You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
qemu/target/hexagon/decode.c

764 lines
23 KiB
C

/*
* Copyright(c) 2019-2023 Qualcomm Innovation Center, Inc. All Rights Reserved.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program 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 General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, see <http://www.gnu.org/licenses/>.
*/
#include "qemu/osdep.h"
#include "iclass.h"
#include "attribs.h"
#include "genptr.h"
#include "decode.h"
#include "insn.h"
#include "printinsn.h"
#include "mmvec/decode_ext_mmvec.h"
#define fZXTN(N, M, VAL) ((VAL) & ((1LL << (N)) - 1))
enum {
EXT_IDX_noext = 0,
EXT_IDX_noext_AFTER = 4,
EXT_IDX_mmvec = 4,
EXT_IDX_mmvec_AFTER = 8,
XX_LAST_EXT_IDX
};
/*
* Certain operand types represent a non-contiguous set of values.
* For example, the compound compare-and-jump instruction can only access
* registers R0-R7 and R16-23.
* This table represents the mapping from the encoding to the actual values.
*/
#define DEF_REGMAP(NAME, ELEMENTS, ...) \
static const unsigned int DECODE_REGISTER_##NAME[ELEMENTS] = \
{ __VA_ARGS__ };
/* Name Num Table */
DEF_REGMAP(R_16, 16, 0, 1, 2, 3, 4, 5, 6, 7, 16, 17, 18, 19, 20, 21, 22, 23)
DEF_REGMAP(R__8, 8, 0, 2, 4, 6, 16, 18, 20, 22)
DEF_REGMAP(R_8, 8, 0, 1, 2, 3, 4, 5, 6, 7)
#define DECODE_MAPPED_REG(OPNUM, NAME) \
insn->regno[OPNUM] = DECODE_REGISTER_##NAME[insn->regno[OPNUM]];
/* Helper functions for decode_*_generated.c.inc */
#define DECODE_MAPPED(NAME) \
static int decode_mapped_reg_##NAME(DisasContext *ctx, int x) \
{ \
return DECODE_REGISTER_##NAME[x]; \
}
DECODE_MAPPED(R_16)
DECODE_MAPPED(R_8)
DECODE_MAPPED(R__8)
/* Helper function for decodetree_trans_funcs_generated.c.inc */
static int shift_left(DisasContext *ctx, int x, int n, int immno)
{
int ret = x;
Insn *insn = ctx->insn;
if (!insn->extension_valid ||
insn->which_extended != immno) {
ret <<= n;
}
return ret;
}
/* Include the generated decoder for 32 bit insn */
#include "decode_normal_generated.c.inc"
#include "decode_hvx_generated.c.inc"
/* Include the generated decoder for 16 bit insn */
#include "decode_subinsn_a_generated.c.inc"
#include "decode_subinsn_l1_generated.c.inc"
#include "decode_subinsn_l2_generated.c.inc"
#include "decode_subinsn_s1_generated.c.inc"
#include "decode_subinsn_s2_generated.c.inc"
/* Include the generated helpers for the decoder */
#include "decodetree_trans_funcs_generated.c.inc"
void decode_send_insn_to(Packet *packet, int start, int newloc)
{
Insn tmpinsn;
int direction;
int i;
if (start == newloc) {
return;
}
if (start < newloc) {
/* Move towards end */
direction = 1;
} else {
/* move towards beginning */
direction = -1;
}
for (i = start; i != newloc; i += direction) {
tmpinsn = packet->insn[i];
packet->insn[i] = packet->insn[i + direction];
packet->insn[i + direction] = tmpinsn;
}
}
/* Fill newvalue registers with the correct regno */
static void
decode_fill_newvalue_regno(Packet *packet)
{
int i, use_regidx, offset, def_idx, dst_idx;
for (i = 1; i < packet->num_insns; i++) {
if (GET_ATTRIB(packet->insn[i].opcode, A_DOTNEWVALUE) &&
!GET_ATTRIB(packet->insn[i].opcode, A_EXTENSION)) {
g_assert(packet->insn[i].new_read_idx != -1);
use_regidx = packet->insn[i].new_read_idx;
/*
* What's encoded at the N-field is the offset to who's producing
* the value. Shift off the LSB which indicates odd/even register,
* then walk backwards and skip over the constant extenders.
