xemu/target/hexagon/cpu.c
Richard Henderson c88da1f3da This series adds support for the Hexagon Vector eXtensions (HVX)
These instructions are documented here
 https://developer.qualcomm.com/downloads/qualcomm-hexagon-v66-hvx-programmer-s-reference-manual
 
 Hexagon HVX is a wide vector engine with 128 byte vectors.
 
 See patch 01 Hexagon HVX README for more information.
 
 *** Changes in v2 ***
 Remove HVX tests from makefile to avoid need for toolchain upgrade
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1
 
 iQEcBAABAgAGBQJhgvvEAAoJEHsCRPsS3kQiMrQH/2ZGLfiWts1Vyi+phoXz24bC
 x/kAarlkCPP9YjiKa1HFCoX8L58O/r0ES6gwe2amXWDQoC+kQYoAJawOtgasGvuj
 ePctEkNLTJ2l5kkh83ITREVDpVXzoiAYSQ5hicJtKyAyfjFrdqC9mhHhRbShGG3+
 KA5JnyczrnuoIzaJxEbDPVaL/3hHThvpxV1bi9CWrEvjqa64Flot1szKAJlCeaTM
 GgersSCodtMyZHtU36ngzPDVrYNtTYZ5OiYkNpdEzHtzpuItq0xVMP0dGn2WGO+1
 0CTnuDmzJXMo+RDTo8pniz5pESgBbD9WMbgFEG3I0xUQIj0MdMkZa+h8lUvhmb4=
 =TK+z
 -----END PGP SIGNATURE-----

Merge remote-tracking branch 'remotes/quic/tags/pull-hex-20211103' into staging

This series adds support for the Hexagon Vector eXtensions (HVX)

These instructions are documented here
https://developer.qualcomm.com/downloads/qualcomm-hexagon-v66-hvx-programmer-s-reference-manual

Hexagon HVX is a wide vector engine with 128 byte vectors.

See patch 01 Hexagon HVX README for more information.

*** Changes in v2 ***
Remove HVX tests from makefile to avoid need for toolchain upgrade

# gpg: Signature made Wed 03 Nov 2021 05:14:44 PM EDT
# gpg:                using RSA key 7B0244FB12DE4422
# gpg: Good signature from "Taylor Simpson (Rock on) <tsimpson@quicinc.com>" [marginal]
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg:          It is not certain that the signature belongs to the owner.
# Primary key fingerprint: 3635 C788 CE62 B91F D4C5  9AB4 7B02 44FB 12DE 4422

* remotes/quic/tags/pull-hex-20211103: (30 commits)
  Hexagon HVX (tests/tcg/hexagon) histogram test
  Hexagon HVX (tests/tcg/hexagon) scatter_gather test
  Hexagon HVX (tests/tcg/hexagon) hvx_misc test
  Hexagon HVX (tests/tcg/hexagon) vector_add_int test
  Hexagon HVX (target/hexagon) import instruction encodings
  Hexagon HVX (target/hexagon) instruction decoding
  Hexagon HVX (target/hexagon) import semantics
  Hexagon HVX (target/hexagon) helper overrides - vector stores
  Hexagon HVX (target/hexagon) helper overrides - vector loads
  Hexagon HVX (target/hexagon) helper overrides - vector splat and abs
  Hexagon HVX (target/hexagon) helper overrides - vector compares
  Hexagon HVX (target/hexagon) helper overrides - vector logical ops
  Hexagon HVX (target/hexagon) helper overrides - vector max/min
  Hexagon HVX (target/hexagon) helper overrides - vector shifts
  Hexagon HVX (target/hexagon) helper overrides - vector add & sub
  Hexagon HVX (target/hexagon) helper overrides - vector assign & cmov
  Hexagon HVX (target/hexagon) helper overrides for histogram instructions
  Hexagon HVX (target/hexagon) helper overrides infrastructure
  Hexagon HVX (target/hexagon) TCG generation
  Hexagon HVX (target/hexagon) helper functions
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-11-04 06:34:36 -04:00

