openocd/src/target/riscv/riscv.c

4176 lines
113 KiB
C

/* SPDX-License-Identifier: GPL-2.0-or-later */
#include <assert.h>
#include <stdlib.h>
#include <time.h>
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "target/target.h"
#include "target/algorithm.h"
#include "target/target_type.h"
#include "log.h"
#include "jtag/jtag.h"
#include "target/register.h"
#include "target/breakpoints.h"
#include "helper/time_support.h"
#include "riscv.h"
#include "gdb_regs.h"
#include "rtos/rtos.h"
#define get_field(reg, mask) (((reg) & (mask)) / ((mask) & ~((mask) << 1)))
#define set_field(reg, mask, val) (((reg) & ~(mask)) | (((val) * ((mask) & ~((mask) << 1))) & (mask)))
#define DIM(x) (sizeof(x)/sizeof(*x))
/* Constants for legacy SiFive hardware breakpoints. */
#define CSR_BPCONTROL_X (1<<0)
#define CSR_BPCONTROL_W (1<<1)
#define CSR_BPCONTROL_R (1<<2)
#define CSR_BPCONTROL_U (1<<3)
#define CSR_BPCONTROL_S (1<<4)
#define CSR_BPCONTROL_H (1<<5)
#define CSR_BPCONTROL_M (1<<6)
#define CSR_BPCONTROL_BPMATCH (0xf<<7)
#define CSR_BPCONTROL_BPACTION (0xff<<11)
#define DEBUG_ROM_START 0x800
#define DEBUG_ROM_RESUME (DEBUG_ROM_START + 4)
#define DEBUG_ROM_EXCEPTION (DEBUG_ROM_START + 8)
#define DEBUG_RAM_START 0x400
#define SETHALTNOT 0x10c
/*** JTAG registers. ***/
#define DTMCONTROL 0x10
#define DTMCONTROL_DBUS_RESET (1<<16)
#define DTMCONTROL_IDLE (7<<10)
#define DTMCONTROL_ADDRBITS (0xf<<4)
#define DTMCONTROL_VERSION (0xf)
#define DBUS 0x11
#define DBUS_OP_START 0
#define DBUS_OP_SIZE 2
typedef enum {
DBUS_OP_NOP = 0,
DBUS_OP_READ = 1,
DBUS_OP_WRITE = 2
} dbus_op_t;
typedef enum {
DBUS_STATUS_SUCCESS = 0,
DBUS_STATUS_FAILED = 2,
DBUS_STATUS_BUSY = 3
} dbus_status_t;
#define DBUS_DATA_START 2
#define DBUS_DATA_SIZE 34
#define DBUS_ADDRESS_START 36
typedef enum slot {
SLOT0,
SLOT1,
SLOT_LAST,
} slot_t;
/*** Debug Bus registers. ***/
#define DMCONTROL 0x10
#define DMCONTROL_INTERRUPT (((uint64_t)1)<<33)
#define DMCONTROL_HALTNOT (((uint64_t)1)<<32)
#define DMCONTROL_BUSERROR (7<<19)
#define DMCONTROL_SERIAL (3<<16)
#define DMCONTROL_AUTOINCREMENT (1<<15)
#define DMCONTROL_ACCESS (7<<12)
#define DMCONTROL_HARTID (0x3ff<<2)
#define DMCONTROL_NDRESET (1<<1)
#define DMCONTROL_FULLRESET 1
#define DMINFO 0x11
#define DMINFO_ABUSSIZE (0x7fU<<25)
#define DMINFO_SERIALCOUNT (0xf<<21)
#define DMINFO_ACCESS128 (1<<20)
#define DMINFO_ACCESS64 (1<<19)
#define DMINFO_ACCESS32 (1<<18)
#define DMINFO_ACCESS16 (1<<17)
#define DMINFO_ACCESS8 (1<<16)
#define DMINFO_DRAMSIZE (0x3f<<10)
#define DMINFO_AUTHENTICATED (1<<5)
#define DMINFO_AUTHBUSY (1<<4)
#define DMINFO_AUTHTYPE (3<<2)
#define DMINFO_VERSION 3
/*** Info about the core being debugged. ***/
#define DBUS_ADDRESS_UNKNOWN 0xffff
#define MAX_HWBPS 16
#define DRAM_CACHE_SIZE 16
uint8_t ir_dtmcontrol[4] = {DTMCONTROL};
struct scan_field select_dtmcontrol = {
.in_value = NULL,
.out_value = ir_dtmcontrol
};
uint8_t ir_dbus[4] = {DBUS};
struct scan_field select_dbus = {
.in_value = NULL,
.out_value = ir_dbus
};
uint8_t ir_idcode[4] = {0x1};
struct scan_field select_idcode = {
.in_value = NULL,
.out_value = ir_idcode
};
bscan_tunnel_type_t bscan_tunnel_type;
int bscan_tunnel_ir_width; /* if zero, then tunneling is not present/active */
static uint8_t bscan_zero[4] = {0};
static uint8_t bscan_one[4] = {1};
uint8_t ir_user4[4] = {0x23};
struct scan_field select_user4 = {
.in_value = NULL,
.out_value = ir_user4
};
uint8_t bscan_tunneled_ir_width[4] = {5}; /* overridden by assignment in riscv_init_target */
struct scan_field _bscan_tunnel_data_register_select_dmi[] = {
{
.num_bits = 3,
.out_value = bscan_zero,
.in_value = NULL,
},
{
.num_bits = 5, /* initialized in riscv_init_target to ir width of DM */
.out_value = ir_dbus,
.in_value = NULL,
},
{
.num_bits = 7,
.out_value = bscan_tunneled_ir_width,
.in_value = NULL,
},
{
.num_bits = 1,
.out_value = bscan_zero,
.in_value = NULL,
}
};
struct scan_field _bscan_tunnel_nested_tap_select_dmi[] = {
{
.num_bits = 1,
.out_value = bscan_zero,
.in_value = NULL,
},
{
.num_bits = 7,
.out_value = bscan_tunneled_ir_width,
.in_value = NULL,
},
{
.num_bits = 0, /* initialized in riscv_init_target to ir width of DM */
.out_value = ir_dbus,
.in_value = NULL,
},
{
.num_bits = 3,
.out_value = bscan_zero,
.in_value = NULL,
}
};
struct scan_field *bscan_tunnel_nested_tap_select_dmi = _bscan_tunnel_nested_tap_select_dmi;
uint32_t bscan_tunnel_nested_tap_select_dmi_num_fields = DIM(_bscan_tunnel_nested_tap_select_dmi);
struct scan_field *bscan_tunnel_data_register_select_dmi = _bscan_tunnel_data_register_select_dmi;
uint32_t bscan_tunnel_data_register_select_dmi_num_fields = DIM(_bscan_tunnel_data_register_select_dmi);
struct trigger {
uint64_t address;
uint32_t length;
uint64_t mask;
uint64_t value;
bool read, write, execute;
int unique_id;
};
/* Wall-clock timeout for a command/access. Settable via RISC-V Target commands.*/
int riscv_command_timeout_sec = DEFAULT_COMMAND_TIMEOUT_SEC;
/* Wall-clock timeout after reset. Settable via RISC-V Target commands.*/
int riscv_reset_timeout_sec = DEFAULT_RESET_TIMEOUT_SEC;
bool riscv_prefer_sba;
bool riscv_enable_virt2phys = true;
bool riscv_ebreakm = true;
bool riscv_ebreaks = true;
bool riscv_ebreaku = true;
bool riscv_enable_virtual;
typedef struct {
uint16_t low, high;
} range_t;
/* In addition to the ones in the standard spec, we'll also expose additional
* CSRs in this list.
* The list is either NULL, or a series of ranges (inclusive), terminated with
* 1,0. */
range_t *expose_csr;
/* Same, but for custom registers. */
range_t *expose_custom;
static enum {
RO_NORMAL,
RO_REVERSED
} resume_order;
virt2phys_info_t sv32 = {
.name = "Sv32",
.va_bits = 32,
.level = 2,
.pte_shift = 2,
.vpn_shift = {12, 22},
.vpn_mask = {0x3ff, 0x3ff},
.pte_ppn_shift = {10, 20},
.pte_ppn_mask = {0x3ff, 0xfff},
.pa_ppn_shift = {12, 22},
.pa_ppn_mask = {0x3ff, 0xfff},
};
virt2phys_info_t sv39 = {
.name = "Sv39",
.va_bits = 39,
.level = 3,
.pte_shift = 3,
.vpn_shift = {12, 21, 30},
.vpn_mask = {0x1ff, 0x1ff, 0x1ff},
.pte_ppn_shift = {10, 19, 28},
.pte_ppn_mask = {0x1ff, 0x1ff, 0x3ffffff},
.pa_ppn_shift = {12, 21, 30},
.pa_ppn_mask = {0x1ff, 0x1ff, 0x3ffffff},
};
virt2phys_info_t sv48 = {
.name = "Sv48",
.va_bits = 48,
.level = 4,
.pte_shift = 3,
.vpn_shift = {12, 21, 30, 39},
.vpn_mask = {0x1ff, 0x1ff, 0x1ff, 0x1ff},
.pte_ppn_shift = {10, 19, 28, 37},
.pte_ppn_mask = {0x1ff, 0x1ff, 0x1ff, 0x1ffff},
.pa_ppn_shift = {12, 21, 30, 39},
.pa_ppn_mask = {0x1ff, 0x1ff, 0x1ff, 0x1ffff},
};
static int riscv_resume_go_all_harts(struct target *target);
void select_dmi_via_bscan(struct target *target)
{
jtag_add_ir_scan(target->tap, &select_user4, TAP_IDLE);
if (bscan_tunnel_type == BSCAN_TUNNEL_DATA_REGISTER)
jtag_add_dr_scan(target->tap, bscan_tunnel_data_register_select_dmi_num_fields,
bscan_tunnel_data_register_select_dmi, TAP_IDLE);
else /* BSCAN_TUNNEL_NESTED_TAP */
jtag_add_dr_scan(target->tap, bscan_tunnel_nested_tap_select_dmi_num_fields,
bscan_tunnel_nested_tap_select_dmi, TAP_IDLE);
}
uint32_t dtmcontrol_scan_via_bscan(struct target *target, uint32_t out)
{
/* On BSCAN TAP: Select IR=USER4, issue tunneled IR scan via BSCAN TAP's DR */
uint8_t tunneled_ir_width[4] = {bscan_tunnel_ir_width};
uint8_t tunneled_dr_width[4] = {32};
uint8_t out_value[5] = {0};
uint8_t in_value[5] = {0};
buf_set_u32(out_value, 0, 32, out);
struct scan_field tunneled_ir[4] = {};
struct scan_field tunneled_dr[4] = {};
if (bscan_tunnel_type == BSCAN_TUNNEL_DATA_REGISTER) {
tunneled_ir[0].num_bits = 3;
tunneled_ir[0].out_value = bscan_zero;
tunneled_ir[0].in_value = NULL;
tunneled_ir[1].num_bits = bscan_tunnel_ir_width;
tunneled_ir[1].out_value = ir_dtmcontrol;
tunneled_ir[1].in_value = NULL;
tunneled_ir[2].num_bits = 7;
tunneled_ir[2].out_value = tunneled_ir_width;
tunneled_ir[2].in_value = NULL;
tunneled_ir[3].num_bits = 1;
tunneled_ir[3].out_value = bscan_zero;
tunneled_ir[3].in_value = NULL;
tunneled_dr[0].num_bits = 3;
tunneled_dr[0].out_value = bscan_zero;
tunneled_dr[0].in_value = NULL;
tunneled_dr[1].num_bits = 32 + 1;
tunneled_dr[1].out_value = out_value;
tunneled_dr[1].in_value = in_value;
tunneled_dr[2].num_bits = 7;
tunneled_dr[2].out_value = tunneled_dr_width;
tunneled_dr[2].in_value = NULL;
tunneled_dr[3].num_bits = 1;
tunneled_dr[3].out_value = bscan_one;
tunneled_dr[3].in_value = NULL;
} else {
/* BSCAN_TUNNEL_NESTED_TAP */
tunneled_ir[3].num_bits = 3;
tunneled_ir[3].out_value = bscan_zero;
tunneled_ir[3].in_value = NULL;
tunneled_ir[2].num_bits = bscan_tunnel_ir_width;
tunneled_ir[2].out_value = ir_dtmcontrol;
tunneled_ir[1].in_value = NULL;
tunneled_ir[1].num_bits = 7;
tunneled_ir[1].out_value = tunneled_ir_width;
tunneled_ir[2].in_value = NULL;
tunneled_ir[0].num_bits = 1;
tunneled_ir[0].out_value = bscan_zero;
tunneled_ir[0].in_value = NULL;
tunneled_dr[3].num_bits = 3;
tunneled_dr[3].out_value = bscan_zero;
tunneled_dr[3].in_value = NULL;
tunneled_dr[2].num_bits = 32 + 1;
tunneled_dr[2].out_value = out_value;
tunneled_dr[2].in_value = in_value;
tunneled_dr[1].num_bits = 7;
tunneled_dr[1].out_value = tunneled_dr_width;
tunneled_dr[1].in_value = NULL;
tunneled_dr[0].num_bits = 1;
tunneled_dr[0].out_value = bscan_one;
tunneled_dr[0].in_value = NULL;
}
jtag_add_ir_scan(target->tap, &select_user4, TAP_IDLE);
jtag_add_dr_scan(target->tap, DIM(tunneled_ir), tunneled_ir, TAP_IDLE);
jtag_add_dr_scan(target->tap, DIM(tunneled_dr), tunneled_dr, TAP_IDLE);
select_dmi_via_bscan(target);
int retval = jtag_execute_queue();
if (retval != ERROR_OK) {
LOG_ERROR("failed jtag scan: %d", retval);
return retval;
}
/* Note the starting offset is bit 1, not bit 0. In BSCAN tunnel, there is a one-bit TCK skew between
output and input */
uint32_t in = buf_get_u32(in_value, 1, 32);
LOG_DEBUG("DTMCS: 0x%x -> 0x%x", out, in);
return in;
}
static uint32_t dtmcontrol_scan(struct target *target, uint32_t out)
{
struct scan_field field;
uint8_t in_value[4];
uint8_t out_value[4] = { 0 };
if (bscan_tunnel_ir_width != 0)
return dtmcontrol_scan_via_bscan(target, out);
buf_set_u32(out_value, 0, 32, out);
jtag_add_ir_scan(target->tap, &select_dtmcontrol, TAP_IDLE);
field.num_bits = 32;
field.out_value = out_value;
field.in_value = in_value;
jtag_add_dr_scan(target->tap, 1, &field, TAP_IDLE);
/* Always return to dbus. */
jtag_add_ir_scan(target->tap, &select_dbus, TAP_IDLE);
int retval = jtag_execute_queue();
if (retval != ERROR_OK) {
LOG_ERROR("failed jtag scan: %d", retval);
return retval;
}
uint32_t in = buf_get_u32(field.in_value, 0, 32);
LOG_DEBUG("DTMCONTROL: 0x%x -> 0x%x", out, in);
return in;
}
static struct target_type *get_target_type(struct target *target)
{
riscv_info_t *info = (riscv_info_t *) target->arch_info;
if (!info) {
LOG_ERROR("Target has not been initialized");
return NULL;
}
switch (info->dtm_version) {
case 0:
return &riscv011_target;
case 1:
return &riscv013_target;
default:
LOG_ERROR("Unsupported DTM version: %d", info->dtm_version);
return NULL;
}
}
static int riscv_init_target(struct command_context *cmd_ctx,
struct target *target)
{
LOG_DEBUG("riscv_init_target()");
target->arch_info = calloc(1, sizeof(riscv_info_t));
if (!target->arch_info)
return ERROR_FAIL;
riscv_info_t *info = (riscv_info_t *) target->arch_info;
riscv_info_init(target, info);
info->cmd_ctx = cmd_ctx;
select_dtmcontrol.num_bits = target->tap->ir_length;
select_dbus.num_bits = target->tap->ir_length;
select_idcode.num_bits = target->tap->ir_length;
if (bscan_tunnel_ir_width != 0) {
select_user4.num_bits = target->tap->ir_length;
bscan_tunneled_ir_width[0] = bscan_tunnel_ir_width;
if (bscan_tunnel_type == BSCAN_TUNNEL_DATA_REGISTER)
bscan_tunnel_data_register_select_dmi[1].num_bits = bscan_tunnel_ir_width;
else /* BSCAN_TUNNEL_NESTED_TAP */
bscan_tunnel_nested_tap_select_dmi[2].num_bits = bscan_tunnel_ir_width;
}