*/
offset = packet->insn[i].regno[use_regidx] >> 1;
def_idx = i - offset;
for (int j = 0; j < offset; j++) {
if (GET_ATTRIB(packet->insn[i - j - 1].opcode, A_IT_EXTENDER)) {
def_idx--;
}
}
/*
* Check for a badly encoded N-field which points to an instruction
* out-of-range
*/
g_assert(!((def_idx < 0) || (def_idx > (packet->num_insns - 1))));
/* Now patch up the consumer with the register number */
g_assert(packet->insn[def_idx].dest_idx != -1);
dst_idx = packet->insn[def_idx].dest_idx;
packet->insn[i].regno[use_regidx] =
packet->insn[def_idx].regno[dst_idx];
/*
* We need to remember who produces this value to later
* check if it was dynamically cancelled
*/
packet->insn[i].new_value_producer_slot =
packet->insn[def_idx].slot;
}
}
}
/* Split CJ into a compare and a jump */
static void decode_split_cmpjump(Packet *pkt)
{
int last, i;
int numinsns = pkt->num_insns;
/*
* First, split all compare-jumps.
* The compare is sent to the end as a new instruction.
* Do it this way so we don't reorder dual jumps. Those need to stay in
* original order.
*/
for (i = 0; i < numinsns; i++) {
/* It's a cmp-jump */
if (GET_ATTRIB(pkt->insn[i].opcode, A_NEWCMPJUMP)) {
last = pkt->num_insns;
pkt->insn[last] = pkt->insn[i]; /* copy the instruction */
pkt->insn[last].part1 = true; /* last insn does the CMP */
pkt->insn[i].part1 = false; /* existing insn does the JUMP */
pkt->num_insns++;
}
}
/* Now re-shuffle all the compares back to the beginning */
for (i = 0; i < pkt->num_insns; i++) {
if (pkt->insn[i].part1) {
decode_send_insn_to(pkt, i, 0);
}
}
}
static bool decode_opcode_can_jump(int opcode)
{
if ((GET_ATTRIB(opcode, A_JUMP)) ||
(GET_ATTRIB(opcode, A_CALL)) ||
(opcode == J2_trap0) ||
(opcode == J2_pause)) {
/* Exception to A_JUMP attribute */
if (opcode == J4_hintjumpr) {
return false;
}
return true;
}
return false;
}
static bool decode_opcode_ends_loop(int opcode)
{
return GET_ATTRIB(opcode, A_HWLOOP0_END) ||
GET_ATTRIB(opcode, A_HWLOOP1_END);
}
/* Set the is_* fields in each instruction */
static void decode_set_insn_attr_fields(Packet *pkt)
{
int i;
int numinsns = pkt->num_insns;
uint16_t opcode;
pkt->pkt_has_cof = false;
pkt->pkt_has_multi_cof = false;
pkt->pkt_has_endloop = false;
pkt->pkt_has_dczeroa = false;
for (i = 0; i < numinsns; i++) {
opcode = pkt->insn[i].opcode;
if (pkt->insn[i].part1) {
continue; /* Skip compare of cmp-jumps */
}
if (GET_ATTRIB(opcode, A_DCZEROA)) {
pkt->pkt_has_dczeroa = true;
}
if (GET_ATTRIB(opcode, A_STORE)) {
if (GET_ATTRIB(opcode, A_SCALAR_STORE) &&
!GET_ATTRIB(opcode, A_MEMSIZE_0B)) {
if (pkt->insn[i].slot == 0) {
pkt->pkt_has_store_s0 = true;
} else {
pkt->pkt_has_store_s1 = true;
}
}
}
if (decode_opcode_can_jump(opcode)) {
if (pkt->pkt_has_cof) {
pkt->pkt_has_multi_cof = true;
}
pkt->pkt_has_cof = true;
}
pkt->insn[i].is_endloop = decode_opcode_ends_loop(opcode);
pkt->pkt_has_endloop |= pkt->insn[i].is_endloop;
if (pkt->pkt_has_endloop) {
if (pkt->pkt_has_cof) {
pkt->pkt_has_multi_cof = true;
}
pkt->pkt_has_cof = true;
}
}
}
/*
* Shuffle for execution
* Move stores to end (in same order as encoding)
* Move compares to beginning (for use by .new insns)
*/
static void decode_shuffle_for_execution(Packet *packet)
{
bool changed = false;
int i;
bool flag; /* flag means we've seen a non-memory instruction */
int n_mems;
int last_insn = packet->num_insns - 1;
/*
* Skip end loops, somehow an end loop is getting in and messing
* up the order
*/
if (decode_opcode_ends_loop(packet->insn[last_insn].opcode)) {
last_insn--;
}
do {
changed = false;
/*
* Stores go last, must not reorder.
