1 /*
2 * RISC-V Vector Extension Internals
3 *
4 * Copyright (c) 2020 T-Head Semiconductor Co., Ltd. All rights reserved.
5 *
6 * This program is free software; you can redistribute it and/or modify it
7 * under the terms and conditions of the GNU General Public License,
8 * version 2 or later, as published by the Free Software Foundation.
9 *
10 * This program is distributed in the hope it will be useful, but WITHOUT
11 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
12 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
13 * more details.
14 *
15 * You should have received a copy of the GNU General Public License along with
16 * this program. If not, see <http://www.gnu.org/licenses/>.
17 */
18
19 #include "qemu/osdep.h"
20 #include "vector_internals.h"
21
22 /* set agnostic elements to 1s */
vext_set_elems_1s(void * base,uint32_t is_agnostic,uint32_t cnt,uint32_t tot)23 void vext_set_elems_1s(void *base, uint32_t is_agnostic, uint32_t cnt,
24 uint32_t tot)
25 {
26 if (is_agnostic == 0) {
27 /* policy undisturbed */
28 return;
29 }
30 if (tot - cnt == 0) {
31 return ;
32 }
33
34 if (HOST_BIG_ENDIAN) {
35 /*
36 * Deal the situation when the elements are insdie
37 * only one uint64 block including setting the
38 * masked-off element.
39 */
40 if (((tot - 1) ^ cnt) < 8) {
41 memset(base + H1(tot - 1), -1, tot - cnt);
42 return;
43 }
44 /*
45 * Otherwise, at least cross two uint64_t blocks.
46 * Set first unaligned block.
47 */
48 if (cnt % 8 != 0) {
49 uint32_t j = ROUND_UP(cnt, 8);
50 memset(base + H1(j - 1), -1, j - cnt);
51 cnt = j;
52 }
53 /* Set other 64bit aligend blocks */
54 }
55 memset(base + cnt, -1, tot - cnt);
56 }
57
do_vext_vv(void * vd,void * v0,void * vs1,void * vs2,CPURISCVState * env,uint32_t desc,opivv2_fn * fn,uint32_t esz)58 void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
59 CPURISCVState *env, uint32_t desc,
60 opivv2_fn *fn, uint32_t esz)
61 {
62 uint32_t vm = vext_vm(desc);
63 uint32_t vl = env->vl;
64 uint32_t total_elems = vext_get_total_elems(env, desc, esz);
65 uint32_t vta = vext_vta(desc);
66 uint32_t vma = vext_vma(desc);
67 uint32_t i;
68
69 VSTART_CHECK_EARLY_EXIT(env);
70
71 for (i = env->vstart; i < vl; i++) {
72 if (!vm && !vext_elem_mask(v0, i)) {
73 /* set masked-off elements to 1s */
74 vext_set_elems_1s(vd, vma, i * esz, (i + 1) * esz);
75 continue;
76 }
77 fn(vd, vs1, vs2, i);
78 }
79 env->vstart = 0;
80 /* set tail elements to 1s */
81 vext_set_elems_1s(vd, vta, vl * esz, total_elems * esz);
82 }
83
do_vext_vx(void * vd,void * v0,target_long s1,void * vs2,CPURISCVState * env,uint32_t desc,opivx2_fn fn,uint32_t esz)84 void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
85 CPURISCVState *env, uint32_t desc,
86 opivx2_fn fn, uint32_t esz)
87 {
88 uint32_t vm = vext_vm(desc);
89 uint32_t vl = env->vl;
90 uint32_t total_elems = vext_get_total_elems(env, desc, esz);
91 uint32_t vta = vext_vta(desc);
92 uint32_t vma = vext_vma(desc);
93 uint32_t i;
94
95 VSTART_CHECK_EARLY_EXIT(env);
96
97 for (i = env->vstart; i < vl; i++) {
98 if (!vm && !vext_elem_mask(v0, i)) {
99 /* set masked-off elements to 1s */
100 vext_set_elems_1s(vd, vma, i * esz, (i + 1) * esz);
101 continue;
102 }
103 fn(vd, s1, vs2, i);
104 }
105 env->vstart = 0;
106 /* set tail elements to 1s */
107 vext_set_elems_1s(vd, vta, vl * esz, total_elems * esz);
108 }
109