riscv_semihosting_init(target);
target->debug_reason = DBG_REASON_DBGRQ;
return ERROR_OK;
}
static void riscv_free_registers(struct target *target)
{
/* Free the shared structure use for most registers. */
if (target->reg_cache) {
if (target->reg_cache->reg_list) {
free(target->reg_cache->reg_list[0].arch_info);
/* Free the ones we allocated separately. */
for (unsigned i = GDB_REGNO_COUNT; i < target->reg_cache->num_regs; i++)
free(target->reg_cache->reg_list[i].arch_info);
free(target->reg_cache->reg_list);
}
free(target->reg_cache);
}
}
static void riscv_deinit_target(struct target *target)
{
LOG_DEBUG("riscv_deinit_target()");
struct target_type *tt = get_target_type(target);
if (tt) {
tt->deinit_target(target);
riscv_info_t *info = (riscv_info_t *) target->arch_info;
free(info->reg_names);
free(info);
}
riscv_free_registers(target);
target->arch_info = NULL;
}
static void trigger_from_breakpoint(struct trigger *trigger,
const struct breakpoint *breakpoint)
{
trigger->address = breakpoint->address;
trigger->length = breakpoint->length;
trigger->mask = ~0LL;
trigger->read = false;
trigger->write = false;
trigger->execute = true;
/* unique_id is unique across both breakpoints and watchpoints. */
trigger->unique_id = breakpoint->unique_id;
}
static int maybe_add_trigger_t1(struct target *target, unsigned hartid,
struct trigger *trigger, uint64_t tdata1)
{
RISCV_INFO(r);
const uint32_t bpcontrol_x = 1<<0;
const uint32_t bpcontrol_w = 1<<1;
const uint32_t bpcontrol_r = 1<<2;
const uint32_t bpcontrol_u = 1<<3;
const uint32_t bpcontrol_s = 1<<4;
const uint32_t bpcontrol_h = 1<<5;
const uint32_t bpcontrol_m = 1<<6;
const uint32_t bpcontrol_bpmatch = 0xf << 7;
const uint32_t bpcontrol_bpaction = 0xff << 11;
if (tdata1 & (bpcontrol_r | bpcontrol_w | bpcontrol_x)) {
/* Trigger is already in use, presumably by user code. */
return ERROR_TARGET_RESOURCE_NOT_AVAILABLE;
}
tdata1 = set_field(tdata1, bpcontrol_r, trigger->read);
tdata1 = set_field(tdata1, bpcontrol_w, trigger->write);
tdata1 = set_field(tdata1, bpcontrol_x, trigger->execute);
tdata1 = set_field(tdata1, bpcontrol_u,
!!(r->misa[hartid] & (1 << ('U' - 'A'))));
tdata1 = set_field(tdata1, bpcontrol_s,
!!(r->misa[hartid] & (1 << ('S' - 'A'))));
tdata1 = set_field(tdata1, bpcontrol_h,
!!(r->misa[hartid] & (1 << ('H' - 'A'))));
tdata1 |= bpcontrol_m;
tdata1 = set_field(tdata1, bpcontrol_bpmatch, 0); /* exact match */
tdata1 = set_field(tdata1, bpcontrol_bpaction, 0); /* cause bp exception */
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TDATA1, tdata1);
riscv_reg_t tdata1_rb;
if (riscv_get_register_on_hart(target, &tdata1_rb, hartid,
GDB_REGNO_TDATA1) != ERROR_OK)
return ERROR_FAIL;
LOG_DEBUG("tdata1=0x%" PRIx64, tdata1_rb);
if (tdata1 != tdata1_rb) {
LOG_DEBUG("Trigger doesn't support what we need; After writing 0x%"
PRIx64 " to tdata1 it contains 0x%" PRIx64,
tdata1, tdata1_rb);
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TDATA1, 0);
return ERROR_TARGET_RESOURCE_NOT_AVAILABLE;
}
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TDATA2, trigger->address);
return ERROR_OK;
}
static int maybe_add_trigger_t2(struct target *target, unsigned hartid,
struct trigger *trigger, uint64_t tdata1)
{
RISCV_INFO(r);
/* tselect is already set */
if (tdata1 & (MCONTROL_EXECUTE | MCONTROL_STORE | MCONTROL_LOAD)) {
/* Trigger is already in use, presumably by user code. */
return ERROR_TARGET_RESOURCE_NOT_AVAILABLE;
}
/* address/data match trigger */
tdata1 |= MCONTROL_DMODE(riscv_xlen(target));
tdata1 = set_field(tdata1, MCONTROL_ACTION,
MCONTROL_ACTION_DEBUG_MODE);
tdata1 = set_field(tdata1, MCONTROL_MATCH, MCONTROL_MATCH_EQUAL);
tdata1 |= MCONTROL_M;
if (r->misa[hartid] & (1 << ('H' - 'A')))
tdata1 |= MCONTROL_H;
if (r->misa[hartid] & (1 << ('S' - 'A')))
tdata1 |= MCONTROL_S;
if (r->misa[hartid] & (1 << ('U' - 'A')))
tdata1 |= MCONTROL_U;
if (trigger->execute)
tdata1 |= MCONTROL_EXECUTE;
if (trigger->read)
tdata1 |= MCONTROL_LOAD;
if (trigger->write)
tdata1 |= MCONTROL_STORE;
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TDATA1, tdata1);
uint64_t tdata1_rb;
int result = riscv_get_register_on_hart(target, &tdata1_rb, hartid, GDB_REGNO_TDATA1);
if (result != ERROR_OK)
return result;
LOG_DEBUG("tdata1=0x%" PRIx64, tdata1_rb);
if (tdata1 != tdata1_rb) {
LOG_DEBUG("Trigger doesn't support what we need; After writing 0x%"
PRIx64 " to tdata1 it contains 0x%" PRIx64,
tdata1, tdata1_rb);
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TDATA1, 0);
return ERROR_TARGET_RESOURCE_NOT_AVAILABLE;
}
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TDATA2, trigger->address);
return ERROR_OK;
}
static int add_trigger(struct target *target, struct trigger *trigger)
{
RISCV_INFO(r);
if (riscv_enumerate_triggers(target) != ERROR_OK)
return ERROR_FAIL;
/* In RTOS mode, we need to set the same trigger in the same slot on every
* hart, to keep up the illusion that each hart is a thread running on the
* same core. */
/* Otherwise, we just set the trigger on the one hart this target deals
* with. */
riscv_reg_t tselect[RISCV_MAX_HARTS];
int first_hart = -1;
for (int hartid = 0; hartid < riscv_count_harts(target); ++hartid) {
if (!riscv_hart_enabled(target, hartid))
continue;
if (first_hart < 0)
first_hart = hartid;
int result = riscv_get_register_on_hart(target, &tselect[hartid],
hartid, GDB_REGNO_TSELECT);
if (result != ERROR_OK)
return result;
}
assert(first_hart >= 0);
unsigned int i;
for (i = 0; i < r->trigger_count[first_hart]; i++) {
if (r->trigger_unique_id[i] != -1)
continue;
riscv_set_register_on_hart(target, first_hart, GDB_REGNO_TSELECT, i);
uint64_t tdata1;
int result = riscv_get_register_on_hart(target, &tdata1, first_hart,
GDB_REGNO_TDATA1);
if (result != ERROR_OK)
return result;
int type = get_field(tdata1, MCONTROL_TYPE(riscv_xlen(target)));
result = ERROR_OK;
for (int hartid = first_hart; hartid < riscv_count_harts(target); ++hartid) {
if (!riscv_hart_enabled(target, hartid))
continue;
if (hartid > first_hart)
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TSELECT, i);
switch (type) {
case 1:
result = maybe_add_trigger_t1(target, hartid, trigger, tdata1);
break;
case 2:
result = maybe_add_trigger_t2(target, hartid, trigger, tdata1);
break;
default:
LOG_DEBUG("trigger %d has unknown type %d", i, type);
continue;
}
if (result != ERROR_OK)
continue;
}
if (result != ERROR_OK)
continue;
LOG_DEBUG("[%d] Using trigger %d (type %d) for bp %d", target->coreid,
i, type, trigger->unique_id);
r->trigger_unique_id[i] = trigger->unique_id;
break;
}
for (int hartid = first_hart; hartid < riscv_count_harts(target); ++hartid) {
if (!riscv_hart_enabled(target, hartid))
continue;
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TSELECT,
tselect[hartid]);
}
if (i >= r->trigger_count[first_hart]) {
LOG_ERROR("Couldn't find an available hardware trigger.");
return ERROR_TARGET_RESOURCE_NOT_AVAILABLE;
}
return ERROR_OK;
}
int riscv_add_breakpoint(struct target *target, struct breakpoint *breakpoint)
{
LOG_DEBUG("[%d] @0x%" TARGET_PRIxADDR, target->coreid, breakpoint->address);
assert(breakpoint);
if (breakpoint->type == BKPT_SOFT) {
/** @todo check RVC for size/alignment */
if (!(breakpoint->length == 4 || breakpoint->length == 2)) {
LOG_ERROR("Invalid breakpoint length %d", breakpoint->length);
return ERROR_FAIL;
}
if (0 != (breakpoint->address % 2)) {
LOG_ERROR("Invalid breakpoint alignment for address 0x%" TARGET_PRIxADDR, breakpoint->address);
return ERROR_FAIL;
}
if (target_read_memory(target, breakpoint->address, 2, breakpoint->length / 2,
breakpoint->orig_instr) != ERROR_OK) {
LOG_ERROR("Failed to read original instruction at 0x%" TARGET_PRIxADDR,
breakpoint->address);
return ERROR_FAIL;
}
uint8_t buff[4] = { 0 };
buf_set_u32(buff, 0, breakpoint->length * CHAR_BIT, breakpoint->length == 4 ? ebreak() : ebreak_c());
int const retval = target_write_memory(target, breakpoint->address, 2, breakpoint->length / 2, buff);
if (retval != ERROR_OK) {
LOG_ERROR("Failed to write %d-byte breakpoint instruction at 0x%"
TARGET_PRIxADDR, breakpoint->length, breakpoint->address);
return ERROR_FAIL;
}
} else if (breakpoint->type == BKPT_HARD) {
struct trigger trigger;
trigger_from_breakpoint(&trigger, breakpoint);
int const result = add_trigger(target, &trigger);
if (result != ERROR_OK)
return result;
} else {
LOG_INFO("OpenOCD only supports hardware and software breakpoints.");
return ERROR_TARGET_RESOURCE_NOT_AVAILABLE;
}
breakpoint->set = true;
return ERROR_OK;
}
static int remove_trigger(struct target *target, struct trigger *trigger)
{
RISCV_INFO(r);
if (riscv_enumerate_triggers(target) != ERROR_OK)
return ERROR_FAIL;
int first_hart = -1;
for (int hartid = 0; hartid < riscv_count_harts(target); ++hartid) {
if (!riscv_hart_enabled(target, hartid))
continue;
if (first_hart < 0) {
first_hart = hartid;
break;
}
}
assert(first_hart >= 0);
unsigned int i;
for (i = 0; i < r->trigger_count[first_hart]; i++) {
if (r->trigger_unique_id[i] == trigger->unique_id)
break;
}
if (i >= r->trigger_count[first_hart]) {
LOG_ERROR("Couldn't find the hardware resources used by hardware "
"trigger.");
return ERROR_FAIL;
}
LOG_DEBUG("[%d] Stop using resource %d for bp %d", target->coreid, i,
trigger->unique_id);
for (int hartid = first_hart; hartid < riscv_count_harts(target); ++hartid) {
if (!riscv_hart_enabled(target, hartid))
continue;
riscv_reg_t tselect;
int result = riscv_get_register_on_hart(target, &tselect, hartid, GDB_REGNO_TSELECT);
if (result != ERROR_OK)
return result;
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TSELECT, i);
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TDATA1, 0);
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TSELECT, tselect);
}
r->trigger_unique_id[i] = -1;
return ERROR_OK;
}
int riscv_remove_breakpoint(struct target *target,
struct breakpoint *breakpoint)
{
if (breakpoint->type == BKPT_SOFT) {
if (target_write_memory(target, breakpoint->address, 2, breakpoint->length / 2,
breakpoint->orig_instr) != ERROR_OK) {
LOG_ERROR("Failed to restore instruction for %d-byte breakpoint at "
"0x%" TARGET_PRIxADDR, breakpoint->length, breakpoint->address);
return ERROR_FAIL;
}
} else if (breakpoint->type == BKPT_HARD) {
struct trigger trigger;
trigger_from_breakpoint(&trigger, breakpoint);
int result = remove_trigger(target, &trigger);
if (result != ERROR_OK)
return result;
} else {
LOG_INFO("OpenOCD only supports hardware and software breakpoints.");
return ERROR_TARGET_RESOURCE_NOT_AVAILABLE;
}
breakpoint->set = false;
return ERROR_OK;
}
static void trigger_from_watchpoint(struct trigger *trigger,
const struct watchpoint *watchpoint)
{
trigger->address = watchpoint->address;
trigger->length = watchpoint->length;
trigger->mask = watchpoint->mask;
trigger->value = watchpoint->value;
trigger->read = (watchpoint->rw == WPT_READ || watchpoint->rw == WPT_ACCESS);
trigger->write = (watchpoint->rw == WPT_WRITE || watchpoint->rw == WPT_ACCESS);
trigger->execute = false;
/* unique_id is unique across both breakpoints and watchpoints. */
trigger->unique_id = watchpoint->unique_id;
}
int riscv_add_watchpoint(struct target *target, struct watchpoint *watchpoint)
{
struct trigger trigger;
trigger_from_watchpoint(&trigger, watchpoint);
int result = add_trigger(target, &trigger);
if (result != ERROR_OK)
return result;
watchpoint->set = true;
return ERROR_OK;
}
int riscv_remove_watchpoint(struct target *target,
struct watchpoint *watchpoint)
{
LOG_DEBUG("[%d] @0x%" TARGET_PRIxADDR, target->coreid, watchpoint->address);
struct trigger trigger;
trigger_from_watchpoint(&trigger, watchpoint);
int result = remove_trigger(target, &trigger);
if (result != ERROR_OK)
return result;
watchpoint->set = false;
return ERROR_OK;
}
/* Sets *hit_watchpoint to the first watchpoint identified as causing the
* current halt.