* Cannot shuffle stores past loads, either.
* Iterate backwards. If we see a non-memory instruction,
* then a store, shuffle the store to the front. Don't shuffle
* stores wrt each other or a load.
*/
for (flag = false, n_mems = 0, i = last_insn; i >= 0; i--) {
int opcode = packet->insn[i].opcode;
if (flag && GET_ATTRIB(opcode, A_STORE)) {
decode_send_insn_to(packet, i, last_insn - n_mems);
n_mems++;
changed = true;
} else if (GET_ATTRIB(opcode, A_STORE)) {
n_mems++;
} else if (GET_ATTRIB(opcode, A_LOAD)) {
/*
* Don't set flag, since we don't want to shuffle a
* store past a load
*/
n_mems++;
} else if (GET_ATTRIB(opcode, A_DOTNEWVALUE)) {
/*
* Don't set flag, since we don't want to shuffle past
* a .new value
*/
} else {
flag = true;
}
}
if (changed) {
continue;
}
/* Compares go first, may be reordered wrt each other */
for (flag = false, i = 0; i < last_insn + 1; i++) {
int opcode = packet->insn[i].opcode;
if (packet->insn[i].has_pred_dest &&
GET_ATTRIB(opcode, A_STORE) == 0) {
/* This should be a compare (not a store conditional) */
if (flag) {
decode_send_insn_to(packet, i, 0);
changed = true;
continue;
}
} else if (GET_ATTRIB(opcode, A_IMPLICIT_WRITES_P3) &&
!decode_opcode_ends_loop(packet->insn[i].opcode)) {
/*
* spNloop instruction
* Don't reorder endloops; they are not valid for .new uses,
* and we want to match HW
*/
if (flag) {
decode_send_insn_to(packet, i, 0);
changed = true;
continue;
}
} else if (GET_ATTRIB(opcode, A_IMPLICIT_WRITES_P0) &&
!GET_ATTRIB(opcode, A_NEWCMPJUMP)) {
if (flag) {
decode_send_insn_to(packet, i, 0);
changed = true;
continue;
}
} else {
flag = true;
}
}
if (changed) {
continue;
}
} while (changed);
/*
* If we have a .new register compare/branch, move that to the very
* very end, past stores
*/
for (i = 0; i < last_insn; i++) {
if (GET_ATTRIB(packet->insn[i].opcode, A_DOTNEWVALUE)) {
decode_send_insn_to(packet, i, last_insn);
break;
}
}
}
static void
apply_extender(Packet *pkt, int i, uint32_t extender)
{
int immed_num;
uint32_t base_immed;
immed_num = pkt->insn[i].which_extended;
base_immed = pkt->insn[i].immed[immed_num];
pkt->insn[i].immed[immed_num] = extender | fZXTN(6, 32, base_immed);
}
static void decode_apply_extenders(Packet *packet)
{
int i;
for (i = 0; i < packet->num_insns; i++) {
if (GET_ATTRIB(packet->insn[i].opcode, A_IT_EXTENDER)) {
packet->insn[i + 1].extension_valid = true;
apply_extender(packet, i + 1, packet->insn[i].immed[0]);
}
}
}
static void decode_remove_extenders(Packet *packet)
{
int i, j;
for (i = 0; i < packet->num_insns; i++) {
if (GET_ATTRIB(packet->insn[i].opcode, A_IT_EXTENDER)) {
/* Remove this one by moving the remaining instructions down */
for (j = i;
(j < packet->num_insns - 1) && (j < INSTRUCTIONS_MAX - 1);
j++) {
packet->insn[j] = packet->insn[j + 1];
}
packet->num_insns--;
}
}
}
static SlotMask get_valid_slots(const Packet *pkt, unsigned int slot)
{
if (GET_ATTRIB(pkt->insn[slot].opcode, A_EXTENSION)) {
return mmvec_ext_decode_find_iclass_slots(pkt->insn[slot].opcode);
} else {
return find_iclass_slots(pkt->insn[slot].opcode,
pkt->insn[slot].iclass);
}
}
/*
* Section 10.3 of the Hexagon V73 Programmer's Reference Manual
*
* A duplex is encoded as a 32-bit instruction with bits [15:14] set to 00.
