1/* 2 * Copyright (c) 2006 - 2008, Intel Corporation. All rights reserved. 3 * This program and the accompanying materials are licensed and made 4available 5 * under the terms and conditions of the BSD License which accompanies 6this 7 * distribution. The full text of the license may be found at 8 * http://opensource.org/licenses/bsd-license.php. 9 * 10 * THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" 11BASIS, 12 * WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR 13 * IMPLIED. 14 */ 15 .text 16 .globl setjmp 17 .type setjmp, @function 18setjmp: 19 alloc loc0 = ar.pfs, 1, 2, 1, 0 20 ;; 21 mov r14 = ar.unat 22 mov r15 = ar.bsp 23 add r10 = 0x10*20, in0 24 ;; 25 stf.spill.nta [in0] = f2, 0x10 26 st8.spill.nta [r10] = r4, 8 27 mov r21 = b1 28 ;; 29 stf.spill.nta [in0] = f3, 0x10 30 st8.spill.nta [r10] = r5, 8 31 mov r22 = b2 32 ;; 33 stf.spill.nta [in0] = f4, 0x10 34 st8.spill.nta [r10] = r6, 8 35 mov r23 = b3 36 ;; 37 stf.spill.nta [in0] = f5, 0x10 38 st8.spill.nta [r10] = r7, 8 39 mov r24 = b4 40 ;; 41 stf.spill.nta [in0] = f16, 0x10 42 st8.spill.nta [r10] = sp, 8 43 mov r25 = b5 44 ;; 45 stf.spill.nta [in0] = f17, 0x10 46 st8.nta [r10] = loc1, 8 47 mov r16 = pr 48 ;; 49 stf.spill.nta [in0] = f18, 0x10 50 st8.nta [r10] = r21, 8 51 mov r17 = ar.lc 52 ;; 53 stf.spill.nta [in0] = f19, 0x10 54 st8.nta [r10] = r22, 8 55 ;; 56 stf.spill.nta [in0] = f20, 0x10 57 st8.nta [r10] = r23, 8 58 ;; 59 stf.spill.nta [in0] = f21, 0x10 60 st8.nta [r10] = r24, 8 61 ;; 62 stf.spill.nta [in0] = f22, 0x10 63 st8.nta [r10] = r25, 8 64 ;; 65 stf.spill.nta [in0] = f23, 0x10 66 mov r18 = ar.unat 67 ;; 68 stf.spill.nta [in0] = f24, 0x10 69 st8.nta [r10] = r14, 8 70 ;; 71 stf.spill.nta [in0] = f25, 0x10 72 st8.nta [r10] = r18, 8 73 ;; 74 stf.spill.nta [in0] = f26, 0x10 75 st8.nta [r10] = loc0, 8 76 ;; 77 stf.spill.nta [in0] = f27, 0x10 78 st8.nta [r10] = r15, 8 79 mov r8 = 0 80 ;; 81 stf.spill.nta [in0] = f28, 0x10 82 mov r19 = ar.fpsr 83 ;; 84 stf.spill.nta [in0] = f29, 0x10 85 st8.nta [r10] = r16, 8 86 mov ar.pfs = loc0 87 ;; 88 stf.spill.nta [in0] = f30, 0x10 89 st8.nta [r10] = r17, 8 90 mov b0 = loc1 91 ;; 92 stf.spill.nta [in0] = f31, 0x10 93 st8.nta [r10] = r19 94 ;; 95 mov ar.unat = r14 96 br.ret.sptk b0 97 ;; 98 99 .globl longjmp 100 .type longjmp, @function 101 .regstk 2, 0, 0, 0 102longjmp: 103 add r10 = 0x10*20 + 8*14, in0 104 movl r2 = ~((((1<<14) - 1) << 16) | 3) 105 ;; 106 ld8.nt1 r14 = [r10], -8*2 107 mov r15 = ar.bspstore 108 ;; 109 ld8.nt1 r17 = [r10], -8 110 mov r16 = ar.rsc 111 cmp.leu p6 = r14, r15 112 ;; 113 ld8.nt1 r18 = [r10], -8 114 ld8.nt1 r25 = [r10], -8 115 and r2 = r16, r2 116 ;; 117 ldf.fill.nt1 f2 = [in0], 0x10 118 ld8.nt1 r24 = [r10], -8 119 mov b5 = r25 120 ;; 121 mov ar.rsc = r2 122 ld8.nt1 r23 = [r10], -8 123 mov b4 = r24 124 ;; 125 ldf.fill.nt1 f3 = [in0], 0x10 126 mov ar.unat = r17 127(p6) br.spnt.many _skip_flushrs 128 ;; 129 flushrs 130 mov r15 = ar.bsp 131 ;; 132_skip_flushrs: 133 mov r31 = ar.rnat 134 loadrs 135 ;; 136 ldf.fill.nt1 f4 = [in0], 0x10 137 ld8.nt1 r22 = [r10], -8 138 dep r2 = -1, r14, 3, 6 139 ;; 140 ldf.fill.nt1 f5 = [in0], 0x10 141 ld8.nt1 r21 = [r10], -8 142 cmp.ltu p6 = r2, r15 143 ;; 144 ld8.nt1 r20 = [r10], -0x10 145(p6) ld8.nta r31 = [r2] 146 mov b3 = r23 147 ;; 148 ldf.fill.nt1 f16 = [in0], 0x10 149 ld8.fill.nt1 r7 = [r10], -8 150 mov b2 = r22 151 ;; 152 ldf.fill.nt1 f17 = [in0], 0x10 153 ld8.fill.nt1 r6 = [r10], -8 154 mov b1 = r21 155 ;; 156 ldf.fill.nt1 f18 = [in0], 0x10 157 ld8.fill.nt1 r5 = [r10], -8 158 mov b0 = r20 159 ;; 160 ldf.fill.nt1 f19 = [in0], 0x10 161 ld8.fill.nt1 r4 = [r10], 8*13 162 ;; 163 ldf.fill.nt1 f20 = [in0], 0x10 164 ld8.nt1 r19 = [r10], 0x10 165 ;; 166 ldf.fill.nt1 f21 = [in0], 0x10 167 ld8.nt1 r26 = [r10], 8 168 mov ar.pfs = r19 169 ;; 170 ldf.fill.nt1 f22 = [in0], 0x10 171 ld8.nt1 r27 = [r10], 8 172 mov pr = r26, -1 173 ;; 174 ldf.fill.nt1 f23 = [in0], 0x10 175 ld8.nt1 r28 = [r10], -17*8 - 0x10 176 mov ar.lc = r27 177 ;; 178 ldf.fill.nt1 f24 = [in0], 0x10 179 ldf.fill.nt1 f25 = [in0], 0x10 180 mov r8 = in1 181 ;; 182 ldf.fill.nt1 f26 = [in0], 0x10 183 ldf.fill.nt1 f31 = [r10], -0x10 184 ;; 185 ldf.fill.nt1 f27 = [in0], 0x10 186 ldf.fill.nt1 f30 = [r10], -0x10 187 ;; 188 ldf.fill.nt1 f28 = [in0] 189 ldf.fill.nt1 f29 = [r10], 0x10*3 + 8*4 190 ;; 191 ld8.fill.nt1 sp = [r10] 192 mov ar.unat = r18 193 ;; 194 mov ar.bspstore = r14 195 mov ar.rnat = r31 196 ;; 197 invala 198 mov ar.rsc = r16 199 br.ret.sptk b0 200