Tom Rini | 10e4779 | 2018-05-06 17:58:06 -0400 | [diff] [blame] | 1 | /* SPDX-License-Identifier: GPL-2.0+ */ |
Phil Edworthy | 8e5ab4b | 2011-06-02 22:15:27 +0000 | [diff] [blame] | 2 | /* Copyright (C) 1994, 1995, 1997, 1998, 1999, 2000, 2001, 2002, 2003, |
| 3 | 2004, 2005, 2006 |
| 4 | Free Software Foundation, Inc. |
Wolfgang Denk | d79de1d | 2013-07-08 09:37:19 +0200 | [diff] [blame] | 5 | */ |
Phil Edworthy | 8e5ab4b | 2011-06-02 22:15:27 +0000 | [diff] [blame] | 6 | |
| 7 | !! libgcc routines for the Renesas / SuperH SH CPUs. |
| 8 | !! Contributed by Steve Chamberlain. |
| 9 | !! sac@cygnus.com |
| 10 | |
| 11 | !! ashiftrt_r4_x, ___ashrsi3, ___ashlsi3, ___lshrsi3 routines |
| 12 | !! recoded in assembly by Toshiyasu Morita |
| 13 | !! tm@netcom.com |
| 14 | |
| 15 | /* SH2 optimizations for ___ashrsi3, ___ashlsi3, ___lshrsi3 and |
| 16 | ELF local label prefixes by J"orn Rennecke |
| 17 | amylaar@cygnus.com */ |
| 18 | |
| 19 | .text |
| 20 | .balign 4 |
| 21 | .global __movmem |
| 22 | .global __movstr |
| 23 | .set __movstr, __movmem |
| 24 | /* This would be a lot simpler if r6 contained the byte count |
| 25 | minus 64, and we wouldn't be called here for a byte count of 64. */ |
| 26 | __movmem: |
| 27 | sts.l pr,@-r15 |
| 28 | shll2 r6 |
| 29 | bsr __movmemSI52+2 |
| 30 | mov.l @(48,r5),r0 |
| 31 | .balign 4 |
| 32 | movmem_loop: /* Reached with rts */ |
| 33 | mov.l @(60,r5),r0 |
| 34 | add #-64,r6 |
| 35 | mov.l r0,@(60,r4) |
| 36 | tst r6,r6 |
| 37 | mov.l @(56,r5),r0 |
| 38 | bt movmem_done |
| 39 | mov.l r0,@(56,r4) |
| 40 | cmp/pl r6 |
| 41 | mov.l @(52,r5),r0 |
| 42 | add #64,r5 |
| 43 | mov.l r0,@(52,r4) |
| 44 | add #64,r4 |
| 45 | bt __movmemSI52 |
| 46 | ! done all the large groups, do the remainder |
| 47 | ! jump to movmem+ |
| 48 | mova __movmemSI4+4,r0 |
| 49 | add r6,r0 |
| 50 | jmp @r0 |
| 51 | movmem_done: ! share slot insn, works out aligned. |
| 52 | lds.l @r15+,pr |
| 53 | mov.l r0,@(56,r4) |
| 54 | mov.l @(52,r5),r0 |
| 55 | rts |
| 56 | mov.l r0,@(52,r4) |
| 57 | .balign 4 |
| 58 | |
| 59 | .global __movmemSI64 |
| 60 | .global __movstrSI64 |
| 61 | .set __movstrSI64, __movmemSI64 |
| 62 | __movmemSI64: |
| 63 | mov.l @(60,r5),r0 |
| 64 | mov.l r0,@(60,r4) |
| 65 | .global __movmemSI60 |
| 66 | .global __movstrSI60 |
| 67 | .set __movstrSI60, __movmemSI60 |
| 68 | __movmemSI60: |
| 69 | mov.l @(56,r5),r0 |
| 70 | mov.l r0,@(56,r4) |
| 71 | .global __movmemSI56 |
| 72 | .global __movstrSI56 |
| 73 | .set __movstrSI56, __movmemSI56 |
| 74 | __movmemSI56: |
| 75 | mov.l @(52,r5),r0 |
| 76 | mov.l r0,@(52,r4) |
| 77 | .global __movmemSI52 |
| 78 | .global __movstrSI52 |
| 79 | .set __movstrSI52, __movmemSI52 |
| 80 | __movmemSI52: |
| 81 | mov.l @(48,r5),r0 |
| 82 | mov.l r0,@(48,r4) |
| 83 | .global __movmemSI48 |
| 84 | .global __movstrSI48 |
| 85 | .set __movstrSI48, __movmemSI48 |
| 86 | __movmemSI48: |
| 87 | mov.l @(44,r5),r0 |
| 88 | mov.l r0,@(44,r4) |
| 89 | .global __movmemSI44 |
| 90 | .global __movstrSI44 |
| 91 | .set __movstrSI44, __movmemSI44 |
| 92 | __movmemSI44: |
| 93 | mov.l @(40,r5),r0 |
| 94 | mov.l r0,@(40,r4) |
| 95 | .global __movmemSI40 |
| 96 | .global __movstrSI40 |
| 97 | .set __movstrSI40, __movmemSI40 |
