ADD dst, dst, t2 + beqz match, both_aligned + ADD src, src, t2 + +src_unaligned_dst_aligned: + SRL t0, len, LOG_NBYTES+2 # +2 for 4 units/iter + beqz t0, cleanup_src_unaligned + and rem, len, (4*NBYTES-1) # rem = len % 4*NBYTES +1: +/* + * Avoid consecutive LD*'s to the same register since some mips + * implementations can't issue them in the same cycle. + * It's OK to load FIRST(N+1) before REST(N) because the two addresses + * are to the same unit (unless src is aligned, but it's not). + */ +EXC( LDFIRST t0, FIRST(0)(src), l_exc) +EXC( LDFIRST t1, FIRST(1)(src), l_exc_copy) + SUB len, len, 4*NBYTES +EXC( LDREST t0, REST(0)(src), l_exc_copy) +EXC( LDREST t1, REST(1)(src), l_exc_copy) +EXC( LDFIRST t2, FIRST(2)(src), l_exc_copy) +EXC( LDFIRST t3, FIRST(3)(src), l_exc_copy) +EXC( LDREST t2, REST(2)(src), l_exc_copy) +EXC( LDREST t3, REST(3)(src), l_exc_copy) + ADD src, src, 4*NBYTES +#ifdef CONFIG_CPU_SB1 + nop # improves slotting +#endif +EXC( STORE t0, UNIT(0)(dst), s_exc) + ADDC(sum, t0) +EXC( STORE t1, UNIT(1)(dst), s_exc) + ADDC(sum, t1) +EXC( STORE t2, UNIT(2)(dst), s_exc) + ADDC(sum, t2) +EXC( STORE t3, UNIT(3)(dst), s_exc) + ADDC(sum, t3) + bne len, rem, 1b + ADD dst, dst, 4*NBYTES + +cleanup_src_unaligned: + beqz len, done + and rem, len, NBYTES-1 # rem = len % NBYTES + beq rem, len, copy_bytes + nop +1: +EXC( LDFIRST t0, FIRST(0)(src), l_exc) +EXC( LDREST t0, REST(0)(src), l_exc_copy) + ADD src, src, NBYTES + SUB len, len, NBYTES +EXC( STORE t0, 0(dst), s_exc) + ADDC(sum, t0) + bne len, rem, 1b + ADD dst, dst, NBYTES + +copy_bytes_checklen: + beqz len, done + nop +copy_bytes: + /* 0 < len < NBYTES */ +#ifdef CONFIG_CPU_LITTLE_ENDIAN +#define SHIFT_START 0 +#define SHIFT_INC 8 +#else +#define SHIFT_START 8*(NBYTES-1) +#define SHIFT_INC -8 +#endif + move t2, zero # partial word + li t3, SHIFT_START # shift +/* use l_exc_copy here to return correct sum on fault */ +#define COPY_BYTE(N) \ +EXC( lbu t0, N(src), l_exc_copy); \ + SUB len, len, 1; \ +EXC( sb t0, N(dst), s_exc); \ + SLLV t0, t0, t3; \ + addu t3, SHIFT_INC; \ + beqz len, copy_bytes_done; \ + or t2, t0 + + COPY_BYTE(0) + COPY_BYTE(1) +#ifdef USE_DOUBLE + COPY_BYTE(2) + COPY_BYTE(3) + COPY_BYTE(4) + COPY_BYTE(5) +#endif +EXC( lbu t0, NBYTES-2(src), l_exc_copy) + SUB len, len, 1 +EXC( sb t0, NBYTES-2(dst), s_exc) + SLLV t0, t0, t3 + or t2, t0 +copy_bytes_done: + ADDC(sum, t2) +done: + /* fold checksum */ +#ifdef USE_DOUBLE + dsll32 v1, sum, 0 + daddu sum, v1 + sltu v1, sum, v1 + dsra32 sum, sum, 0 + addu sum, v1 +#endif + sll v1, sum, 16 + addu sum, v1 + sltu v1, sum, v1 + srl sum, sum, 16 + addu sum, v1 + + /* odd buffer alignment? */ + beqz odd, 1f + nop + sll v1, sum, 8 + srl sum, sum, 8 + or sum, v1 + andi sum, 0xffff +1: + .set reorder + ADDC(sum, psum) + jr ra + .set noreorder + +l_exc_copy: + /* + * Copy bytes from src until faulting load address (or until a + * lb faults) + * + * When reached by a faulting LDFIRST/LDREST, THREAD_BUADDR($28) + * may be more than a byte beyond the last address. + * Hence, the lb below may get an exception. + * + * Assumes src < THREAD_BUADDR($28) + */ + LOAD t0, TI_TASK($28) + li t2, SHIFT_START + LOAD t0, THREAD_BUADDR(t0) +1: +EXC( lbu t1, 0(src), l_exc) + ADD src, src, 1 + sb t1, 0(dst) # can't fault -- we're copy_from_user + SLLV t1, t1, t2 + addu t2, SHIFT_INC + ADDC(sum, t1) + bne src, t0, 1b + ADD dst, dst, 1 +l_exc: + LOAD t0, TI_TASK($28) + nop + LOAD