*
* The GDB server uses this information to tell GDB what data address has
* been hit, which enables GDB to print the hit variable along with its old
* and new value. */
int riscv_hit_watchpoint(struct target *target, struct watchpoint **hit_watchpoint)
{
struct watchpoint *wp = target->watchpoints;
if (riscv_rtos_enabled(target))
riscv_set_current_hartid(target, target->rtos->current_thread - 1);
LOG_DEBUG("Current hartid = %d", riscv_current_hartid(target));
/*TODO instead of disassembling the instruction that we think caused the
* trigger, check the hit bit of each watchpoint first. The hit bit is
* simpler and more reliable to check but as it is optional and relatively
* new, not all hardware will implement it */
riscv_reg_t dpc;
riscv_get_register(target, &dpc, GDB_REGNO_DPC);
const uint8_t length = 4;
LOG_DEBUG("dpc is 0x%" PRIx64, dpc);
/* fetch the instruction at dpc */
uint8_t buffer[length];
if (target_read_buffer(target, dpc, length, buffer) != ERROR_OK) {
LOG_ERROR("Failed to read instruction at dpc 0x%" PRIx64, dpc);
return ERROR_FAIL;
}
uint32_t instruction = 0;
for (int i = 0; i < length; i++) {
LOG_DEBUG("Next byte is %x", buffer[i]);
instruction += (buffer[i] << 8 * i);
}
LOG_DEBUG("Full instruction is %x", instruction);
/* find out which memory address is accessed by the instruction at dpc */
/* opcode is first 7 bits of the instruction */
uint8_t opcode = instruction & 0x7F;
uint32_t rs1;
int16_t imm;
riscv_reg_t mem_addr;
if (opcode == MATCH_LB || opcode == MATCH_SB) {
rs1 = (instruction & 0xf8000) >> 15;
riscv_get_register(target, &mem_addr, rs1);
if (opcode == MATCH_SB) {
LOG_DEBUG("%x is store instruction", instruction);
imm = ((instruction & 0xf80) >> 7) | ((instruction & 0xfe000000) >> 20);
} else {
LOG_DEBUG("%x is load instruction", instruction);
imm = (instruction & 0xfff00000) >> 20;
}
/* sign extend 12-bit imm to 16-bits */
if (imm & (1 << 11))
imm |= 0xf000;
mem_addr += imm;
LOG_DEBUG("memory address=0x%" PRIx64, mem_addr);
} else {
LOG_DEBUG("%x is not a RV32I load or store", instruction);
return ERROR_FAIL;
}
while (wp) {
/*TODO support length/mask */
if (wp->address == mem_addr) {
*hit_watchpoint = wp;
LOG_DEBUG("Hit address=%" TARGET_PRIxADDR, wp->address);
return ERROR_OK;
}
wp = wp->next;
}
/* No match found - either we hit a watchpoint caused by an instruction that
* this function does not yet disassemble, or we hit a breakpoint.
*
* OpenOCD will behave as if this function had never been implemented i.e.
* report the halt to GDB with no address information. */
return ERROR_FAIL;
}
static int oldriscv_step(struct target *target, int current, uint32_t address,
int handle_breakpoints)
{
struct target_type *tt = get_target_type(target);
return tt->step(target, current, address, handle_breakpoints);
}
static int old_or_new_riscv_step(struct target *target, int current,
target_addr_t address, int handle_breakpoints)
{
RISCV_INFO(r);
LOG_DEBUG("handle_breakpoints=%d", handle_breakpoints);
if (r->is_halted == NULL)
return oldriscv_step(target, current, address, handle_breakpoints);
else
return riscv_openocd_step(target, current, address, handle_breakpoints);
}
static int riscv_examine(struct target *target)
{
LOG_DEBUG("riscv_examine()");
if (target_was_examined(target)) {
LOG_DEBUG("Target was already examined.");
return ERROR_OK;
}
/* Don't need to select dbus, since the first thing we do is read dtmcontrol. */
riscv_info_t *info = (riscv_info_t *) target->arch_info;
uint32_t dtmcontrol = dtmcontrol_scan(target, 0);
LOG_DEBUG("dtmcontrol=0x%x", dtmcontrol);
info->dtm_version = get_field(dtmcontrol, DTMCONTROL_VERSION);
LOG_DEBUG(" version=0x%x", info->dtm_version);
struct target_type *tt = get_target_type(target);
if (tt == NULL)
return ERROR_FAIL;
int result = tt->init_target(info->cmd_ctx, target);
if (result != ERROR_OK)
return result;
return tt->examine(target);
}
static int oldriscv_poll(struct target *target)
{
struct target_type *tt = get_target_type(target);
return tt->poll(target);
}
static int old_or_new_riscv_poll(struct target *target)
{
RISCV_INFO(r);
if (r->is_halted == NULL)
return oldriscv_poll(target);
else
return riscv_openocd_poll(target);
}
int halt_prep(struct target *target)
{
RISCV_INFO(r);
for (int i = 0; i < riscv_count_harts(target); ++i) {
if (!riscv_hart_enabled(target, i))
continue;
LOG_DEBUG("[%s] prep hart, debug_reason=%d", target_name(target),
target->debug_reason);
if (riscv_set_current_hartid(target, i) != ERROR_OK)
return ERROR_FAIL;
if (riscv_is_halted(target)) {
LOG_DEBUG("Hart %d is already halted (reason=%d).", i,
target->debug_reason);
} else {
if (r->halt_prep(target) != ERROR_OK)
return ERROR_FAIL;
r->prepped = true;
}
}
return ERROR_OK;
}
int riscv_halt_go_all_harts(struct target *target)
{
RISCV_INFO(r);
for (int i = 0; i < riscv_count_harts(target); ++i) {
if (!riscv_hart_enabled(target, i))
continue;
if (riscv_set_current_hartid(target, i) != ERROR_OK)
return ERROR_FAIL;
if (riscv_is_halted(target)) {
LOG_DEBUG("Hart %d is already halted.", i);
} else {
if (r->halt_go(target) != ERROR_OK)
return ERROR_FAIL;
}
}
riscv_invalidate_register_cache(target);
return ERROR_OK;
}
int halt_go(struct target *target)
{
riscv_info_t *r = riscv_info(target);
int result;
if (r->is_halted == NULL) {
struct target_type *tt = get_target_type(target);
result = tt->halt(target);
} else {
result = riscv_halt_go_all_harts(target);
}
target->state = TARGET_HALTED;
if (target->debug_reason == DBG_REASON_NOTHALTED)
target->debug_reason = DBG_REASON_DBGRQ;
return result;
}
static int halt_finish(struct target *target)
{
return target_call_event_callbacks(target, TARGET_EVENT_HALTED);
}
int riscv_halt(struct target *target)
{
RISCV_INFO(r);
if (r->is_halted == NULL) {
struct target_type *tt = get_target_type(target);
return tt->halt(target);
}
LOG_DEBUG("[%d] halting all harts", target->coreid);
int result = ERROR_OK;
if (target->smp) {
for (struct target_list *tlist = target->head; tlist; tlist = tlist->next) {
struct target *t = tlist->target;
if (halt_prep(t) != ERROR_OK)
result = ERROR_FAIL;
}
for (struct target_list *tlist = target->head; tlist; tlist = tlist->next) {
struct target *t = tlist->target;
riscv_info_t *i = riscv_info(t);
if (i->prepped) {
if (halt_go(t) != ERROR_OK)
result = ERROR_FAIL;
}
}
for (struct target_list *tlist = target->head; tlist; tlist = tlist->next) {
struct target *t = tlist->target;
if (halt_finish(t) != ERROR_OK)
return ERROR_FAIL;
}
} else {
if (halt_prep(target) != ERROR_OK)
result = ERROR_FAIL;
if (halt_go(target) != ERROR_OK)
result = ERROR_FAIL;
if (halt_finish(target) != ERROR_OK)
return ERROR_FAIL;
}
if (riscv_rtos_enabled(target)) {
if (r->rtos_hartid != -1) {
LOG_DEBUG("halt requested on RTOS hartid %d", r->rtos_hartid);
target->rtos->current_threadid = r->rtos_hartid + 1;
target->rtos->current_thread = r->rtos_hartid + 1;
} else
LOG_DEBUG("halt requested, but no known RTOS hartid");
}
return result;
}
static int riscv_assert_reset(struct target *target)
{
LOG_DEBUG("[%d]", target->coreid);
struct target_type *tt = get_target_type(target);
riscv_invalidate_register_cache(target);
return tt->assert_reset(target);
}
static int riscv_deassert_reset(struct target *target)
{
LOG_DEBUG("[%d]", target->coreid);
struct target_type *tt = get_target_type(target);
return tt->deassert_reset(target);
}
int riscv_resume_prep_all_harts(struct target *target)
{
RISCV_INFO(r);
for (int i = 0; i < riscv_count_harts(target); ++i) {
if (!riscv_hart_enabled(target, i))
continue;
LOG_DEBUG("prep hart %d", i);
if (riscv_set_current_hartid(target, i) != ERROR_OK)
return ERROR_FAIL;
if (riscv_is_halted(target)) {
if (r->resume_prep(target) != ERROR_OK)
return ERROR_FAIL;
} else {
LOG_DEBUG(" hart %d requested resume, but was already resumed", i);
}
}
LOG_DEBUG("[%d] mark as prepped", target->coreid);
r->prepped = true;
return ERROR_OK;
}
/* state must be riscv_reg_t state[RISCV_MAX_HWBPS] = {0}; */
static int disable_triggers(struct target *target, riscv_reg_t *state)
{
RISCV_INFO(r);
LOG_DEBUG("deal with triggers");
if (riscv_enumerate_triggers(target) != ERROR_OK)
return ERROR_FAIL;
int hartid = riscv_current_hartid(target);
if (r->manual_hwbp_set) {
/* Look at every trigger that may have been set. */
riscv_reg_t tselect;
if (riscv_get_register(target, &tselect, GDB_REGNO_TSELECT) != ERROR_OK)
return ERROR_FAIL;
for (unsigned t = 0; t < r->trigger_count[hartid]; t++) {
if (riscv_set_register(target, GDB_REGNO_TSELECT, t) != ERROR_OK)
return ERROR_FAIL;
riscv_reg_t tdata1;
if (riscv_get_register(target, &tdata1, GDB_REGNO_TDATA1) != ERROR_OK)
return ERROR_FAIL;
if (tdata1 & MCONTROL_DMODE(riscv_xlen(target))) {
state[t] = tdata1;
if (riscv_set_register(target, GDB_REGNO_TDATA1, 0) != ERROR_OK)
return ERROR_FAIL;
}
}
if (riscv_set_register(target, GDB_REGNO_TSELECT, tselect) != ERROR_OK)
return ERROR_FAIL;
} else {
/* Just go through the triggers we manage. */
struct watchpoint *watchpoint = target->watchpoints;
int i = 0;
while (watchpoint) {
LOG_DEBUG("watchpoint %d: set=%d", i, watchpoint->set);
state[i] = watchpoint->set;
if (watchpoint->set) {
if (riscv_remove_watchpoint(target, watchpoint) != ERROR_OK)
return ERROR_FAIL;
}
watchpoint = watchpoint->next;
i++;
}
}
return ERROR_OK;
}
static int enable_triggers(struct target *target, riscv_reg_t *state)
{
RISCV_INFO(r);
int hartid = riscv_current_hartid(target);
if (r->manual_hwbp_set) {
/* Look at every trigger that may have been set. */
riscv_reg_t tselect;
if (riscv_get_register(target, &tselect, GDB_REGNO_TSELECT) != ERROR_OK)
return ERROR_FAIL;
for (unsigned t = 0; t < r->trigger_count[hartid]; t++) {
if (state[t] != 0) {
if (riscv_set_register(target, GDB_REGNO_TSELECT, t) != ERROR_OK)
return ERROR_FAIL;
if (riscv_set_register(target, GDB_REGNO_TDATA1, state[t]) != ERROR_OK)
return ERROR_FAIL;
}
}
if (riscv_set_register(target, GDB_REGNO_TSELECT, tselect) != ERROR_OK)
return ERROR_FAIL;
} else {
struct watchpoint *watchpoint = target->watchpoints;
int i = 0;
while (watchpoint) {
LOG_DEBUG("watchpoint %d: cleared=%" PRId64, i, state[i]);
if (state[i]) {
if (riscv_add_watchpoint(target, watchpoint) != ERROR_OK)
return ERROR_FAIL;
}
watchpoint = watchpoint->next;
i++;
}
}
return ERROR_OK;
}
/**
* Get everything ready to resume.
*/
static int resume_prep(struct target *target, int current,
target_addr_t address, int handle_breakpoints, int debug_execution)
{
RISCV_INFO(r);
LOG_DEBUG("[%d]", target->coreid);
if (!current)
riscv_set_register(target, GDB_REGNO_PC, address);
if (target->debug_reason == DBG_REASON_WATCHPOINT) {
/* To be able to run off a trigger, disable all the triggers, step, and
* then resume as usual. */
riscv_reg_t trigger_state[RISCV_MAX_HWBPS] = {0};
if (disable_triggers(target, trigger_state) != ERROR_OK)
return ERROR_FAIL;
if (old_or_new_riscv_step(target, true, 0, false) != ERROR_OK)
return ERROR_FAIL;
if (enable_triggers(target, trigger_state) != ERROR_OK)
return ERROR_FAIL;
}
if (r->is_halted) {
if (riscv_resume_prep_all_harts(target) != ERROR_OK)
return ERROR_FAIL;
}
LOG_DEBUG("[%d] mark as prepped", target->coreid);
r->prepped = true;
return ERROR_OK;
}
/**
* Resume all the harts that have been prepped, as close to instantaneous as
* possible.
*/
static int resume_go(struct target *target, int current,
target_addr_t address, int handle_breakpoints, int debug_execution)
{
riscv_info_t *r = riscv_info(target);
int result;
if (r->is_halted == NULL) {
struct target_type *tt = get_target_type(target);
result = tt->resume(target, current, address, handle_breakpoints,
debug_execution);
} else {
result = riscv_resume_go_all_harts(target);
}
return result;
}
static int resume_finish(struct target *target)
{
register_cache_invalidate(target->reg_cache);
target->state = TARGET_RUNNING;
target->debug_reason = DBG_REASON_NOTHALTED;
return target_call_event_callbacks(target, TARGET_EVENT_RESUMED);
}
/**
* @par single_hart When true, only resume a single hart even if SMP is
* configured. This is used to run algorithms on just one hart.