* The sub-instructions that comprise a duplex are encoded as 13-bit fields
* in the duplex.
*
* Per table 10-4, the 4-bit duplex iclass is encoded in bits 31:29, 13
*/
static uint32_t get_duplex_iclass(uint32_t encoding)
{
uint32_t iclass = extract32(encoding, 13, 1);
iclass = deposit32(iclass, 1, 3, extract32(encoding, 29, 3));
return iclass;
}
/*
* Per table 10-5, the duplex ICLASS field values that specify the group of
* each sub-instruction in a duplex
*
* This table points to the decode instruction for each entry in the table
*/
typedef bool (*subinsn_decode_func)(DisasContext *ctx, uint16_t insn);
typedef struct {
subinsn_decode_func decode_slot0_subinsn;
subinsn_decode_func decode_slot1_subinsn;
} subinsn_decode_groups;
static const subinsn_decode_groups decode_groups[16] = {
[0x0] = { decode_subinsn_l1, decode_subinsn_l1 },
[0x1] = { decode_subinsn_l2, decode_subinsn_l1 },
[0x2] = { decode_subinsn_l2, decode_subinsn_l2 },
[0x3] = { decode_subinsn_a, decode_subinsn_a },
[0x4] = { decode_subinsn_l1, decode_subinsn_a },
[0x5] = { decode_subinsn_l2, decode_subinsn_a },
[0x6] = { decode_subinsn_s1, decode_subinsn_a },
[0x7] = { decode_subinsn_s2, decode_subinsn_a },
[0x8] = { decode_subinsn_s1, decode_subinsn_l1 },
[0x9] = { decode_subinsn_s1, decode_subinsn_l2 },
[0xa] = { decode_subinsn_s1, decode_subinsn_s1 },
[0xb] = { decode_subinsn_s2, decode_subinsn_s1 },
[0xc] = { decode_subinsn_s2, decode_subinsn_l1 },
[0xd] = { decode_subinsn_s2, decode_subinsn_l2 },
[0xe] = { decode_subinsn_s2, decode_subinsn_s2 },
[0xf] = { NULL, NULL }, /* Reserved */
};
static uint16_t get_slot0_subinsn(uint32_t encoding)
{
return extract32(encoding, 0, 13);
}
static uint16_t get_slot1_subinsn(uint32_t encoding)
{
return extract32(encoding, 16, 13);
}
static unsigned int
decode_insns(DisasContext *ctx, Insn *insn, uint32_t encoding)
{
if (parse_bits(encoding) != 0) {
if (decode_normal(ctx, encoding) ||
decode_hvx(ctx, encoding)) {
insn->generate = opcode_genptr[insn->opcode];
insn->iclass = iclass_bits(encoding);
return 1;
}
g_assert_not_reached();
} else {
uint32_t iclass = get_duplex_iclass(encoding);
unsigned int slot0_subinsn = get_slot0_subinsn(encoding);
unsigned int slot1_subinsn = get_slot1_subinsn(encoding);
subinsn_decode_func decode_slot0_subinsn =
decode_groups[iclass].decode_slot0_subinsn;
subinsn_decode_func decode_slot1_subinsn =
decode_groups[iclass].decode_slot1_subinsn;
/* The slot1 subinsn needs to be in the packet first */
if (decode_slot1_subinsn(ctx, slot1_subinsn)) {
insn->generate = opcode_genptr[insn->opcode];
insn->iclass = iclass_bits(encoding);
ctx->insn = ++insn;
if (decode_slot0_subinsn(ctx, slot0_subinsn)) {
insn->generate = opcode_genptr[insn->opcode];
insn->iclass = iclass_bits(encoding);
return 2;
}
}
g_assert_not_reached();
}
}
static void decode_add_endloop_insn(Insn *insn, int loopnum)
{