t0, THREAD_BUADDR(t0) # t0 is just past last good address + nop + SUB len, AT, t0 # len number of uncopied bytes + /* + * Here's where we rely on src and dst being incremented in tandem, + * See (3) above. + * dst += (fault addr - src) to put dst at first byte to clear + */ + ADD dst, t0 # compute start address in a1 + SUB dst, src + /* + * Clear len bytes starting at dst. Can't call __bzero because it + * might modify len. An inefficient loop for these rare times... + */ + beqz len, done + SUB src, len, 1 +1: sb zero, 0(dst) + ADD dst, dst, 1 + bnez src, 1b + SUB src, src, 1 + li v1, -EFAULT + b done + sw v1, (errptr) + +s_exc: + li v0, -1 /* invalid checksum */ + li v1, -EFAULT + jr ra + sw v1, (errptr) + END(__csum_partial_copy_user) diff --git a/arch/mips/lib/csum_partial_copy.c b/arch/mips/lib/csum_partial_copy.c deleted file mode 100644 index 1720f2ceeeae..000000000000 --- a/arch/mips/lib/csum_partial_copy.c +++ /dev/null @@ -1,49 +0,0 @@ -/* - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * Copyright (C) 1994, 1995 Waldorf Electronics GmbH - * Copyright (C) 1998, 1999 Ralf Baechle - */ -#include <linux/kernel.h> -#include <linux/types.h> -#include <asm/byteorder.h> -#include <asm/string.h> -#include <asm/uaccess.h> -#include <net/checksum.h> - -/* - * copy while checksumming, otherwise like csum_partial - */ -__wsum csum_partial_copy_nocheck(const void *src, - void *dst, int len, __wsum sum) -{ - /* - * It's 2:30 am and I don't feel like doing it real ... - * This is lots slower than the real thing (tm) - */ - sum = csum_partial(src, len, sum); - memcpy(dst, src, len); - - return sum; -} - -/* - * Copy from userspace and compute checksum. If we catch an exception - * then zero the rest of the buffer. - */ -__wsum csum_partial_copy_from_user (const void __user *src, - void *dst, int len, __wsum sum, int *err_ptr) -{ - int missing; - - might_sleep(); - missing = copy_from_user(dst, src, len); - if (missing) { - memset(dst + len - missing, 0, missing); - *err_ptr = -EFAULT; - } - - return csum_partial(dst, len, sum); -} diff --git a/arch/mips/lib/memset.S b/arch/mips/lib/memset.S new file mode 100644 index 000000000000..3f8b8b3d0b23 --- /dev/null +++ b/arch/mips/lib/memset.S @@ -0,0 +1,166 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 1998, 1999, 2000 by Ralf Baechle + * Copyright (C) 1999, 2000 Silicon Graphics, Inc. + */ +#include <asm/asm.h> +#include <asm/asm-offsets.h> +#include <asm/regdef.h> + +#if LONGSIZE == 4 +#define LONG_S_L swl +#define LONG_S_R swr +#else +#define LONG_S_L sdl +#define LONG_S_R sdr +#endif + +#define EX(insn,reg,addr,handler) \ +9: insn reg, addr; \ + .section __ex_table,"a"; \ + PTR 9b, handler; \ + .previous + + .macro f_fill64 dst, offset, val, fixup + EX(LONG_S, \val, (\offset + 0 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 1 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 2 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 3 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 4 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 5 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 6 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 7 * LONGSIZE)(\dst), \fixup) +#if LONGSIZE == 4 + EX(LONG_S, \val, (\offset + 8 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 9 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 10 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 11 