*/
int riscv_resume(
struct target *target,
int current,
target_addr_t address,
int handle_breakpoints,
int debug_execution,
bool single_hart)
{
LOG_DEBUG("handle_breakpoints=%d", handle_breakpoints);
int result = ERROR_OK;
if (target->smp && !single_hart) {
for (struct target_list *tlist = target->head; tlist; tlist = tlist->next) {
struct target *t = tlist->target;
if (resume_prep(t, current, address, handle_breakpoints,
debug_execution) != ERROR_OK)
result = ERROR_FAIL;
}
for (struct target_list *tlist = target->head; tlist; tlist = tlist->next) {
struct target *t = tlist->target;
riscv_info_t *i = riscv_info(t);
if (i->prepped) {
if (resume_go(t, current, address, handle_breakpoints,
debug_execution) != ERROR_OK)
result = ERROR_FAIL;
}
}
for (struct target_list *tlist = target->head; tlist; tlist = tlist->next) {
struct target *t = tlist->target;
if (resume_finish(t) != ERROR_OK)
return ERROR_FAIL;
}
} else {
if (resume_prep(target, current, address, handle_breakpoints,
debug_execution) != ERROR_OK)
result = ERROR_FAIL;
if (resume_go(target, current, address, handle_breakpoints,
debug_execution) != ERROR_OK)
result = ERROR_FAIL;
if (resume_finish(target) != ERROR_OK)
return ERROR_FAIL;
}
return result;
}
static int riscv_target_resume(struct target *target, int current, target_addr_t address,
int handle_breakpoints, int debug_execution)
{
return riscv_resume(target, current, address, handle_breakpoints,
debug_execution, false);
}
static int riscv_select_current_hart(struct target *target)
{
RISCV_INFO(r);
if (riscv_rtos_enabled(target)) {
if (r->rtos_hartid == -1)
r->rtos_hartid = target->rtos->current_threadid - 1;
return riscv_set_current_hartid(target, r->rtos_hartid);
} else
return riscv_set_current_hartid(target, target->coreid);
}
static int riscv_mmu(struct target *target, int *enabled)
{
if (!riscv_enable_virt2phys) {
*enabled = 0;
return ERROR_OK;
}
if (riscv_rtos_enabled(target))
riscv_set_current_hartid(target, target->rtos->current_thread - 1);
/* Don't use MMU in explicit or effective M (machine) mode */
riscv_reg_t priv;
if (riscv_get_register(target, &priv, GDB_REGNO_PRIV) != ERROR_OK) {
LOG_ERROR("Failed to read priv register.");
return ERROR_FAIL;
}
riscv_reg_t mstatus;
if (riscv_get_register(target, &mstatus, GDB_REGNO_MSTATUS) != ERROR_OK) {
LOG_ERROR("Failed to read mstatus register.");
return ERROR_FAIL;
}
if ((get_field(mstatus, MSTATUS_MPRV) ? get_field(mstatus, MSTATUS_MPP) : priv) == PRV_M) {
LOG_DEBUG("SATP/MMU ignored in Machine mode (mstatus=0x%" PRIx64 ").", mstatus);
*enabled = 0;
return ERROR_OK;
}
riscv_reg_t satp;
if (riscv_get_register(target, &satp, GDB_REGNO_SATP) != ERROR_OK) {
LOG_DEBUG("Couldn't read SATP.");
/* If we can't read SATP, then there must not be an MMU. */
*enabled = 0;
return ERROR_OK;
}
if (get_field(satp, RISCV_SATP_MODE(riscv_xlen(target))) == SATP_MODE_OFF) {
LOG_DEBUG("MMU is disabled.");
*enabled = 0;
} else {
LOG_DEBUG("MMU is enabled.");
*enabled = 1;
}
return ERROR_OK;
}
static int riscv_address_translate(struct target *target,
target_addr_t virtual, target_addr_t *physical)
{
RISCV_INFO(r);
riscv_reg_t satp_value;
int mode;
uint64_t ppn_value;
target_addr_t table_address;
virt2phys_info_t *info;
uint64_t pte;
int i;
if (riscv_rtos_enabled(target))
riscv_set_current_hartid(target, target->rtos->current_thread - 1);
int result = riscv_get_register(target, &satp_value, GDB_REGNO_SATP);
if (result != ERROR_OK)
return result;
unsigned xlen = riscv_xlen(target);
mode = get_field(satp_value, RISCV_SATP_MODE(xlen));
switch (mode) {
case SATP_MODE_SV32:
info = &sv32;
break;
case SATP_MODE_SV39:
info = &sv39;
break;
case SATP_MODE_SV48:
info = &sv48;
break;
case SATP_MODE_OFF:
LOG_ERROR("No translation or protection." \
" (satp: 0x%" PRIx64 ")", satp_value);
return ERROR_FAIL;
default:
LOG_ERROR("The translation mode is not supported." \
" (satp: 0x%" PRIx64 ")", satp_value);
return ERROR_FAIL;
}
LOG_DEBUG("virtual=0x%" TARGET_PRIxADDR "; mode=%s", virtual, info->name);
/* verify bits xlen-1:va_bits-1 are all equal */
target_addr_t mask = ((target_addr_t)1 << (xlen - (info->va_bits - 1))) - 1;
target_addr_t masked_msbs = (virtual >> (info->va_bits - 1)) & mask;
if (masked_msbs != 0 && masked_msbs != mask) {
LOG_ERROR("Virtual address 0x%" TARGET_PRIxADDR " is not sign-extended "
"for %s mode.", virtual, info->name);
return ERROR_FAIL;
}
ppn_value = get_field(satp_value, RISCV_SATP_PPN(xlen));
table_address = ppn_value << RISCV_PGSHIFT;
i = info->level - 1;
while (i >= 0) {
uint64_t vpn = virtual >> info->vpn_shift[i];
vpn &= info->vpn_mask[i];
target_addr_t pte_address = table_address +
(vpn << info->pte_shift);
uint8_t buffer[8];
assert(info->pte_shift <= 3);
int retval = r->read_memory(target, pte_address,
4, (1 << info->pte_shift) / 4, buffer, 4);
if (retval != ERROR_OK)
return ERROR_FAIL;
if (info->pte_shift == 2)
pte = buf_get_u32(buffer, 0, 32);
else
pte = buf_get_u64(buffer, 0, 64);
LOG_DEBUG("i=%d; PTE @0x%" TARGET_PRIxADDR " = 0x%" PRIx64, i,
pte_address, pte);
if (!(pte & PTE_V) || (!(pte & PTE_R) && (pte & PTE_W)))
return ERROR_FAIL;
if ((pte & PTE_R) || (pte & PTE_X)) /* Found leaf PTE. */
break;
i--;
if (i < 0)
break;
ppn_value = pte >> PTE_PPN_SHIFT;
table_address = ppn_value << RISCV_PGSHIFT;
}
if (i < 0) {
LOG_ERROR("Couldn't find the PTE.");
return ERROR_FAIL;
}
/* Make sure to clear out the high bits that may be set. */
*physical = virtual & (((target_addr_t)1 << info->va_bits) - 1);
while (i < info->level) {
ppn_value = pte >> info->pte_ppn_shift[i];
ppn_value &= info->pte_ppn_mask[i];
*physical &= ~(((target_addr_t)info->pa_ppn_mask[i]) <<
info->pa_ppn_shift[i]);
*physical |= (ppn_value << info->pa_ppn_shift[i]);
i++;
}
LOG_DEBUG("0x%" TARGET_PRIxADDR " -> 0x%" TARGET_PRIxADDR, virtual,
*physical);
return ERROR_OK;
}
static int riscv_virt2phys(struct target *target, target_addr_t virtual, target_addr_t *physical)
{
int enabled;
if (riscv_mmu(target, &enabled) == ERROR_OK) {
if (!enabled)
return ERROR_FAIL;
if (riscv_address_translate(target, virtual, physical) == ERROR_OK)
return ERROR_OK;
}
return ERROR_FAIL;
}
static int riscv_read_phys_memory(struct target *target, target_addr_t phys_address,
uint32_t size, uint32_t count, uint8_t *buffer)
{
RISCV_INFO(r);
if (riscv_select_current_hart(target) != ERROR_OK)
return ERROR_FAIL;
return r->read_memory(target, phys_address, size, count, buffer, size);
}
static int riscv_read_memory(struct target *target, target_addr_t address,
uint32_t size, uint32_t count, uint8_t *buffer)
{
if (count == 0) {
LOG_WARNING("0-length read from 0x%" TARGET_PRIxADDR, address);
return ERROR_OK;
}
if (riscv_select_current_hart(target) != ERROR_OK)
return ERROR_FAIL;
target_addr_t physical_addr;
if (target->type->virt2phys(target, address, &physical_addr) == ERROR_OK)
address = physical_addr;
RISCV_INFO(r);
return r->read_memory(target, address, size, count, buffer, size);
}
static int riscv_write_phys_memory(struct target *target, target_addr_t phys_address,
uint32_t size, uint32_t count, const uint8_t *buffer)
{
if (riscv_select_current_hart(target) != ERROR_OK)
return ERROR_FAIL;
struct target_type *tt = get_target_type(target);
return tt->write_memory(target, phys_address, size, count, buffer);
}
static int riscv_write_memory(struct target *target, target_addr_t address,
uint32_t size, uint32_t count, const uint8_t *buffer)
{
if (count == 0) {
LOG_WARNING("0-length write to 0x%" TARGET_PRIxADDR, address);
return ERROR_OK;
}
if (riscv_select_current_hart(target) != ERROR_OK)
return ERROR_FAIL;
target_addr_t physical_addr;
if (target->type->virt2phys(target, address, &physical_addr) == ERROR_OK)
address = physical_addr;
struct target_type *tt = get_target_type(target);
return tt->write_memory(target, address, size, count, buffer);
}
static int riscv_get_gdb_reg_list_internal(struct target *target,
struct reg **reg_list[], int *reg_list_size,
enum target_register_class reg_class, bool read)
{
RISCV_INFO(r);
LOG_DEBUG("rtos_hartid=%d, current_hartid=%d, reg_class=%d, read=%d",
r->rtos_hartid, r->current_hartid, reg_class, read);
if (!target->reg_cache) {
LOG_ERROR("Target not initialized. Return ERROR_FAIL.");
return ERROR_FAIL;
}
if (riscv_select_current_hart(target) != ERROR_OK)
return ERROR_FAIL;
switch (reg_class) {
case REG_CLASS_GENERAL:
*reg_list_size = 33;
break;
case REG_CLASS_ALL:
*reg_list_size = target->reg_cache->num_regs;
break;
default:
LOG_ERROR("Unsupported reg_class: %d", reg_class);
return ERROR_FAIL;
}
*reg_list = calloc(*reg_list_size, sizeof(struct reg *));
if (!*reg_list)
return ERROR_FAIL;
for (int i = 0; i < *reg_list_size; i++) {
assert(!target->reg_cache->reg_list[i].valid ||
target->reg_cache->reg_list[i].size > 0);
(*reg_list)[i] = &target->reg_cache->reg_list[i];
if (read &&
target->reg_cache->reg_list[i].exist &&
!target->reg_cache->reg_list[i].valid) {
if (target->reg_cache->reg_list[i].type->get(
&target->reg_cache->reg_list[i]) != ERROR_OK)
return ERROR_FAIL;
}
}
return ERROR_OK;
}
static int riscv_get_gdb_reg_list_noread(struct target *target,
struct reg **reg_list[], int *reg_list_size,
enum target_register_class reg_class)
{
return riscv_get_gdb_reg_list_internal(target, reg_list, reg_list_size,
reg_class, false);
}
static int riscv_get_gdb_reg_list(struct target *target,
struct reg **reg_list[], int *reg_list_size,
enum target_register_class reg_class)
{
return riscv_get_gdb_reg_list_internal(target, reg_list, reg_list_size,
reg_class, true);
}
static int riscv_arch_state(struct target *target)
{
struct target_type *tt = get_target_type(target);
return tt->arch_state(target);
}
/* Algorithm must end with a software breakpoint instruction. */
static int riscv_run_algorithm(struct target *target, int num_mem_params,
struct mem_param *mem_params, int num_reg_params,
struct reg_param *reg_params, target_addr_t entry_point,
target_addr_t exit_point, int timeout_ms, void *arch_info)
{
riscv_info_t *info = (riscv_info_t *) target->arch_info;
int hartid = riscv_current_hartid(target);
if (num_mem_params > 0) {
LOG_ERROR("Memory parameters are not supported for RISC-V algorithms.");
return ERROR_FAIL;
}
if (target->state != TARGET_HALTED) {
LOG_WARNING("target not halted");
return ERROR_TARGET_NOT_HALTED;
}
/* Save registers */
struct reg *reg_pc = register_get_by_name(target->reg_cache, "pc", 1);
if (!reg_pc || reg_pc->type->get(reg_pc) != ERROR_OK)
return ERROR_FAIL;
uint64_t saved_pc = buf_get_u64(reg_pc->value, 0, reg_pc->size);
LOG_DEBUG("saved_pc=0x%" PRIx64, saved_pc);
uint64_t saved_regs[32];
for (int i = 0; i < num_reg_params; i++) {
LOG_DEBUG("save %s", reg_params[i].reg_name);
struct reg *r = register_get_by_name(target->reg_cache, reg_params[i].reg_name, 0);
if (!r) {
LOG_ERROR("Couldn't find register named '%s'", reg_params[i].reg_name);
return ERROR_FAIL;
}
if (r->size != reg_params[i].size) {
LOG_ERROR("Register %s is %d bits instead of %d bits.",
reg_params[i].reg_name, r->size, reg_params[i].size);
return ERROR_FAIL;
}
if (r->number > GDB_REGNO_XPR31) {
LOG_ERROR("Only GPRs can be use as argument registers.");
return ERROR_FAIL;
}
if (r->type->get(r) != ERROR_OK)
return ERROR_FAIL;
saved_regs[r->number] = buf_get_u64(r->value, 0, r->size);
if (reg_params[i].direction == PARAM_OUT || reg_params[i].direction == PARAM_IN_OUT) {
if (r->type->set(r, reg_params[i].value) != ERROR_OK)
return ERROR_FAIL;
}
}
/* Disable Interrupts before attempting to run the algorithm. */
uint64_t current_mstatus;
uint8_t mstatus_bytes[8] = { 0 };
LOG_DEBUG("Disabling Interrupts");
struct reg *reg_mstatus = register_get_by_name(target->reg_cache,
"mstatus", 1);
if (!reg_mstatus) {
LOG_ERROR("Couldn't find mstatus!");
return ERROR_FAIL;
}
