1/*- 2 * Copyright (c) 2011 The NetBSD Foundation, Inc. 3 * All rights reserved. 4 * 5 * This code is derived from software contributed to The NetBSD Foundation 6 * by Matt Thomas of 3am Software Foundry. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 17 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 18 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 19 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 20 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 21 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 22 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 23 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 24 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 25 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 26 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 27 * POSSIBILITY OF SUCH DAMAGE. 28 */ 29 30#include <machine/asm.h> 31#include "assym.h" 32 33RCSID("$NetBSD: spe_subr.S,v 1.4 2014/08/10 18:08:33 joerg Exp $") 34 35 .text 36#ifndef __clang__ 37 .machine e500x2 38#endif 39/* 40 * Lintstub: void vec_load_from_vreg(const struct vreg *); 41 */ 42ENTRY(vec_load_from_vreg) 43 /* 44 * Load and initialize the accumulator. Don't care about trashing 45 * r0's high half since we are about to load it. 46 */ 47 evldd %r0,(8 << 4)(%r3) 48 evmra %r0,%r0 49 50 /* 51 * SPEFSCR is in the trapframe. 52 */ 53 54 /* 55 * Since the high part of the register is saved in vreg (because we 56 * are loading it), and we know SPE is on (to load it), just load two 57 * high parts at one time and merge appropriately. For each even-odd 58 * register pair, the evldw will load high part for the even register 59 * in the high 32 bits and the high part for the odd regiter in the 60 * low 32 bits. 61 * 62 * The evmergelo will move the lower half of r0 to the high half of the 63 * destination register and evmergehilo will merge the high half of r0 64 * and lower half of the other register. 65 */ 66 evldw %r0,(30 << 2)(%r3) 67 evmergelo %r31,%r0,%r31; evmergehilo %r30,%r0,%r30 68 evldw %r0,(28 << 2)(%r3) 69 evmergelo %r29,%r0,%r29; evmergehilo %r28,%r0,%r28 70 evldw %r0,(26 << 2)(%r3) 71 evmergelo %r27,%r0,%r27; evmergehilo %r26,%r0,%r26 72 evldw %r0,(24 << 2)(%r3) 73 evmergelo %r25,%r0,%r25; evmergehilo %r24,%r0,%r24 74 evldw %r0,(22 << 2)(%r3) 75 evmergelo %r23,%r0,%r23; evmergehilo %r22,%r0,%r22 76 evldw %r0,(20 << 2)(%r3) 77 evmergelo %r21,%r0,%r21; evmergehilo %r20,%r0,%r20 78 evldw %r0,(18 << 2)(%r3) 79 evmergelo %r19,%r0,%r19; evmergehilo %r18,%r0,%r18 80 evldw %r0,(16 << 2)(%r3) 81 evmergelo %r17,%r0,%r17; evmergehilo %r16,%r0,%r16 82 evldw %r0,(14 << 2)(%r3) 83 evmergelo %r15,%r0,%r15; evmergehilo %r14,%r0,%r14 84 evldw %r0,(12 << 2)(%r3) 85 evmergelo %r13,%r0,%r13; evmergehilo %r12,%r0,%r12 86 /* 87 * Done with callee-saved registers. For caller-saved we can just 88 * trash the register contents. So load the two words with the high 89 * halves into the appropriate register and merge the half for the 90 * next register into it. 91 */ 92 evldw %r10,(10 << 2)(%r3) 93 evmergelo %r11,%r10,%r11 94 evldw %r8,(8 << 2)(%r3) 95 evmergelo %r9,%r8,%r9 96 evldw %r6,(6 << 2)(%r3) 97 evmergelo %r7,%r6,%r7 98 evldw %r4,(4 << 2)(%r3) 99 evmergelo %r5,%r4,%r5 100 101 /* 102 * R2 isn't a callee-saved, so load into r0 because we still need r3 103 */ 104 evldw %r0,(2 << 2)(%r3) 105 evmergelo %r3,%r0,%r3; evmergehilo %r2,%r0,%r2 106 evldd %r0,(0 << 2)(%r3) 107 evmergelo %r1,%r0,%r1 /* why bother? */ 108 109 blr 110END(vec_load_from_vreg) 111 112/* 113 * Lintstub: void vec_unload_to_vreg(struct vreg *); 114 */ 115ENTRY(vec_unload_to_vreg) 116 evmergehi %r4,%r4,%r0; /* save r0's high part in lo r4 */ 117 evmergehi %r0,%r0,%r1; evstdw %r0,(0 << 2)(%r3) 118 evmergehi %r0,%r2,%r3; evstdw %r0,(2 << 2)(%r3) 119 evmergehi %r0,%r4,%r5; evstdw %r0,(4 << 2)(%r3) 120 evmergehi %r0,%r6,%r7; evstdw %r0,(6 << 2)(%r3) 121 evmergehi %r0,%r8,%r9; evstdw %r0,(8 << 2)(%r3) 122 evmergehi %r0,%r10,%r11; evstdw %r0,(10 << 2)(%r3) 123 evmergehi %r0,%r12,%r13; evstdw %r0,(12 << 2)(%r3) 124 evmergehi %r0,%r14,%r15; evstdw %r0,(14 << 2)(%r3) 125 evmergehi %r0,%r16,%r17; evstdw %r0,(16 << 2)(%r3) 126 evmergehi %r0,%r18,%r19; evstdw %r0,(18 << 2)(%r3) 127 evmergehi %r0,%r20,%r21; evstdw %r0,(20 << 2)(%r3) 128 evmergehi %r0,%r22,%r23; evstdw %r0,(22 << 2)(%r3) 129 evmergehi %r0,%r24,%r25; evstdw %r0,(24 << 2)(%r3) 130 evmergehi %r0,%r26,%r27; evstdw %r0,(26 << 2)(%r3) 131 evmergehi %r0,%r28,%r29; evstdw %r0,(28 << 2)(%r3) 132 evmergehi %r0,%r30,%r31; evstdw %r0,(30 << 2)(%r3) 133 134 /* 135 * Now save the accumulator. 136 */ 137 evxor %r0,%r0,%r0 /* zero r0 */ 138 evaddumiaaw %r0,%r0 /* r0 = accum + r0 */ 139 evstdd %r0,(8 << 4)(%r3) /* store it */ 140 141 evmergelo %r0,%r4,%r0 /* retore r0's high half */ 142 143 /* 144 * The SPEFSCR will be restored when the exception returns. 145 */ 146 blr 147END(vec_unload_to_vreg) 148