369 lines
11 KiB
C

/*
* Copyright(c) 2019-2021 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 "qemu/qemu-print.h"
#include "cpu.h"
#include "internal.h"
#include "exec/exec-all.h"
#include "qapi/error.h"
#include "hw/qdev-properties.h"
#include "fpu/softfloat-helpers.h"
static void hexagon_v67_cpu_init(Object *obj)
{
}
static ObjectClass *hexagon_cpu_class_by_name(const char *cpu_model)
{
ObjectClass *oc;
char *typename;
char **cpuname;
cpuname = g_strsplit(cpu_model, ",", 1);
typename = g_strdup_printf(HEXAGON_CPU_TYPE_NAME("%s"), cpuname[0]);
oc = object_class_by_name(typename);
g_strfreev(cpuname);
g_free(typename);
if (!oc || !object_class_dynamic_cast(oc, TYPE_HEXAGON_CPU) ||
object_class_is_abstract(oc)) {
return NULL;
}
return oc;
}
static Property hexagon_lldb_compat_property =
DEFINE_PROP_BOOL("lldb-compat", HexagonCPU, lldb_compat, false);
static Property hexagon_lldb_stack_adjust_property =
DEFINE_PROP_UNSIGNED("lldb-stack-adjust", HexagonCPU, lldb_stack_adjust,
0, qdev_prop_uint32, target_ulong);
const char * const hexagon_regnames[TOTAL_PER_THREAD_REGS] = {
"r0", "r1", "r2", "r3", "r4", "r5", "r6", "r7",
"r8", "r9", "r10", "r11", "r12", "r13", "r14", "r15",
"r16", "r17", "r18", "r19", "r20", "r21", "r22", "r23",
"r24", "r25", "r26", "r27", "r28", "r29", "r30", "r31",
"sa0", "lc0", "sa1", "lc1", "p3_0", "c5", "m0", "m1",
"usr", "pc", "ugp", "gp", "cs0", "cs1", "c14", "c15",
"c16", "c17", "c18", "c19", "pkt_cnt", "insn_cnt", "hvx_cnt", "c23",
"c24", "c25", "c26", "c27", "c28", "c29", "c30", "c31",
};
/*
* One of the main debugging techniques is to use "-d cpu" and compare against
* LLDB output when single stepping. However, the target and qemu put the
* stacks at different locations. This is used to compensate so the diff is
* cleaner.
*/
static target_ulong adjust_stack_ptrs(CPUHexagonState *env, target_ulong addr)
{
HexagonCPU *cpu = env_archcpu(env);
target_ulong stack_adjust = cpu->lldb_stack_adjust;
target_ulong stack_start = env->stack_start;
target_ulong stack_size = 0x10000;
if (stack_adjust == 0) {
return addr;
}
if (stack_start + 0x1000 >= addr && addr >= (stack_start - stack_size)) {
return addr - stack_adjust;
}
return addr;
}
/* HEX_REG_P3_0 (aka C4) is an alias for the predicate registers */
static target_ulong read_p3_0(CPUHexagonState *env)
{
int32_t control_reg = 0;
int i;
for (i = NUM_PREGS - 1; i >= 0; i--) {
control_reg <<= 8;
control_reg |= env->pred[i] & 0xff;
}
return control_reg;
}
static void print_reg(FILE *f, CPUHexagonState *env, int regnum)
{
target_ulong value;
if (regnum == HEX_REG_P3_0) {
value = read_p3_0(env);
} else {
value = regnum < 32 ? adjust_stack_ptrs(env, env->gpr[regnum])
: env->gpr[regnum];
}
qemu_fprintf(f, " %s = 0x" TARGET_FMT_lx "\n",
hexagon_regnames[regnum], value);
}
static void print_vreg(FILE *f, CPUHexagonState *env, int regnum,
bool skip_if_zero)
{
if (skip_if_zero) {
bool nonzero_found = false;
for (int i = 0; i < MAX_VEC_SIZE_BYTES; i++) {