reg_mstatus->type->get(reg_mstatus);
current_mstatus = buf_get_u64(reg_mstatus->value, 0, reg_mstatus->size);
uint64_t ie_mask = MSTATUS_MIE | MSTATUS_HIE | MSTATUS_SIE | MSTATUS_UIE;
buf_set_u64(mstatus_bytes, 0, info->xlen[0], set_field(current_mstatus,
ie_mask, 0));
reg_mstatus->type->set(reg_mstatus, mstatus_bytes);
/* Run algorithm */
LOG_DEBUG("resume at 0x%" TARGET_PRIxADDR, entry_point);
if (riscv_resume(target, 0, entry_point, 0, 0, true) != ERROR_OK)
return ERROR_FAIL;
int64_t start = timeval_ms();
while (target->state != TARGET_HALTED) {
LOG_DEBUG("poll()");
int64_t now = timeval_ms();
if (now - start > timeout_ms) {
LOG_ERROR("Algorithm timed out after %" PRId64 " ms.", now - start);
riscv_halt(target);
old_or_new_riscv_poll(target);
enum gdb_regno regnums[] = {
GDB_REGNO_RA, GDB_REGNO_SP, GDB_REGNO_GP, GDB_REGNO_TP,
GDB_REGNO_T0, GDB_REGNO_T1, GDB_REGNO_T2, GDB_REGNO_FP,
GDB_REGNO_S1, GDB_REGNO_A0, GDB_REGNO_A1, GDB_REGNO_A2,
GDB_REGNO_A3, GDB_REGNO_A4, GDB_REGNO_A5, GDB_REGNO_A6,
GDB_REGNO_A7, GDB_REGNO_S2, GDB_REGNO_S3, GDB_REGNO_S4,
GDB_REGNO_S5, GDB_REGNO_S6, GDB_REGNO_S7, GDB_REGNO_S8,
GDB_REGNO_S9, GDB_REGNO_S10, GDB_REGNO_S11, GDB_REGNO_T3,
GDB_REGNO_T4, GDB_REGNO_T5, GDB_REGNO_T6,
GDB_REGNO_PC,
GDB_REGNO_MSTATUS, GDB_REGNO_MEPC, GDB_REGNO_MCAUSE,
};
for (unsigned i = 0; i < DIM(regnums); i++) {
enum gdb_regno regno = regnums[i];
riscv_reg_t reg_value;
if (riscv_get_register(target, &reg_value, regno) != ERROR_OK)
break;
LOG_ERROR("%s = 0x%" PRIx64, gdb_regno_name(regno), reg_value);
}
return ERROR_TARGET_TIMEOUT;
}
int result = old_or_new_riscv_poll(target);
if (result != ERROR_OK)
return result;
}
/* The current hart id might have been changed in poll(). */
if (riscv_set_current_hartid(target, hartid) != ERROR_OK)
return ERROR_FAIL;
if (reg_pc->type->get(reg_pc) != ERROR_OK)
return ERROR_FAIL;
uint64_t final_pc = buf_get_u64(reg_pc->value, 0, reg_pc->size);
if (exit_point && final_pc != exit_point) {
LOG_ERROR("PC ended up at 0x%" PRIx64 " instead of 0x%"
TARGET_PRIxADDR, final_pc, exit_point);
return ERROR_FAIL;
}
/* Restore Interrupts */
LOG_DEBUG("Restoring Interrupts");
buf_set_u64(mstatus_bytes, 0, info->xlen[0], current_mstatus);
reg_mstatus->type->set(reg_mstatus, mstatus_bytes);
/* Restore registers */
uint8_t buf[8] = { 0 };
buf_set_u64(buf, 0, info->xlen[0], saved_pc);
if (reg_pc->type->set(reg_pc, buf) != ERROR_OK)
return ERROR_FAIL;
for (int i = 0; i < num_reg_params; i++) {
if (reg_params[i].direction == PARAM_IN ||
reg_params[i].direction == PARAM_IN_OUT) {
struct reg *r = register_get_by_name(target->reg_cache, reg_params[i].reg_name, 0);
if (r->type->get(r) != ERROR_OK) {
LOG_ERROR("get(%s) failed", r->name);
return ERROR_FAIL;
}
buf_cpy(r->value, reg_params[i].value, reg_params[i].size);
}
LOG_DEBUG("restore %s", reg_params[i].reg_name);
struct reg *r = register_get_by_name(target->reg_cache, reg_params[i].reg_name, 0);
buf_set_u64(buf, 0, info->xlen[0], saved_regs[r->number]);
if (r->type->set(r, buf) != ERROR_OK) {
LOG_ERROR("set(%s) failed", r->name);
return ERROR_FAIL;
}
}
return ERROR_OK;
}
static int riscv_checksum_memory(struct target *target,
target_addr_t address, uint32_t count,
uint32_t *checksum)
{
return ERROR_FAIL;
}
/*** OpenOCD Helper Functions ***/
enum riscv_poll_hart {
RPH_NO_CHANGE,
RPH_DISCOVERED_HALTED,
RPH_DISCOVERED_RUNNING,
RPH_ERROR
};
static enum riscv_poll_hart riscv_poll_hart(struct target *target, int hartid)
{
RISCV_INFO(r);
if (riscv_set_current_hartid(target, hartid) != ERROR_OK)
return RPH_ERROR;
LOG_DEBUG("polling hart %d, target->state=%d", hartid, target->state);
/* If OpenOCD thinks we're running but this hart is halted then it's time
* to raise an event. */
bool halted = riscv_is_halted(target);
if (target->state != TARGET_HALTED && halted) {
LOG_DEBUG(" triggered a halt");
r->on_halt(target);
return RPH_DISCOVERED_HALTED;
} else if (target->state != TARGET_RUNNING && !halted) {
LOG_DEBUG(" triggered running");
target->state = TARGET_RUNNING;
target->debug_reason = DBG_REASON_NOTHALTED;
return RPH_DISCOVERED_RUNNING;
}
return RPH_NO_CHANGE;
}
int set_debug_reason(struct target *target, enum riscv_halt_reason halt_reason)
{
switch (halt_reason) {
case RISCV_HALT_BREAKPOINT:
target->debug_reason = DBG_REASON_BREAKPOINT;
break;
case RISCV_HALT_TRIGGER:
target->debug_reason = DBG_REASON_WATCHPOINT;
break;
case RISCV_HALT_INTERRUPT:
case RISCV_HALT_GROUP:
target->debug_reason = DBG_REASON_DBGRQ;
break;
case RISCV_HALT_SINGLESTEP:
target->debug_reason = DBG_REASON_SINGLESTEP;
break;
case RISCV_HALT_UNKNOWN:
target->debug_reason = DBG_REASON_UNDEFINED;
break;
case RISCV_HALT_ERROR:
return ERROR_FAIL;
}
LOG_DEBUG("[%s] debug_reason=%d", target_name(target), target->debug_reason);
return ERROR_OK;
}
/*** OpenOCD Interface ***/
int riscv_openocd_poll(struct target *target)
{
LOG_DEBUG("polling all harts");
int halted_hart = -1;
if (riscv_rtos_enabled(target)) {
/* Check every hart for an event. */
for (int i = 0; i < riscv_count_harts(target); ++i) {
enum riscv_poll_hart out = riscv_poll_hart(target, i);
switch (out) {
case RPH_NO_CHANGE:
case RPH_DISCOVERED_RUNNING:
continue;
case RPH_DISCOVERED_HALTED:
halted_hart = i;
break;
case RPH_ERROR:
return ERROR_FAIL;
}
}
if (halted_hart == -1) {
LOG_DEBUG(" no harts just halted, target->state=%d", target->state);
return ERROR_OK;
}
LOG_DEBUG(" hart %d halted", halted_hart);
target->state = TARGET_HALTED;
enum riscv_halt_reason halt_reason = riscv_halt_reason(target, halted_hart);
if (set_debug_reason(target, halt_reason) != ERROR_OK)
return ERROR_FAIL;
target->rtos->current_threadid = halted_hart + 1;
target->rtos->current_thread = halted_hart + 1;
riscv_set_rtos_hartid(target, halted_hart);
/* If we're here then at least one hart triggered. That means we want
* to go and halt _every_ hart (configured with -rtos riscv) in the
* system, as that's the invariant we hold here. Some harts might have
* already halted (as we're either in single-step mode or they also
* triggered a breakpoint), so don't attempt to halt those harts.
* riscv_halt() will do all that for us. */
riscv_halt(target);
} else if (target->smp) {
unsigned halts_discovered = 0;
unsigned total_targets = 0;
bool newly_halted[RISCV_MAX_HARTS] = {0};
unsigned should_remain_halted = 0;
unsigned should_resume = 0;
unsigned i = 0;
for (struct target_list *list = target->head; list != NULL;
list = list->next, i++) {
total_targets++;
struct target *t = list->target;
riscv_info_t *r = riscv_info(t);
assert(i < DIM(newly_halted));
enum riscv_poll_hart out = riscv_poll_hart(t, r->current_hartid);
switch (out) {
case RPH_NO_CHANGE:
break;
case RPH_DISCOVERED_RUNNING:
t->state = TARGET_RUNNING;
t->debug_reason = DBG_REASON_NOTHALTED;
break;
case RPH_DISCOVERED_HALTED:
halts_discovered++;
newly_halted[i] = true;
t->state = TARGET_HALTED;
enum riscv_halt_reason halt_reason =
riscv_halt_reason(t, r->current_hartid);
if (set_debug_reason(t, halt_reason) != ERROR_OK)
return ERROR_FAIL;
if (halt_reason == RISCV_HALT_BREAKPOINT) {
int retval;
switch (riscv_semihosting(t, &retval)) {
case SEMI_NONE:
case SEMI_WAITING:
/* This hart should remain halted. */
should_remain_halted++;
break;
case SEMI_HANDLED:
/* This hart should be resumed, along with any other
* harts that halted due to haltgroups. */
should_resume++;
break;
case SEMI_ERROR:
return retval;
}
} else if (halt_reason != RISCV_HALT_GROUP) {
should_remain_halted++;
}
break;
case RPH_ERROR:
return ERROR_FAIL;
}
}
LOG_DEBUG("should_remain_halted=%d, should_resume=%d",
should_remain_halted, should_resume);
if (should_remain_halted && should_resume) {
LOG_WARNING("%d harts should remain halted, and %d should resume.",
should_remain_halted, should_resume);
}
if (should_remain_halted) {
LOG_DEBUG("halt all");
riscv_halt(target);
} else if (should_resume) {
LOG_DEBUG("resume all");
riscv_resume(target, true, 0, 0, 0, false);
}
return ERROR_OK;
} else {
enum riscv_poll_hart out = riscv_poll_hart(target,
riscv_current_hartid(target));
if (out == RPH_NO_CHANGE || out == RPH_DISCOVERED_RUNNING)
return ERROR_OK;
else if (out == RPH_ERROR)
return ERROR_FAIL;
halted_hart = riscv_current_hartid(target);
LOG_DEBUG(" hart %d halted", halted_hart);
enum riscv_halt_reason halt_reason = riscv_halt_reason(target, halted_hart);
if (set_debug_reason(target, halt_reason) != ERROR_OK)
return ERROR_FAIL;
target->state = TARGET_HALTED;
}
if (target->debug_reason == DBG_REASON_BREAKPOINT) {
int retval;
switch (riscv_semihosting(target, &retval)) {
case SEMI_NONE:
case SEMI_WAITING:
target_call_event_callbacks(target, TARGET_EVENT_HALTED);
break;
case SEMI_HANDLED:
if (riscv_resume(target, true, 0, 0, 0, false) != ERROR_OK)
return ERROR_FAIL;
break;
case SEMI_ERROR:
return retval;
}
} else {
target_call_event_callbacks(target, TARGET_EVENT_HALTED);
}
return ERROR_OK;
}
int riscv_openocd_step(struct target *target, int current,
target_addr_t address, int handle_breakpoints)
{
LOG_DEBUG("stepping rtos hart");
if (!current)
riscv_set_register(target, GDB_REGNO_PC, address);
riscv_reg_t trigger_state[RISCV_MAX_HWBPS] = {0};
if (disable_triggers(target, trigger_state) != ERROR_OK)
return ERROR_FAIL;
int out = riscv_step_rtos_hart(target);
if (out != ERROR_OK) {
LOG_ERROR("unable to step rtos hart");
return out;
}
register_cache_invalidate(target->reg_cache);
if (enable_triggers(target, trigger_state) != ERROR_OK)
return ERROR_FAIL;
target->state = TARGET_RUNNING;
target_call_event_callbacks(target, TARGET_EVENT_RESUMED);
target->state = TARGET_HALTED;
target->debug_reason = DBG_REASON_SINGLESTEP;
target_call_event_callbacks(target, TARGET_EVENT_HALTED);
return out;
}
/* Command Handlers */
COMMAND_HANDLER(riscv_set_command_timeout_sec)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
int timeout = atoi(CMD_ARGV[0]);
if (timeout <= 0) {
LOG_ERROR("%s is not a valid integer argument for command.", CMD_ARGV[0]);
return ERROR_FAIL;
}
riscv_command_timeout_sec = timeout;
return ERROR_OK;
}
COMMAND_HANDLER(riscv_set_reset_timeout_sec)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
int timeout = atoi(CMD_ARGV[0]);
if (timeout <= 0) {
LOG_ERROR("%s is not a valid integer argument for command.", CMD_ARGV[0]);
return ERROR_FAIL;
}
riscv_reset_timeout_sec = timeout;
return ERROR_OK;
}
COMMAND_HANDLER(riscv_test_compliance) {
struct target *target = get_current_target(CMD_CTX);
RISCV_INFO(r);
if (CMD_ARGC > 0) {
LOG_ERROR("Command does not take any parameters.");
return ERROR_COMMAND_SYNTAX_ERROR;
}
if (r->test_compliance) {
return r->test_compliance(target);
} else {
LOG_ERROR("This target does not support this command (may implement an older version of the spec).");
return ERROR_FAIL;
}
}
COMMAND_HANDLER(riscv_set_prefer_sba)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
COMMAND_PARSE_ON_OFF(CMD_ARGV[0], riscv_prefer_sba);
return ERROR_OK;
}
COMMAND_HANDLER(riscv_set_enable_virtual)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
COMMAND_PARSE_ON_OFF(CMD_ARGV[0], riscv_enable_virtual);
return ERROR_OK;
}
void parse_error(const char *string, char c, unsigned position)
{
char buf[position+2];
for (unsigned i = 0; i < position; i++)
buf[i] = ' ';
buf[position] = '^';
buf[position + 1] = 0;
LOG_ERROR("Parse error at character %c in:", c);
LOG_ERROR("%s", string);
LOG_ERROR("%s", buf);
}
int parse_ranges(range_t **ranges, const char **argv)
{
for (unsigned pass = 0; pass < 2; pass++) {
unsigned range = 0;
unsigned low = 0;
bool parse_low = true;
unsigned high = 0;
for (unsigned i = 0; i == 0 || argv[0][i-1]; i++) {
char c = argv[0][i];
if (isspace(c)) {
/* Ignore whitespace. */
continue;
}
if (parse_low) {
if (isdigit(c)) {
low *= 10;
low += c - '0';
} else if (c == '-') {
parse_low = false;
