helper-a64.c 12.6 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
/*
 *  AArch64 specific helpers
 *
 *  Copyright (c) 2013 Alexander Graf <agraf@suse.de>
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2 of the License, or (at your option) any later version.
 *
 * This library is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with this library; if not, see <http://www.gnu.org/licenses/>.
 */

20
#include "qemu/osdep.h"
21 22
#include "cpu.h"
#include "exec/gdbstub.h"
23
#include "exec/helper-proto.h"
24
#include "qemu/host-utils.h"
25
#include "qemu/log.h"
26 27
#include "sysemu/sysemu.h"
#include "qemu/bitops.h"
28
#include "internals.h"
29 30
#include "qemu/crc32c.h"
#include <zlib.h> /* For crc32 */
31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

/* C2.4.7 Multiply and divide */
/* special cases for 0 and LLONG_MIN are mandated by the standard */
uint64_t HELPER(udiv64)(uint64_t num, uint64_t den)
{
    if (den == 0) {
        return 0;
    }
    return num / den;
}

int64_t HELPER(sdiv64)(int64_t num, int64_t den)
{
    if (den == 0) {
        return 0;
    }
    if (num == LLONG_MIN && den == -1) {
        return LLONG_MIN;
    }
    return num / den;
}
52 53 54 55 56

uint64_t HELPER(clz64)(uint64_t x)
{
    return clz64(x);
}
57

58 59 60 61 62 63 64 65 66 67
uint64_t HELPER(cls64)(uint64_t x)
{
    return clrsb64(x);
}

uint32_t HELPER(cls32)(uint32_t x)
{
    return clrsb32(x);
}

68 69 70 71 72
uint32_t HELPER(clz32)(uint32_t x)
{
    return clz32(x);
}

73 74
uint64_t HELPER(rbit64)(uint64_t x)
{
75
    return revbit64(x);
76
}
77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121

/* Convert a softfloat float_relation_ (as returned by
 * the float*_compare functions) to the correct ARM
 * NZCV flag state.
 */
static inline uint32_t float_rel_to_flags(int res)
{
    uint64_t flags;
    switch (res) {
    case float_relation_equal:
        flags = PSTATE_Z | PSTATE_C;
        break;
    case float_relation_less:
        flags = PSTATE_N;
        break;
    case float_relation_greater:
        flags = PSTATE_C;
        break;
    case float_relation_unordered:
    default:
        flags = PSTATE_C | PSTATE_V;
        break;
    }
    return flags;
}

uint64_t HELPER(vfp_cmps_a64)(float32 x, float32 y, void *fp_status)
{
    return float_rel_to_flags(float32_compare_quiet(x, y, fp_status));
}

uint64_t HELPER(vfp_cmpes_a64)(float32 x, float32 y, void *fp_status)
{
    return float_rel_to_flags(float32_compare(x, y, fp_status));
}

uint64_t HELPER(vfp_cmpd_a64)(float64 x, float64 y, void *fp_status)
{
    return float_rel_to_flags(float64_compare_quiet(x, y, fp_status));
}

uint64_t HELPER(vfp_cmped_a64)(float64 x, float64 y, void *fp_status)
{
    return float_rel_to_flags(float64_compare(x, y, fp_status));
}
122

123 124 125 126
float32 HELPER(vfp_mulxs)(float32 a, float32 b, void *fpstp)
{
    float_status *fpst = fpstp;

127 128 129
    a = float32_squash_input_denormal(a, fpst);
    b = float32_squash_input_denormal(b, fpst);

130 131 132 133 134 135 136 137 138 139 140 141 142
    if ((float32_is_zero(a) && float32_is_infinity(b)) ||
        (float32_is_infinity(a) && float32_is_zero(b))) {
        /* 2.0 with the sign bit set to sign(A) XOR sign(B) */
        return make_float32((1U << 30) |
                            ((float32_val(a) ^ float32_val(b)) & (1U << 31)));
    }
    return float32_mul(a, b, fpst);
}

float64 HELPER(vfp_mulxd)(float64 a, float64 b, void *fpstp)
{
    float_status *fpst = fpstp;