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 12 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 13 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 14 * LONGSIZE)(\dst), \fixup) + EX(LONG_S, \val, (\offset + 15 * LONGSIZE)(\dst), \fixup) +#endif + .endm + +/* + * memset(void *s, int c, size_t n) + * + * a0: start of area to clear + * a1: char to fill with + * a2: size of area to clear + */ + .set noreorder + .align 5 +LEAF(memset) + beqz a1, 1f + move v0, a0 /* result */ + + andi a1, 0xff /* spread fillword */ + LONG_SLL t1, a1, 8 + or a1, t1 + LONG_SLL t1, a1, 16 +#if LONGSIZE == 8 + or a1, t1 + LONG_SLL t1, a1, 32 +#endif + or a1, t1 +1: + +FEXPORT(__bzero) + sltiu t0, a2, LONGSIZE /* very small region? */ + bnez t0, small_memset + andi t0, a0, LONGMASK /* aligned? */ + + beqz t0, 1f + PTR_SUBU t0, LONGSIZE /* alignment in bytes */ + +#ifdef __MIPSEB__ + EX(LONG_S_L, a1, (a0), first_fixup) /* make word/dword aligned */ +#endif +#ifdef __MIPSEL__ + EX(LONG_S_R, a1, (a0), first_fixup) /* make word/dword aligned */ +#endif + PTR_SUBU a0, t0 /* long align ptr */ + PTR_ADDU a2, t0 /* correct size */ + +1: ori t1, a2, 0x3f /* # of full blocks */ + xori t1, 0x3f + beqz t1, memset_partial /* no block to fill */ + andi t0, a2, 0x40-LONGSIZE + + PTR_ADDU t1, a0 /* end address */ + .set reorder +1: PTR_ADDIU a0, 64 + f_fill64 a0, -64, a1, fwd_fixup + bne t1, a0, 1b + .set noreorder + +memset_partial: + PTR_LA t1, 2f /* where to start */ +#if LONGSIZE == 4 + PTR_SUBU t1, t0 +#else + .set noat + LONG_SRL AT, t0, 1 + PTR_SUBU t1, AT + .set noat +#endif + jr t1 + PTR_ADDU a0, t0 /* dest ptr */ + + .set push + .set noreorder + .set nomacro + f_fill64 a0, -64, a1, partial_fixup /* ... but first do longs ... */ +2: .set pop + andi a2, LONGMASK /* At most one long to go */ + + beqz a2, 1f + PTR_ADDU a0, a2 /* What's left */ +#ifdef __MIPSEB__ + EX(LONG_S_R, a1, -1(a0), last_fixup) +#endif +#ifdef __MIPSEL__ + EX(LONG_S_L, a1, -1(a0), last_fixup) +#endif +1: jr ra + move a2, zero + +small_memset: + beqz a2, 2f + PTR_ADDU t1, a0, a2 + +1: PTR_ADDIU a0, 1 /* fill bytewise */ + bne t1, a0, 1b + sb a1, -1(a0) + +2: jr ra /* done */ + move a2, zero + END(memset) + +first_fixup: + jr ra + nop + +fwd_fixup: + PTR_L t0, TI_TASK($28) + LONG_L t0, THREAD_BUADDR(t0) + andi a2, 0x3f + LONG_ADDU a2, t1 + jr ra + LONG_SUBU a2, t0 + +partial_fixup: + PTR_L t0, TI_TASK($28) + LONG_L t0, THREAD_BUADDR(t0) + andi a2, LONGMASK + LONG_ADDU a2, t1 + jr ra + LONG_SUBU a2, t0 + +last_fixup: + jr ra + andi v1, a2, LONGMASK diff --git a/arch/mips/lib/uncached.c b/arch/mips/lib/uncached.c index 98ce89f8068b..2388f7f3ffde 100644 --- a/arch/mips/lib/uncached.c +++ b/arch/mips/lib/uncached.c @@ -44,20 +44,24 @@ unsigned long __init run_uncached(void *func) if (sp >= (long)CKSEG0 && sp < (long)CKSEG2) usp = CKSEG1ADDR(sp); +#ifdef CONFIG_64BIT else if ((long long)sp >= (long long)PHYS_TO_XKPHYS(0LL, 0) && (long long)sp < (long long)PHYS_TO_XKPHYS(8LL, 0)) usp = PHYS_TO_XKPHYS((long long)K_CALG_UNCACHED, XKPHYS_TO_PHYS((long long)sp)); +#endif else { BUG(); usp = sp; } if (lfunc >= (long)CKSEG0 && lfunc < (long)CKSEG2) ufunc = CKSEG1ADDR(lfunc); +#ifdef CONFIG_64BIT else if ((long long)lfunc >= (long long)PHYS_TO_XKPHYS(0LL, 0) && (long long)lfunc < (long long)PHYS_TO_XKPHYS(8LL, 0)) ufunc = PHYS_TO_XKPHYS((long long)K_CALG_UNCACHED, XKPHYS_TO_PHYS((long long)lfunc)); +#endif else { BUG(); ufunc = lfunc; |