} else if (c == ',' || c == 0) {
if (pass == 1) {
(*ranges)[range].low = low;
(*ranges)[range].high = low;
}
low = 0;
range++;
} else {
parse_error(argv[0], c, i);
return ERROR_COMMAND_SYNTAX_ERROR;
}
} else {
if (isdigit(c)) {
high *= 10;
high += c - '0';
} else if (c == ',' || c == 0) {
parse_low = true;
if (pass == 1) {
(*ranges)[range].low = low;
(*ranges)[range].high = high;
}
low = 0;
high = 0;
range++;
} else {
parse_error(argv[0], c, i);
return ERROR_COMMAND_SYNTAX_ERROR;
}
}
}
if (pass == 0) {
free(*ranges);
*ranges = calloc(range + 2, sizeof(range_t));
if (!*ranges)
return ERROR_FAIL;
} else {
(*ranges)[range].low = 1;
(*ranges)[range].high = 0;
}
}
return ERROR_OK;
}
COMMAND_HANDLER(riscv_set_expose_csrs)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
return parse_ranges(&expose_csr, CMD_ARGV);
}
COMMAND_HANDLER(riscv_set_expose_custom)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
return parse_ranges(&expose_custom, CMD_ARGV);
}
COMMAND_HANDLER(riscv_authdata_read)
{
if (CMD_ARGC != 0) {
LOG_ERROR("Command takes no parameters");
return ERROR_COMMAND_SYNTAX_ERROR;
}
struct target *target = get_current_target(CMD_CTX);
if (!target) {
LOG_ERROR("target is NULL!");
return ERROR_FAIL;
}
RISCV_INFO(r);
if (!r) {
LOG_ERROR("riscv_info is NULL!");
return ERROR_FAIL;
}
if (r->authdata_read) {
uint32_t value;
if (r->authdata_read(target, &value) != ERROR_OK)
return ERROR_FAIL;
command_print(CMD, "0x%" PRIx32, value);
return ERROR_OK;
} else {
LOG_ERROR("authdata_read is not implemented for this target.");
return ERROR_FAIL;
}
}
COMMAND_HANDLER(riscv_authdata_write)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 argument");
return ERROR_COMMAND_SYNTAX_ERROR;
}
struct target *target = get_current_target(CMD_CTX);
RISCV_INFO(r);
uint32_t value;
COMMAND_PARSE_NUMBER(u32, CMD_ARGV[0], value);
if (r->authdata_write) {
return r->authdata_write(target, value);
} else {
LOG_ERROR("authdata_write is not implemented for this target.");
return ERROR_FAIL;
}
}
COMMAND_HANDLER(riscv_dmi_read)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
struct target *target = get_current_target(CMD_CTX);
if (!target) {
LOG_ERROR("target is NULL!");
return ERROR_FAIL;
}
RISCV_INFO(r);
if (!r) {
LOG_ERROR("riscv_info is NULL!");
return ERROR_FAIL;
}
if (r->dmi_read) {
uint32_t address, value;
COMMAND_PARSE_NUMBER(u32, CMD_ARGV[0], address);
if (r->dmi_read(target, &value, address) != ERROR_OK)
return ERROR_FAIL;
command_print(CMD, "0x%" PRIx32, value);
return ERROR_OK;
} else {
LOG_ERROR("dmi_read is not implemented for this target.");
return ERROR_FAIL;
}
}
COMMAND_HANDLER(riscv_dmi_write)
{
if (CMD_ARGC != 2) {
LOG_ERROR("Command takes exactly 2 arguments");
return ERROR_COMMAND_SYNTAX_ERROR;
}
struct target *target = get_current_target(CMD_CTX);
RISCV_INFO(r);
uint32_t address, value;
COMMAND_PARSE_NUMBER(u32, CMD_ARGV[0], address);
COMMAND_PARSE_NUMBER(u32, CMD_ARGV[1], value);
if (r->dmi_write) {
return r->dmi_write(target, address, value);
} else {
LOG_ERROR("dmi_write is not implemented for this target.");
return ERROR_FAIL;
}
}
COMMAND_HANDLER(riscv_test_sba_config_reg)
{
if (CMD_ARGC != 4) {
LOG_ERROR("Command takes exactly 4 arguments");
return ERROR_COMMAND_SYNTAX_ERROR;
}
struct target *target = get_current_target(CMD_CTX);
RISCV_INFO(r);
target_addr_t legal_address;
uint32_t num_words;
target_addr_t illegal_address;
bool run_sbbusyerror_test;
COMMAND_PARSE_NUMBER(target_addr, CMD_ARGV[0], legal_address);
COMMAND_PARSE_NUMBER(u32, CMD_ARGV[1], num_words);
COMMAND_PARSE_NUMBER(target_addr, CMD_ARGV[2], illegal_address);
COMMAND_PARSE_ON_OFF(CMD_ARGV[3], run_sbbusyerror_test);
if (r->test_sba_config_reg) {
return r->test_sba_config_reg(target, legal_address, num_words,
illegal_address, run_sbbusyerror_test);
} else {
LOG_ERROR("test_sba_config_reg is not implemented for this target.");
return ERROR_FAIL;
}
}
COMMAND_HANDLER(riscv_reset_delays)
{
int wait = 0;
if (CMD_ARGC > 1) {
LOG_ERROR("Command takes at most one argument");
return ERROR_COMMAND_SYNTAX_ERROR;
}
if (CMD_ARGC == 1)
COMMAND_PARSE_NUMBER(int, CMD_ARGV[0], wait);
struct target *target = get_current_target(CMD_CTX);
RISCV_INFO(r);
r->reset_delays_wait = wait;
return ERROR_OK;
}
COMMAND_HANDLER(riscv_set_ir)
{
if (CMD_ARGC != 2) {
LOG_ERROR("Command takes exactly 2 arguments");
return ERROR_COMMAND_SYNTAX_ERROR;
}
uint32_t value;
COMMAND_PARSE_NUMBER(u32, CMD_ARGV[1], value);
if (!strcmp(CMD_ARGV[0], "idcode"))
buf_set_u32(ir_idcode, 0, 32, value);
else if (!strcmp(CMD_ARGV[0], "dtmcs"))
buf_set_u32(ir_dtmcontrol, 0, 32, value);
else if (!strcmp(CMD_ARGV[0], "dmi"))
buf_set_u32(ir_dbus, 0, 32, value);
else
return ERROR_FAIL;
return ERROR_OK;
}
COMMAND_HANDLER(riscv_resume_order)
{
if (CMD_ARGC > 1) {
LOG_ERROR("Command takes at most one argument");
return ERROR_COMMAND_SYNTAX_ERROR;
}
if (!strcmp(CMD_ARGV[0], "normal")) {
resume_order = RO_NORMAL;
} else if (!strcmp(CMD_ARGV[0], "reversed")) {
resume_order = RO_REVERSED;
} else {
LOG_ERROR("Unsupported resume order: %s", CMD_ARGV[0]);
return ERROR_FAIL;
}
return ERROR_OK;
}
COMMAND_HANDLER(riscv_use_bscan_tunnel)
{
int irwidth = 0;
int tunnel_type = BSCAN_TUNNEL_NESTED_TAP;
if (CMD_ARGC > 2) {
LOG_ERROR("Command takes at most two arguments");
return ERROR_COMMAND_SYNTAX_ERROR;
} else if (CMD_ARGC == 1) {
COMMAND_PARSE_NUMBER(int, CMD_ARGV[0], irwidth);
} else if (CMD_ARGC == 2) {
COMMAND_PARSE_NUMBER(int, CMD_ARGV[0], irwidth);
COMMAND_PARSE_NUMBER(int, CMD_ARGV[1], tunnel_type);
}
if (tunnel_type == BSCAN_TUNNEL_NESTED_TAP)
LOG_INFO("Nested Tap based Bscan Tunnel Selected");
else if (tunnel_type == BSCAN_TUNNEL_DATA_REGISTER)
LOG_INFO("Simple Register based Bscan Tunnel Selected");
else
LOG_INFO("Invalid Tunnel type selected ! : selecting default Nested Tap Type");
bscan_tunnel_type = tunnel_type;
bscan_tunnel_ir_width = irwidth;
return ERROR_OK;
}
COMMAND_HANDLER(riscv_set_enable_virt2phys)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
COMMAND_PARSE_ON_OFF(CMD_ARGV[0], riscv_enable_virt2phys);
return ERROR_OK;
}
COMMAND_HANDLER(riscv_set_ebreakm)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
COMMAND_PARSE_ON_OFF(CMD_ARGV[0], riscv_ebreakm);
return ERROR_OK;
}
COMMAND_HANDLER(riscv_set_ebreaks)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
COMMAND_PARSE_ON_OFF(CMD_ARGV[0], riscv_ebreaks);
return ERROR_OK;
}
COMMAND_HANDLER(riscv_set_ebreaku)
{
if (CMD_ARGC != 1) {
LOG_ERROR("Command takes exactly 1 parameter");
return ERROR_COMMAND_SYNTAX_ERROR;
}
COMMAND_PARSE_ON_OFF(CMD_ARGV[0], riscv_ebreaku);
return ERROR_OK;
}
static const struct command_registration riscv_exec_command_handlers[] = {
{
.name = "test_compliance",
.handler = riscv_test_compliance,
.usage = "",
.mode = COMMAND_EXEC,
.help = "Runs a basic compliance test suite against the RISC-V Debug Spec."
},
{
.name = "set_command_timeout_sec",
.handler = riscv_set_command_timeout_sec,
.mode = COMMAND_ANY,
.usage = "[sec]",
.help = "Set the wall-clock timeout (in seconds) for individual commands"
},
{
.name = "set_reset_timeout_sec",
.handler = riscv_set_reset_timeout_sec,
.mode = COMMAND_ANY,
.usage = "[sec]",
.help = "Set the wall-clock timeout (in seconds) after reset is deasserted"
},
{
.name = "set_prefer_sba",
.handler = riscv_set_prefer_sba,
.mode = COMMAND_ANY,
.usage = "on|off",
.help = "When on, prefer to use System Bus Access to access memory. "
"When off (default), prefer to use the Program Buffer to access memory."
},
{
.name = "set_enable_virtual",
.handler = riscv_set_enable_virtual,
.mode = COMMAND_ANY,
.usage = "on|off",
.help = "When on, memory accesses are performed on physical or virtual "
"memory depending on the current system configuration. "
"When off (default), all memory accessses are performed on physical memory."
},
{
.name = "expose_csrs",
.handler = riscv_set_expose_csrs,
.mode = COMMAND_ANY,
.usage = "n0[-m0][,n1[-m1]]...",
.help = "Configure a list of inclusive ranges for CSRs to expose in "
"addition to the standard ones. This must be executed before "
"`init`."
},
{
.name = "expose_custom",
.handler = riscv_set_expose_custom,
.mode = COMMAND_ANY,
.usage = "n0[-m0][,n1[-m1]]...",
.help = "Configure a list of inclusive ranges for custom registers to "
"expose. custom0 is accessed as abstract register number 0xc000, "
"etc. This must be executed before `init`."
},
{
.name = "authdata_read",
.handler = riscv_authdata_read,
.usage = "",
.mode = COMMAND_ANY,
.help = "Return the 32-bit value read from authdata."
},
{
.name = "authdata_write",
.handler = riscv_authdata_write,
.mode = COMMAND_ANY,
.usage = "value",
.help = "Write the 32-bit value to authdata."
},
{
.name = "dmi_read",
.handler = riscv_dmi_read,
.mode = COMMAND_ANY,
.usage = "address",
.help = "Perform a 32-bit DMI read at address, returning the value."
},
{
.name = "dmi_write",
.handler = riscv_dmi_write,
.mode = COMMAND_ANY,
.usage = "address value",
.help = "Perform a 32-bit DMI write of value at address."
},
{
.name = "test_sba_config_reg",
.handler = riscv_test_sba_config_reg,
.mode = COMMAND_ANY,
.usage = "legal_address num_words "
"illegal_address run_sbbusyerror_test[on/off]",
.help = "Perform a series of tests on the SBCS register. "
"Inputs are a legal, 128-byte aligned address and a number of words to "
"read/write starting at that address (i.e., address range [legal address, "
"legal_address+word_size*num_words) must be legally readable/writable), "
"an illegal, 128-byte aligned address for error flag/handling cases, "
"and whether sbbusyerror test should be run."
},
{
.name = "reset_delays",
.handler = riscv_reset_delays,
.mode = COMMAND_ANY,
.usage = "[wait]",
.help = "OpenOCD learns how many Run-Test/Idle cycles are required "
"between scans to avoid encountering the target being busy. This "
"command resets those learned values after `wait` scans. It's only "
"useful for testing OpenOCD itself."
},
{
.name = "resume_order",
.handler = riscv_resume_order,
.mode = COMMAND_ANY,
.usage = "normal|reversed",
.help = "Choose the order that harts are resumed in when `hasel` is not "
"supported. Normal order is from lowest hart index to highest. "
"Reversed order is from highest hart index to lowest."
},
{
.name = "set_ir",
.handler = riscv_set_ir,
.mode = COMMAND_ANY,
.usage = "[idcode|dtmcs|dmi] value",
.help = "Set IR value for specified JTAG register."
},
{
.name = "use_bscan_tunnel",
.handler = riscv_use_bscan_tunnel,
.mode = COMMAND_ANY,
.usage = "value [type]",
.help = "Enable or disable use of a BSCAN tunnel to reach DM. Supply "
"the width of the DM transport TAP's instruction register to "
"enable. Supply a value of 0 to disable. Pass A second argument "
"(optional) to indicate Bscan Tunnel Type {0:(default) NESTED_TAP , "
"1: DATA_REGISTER}"
},
{
.name = "set_enable_virt2phys",
.handler = riscv_set_enable_virt2phys,
.mode = COMMAND_ANY,
.usage = "on|off",
.help = "When on (default), enable translation from virtual address to "
"physical address."
},
{
.name = "set_ebreakm",
.handler = riscv_set_ebreakm,
.mode = COMMAND_ANY,
.usage = "on|off",
.help = "Control dcsr.ebreakm. When off, M-mode ebreak instructions "
"don't trap to OpenOCD. Defaults to on."
},
{
.name = "set_ebreaks",
.handler = riscv_set_ebreaks,
.mode = COMMAND_ANY,
.usage = "on|off",
.help = "Control dcsr.ebreaks. When off, S-mode ebreak instructions "
"don't trap to OpenOCD. Defaults to on."
},
{
.name = "set_ebreaku",
.handler = riscv_set_ebreaku,
.mode = COMMAND_ANY,
.usage = "on|off",
.help = "Control dcsr.ebreaku. When off, U-mode ebreak instructions "
"don't trap to OpenOCD. Defaults to on."
},
COMMAND_REGISTRATION_DONE
};
/*
* To be noted that RISC-V targets use the same semihosting commands as
* ARM targets.