143 144 145
    a = float64_squash_input_denormal(a, fpst);
    b = float64_squash_input_denormal(b, fpst);

146 147 148 149 150 151 152 153 154
    if ((float64_is_zero(a) && float64_is_infinity(b)) ||
        (float64_is_infinity(a) && float64_is_zero(b))) {
        /* 2.0 with the sign bit set to sign(A) XOR sign(B) */
        return make_float64((1ULL << 62) |
                            ((float64_val(a) ^ float64_val(b)) & (1ULL << 63)));
    }
    return float64_mul(a, b, fpst);
}

155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184
uint64_t HELPER(simd_tbl)(CPUARMState *env, uint64_t result, uint64_t indices,
                          uint32_t rn, uint32_t numregs)
{
    /* Helper function for SIMD TBL and TBX. We have to do the table
     * lookup part for the 64 bits worth of indices we're passed in.
     * result is the initial results vector (either zeroes for TBL
     * or some guest values for TBX), rn the register number where
     * the table starts, and numregs the number of registers in the table.
     * We return the results of the lookups.
     */
    int shift;

    for (shift = 0; shift < 64; shift += 8) {
        int index = extract64(indices, shift, 8);
        if (index < 16 * numregs) {
            /* Convert index (a byte offset into the virtual table
             * which is a series of 128-bit vectors concatenated)
             * into the correct vfp.regs[] element plus a bit offset
             * into that element, bearing in mind that the table
             * can wrap around from V31 to V0.
             */
            int elt = (rn * 2 + (index >> 3)) % 64;
            int bitidx = (index & 7) * 8;
            uint64_t val = extract64(env->vfp.regs[elt], bitidx, 8);

            result = deposit64(result, shift, 8, val);
        }
    }
    return result;
}
185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203

/* 64bit/double versions of the neon float compare functions */
uint64_t HELPER(neon_ceq_f64)(float64 a, float64 b, void *fpstp)
{
    float_status *fpst = fpstp;
    return -float64_eq_quiet(a, b, fpst);
}

uint64_t HELPER(neon_cge_f64)(float64 a, float64 b, void *fpstp)
{
    float_status *fpst = fpstp;
    return -float64_le(b, a, fpst);
}

uint64_t HELPER(neon_cgt_f64)(float64 a, float64 b, void *fpstp)
{
    float_status *fpst = fpstp;
    return -float64_lt(b, a, fpst);
}
204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220

/* Reciprocal step and sqrt step. Note that unlike the A32/T32
 * versions, these do a fully fused multiply-add or
 * multiply-add-and-halve.
 */
#define float32_two make_float32(0x40000000)
#define float32_three make_float32(0x40400000)
#define float32_one_point_five make_float32(0x3fc00000)

#define float64_two make_float64(0x4000000000000000ULL)
#define float64_three make_float64(0x4008000000000000ULL)
#define float64_one_point_five make_float64(0x3FF8000000000000ULL)

float32 HELPER(recpsf_f32)(float32 a, float32 b, void *fpstp)
{
    float_status *fpst = fpstp;

221 222 223
    a = float32_squash_input_denormal(a, fpst);
    b = float32_squash_input_denormal(b, fpst);

224 225 226 227 228 229 230 231 232 233 234 235
    a = float32_chs(a);
    if ((float32_is_infinity(a) && float32_is_zero(b)) ||
        (float32_is_infinity(b) && float32_is_zero(a))) {
        return float32_two;
    }
    return float32_muladd(a, b, float32_two, 0, fpst);
}

float64 HELPER(recpsf_f64)(float64 a, float64 b, void *fpstp)
{
    float_status *fpst = fpstp;