if (env->VRegs[regnum].ub[i] != 0) {
nonzero_found = true;
break;
}
}
if (!nonzero_found) {
return;
}
}
qemu_fprintf(f, " v%d = ( ", regnum);
qemu_fprintf(f, "0x%02x", env->VRegs[regnum].ub[MAX_VEC_SIZE_BYTES - 1]);
for (int i = MAX_VEC_SIZE_BYTES - 2; i >= 0; i--) {
qemu_fprintf(f, ", 0x%02x", env->VRegs[regnum].ub[i]);
}
qemu_fprintf(f, " )\n");
}
void hexagon_debug_vreg(CPUHexagonState *env, int regnum)
{
print_vreg(stdout, env, regnum, false);
}
static void print_qreg(FILE *f, CPUHexagonState *env, int regnum,
bool skip_if_zero)
{
if (skip_if_zero) {
bool nonzero_found = false;
for (int i = 0; i < MAX_VEC_SIZE_BYTES / 8; i++) {
if (env->QRegs[regnum].ub[i] != 0) {
nonzero_found = true;
break;
}
}
if (!nonzero_found) {
return;
}
}
qemu_fprintf(f, " q%d = ( ", regnum);
qemu_fprintf(f, "0x%02x",
env->QRegs[regnum].ub[MAX_VEC_SIZE_BYTES / 8 - 1]);
for (int i = MAX_VEC_SIZE_BYTES / 8 - 2; i >= 0; i--) {
qemu_fprintf(f, ", 0x%02x", env->QRegs[regnum].ub[i]);
}
qemu_fprintf(f, " )\n");
}
void hexagon_debug_qreg(CPUHexagonState *env, int regnum)
{
print_qreg(stdout, env, regnum, false);
}
static void hexagon_dump(CPUHexagonState *env, FILE *f, int flags)
{
HexagonCPU *cpu = env_archcpu(env);
if (cpu->lldb_compat) {
/*
* When comparing with LLDB, it doesn't step through single-cycle
* hardware loops the same way. So, we just skip them here
*/
if (env->gpr[HEX_REG_PC] == env->last_pc_dumped) {
return;
}
env->last_pc_dumped = env->gpr[HEX_REG_PC];
}
qemu_fprintf(f, "General Purpose Registers = {\n");
for (int i = 0; i < 32; i++) {
print_reg(f, env, i);
}
print_reg(f, env, HEX_REG_SA0);
print_reg(f, env, HEX_REG_LC0);
print_reg(f, env, HEX_REG_SA1);
print_reg(f, env, HEX_REG_LC1);
print_reg(f, env, HEX_REG_M0);
print_reg(f, env, HEX_REG_M1);
print_reg(f, env, HEX_REG_USR);
print_reg(f, env, HEX_REG_P3_0);
print_reg(f, env, HEX_REG_GP);
print_reg(f, env, HEX_REG_UGP);
print_reg(f, env, HEX_REG_PC);
#ifdef CONFIG_USER_ONLY
/*
* Not modelled in user mode, print junk to minimize the diff's
* with LLDB output
*/
qemu_fprintf(f, " cause = 0x000000db\n");
qemu_fprintf(f, " badva = 0x00000000\n");
qemu_fprintf(f, " cs0 = 0x00000000\n");
qemu_fprintf(f, " cs1 = 0x00000000\n");
#else
print_reg(f, env, HEX_REG_CAUSE);
print_reg(f, env, HEX_REG_BADVA);
print_reg(f, env, HEX_REG_CS0);
print_reg(f, env, HEX_REG_CS1);
#endif
qemu_fprintf(f, "}\n");
if (flags & CPU_DUMP_FPU) {
qemu_fprintf(f, "Vector Registers = {\n");
for (int i = 0; i < NUM_VREGS; i++) {
print_vreg(f, env, i, true);
}
for (int i = 0; i < NUM_QREGS; i++) {
print_qreg(f, env, i, true);
}
qemu_fprintf(f, "}\n");
}
}
static void hexagon_dump_state(CPUState *cs, FILE *f, int flags)
{
HexagonCPU *cpu = HEXAGON_CPU(cs);
CPUHexagonState *env = &cpu->env;
hexagon_dump(env, f, flags);
}
void hexagon_debug(CPUHexagonState *env)
{
hexagon_dump(env, stdout, CPU_DUMP_FPU);
}
static void hexagon_cpu_set_pc(CPUState *cs, vaddr value)
{
HexagonCPU *cpu = HEXAGON_CPU(cs);