*
* The main reason is compatibility with existing tools. For example the
* Eclipse OpenOCD/SEGGER J-Link/QEMU plug-ins have several widgets to
* configure semihosting, which generate commands like `arm semihosting
* enable`.
* A secondary reason is the fact that the protocol used is exactly the
* one specified by ARM. If RISC-V will ever define its own semihosting
* protocol, then a command like `riscv semihosting enable` will make
* sense, but for now all semihosting commands are prefixed with `arm`.
*/
extern const struct command_registration semihosting_common_handlers[];
const struct command_registration riscv_command_handlers[] = {
{
.name = "riscv",
.mode = COMMAND_ANY,
.help = "RISC-V Command Group",
.usage = "",
.chain = riscv_exec_command_handlers
},
{
.name = "arm",
.mode = COMMAND_ANY,
.help = "ARM Command Group",
.usage = "",
.chain = semihosting_common_handlers
},
COMMAND_REGISTRATION_DONE
};
static unsigned riscv_xlen_nonconst(struct target *target)
{
return riscv_xlen(target);
}
struct target_type riscv_target = {
.name = "riscv",
.init_target = riscv_init_target,
.deinit_target = riscv_deinit_target,
.examine = riscv_examine,
/* poll current target status */
.poll = old_or_new_riscv_poll,
.halt = riscv_halt,
.resume = riscv_target_resume,
.step = old_or_new_riscv_step,
.assert_reset = riscv_assert_reset,
.deassert_reset = riscv_deassert_reset,
.read_memory = riscv_read_memory,
.write_memory = riscv_write_memory,
.read_phys_memory = riscv_read_phys_memory,
.write_phys_memory = riscv_write_phys_memory,
.checksum_memory = riscv_checksum_memory,
.mmu = riscv_mmu,
.virt2phys = riscv_virt2phys,
.get_gdb_reg_list = riscv_get_gdb_reg_list,
.get_gdb_reg_list_noread = riscv_get_gdb_reg_list_noread,
.add_breakpoint = riscv_add_breakpoint,
.remove_breakpoint = riscv_remove_breakpoint,
.add_watchpoint = riscv_add_watchpoint,
.remove_watchpoint = riscv_remove_watchpoint,
.hit_watchpoint = riscv_hit_watchpoint,
.arch_state = riscv_arch_state,
.run_algorithm = riscv_run_algorithm,
.commands = riscv_command_handlers,
.address_bits = riscv_xlen_nonconst,
};
/*** RISC-V Interface ***/
void riscv_info_init(struct target *target, riscv_info_t *r)
{
memset(r, 0, sizeof(*r));
r->dtm_version = 1;
r->registers_initialized = false;
r->current_hartid = target->coreid;
memset(r->trigger_unique_id, 0xff, sizeof(r->trigger_unique_id));
for (size_t h = 0; h < RISCV_MAX_HARTS; ++h)
r->xlen[h] = -1;
}
static int riscv_resume_go_all_harts(struct target *target)
{
RISCV_INFO(r);
/* Dummy variables to make mingw32-gcc happy. */
int first = 0;
int last = 1;
int step = 1;
switch (resume_order) {
case RO_NORMAL:
first = 0;
last = riscv_count_harts(target) - 1;
step = 1;
break;
case RO_REVERSED:
first = riscv_count_harts(target) - 1;
last = 0;
step = -1;
break;
default:
assert(0);
}
for (int i = first; i != last + step; i += step) {
if (!riscv_hart_enabled(target, i))
continue;
LOG_DEBUG("resuming hart %d", i);
if (riscv_set_current_hartid(target, i) != ERROR_OK)
return ERROR_FAIL;
if (riscv_is_halted(target)) {
if (r->resume_go(target) != ERROR_OK)
return ERROR_FAIL;
} else {
LOG_DEBUG(" hart %d requested resume, but was already resumed", i);
}
}
riscv_invalidate_register_cache(target);
return ERROR_OK;
}
int riscv_step_rtos_hart(struct target *target)
{
RISCV_INFO(r);
int hartid = r->current_hartid;
if (riscv_rtos_enabled(target)) {
hartid = r->rtos_hartid;
if (hartid == -1) {
LOG_DEBUG("GDB has asked me to step \"any\" thread, so I'm stepping hart 0.");
hartid = 0;
}
}
if (riscv_set_current_hartid(target, hartid) != ERROR_OK)
return ERROR_FAIL;
LOG_DEBUG("stepping hart %d", hartid);
if (!riscv_is_halted(target)) {
LOG_ERROR("Hart isn't halted before single step!");
return ERROR_FAIL;
}
riscv_invalidate_register_cache(target);
r->on_step(target);
if (r->step_current_hart(target) != ERROR_OK)
return ERROR_FAIL;
riscv_invalidate_register_cache(target);
r->on_halt(target);
if (!riscv_is_halted(target)) {
LOG_ERROR("Hart was not halted after single step!");
return ERROR_FAIL;
}
return ERROR_OK;
}
bool riscv_supports_extension(struct target *target, int hartid, char letter)
{
RISCV_INFO(r);
unsigned num;
if (letter >= 'a' && letter <= 'z')
num = letter - 'a';
else if (letter >= 'A' && letter <= 'Z')
num = letter - 'A';
else
return false;
return r->misa[hartid] & (1 << num);
}
unsigned riscv_xlen(const struct target *target)
{
return riscv_xlen_of_hart(target, riscv_current_hartid(target));
}
int riscv_xlen_of_hart(const struct target *target, int hartid)
{
RISCV_INFO(r);
assert(r->xlen[hartid] != -1);
return r->xlen[hartid];
}
bool riscv_rtos_enabled(const struct target *target)
{
return false;
}
int riscv_set_current_hartid(struct target *target, int hartid)
{
RISCV_INFO(r);
if (!r->select_current_hart)
return ERROR_OK;
int previous_hartid = riscv_current_hartid(target);
r->current_hartid = hartid;
assert(riscv_hart_enabled(target, hartid));
LOG_DEBUG("setting hartid to %d, was %d", hartid, previous_hartid);
if (r->select_current_hart(target) != ERROR_OK)
return ERROR_FAIL;
/* This might get called during init, in which case we shouldn't be
* setting up the register cache. */
if (target_was_examined(target) && riscv_rtos_enabled(target))
riscv_invalidate_register_cache(target);
return ERROR_OK;
}
void riscv_invalidate_register_cache(struct target *target)
{
RISCV_INFO(r);
LOG_DEBUG("[%d]", target->coreid);
register_cache_invalidate(target->reg_cache);
for (size_t i = 0; i < target->reg_cache->num_regs; ++i) {
struct reg *reg = &target->reg_cache->reg_list[i];
reg->valid = false;
}
r->registers_initialized = true;
}
int riscv_current_hartid(const struct target *target)
{
RISCV_INFO(r);
return r->current_hartid;
}
void riscv_set_all_rtos_harts(struct target *target)
{
RISCV_INFO(r);
r->rtos_hartid = -1;
}
void riscv_set_rtos_hartid(struct target *target, int hartid)
{
LOG_DEBUG("setting RTOS hartid %d", hartid);
RISCV_INFO(r);
r->rtos_hartid = hartid;
}
int riscv_count_harts(struct target *target)
{
if (target == NULL)
return 1;
RISCV_INFO(r);
if (r == NULL || r->hart_count == NULL)
return 1;
return r->hart_count(target);
}
bool riscv_has_register(struct target *target, int hartid, int regid)
{
return 1;
}
/**
* If write is true:
* return true iff we are guaranteed that the register will contain exactly
* the value we just wrote when it's read.
* If write is false:
* return true iff we are guaranteed that the register will read the same
* value in the future as the value we just read.
*/
static bool gdb_regno_cacheable(enum gdb_regno regno, bool write)
{
/* GPRs, FPRs, vector registers are just normal data stores. */
if (regno <= GDB_REGNO_XPR31 ||
(regno >= GDB_REGNO_FPR0 && regno <= GDB_REGNO_FPR31) ||
(regno >= GDB_REGNO_V0 && regno <= GDB_REGNO_V31))
return true;
/* Most CSRs won't change value on us, but we can't assume it about rbitrary
* CSRs. */
switch (regno) {
case GDB_REGNO_DPC:
return true;
case GDB_REGNO_VSTART:
case GDB_REGNO_VXSAT:
case GDB_REGNO_VXRM:
case GDB_REGNO_VLENB:
case GDB_REGNO_VL:
case GDB_REGNO_VTYPE:
case GDB_REGNO_MISA:
case GDB_REGNO_DCSR:
case GDB_REGNO_DSCRATCH0:
case GDB_REGNO_MSTATUS:
case GDB_REGNO_MEPC:
case GDB_REGNO_MCAUSE:
case GDB_REGNO_SATP:
/*
* WARL registers might not contain the value we just wrote, but
* these ones won't spontaneously change their value either. *
*/
return !write;
case GDB_REGNO_TSELECT: /* I think this should be above, but then it doesn't work. */
case GDB_REGNO_TDATA1: /* Changes value when tselect is changed. */
case GDB_REGNO_TDATA2: /* Changse value when tselect is changed. */
default:
return false;
}
}
/**
* This function is called when the debug user wants to change the value of a
* register. The new value may be cached, and may not be written until the hart
* is resumed. */
int riscv_set_register(struct target *target, enum gdb_regno r, riscv_reg_t v)
{
return riscv_set_register_on_hart(target, riscv_current_hartid(target), r, v);
}
int riscv_set_register_on_hart(struct target *target, int hartid,
enum gdb_regno regid, uint64_t value)
{
RISCV_INFO(r);
LOG_DEBUG("{%d} %s <- %" PRIx64, hartid, gdb_regno_name(regid), value);
assert(r->set_register);
/* TODO: Hack to deal with gdb that thinks these registers still exist. */
if (regid > GDB_REGNO_XPR15 && regid <= GDB_REGNO_XPR31 && value == 0 &&
riscv_supports_extension(target, hartid, 'E'))
return ERROR_OK;
struct reg *reg = &target->reg_cache->reg_list[regid];
buf_set_u64(reg->value, 0, reg->size, value);
int result = r->set_register(target, hartid, regid, value);
if (result == ERROR_OK)
reg->valid = gdb_regno_cacheable(regid, true);
else
reg->valid = false;
LOG_DEBUG("[%s]{%d} wrote 0x%" PRIx64 " to %s valid=%d",
target_name(target), hartid, value, reg->name, reg->valid);
return result;
}
int riscv_get_register(struct target *target, riscv_reg_t *value,
enum gdb_regno r)
{
return riscv_get_register_on_hart(target, value,
riscv_current_hartid(target), r);
}
int riscv_get_register_on_hart(struct target *target, riscv_reg_t *value,
int hartid, enum gdb_regno regid)
{
RISCV_INFO(r);
struct reg *reg = &target->reg_cache->reg_list[regid];
if (!reg->exist) {
LOG_DEBUG("[%s]{%d} %s does not exist.",
target_name(target), hartid, gdb_regno_name(regid));
return ERROR_FAIL;
}
if (reg && reg->valid && hartid == riscv_current_hartid(target)) {
*value = buf_get_u64(reg->value, 0, reg->size);
LOG_DEBUG("{%d} %s: %" PRIx64 " (cached)", hartid,
gdb_regno_name(regid), *value);
return ERROR_OK;
}
/* TODO: Hack to deal with gdb that thinks these registers still exist. */
if (regid > GDB_REGNO_XPR15 && regid <= GDB_REGNO_XPR31 &&
riscv_supports_extension(target, hartid, 'E')) {
*value = 0;
return ERROR_OK;
}
int result = r->get_register(target, value, hartid, regid);
if (result == ERROR_OK)
reg->valid = gdb_regno_cacheable(regid, false);
LOG_DEBUG("{%d} %s: %" PRIx64, hartid, gdb_regno_name(regid), *value);
return result;
}
bool riscv_is_halted(struct target *target)
{
RISCV_INFO(r);
assert(r->is_halted);
return r->is_halted(target);
}
enum riscv_halt_reason riscv_halt_reason(struct target *target, int hartid)
{
RISCV_INFO(r);
if (riscv_set_current_hartid(target, hartid) != ERROR_OK)
return RISCV_HALT_ERROR;
if (!riscv_is_halted(target)) {
LOG_ERROR("Hart is not halted!");
return RISCV_HALT_UNKNOWN;
}
return r->halt_reason(target);
}
size_t riscv_debug_buffer_size(struct target *target)
{
RISCV_INFO(r);
return r->debug_buffer_size[riscv_current_hartid(target)];
}
int riscv_write_debug_buffer(struct target *target, int index, riscv_insn_t insn)
{
RISCV_INFO(r);
r->write_debug_buffer(target, index, insn);
return ERROR_OK;
}
riscv_insn_t riscv_read_debug_buffer(struct target *target, int index)
{
RISCV_INFO(r);
return r->read_debug_buffer(target, index);
}
int riscv_execute_debug_buffer(struct target *target)
{
RISCV_INFO(r);
return r->execute_debug_buffer(target);
}
void riscv_fill_dmi_write_u64(struct target *target, char *buf, int a, uint64_t d)
{
RISCV_INFO(r);
r->fill_dmi_write_u64(target, buf, a, d);
}
void riscv_fill_dmi_read_u64(struct target *target, char *buf, int a)
{
RISCV_INFO(r);
r->fill_dmi_read_u64(target, buf, a);
}
void riscv_fill_dmi_nop_u64(struct target *target, char *buf)
{
RISCV_INFO(r);
r->fill_dmi_nop_u64(target, buf);
}
int riscv_dmi_write_u64_bits(struct target *target)
{
RISCV_INFO(r);
return r->dmi_write_u64_bits(target);
}
bool riscv_hart_enabled(struct target *target, int hartid)
{
/* FIXME: Add a hart mask to the RTOS. */
if (riscv_rtos_enabled(target))
return hartid < riscv_count_harts(target);
return hartid == target->coreid;
}
/**
* Count triggers, and initialize trigger_count for each hart.
* trigger_count is initialized even if this function fails to discover
* something.
* Disable any hardware triggers that have dmode set. We can't have set them
* ourselves. Maybe they're left over from some killed debug session.