236 237 238
    a = float64_squash_input_denormal(a, fpst);
    b = float64_squash_input_denormal(b, fpst);

239 240 241 242 243 244 245 246 247 248 249 250
    a = float64_chs(a);
    if ((float64_is_infinity(a) && float64_is_zero(b)) ||
        (float64_is_infinity(b) && float64_is_zero(a))) {
        return float64_two;
    }
    return float64_muladd(a, b, float64_two, 0, fpst);
}

float32 HELPER(rsqrtsf_f32)(float32 a, float32 b, void *fpstp)
{
    float_status *fpst = fpstp;

251 252 253
    a = float32_squash_input_denormal(a, fpst);
    b = float32_squash_input_denormal(b, fpst);

254 255 256 257 258 259 260 261 262 263 264 265
    a = float32_chs(a);
    if ((float32_is_infinity(a) && float32_is_zero(b)) ||
        (float32_is_infinity(b) && float32_is_zero(a))) {
        return float32_one_point_five;
    }
    return float32_muladd(a, b, float32_three, float_muladd_halve_result, fpst);
}

float64 HELPER(rsqrtsf_f64)(float64 a, float64 b, void *fpstp)
{
    float_status *fpst = fpstp;

266 267 268
    a = float64_squash_input_denormal(a, fpst);
    b = float64_squash_input_denormal(b, fpst);

269 270 271 272 273 274 275
    a = float64_chs(a);
    if ((float64_is_infinity(a) && float64_is_zero(b)) ||
        (float64_is_infinity(b) && float64_is_zero(a))) {
        return float64_one_point_five;
    }
    return float64_muladd(a, b, float64_three, float_muladd_halve_result, fpst);
}
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336

/* Pairwise long add: add pairs of adjacent elements into
 * double-width elements in the result (eg _s8 is an 8x8->16 op)
 */
uint64_t HELPER(neon_addlp_s8)(uint64_t a)
{
    uint64_t nsignmask = 0x0080008000800080ULL;
    uint64_t wsignmask = 0x8000800080008000ULL;
    uint64_t elementmask = 0x00ff00ff00ff00ffULL;
    uint64_t tmp1, tmp2;
    uint64_t res, signres;

    /* Extract odd elements, sign extend each to a 16 bit field */
    tmp1 = a & elementmask;
    tmp1 ^= nsignmask;
    tmp1 |= wsignmask;
    tmp1 = (tmp1 - nsignmask) ^ wsignmask;
    /* Ditto for the even elements */
    tmp2 = (a >> 8) & elementmask;
    tmp2 ^= nsignmask;
    tmp2 |= wsignmask;
    tmp2 = (tmp2 - nsignmask) ^ wsignmask;

    /* calculate the result by summing bits 0..14, 16..22, etc,
     * and then adjusting the sign bits 15, 23, etc manually.
     * This ensures the addition can't overflow the 16 bit field.
     */
    signres = (tmp1 ^ tmp2) & wsignmask;
    res = (tmp1 & ~wsignmask) + (tmp2 & ~wsignmask);
    res ^= signres;

    return res;
}

uint64_t HELPER(neon_addlp_u8)(uint64_t a)
{
    uint64_t tmp;

    tmp = a & 0x00ff00ff00ff00ffULL;
    tmp += (a >> 8) & 0x00ff00ff00ff00ffULL;
    return tmp;
}

uint64_t HELPER(neon_addlp_s16)(uint64_t a)
{
    int32_t reslo, reshi;

    reslo = (int32_t)(int16_t)a + (int32_t)(int16_t)(a >> 16);
    reshi = (int32_t)(int16_t)(a >> 32) + (int32_t)(int16_t)(a >> 48);

    return (uint32_t)reslo | (((uint64_t)reshi) << 32);
}

uint64_t HELPER(neon_addlp_u16)(uint64_t a)
{
    uint64_t tmp;

    tmp = a & 0x0000ffff0000ffffULL;
    tmp += (a >> 16) & 0x0000ffff0000ffffULL;
    return tmp;
}
337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395