CPUHexagonState *env = &cpu->env;
env->gpr[HEX_REG_PC] = value;
}
static void hexagon_cpu_synchronize_from_tb(CPUState *cs,
const TranslationBlock *tb)
{
HexagonCPU *cpu = HEXAGON_CPU(cs);
CPUHexagonState *env = &cpu->env;
env->gpr[HEX_REG_PC] = tb->pc;
}
static bool hexagon_cpu_has_work(CPUState *cs)
{
return true;
}
void restore_state_to_opc(CPUHexagonState *env, TranslationBlock *tb,
target_ulong *data)
{
env->gpr[HEX_REG_PC] = data[0];
}
static void hexagon_cpu_reset(DeviceState *dev)
{
CPUState *cs = CPU(dev);
HexagonCPU *cpu = HEXAGON_CPU(cs);
HexagonCPUClass *mcc = HEXAGON_CPU_GET_CLASS(cpu);
CPUHexagonState *env = &cpu->env;
mcc->parent_reset(dev);
set_default_nan_mode(1, &env->fp_status);
set_float_detect_tininess(float_tininess_before_rounding, &env->fp_status);
}
static void hexagon_cpu_disas_set_info(CPUState *s, disassemble_info *info)
{
info->print_insn = print_insn_hexagon;
}
static void hexagon_cpu_realize(DeviceState *dev, Error **errp)
{
CPUState *cs = CPU(dev);
HexagonCPUClass *mcc = HEXAGON_CPU_GET_CLASS(dev);
Error *local_err = NULL;
cpu_exec_realizefn(cs, &local_err);
if (local_err != NULL) {
error_propagate(errp, local_err);
return;
}
qemu_init_vcpu(cs);
cpu_reset(cs);
mcc->parent_realize(dev, errp);
}
static void hexagon_cpu_init(Object *obj)
{
HexagonCPU *cpu = HEXAGON_CPU(obj);
cpu_set_cpustate_pointers(cpu);
qdev_property_add_static(DEVICE(obj), &hexagon_lldb_compat_property);
qdev_property_add_static(DEVICE(obj), &hexagon_lldb_stack_adjust_property);
}
#include "hw/core/tcg-cpu-ops.h"
static const struct TCGCPUOps hexagon_tcg_ops = {
.initialize = hexagon_translate_init,
.synchronize_from_tb = hexagon_cpu_synchronize_from_tb,
};
static void hexagon_cpu_class_init(ObjectClass *c, void *data)
{
HexagonCPUClass *mcc = HEXAGON_CPU_CLASS(c);
CPUClass *cc = CPU_CLASS(c);
DeviceClass *dc = DEVICE_CLASS(c);
device_class_set_parent_realize(dc, hexagon_cpu_realize,
&mcc->parent_realize);
device_class_set_parent_reset(dc, hexagon_cpu_reset, &mcc->parent_reset);
cc->class_by_name = hexagon_cpu_class_by_name;
cc->has_work = hexagon_cpu_has_work;
cc->dump_state = hexagon_dump_state;
cc->set_pc = hexagon_cpu_set_pc;
cc->gdb_read_register = hexagon_gdb_read_register;
cc->gdb_write_register = hexagon_gdb_write_register;
cc->gdb_num_core_regs = TOTAL_PER_THREAD_REGS + NUM_VREGS + NUM_QREGS;
cc->gdb_stop_before_watchpoint = true;
cc->disas_set_info = hexagon_cpu_disas_set_info;
cc->tcg_ops = &hexagon_tcg_ops;
}
#define DEFINE_CPU(type_name, initfn) \
{ \
.name = type_name, \
.parent = TYPE_HEXAGON_CPU, \
.instance_init = initfn \
}
static const TypeInfo hexagon_cpu_type_infos[] = {
{
.name = TYPE_HEXAGON_CPU,
.parent = TYPE_CPU,
.instance_size = sizeof(HexagonCPU),
.instance_init = hexagon_cpu_init,
.abstract = true,
.class_size = sizeof(HexagonCPUClass),
.class_init = hexagon_cpu_class_init,
},
DEFINE_CPU(TYPE_HEXAGON_CPU_V67, hexagon_v67_cpu_init),
};
DEFINE_TYPES(hexagon_cpu_type_infos)