| 98 | __movmemSI40: |
| 99 | mov.l @(36,r5),r0 |
| 100 | mov.l r0,@(36,r4) |
| 101 | .global __movmemSI36 |
| 102 | .global __movstrSI36 |
| 103 | .set __movstrSI36, __movmemSI36 |
| 104 | __movmemSI36: |
| 105 | mov.l @(32,r5),r0 |
| 106 | mov.l r0,@(32,r4) |
| 107 | .global __movmemSI32 |
| 108 | .global __movstrSI32 |
| 109 | .set __movstrSI32, __movmemSI32 |
| 110 | __movmemSI32: |
| 111 | mov.l @(28,r5),r0 |
| 112 | mov.l r0,@(28,r4) |
| 113 | .global __movmemSI28 |
| 114 | .global __movstrSI28 |
| 115 | .set __movstrSI28, __movmemSI28 |
| 116 | __movmemSI28: |
| 117 | mov.l @(24,r5),r0 |
| 118 | mov.l r0,@(24,r4) |
| 119 | .global __movmemSI24 |
| 120 | .global __movstrSI24 |
| 121 | .set __movstrSI24, __movmemSI24 |
| 122 | __movmemSI24: |
| 123 | mov.l @(20,r5),r0 |
| 124 | mov.l r0,@(20,r4) |
| 125 | .global __movmemSI20 |
| 126 | .global __movstrSI20 |
| 127 | .set __movstrSI20, __movmemSI20 |
| 128 | __movmemSI20: |
| 129 | mov.l @(16,r5),r0 |
| 130 | mov.l r0,@(16,r4) |
| 131 | .global __movmemSI16 |
| 132 | .global __movstrSI16 |
| 133 | .set __movstrSI16, __movmemSI16 |
| 134 | __movmemSI16: |
| 135 | mov.l @(12,r5),r0 |
| 136 | mov.l r0,@(12,r4) |
| 137 | .global __movmemSI12 |
| 138 | .global __movstrSI12 |
| 139 | .set __movstrSI12, __movmemSI12 |
| 140 | __movmemSI12: |
| 141 | mov.l @(8,r5),r0 |
| 142 | mov.l r0,@(8,r4) |
| 143 | .global __movmemSI8 |
| 144 | .global __movstrSI8 |
| 145 | .set __movstrSI8, __movmemSI8 |
| 146 | __movmemSI8: |
| 147 | mov.l @(4,r5),r0 |
| 148 | mov.l r0,@(4,r4) |
| 149 | .global __movmemSI4 |
| 150 | .global __movstrSI4 |
| 151 | .set __movstrSI4, __movmemSI4 |
| 152 | __movmemSI4: |
| 153 | mov.l @(0,r5),r0 |
| 154 | rts |
| 155 | mov.l r0,@(0,r4) |
| 156 | |
| 157 | .global __movmem_i4_even |
| 158 | .global __movstr_i4_even |
| 159 | .set __movstr_i4_even, __movmem_i4_even |
| 160 | |
| 161 | .global __movmem_i4_odd |
| 162 | .global __movstr_i4_odd |
| 163 | .set __movstr_i4_odd, __movmem_i4_odd |
| 164 | |
| 165 | .global __movmemSI12_i4 |
| 166 | .global __movstrSI12_i4 |
| 167 | .set __movstrSI12_i4, __movmemSI12_i4 |
| 168 | |
| 169 | .p2align 5 |
| 170 | L_movmem_2mod4_end: |
| 171 | mov.l r0,@(16,r4) |
| 172 | rts |
| 173 | mov.l r1,@(20,r4) |
| 174 | |
| 175 | .p2align 2 |
| 176 | |
| 177 | __movmem_i4_even: |
| 178 | mov.l @r5+,r0 |
| 179 | bra L_movmem_start_even |
| 180 | mov.l @r5+,r1 |
| 181 | |
| 182 | __movmem_i4_odd: |
| 183 | mov.l @r5+,r1 |
| 184 | add #-4,r4 |
| 185 | mov.l @r5+,r2 |
| 186 | mov.l @r5+,r3 |
| 187 | mov.l r1,@(4,r4) |
| 188 | mov.l r2,@(8,r4) |
| 189 | |
| 190 | L_movmem_loop: |
| 191 | mov.l r3,@(12,r4) |
| 192 | dt r6 |
| 193 | mov.l @r5+,r0 |
| 194 | bt/s L_movmem_2mod4_end |
| 195 | mov.l @r5+,r1 |
| 196 | add #16,r4 |
| 197 | L_movmem_start_even: |
| 198 | mov.l @r5+,r2 |
| 199 | mov.l @r5+,r3 |
| 200 | mov.l r0,@r4 |
| 201 | dt r6 |
| 202 | mov.l r1,@(4,r4) |
| 203 | bf/s L_movmem_loop |
| 204 | mov.l r2,@(8,r4) |
| 205 | rts |
| 206 | mov.l r3,@(12,r4) |
| 207 | |
| 208 | .p2align 4 |
| 209 | __movmemSI12_i4: |
| 210 | mov.l @r5,r0 |
| 211 | mov.l @(4,r5),r1 |
| 212 | mov.l @(8,r5),r2 |
| 213 | mov.l r0,@r4 |
| 214 | mov.l r1,@(4,r4) |
| 215 | rts |
| 216 | mov.l r2,@(8,r4) |