/* Floating-point reciprocal exponent - see FPRecpX in ARM ARM */
float32 HELPER(frecpx_f32)(float32 a, void *fpstp)
{
    float_status *fpst = fpstp;
    uint32_t val32, sbit;
    int32_t exp;

    if (float32_is_any_nan(a)) {
        float32 nan = a;
        if (float32_is_signaling_nan(a)) {
            float_raise(float_flag_invalid, fpst);
            nan = float32_maybe_silence_nan(a);
        }
        if (fpst->default_nan_mode) {
            nan = float32_default_nan;
        }
        return nan;
    }

    val32 = float32_val(a);
    sbit = 0x80000000ULL & val32;
    exp = extract32(val32, 23, 8);

    if (exp == 0) {
        return make_float32(sbit | (0xfe << 23));
    } else {
        return make_float32(sbit | (~exp & 0xff) << 23);
    }
}

float64 HELPER(frecpx_f64)(float64 a, void *fpstp)
{
    float_status *fpst = fpstp;
    uint64_t val64, sbit;
    int64_t exp;

    if (float64_is_any_nan(a)) {
        float64 nan = a;
        if (float64_is_signaling_nan(a)) {
            float_raise(float_flag_invalid, fpst);
            nan = float64_maybe_silence_nan(a);
        }
        if (fpst->default_nan_mode) {
            nan = float64_default_nan;
        }
        return nan;
    }

    val64 = float64_val(a);
    sbit = 0x8000000000000000ULL & val64;
    exp = extract64(float64_val(a), 52, 11);

    if (exp == 0) {
        return make_float64(sbit | (0x7feULL << 52));
    } else {
        return make_float64(sbit | (~exp & 0x7ffULL) << 52);
    }
}
396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418

float32 HELPER(fcvtx_f64_to_f32)(float64 a, CPUARMState *env)
{
    /* Von Neumann rounding is implemented by using round-to-zero
     * and then setting the LSB of the result if Inexact was raised.
     */
    float32 r;
    float_status *fpst = &env->vfp.fp_status;
    float_status tstat = *fpst;
    int exflags;

    set_float_rounding_mode(float_round_to_zero, &tstat);
    set_float_exception_flags(0, &tstat);
    r = float64_to_float32(a, &tstat);
    r = float32_maybe_silence_nan(r);
    exflags = get_float_exception_flags(&tstat);
    if (exflags & float_flag_inexact) {
        r = make_float32(float32_val(r) | 1);
    }
    exflags |= get_float_exception_flags(fpst);
    set_float_exception_flags(exflags, fpst);
    return r;
}
419

420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446
/* 64-bit versions of the CRC helpers. Note that although the operation
 * (and the prototypes of crc32c() and crc32() mean that only the bottom
 * 32 bits of the accumulator and result are used, we pass and return
 * uint64_t for convenience of the generated code. Unlike the 32-bit
 * instruction set versions, val may genuinely have 64 bits of data in it.
 * The upper bytes of val (above the number specified by 'bytes') must have
 * been zeroed out by the caller.
 */
uint64_t HELPER(crc32_64)(uint64_t acc, uint64_t val, uint32_t bytes)
{
    uint8_t buf[8];

    stq_le_p(buf, val);

    /* zlib crc32 converts the accumulator and output to one's complement.  */
    return crc32(acc ^ 0xffffffff, buf, bytes) ^ 0xffffffff;
}

uint64_t HELPER(crc32c_64)(uint64_t acc, uint64_t val, uint32_t bytes)
{
    uint8_t buf[8];

    stq_le_p(buf, val);

    /* Linux crc32c converts the output to one's complement.  */
    return crc32c(acc, buf, bytes) ^ 0xffffffff;
}