if (loopnum == 10) {
insn->opcode = J2_endloop01;
insn->generate = opcode_genptr[J2_endloop01];
} else if (loopnum == 1) {
insn->opcode = J2_endloop1;
insn->generate = opcode_genptr[J2_endloop1];
} else if (loopnum == 0) {
insn->opcode = J2_endloop0;
insn->generate = opcode_genptr[J2_endloop0];
} else {
g_assert_not_reached();
}
}
static bool decode_parsebits_is_loopend(uint32_t encoding32)
{
uint32_t bits = parse_bits(encoding32);
return bits == 0x2;
}
static bool has_valid_slot_assignment(Packet *pkt)
{
int used_slots = 0;
for (int i = 0; i < pkt->num_insns; i++) {
int slot_mask;
Insn *insn = &pkt->insn[i];
if (decode_opcode_ends_loop(insn->opcode)) {
/* We overload slot 0 for endloop. */
continue;
}
slot_mask = 1 << insn->slot;
if (used_slots & slot_mask) {
return false;
}
used_slots |= slot_mask;
}
return true;
}
static bool
decode_set_slot_number(Packet *pkt)
{
int slot;
int i;
bool hit_mem_insn = false;
bool hit_duplex = false;
bool slot0_found = false;
bool slot1_found = false;
int slot1_iidx = 0;
/*
* The slots are encoded in reverse order
* For each instruction, count down until you find a suitable slot
*/
for (i = 0, slot = 3; i < pkt->num_insns; i++) {
SlotMask valid_slots = get_valid_slots(pkt, i);
while (!(valid_slots & (1 << slot))) {
slot--;
}
pkt->insn[i].slot = slot;
if (slot) {
/* I've assigned the slot, now decrement it for the next insn */
slot--;
}
}
/* Fix the exceptions - mem insns to slot 0,1 */
for (i = pkt->num_insns - 1; i >= 0; i--) {
/* First memory instruction always goes to slot 0 */
if ((GET_ATTRIB(pkt->insn[i].opcode, A_MEMLIKE) ||
GET_ATTRIB(pkt->insn[i].opcode, A_MEMLIKE_PACKET_RULES)) &&
!hit_mem_insn) {
hit_mem_insn = true;
pkt->insn[i].slot = 0;
continue;
}
/* Next memory instruction always goes to slot 1 */
if ((GET_ATTRIB(pkt->insn[i].opcode, A_MEMLIKE) ||
GET_ATTRIB(pkt->insn[i].opcode, A_MEMLIKE_PACKET_RULES)) &&
hit_mem_insn) {
pkt->insn[i].slot = 1;
}
}
/* Fix the exceptions - duplex always slot 0,1 */
for (i = pkt->num_insns - 1; i >= 0; i--) {
/* First subinsn always goes to slot 0 */
if (GET_ATTRIB(pkt->insn[i].opcode, A_SUBINSN) && !hit_duplex) {
hit_duplex = true;
pkt->insn[i].slot = 0;
continue;
}
/* Next subinsn always goes to slot 1 */
if (GET_ATTRIB(pkt->insn[i].opcode, A_SUBINSN) && hit_duplex) {
pkt->insn[i].slot = 1;
}
}
/* Fix the exceptions - slot 1 is never empty, always aligns to slot 0 */
for (i = pkt->num_insns - 1; i >= 0; i--) {
/* Is slot0 used? */
if (pkt->insn[i].slot == 0) {
bool is_endloop = (pkt->insn[i].opcode == J2_endloop01);
is_endloop |= (pkt->insn[i].opcode == J2_endloop0);
is_endloop |= (pkt->insn[i].opcode == J2_endloop1);
/*
* Make sure it's not endloop since, we're overloading
* slot0 for endloop
*/
if (!is_endloop) {
slot0_found = true;
}
}
/* Is slot1 used? */
if (pkt->insn[i].slot == 1) {
slot1_found = true;
slot1_iidx = i;
}
}
/* Is slot0 empty and slot1 used? */
if ((!slot0_found) && slot1_found) {
/* Then push it to slot0 */
pkt->insn[slot1_iidx].slot = 0;
}
return has_valid_slot_assignment(pkt);
}
/*
* decode_packet
* Decodes packet with given words
* Returns 0 on insufficient words,
* or number of words used on success
*/
int decode_packet(DisasContext *ctx, int max_words, const uint32_t *words,
Packet *pkt, bool disas_only)
{
int num_insns = 0;
int words_read = 0;
bool end_of_packet = false;
int new_insns = 0;
int i;
uint32_t encoding32;
/* Initialize */
memset(pkt, 0, sizeof(*pkt));
/* Try to build packet */
while (!end_of_packet && (words_read < max_words)) {
Insn *insn = &pkt->insn[num_insns];
ctx->insn = insn;
encoding32 = words[words_read];
end_of_packet = is_packet_end(encoding32);
new_insns = decode_insns(ctx, insn, encoding32);
g_assert(new_insns > 0);
/*
* If we saw an extender, mark next word extended so immediate
* decode works
*/
if (pkt->insn[num_insns].opcode == A4_ext) {
pkt->insn[num_insns + 1].extension_valid = true;
}
num_insns += new_insns;
words_read++;
}
pkt->num_insns = num_insns;
if (!end_of_packet) {
/* Ran out of words! */
return 0;
}
pkt->encod_pkt_size_in_bytes = words_read * 4;
pkt->pkt_has_hvx = false;
for (i = 0; i < num_insns; i++) {
pkt->pkt_has_hvx |=
GET_ATTRIB(pkt->insn[i].opcode, A_CVI);
}
/*
* Check for :endloop in the parse bits
* Section 10.6 of the Programmer's Reference describes the encoding
* The end of hardware loop 0 can be encoded with 2 words
* The end of hardware loop 1 needs 3 words
*/
if ((words_read == 2) && (decode_parsebits_is_loopend(words[0]))) {
decode_add_endloop_insn(&pkt->insn[pkt->num_insns++], 0);
}
if (words_read >= 3) {
bool has_loop0, has_loop1;
has_loop0 = decode_parsebits_is_loopend(words[0]);
has_loop1 = decode_parsebits_is_loopend(words[1]);
if (has_loop0 && has_loop1) {
decode_add_endloop_insn(&pkt->insn[pkt->num_insns++], 10);
} else if (has_loop1) {
decode_add_endloop_insn(&pkt->insn[pkt->num_insns++], 1);
} else if (has_loop0) {
decode_add_endloop_insn(&pkt->insn[pkt->num_insns++], 0);
}
}
decode_apply_extenders(pkt);
if (!disas_only) {
decode_remove_extenders(pkt);
if (!decode_set_slot_number(pkt)) {
/* Invalid packet */
return 0;
}
}
decode_fill_newvalue_regno(pkt);
if (pkt->pkt_has_hvx) {
mmvec_ext_decode_checks(pkt, disas_only);
}
if (!disas_only) {
decode_shuffle_for_execution(pkt);
decode_split_cmpjump(pkt);
decode_set_insn_attr_fields(pkt);
}
return words_read;
}
/* Used for "-d in_asm" logging */
int disassemble_hexagon(uint32_t *words, int nwords, bfd_vma pc,
GString *buf)
{
DisasContext ctx;
Packet pkt;
memset(&ctx, 0, sizeof(DisasContext));
ctx.pkt = &pkt;
if (decode_packet(&ctx, nwords, words, &pkt, true) > 0) {
snprint_a_pkt_disas(buf, &pkt, words, pc);
return pkt.encod_pkt_size_in_bytes;
} else {
g_string_assign(buf, "<invalid>");
return 0;
}
}