* */
int riscv_enumerate_triggers(struct target *target)
{
RISCV_INFO(r);
if (r->triggers_enumerated)
return ERROR_OK;
r->triggers_enumerated = true; /* At the very least we tried. */
for (int hartid = 0; hartid < riscv_count_harts(target); ++hartid) {
if (!riscv_hart_enabled(target, hartid))
continue;
riscv_reg_t tselect;
int result = riscv_get_register_on_hart(target, &tselect, hartid,
GDB_REGNO_TSELECT);
if (result != ERROR_OK)
return result;
for (unsigned t = 0; t < RISCV_MAX_TRIGGERS; ++t) {
r->trigger_count[hartid] = t;
/* If we can't write tselect, then this hart does not support triggers. */
if (riscv_set_register_on_hart(target, hartid, GDB_REGNO_TSELECT, t) != ERROR_OK)
break;
uint64_t tselect_rb;
result = riscv_get_register_on_hart(target, &tselect_rb, hartid,
GDB_REGNO_TSELECT);
if (result != ERROR_OK)
return result;
/* Mask off the top bit, which is used as tdrmode in old
* implementations. */
tselect_rb &= ~(1ULL << (riscv_xlen(target)-1));
if (tselect_rb != t)
break;
uint64_t tdata1;
result = riscv_get_register_on_hart(target, &tdata1, hartid,
GDB_REGNO_TDATA1);
if (result != ERROR_OK)
return result;
int type = get_field(tdata1, MCONTROL_TYPE(riscv_xlen(target)));
if (type == 0)
break;
switch (type) {
case 1:
/* On these older cores we don't support software using
* triggers. */
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TDATA1, 0);
break;
case 2:
if (tdata1 & MCONTROL_DMODE(riscv_xlen(target)))
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TDATA1, 0);
break;
}
}
riscv_set_register_on_hart(target, hartid, GDB_REGNO_TSELECT, tselect);
LOG_INFO("[%d] Found %d triggers", hartid, r->trigger_count[hartid]);
}
return ERROR_OK;
}
const char *gdb_regno_name(enum gdb_regno regno)
{
static char buf[32];
switch (regno) {
case GDB_REGNO_ZERO:
return "zero";
case GDB_REGNO_RA:
return "ra";
case GDB_REGNO_SP:
return "sp";
case GDB_REGNO_GP:
return "gp";
case GDB_REGNO_TP:
return "tp";
case GDB_REGNO_T0:
return "t0";
case GDB_REGNO_T1:
return "t1";
case GDB_REGNO_T2:
return "t2";
case GDB_REGNO_S0:
return "s0";
case GDB_REGNO_S1:
return "s1";
case GDB_REGNO_A0:
return "a0";
case GDB_REGNO_A1:
return "a1";
case GDB_REGNO_A2:
return "a2";
case GDB_REGNO_A3:
return "a3";
case GDB_REGNO_A4:
return "a4";
case GDB_REGNO_A5:
return "a5";
case GDB_REGNO_A6:
return "a6";
case GDB_REGNO_A7:
return "a7";
case GDB_REGNO_S2:
return "s2";
case GDB_REGNO_S3:
return "s3";
case GDB_REGNO_S4:
return "s4";
case GDB_REGNO_S5:
return "s5";
case GDB_REGNO_S6:
return "s6";
case GDB_REGNO_S7:
return "s7";
case GDB_REGNO_S8:
return "s8";
case GDB_REGNO_S9:
return "s9";
case GDB_REGNO_S10:
return "s10";
case GDB_REGNO_S11:
return "s11";
case GDB_REGNO_T3:
return "t3";
case GDB_REGNO_T4:
return "t4";
case GDB_REGNO_T5:
return "t5";
case GDB_REGNO_T6:
return "t6";
case GDB_REGNO_PC:
return "pc";
case GDB_REGNO_FPR0:
return "fpr0";
case GDB_REGNO_FPR31:
return "fpr31";
case GDB_REGNO_CSR0:
return "csr0";
case GDB_REGNO_TSELECT:
return "tselect";
case GDB_REGNO_TDATA1:
return "tdata1";
case GDB_REGNO_TDATA2:
return "tdata2";
case GDB_REGNO_MISA:
return "misa";
case GDB_REGNO_DPC:
return "dpc";
case GDB_REGNO_DCSR:
return "dcsr";
case GDB_REGNO_DSCRATCH0:
return "dscratch0";
case GDB_REGNO_MSTATUS:
return "mstatus";
case GDB_REGNO_MEPC:
return "mepc";
case GDB_REGNO_MCAUSE:
return "mcause";
case GDB_REGNO_PRIV:
return "priv";
case GDB_REGNO_SATP:
return "satp";
case GDB_REGNO_VTYPE:
return "vtype";
case GDB_REGNO_VL:
return "vl";
case GDB_REGNO_V0:
return "v0";
case GDB_REGNO_V1:
return "v1";
case GDB_REGNO_V2:
return "v2";
case GDB_REGNO_V3:
return "v3";
case GDB_REGNO_V4:
return "v4";
case GDB_REGNO_V5:
return "v5";
case GDB_REGNO_V6:
return "v6";
case GDB_REGNO_V7:
return "v7";
case GDB_REGNO_V8:
return "v8";
case GDB_REGNO_V9:
return "v9";
case GDB_REGNO_V10:
return "v10";
case GDB_REGNO_V11:
return "v11";
case GDB_REGNO_V12:
return "v12";
case GDB_REGNO_V13:
return "v13";
case GDB_REGNO_V14:
return "v14";
case GDB_REGNO_V15:
return "v15";
case GDB_REGNO_V16:
return "v16";
case GDB_REGNO_V17:
return "v17";
case GDB_REGNO_V18:
return "v18";
case GDB_REGNO_V19:
return "v19";
case GDB_REGNO_V20:
return "v20";
case GDB_REGNO_V21:
return "v21";
case GDB_REGNO_V22:
return "v22";
case GDB_REGNO_V23:
return "v23";
case GDB_REGNO_V24:
return "v24";
case GDB_REGNO_V25:
return "v25";
case GDB_REGNO_V26:
return "v26";
case GDB_REGNO_V27:
return "v27";
case GDB_REGNO_V28:
return "v28";
case GDB_REGNO_V29:
return "v29";
case GDB_REGNO_V30:
return "v30";
case GDB_REGNO_V31:
return "v31";
default:
if (regno <= GDB_REGNO_XPR31)
sprintf(buf, "x%d", regno - GDB_REGNO_ZERO);
else if (regno >= GDB_REGNO_CSR0 && regno <= GDB_REGNO_CSR4095)
sprintf(buf, "csr%d", regno - GDB_REGNO_CSR0);
else if (regno >= GDB_REGNO_FPR0 && regno <= GDB_REGNO_FPR31)
sprintf(buf, "f%d", regno - GDB_REGNO_FPR0);
else
sprintf(buf, "gdb_regno_%d", regno);
return buf;
}
}
static int register_get(struct reg *reg)
{
riscv_reg_info_t *reg_info = reg->arch_info;
struct target *target = reg_info->target;
RISCV_INFO(r);
if (reg->number >= GDB_REGNO_V0 && reg->number <= GDB_REGNO_V31) {
if (!r->get_register_buf) {
LOG_ERROR("Reading register %s not supported on this RISC-V target.",
gdb_regno_name(reg->number));
return ERROR_FAIL;
}
if (r->get_register_buf(target, reg->value, reg->number) != ERROR_OK)
return ERROR_FAIL;
} else {
uint64_t value;
int result = riscv_get_register(target, &value, reg->number);
if (result != ERROR_OK)
return result;
buf_set_u64(reg->value, 0, reg->size, value);
}
reg->valid = gdb_regno_cacheable(reg->number, false);
char *str = buf_to_hex_str(reg->value, reg->size);
LOG_DEBUG("[%d]{%d} read 0x%s from %s (valid=%d)", target->coreid,
riscv_current_hartid(target), str, reg->name, reg->valid);
free(str);
return ERROR_OK;
}
static int register_set(struct reg *reg, uint8_t *buf)
{
riscv_reg_info_t *reg_info = reg->arch_info;
struct target *target = reg_info->target;
RISCV_INFO(r);
char *str = buf_to_hex_str(buf, reg->size);
LOG_DEBUG("[%d]{%d} write 0x%s to %s (valid=%d)", target->coreid,
riscv_current_hartid(target), str, reg->name, reg->valid);
free(str);
memcpy(reg->value, buf, DIV_ROUND_UP(reg->size, 8));
reg->valid = gdb_regno_cacheable(reg->number, true);
if (reg->number == GDB_REGNO_TDATA1 ||
reg->number == GDB_REGNO_TDATA2) {
r->manual_hwbp_set = true;
/* When enumerating triggers, we clear any triggers with DMODE set,
* assuming they were left over from a previous debug session. So make
* sure that is done before a user might be setting their own triggers.
*/
if (riscv_enumerate_triggers(target) != ERROR_OK)
return ERROR_FAIL;
}
if (reg->number >= GDB_REGNO_V0 && reg->number <= GDB_REGNO_V31) {
if (!r->set_register_buf) {
LOG_ERROR("Writing register %s not supported on this RISC-V target.",
gdb_regno_name(reg->number));
return ERROR_FAIL;
}
if (r->set_register_buf(target, reg->number, reg->value) != ERROR_OK)
return ERROR_FAIL;
} else {
uint64_t value = buf_get_u64(buf, 0, reg->size);
if (riscv_set_register(target, reg->number, value) != ERROR_OK)
return ERROR_FAIL;
}
return ERROR_OK;
}
static struct reg_arch_type riscv_reg_arch_type = {
.get = register_get,
.set = register_set
};
struct csr_info {
unsigned number;
const char *name;
};
static int cmp_csr_info(const void *p1, const void *p2)
{
return (int) (((struct csr_info *)p1)->number) - (int) (((struct csr_info *)p2)->number);
}
int riscv_init_registers(struct target *target)
{
RISCV_INFO(info);
riscv_free_registers(target);
target->reg_cache = calloc(1, sizeof(*target->reg_cache));
if (!target->reg_cache)
return ERROR_FAIL;
target->reg_cache->name = "RISC-V Registers";
target->reg_cache->num_regs = GDB_REGNO_COUNT;
if (expose_custom) {
for (unsigned i = 0; expose_custom[i].low <= expose_custom[i].high; i++) {
for (unsigned number = expose_custom[i].low;
number <= expose_custom[i].high;
number++)
target->reg_cache->num_regs++;
}
}
LOG_DEBUG("create register cache for %d registers",
target->reg_cache->num_regs);
target->reg_cache->reg_list =
calloc(target->reg_cache->num_regs, sizeof(struct reg));
if (!target->reg_cache->reg_list)
return ERROR_FAIL;
const unsigned int max_reg_name_len = 12;
free(info->reg_names);
info->reg_names =
calloc(target->reg_cache->num_regs, max_reg_name_len);
if (!info->reg_names)
return ERROR_FAIL;
char *reg_name = info->reg_names;
int hartid = riscv_current_hartid(target);
static struct reg_feature feature_cpu = {
.name = "org.gnu.gdb.riscv.cpu"
};
static struct reg_feature feature_fpu = {
.name = "org.gnu.gdb.riscv.fpu"
};
static struct reg_feature feature_csr = {
.name = "org.gnu.gdb.riscv.csr"
};
static struct reg_feature feature_vector = {
.name = "org.gnu.gdb.riscv.vector"
};
static struct reg_feature feature_virtual = {
.name = "org.gnu.gdb.riscv.virtual"
};
static struct reg_feature feature_custom = {
.name = "org.gnu.gdb.riscv.custom"
};
/* These types are built into gdb. */
static struct reg_data_type type_ieee_single = { .type = REG_TYPE_IEEE_SINGLE, .id = "ieee_single" };
static struct reg_data_type type_ieee_double = { .type = REG_TYPE_IEEE_DOUBLE, .id = "ieee_double" };
static struct reg_data_type_union_field single_double_fields[] = {
{"float", &type_ieee_single, single_double_fields + 1},
{"double", &type_ieee_double, NULL},
};
static struct reg_data_type_union single_double_union = {
.fields = single_double_fields
};
static struct reg_data_type type_ieee_single_double = {
.type = REG_TYPE_ARCH_DEFINED,
.id = "FPU_FD",
.type_class = REG_TYPE_CLASS_UNION,
.reg_type_union = &single_double_union
};
static struct reg_data_type type_uint8 = { .type = REG_TYPE_UINT8, .id = "uint8" };
static struct reg_data_type type_uint16 = { .type = REG_TYPE_UINT16, .id = "uint16" };
static struct reg_data_type type_uint32 = { .type = REG_TYPE_UINT32, .id = "uint32" };
static struct reg_data_type type_uint64 = { .type = REG_TYPE_UINT64, .id = "uint64" };
static struct reg_data_type type_uint128 = { .type = REG_TYPE_UINT128, .id = "uint128" };
/* This is roughly the XML we want:
* <vector id="bytes" type="uint8" count="16"/>
* <vector id="shorts" type="uint16" count="8"/>
* <vector id="words" type="uint32" count="4"/>
* <vector id="longs" type="uint64" count="2"/>
* <vector id="quads" type="uint128" count="1"/>
* <union id="riscv_vector_type">
* <field name="b" type="bytes"/>
* <field name="s" type="shorts"/>
* <field name="w" type="words"/>
* <field name="l" type="longs"/>
* <field name="q" type="quads"/>
* </union>
*/
info->vector_uint8.type = &type_uint8;
info->vector_uint8.count = info->vlenb[hartid];
info->type_uint8_vector.type = REG_TYPE_ARCH_DEFINED;
info->type_uint8_vector.id = "bytes";
info->type_uint8_vector.type_class = REG_TYPE_CLASS_VECTOR;
info->type_uint8_vector.reg_type_vector = &info->vector_uint8;
info->vector_uint16.type = &type_uint16;
info->vector_uint16.count = info->vlenb[hartid] / 2;
info->type_uint16_vector.type = REG_TYPE_ARCH_DEFINED;
info->type_uint16_vector.id = "shorts";
info->type_uint16_vector.type_class = REG_TYPE_CLASS_VECTOR;
info->type_uint16_vector.reg_type_vector = &info->vector_uint16;
info->vector_uint32.type = &type_uint32;
info->vector_uint32.count = info->vlenb[hartid] / 4;
info->type_uint32_vector.type = REG_TYPE_ARCH_DEFINED;
info->type_uint32_vector.id = "words";
info->type_uint32_vector.type_class = REG_TYPE_CLASS_VECTOR;
info->type_uint32_vector.reg_type_vector = &info->vector_uint32;
info->vector_uint64.type = &type_uint64;
info->vector_uint64.count = info->vlenb[hartid] / 8;
info->type_uint64_vector.type = REG_TYPE_ARCH_DEFINED;
info->type_uint64_vector.id = "longs";
info->type_uint64_vector.type_class = REG_TYPE_CLASS_VECTOR;
info->type_uint64_vector.reg_type_vector = &info->vector_uint64;
info->vector_uint128.type = &type_uint128;
info->vector_uint128.count = info->vlenb[hartid] / 16;
info->type_uint128_vector.type = REG_TYPE_ARCH_DEFINED;
info->type_uint128_vector.id = "quads";
info->type_uint128_vector.type_class = REG_TYPE_CLASS_VECTOR;
info->type_uint128_vector.reg_type_vector = &info->vector_uint128;
info->vector_fields[0].name = "b";
info->vector_fields[0].type = &info->type_uint8_vector;
if (info->vlenb[hartid] >= 2) {
info->vector_fields[0].next = info->vector_fields + 1;
info->vector_fields[1].name = "s";
info->vector_fields[1].type = &info->type_uint16_vector;
} else {
info->vector_fields[0].next = NULL;
}
if (info->vlenb[hartid] >= 4) {
info->vector_fields[1].next = info->vector_fields + 2;
info->vector_fields[2].name = "w";
info->vector_fields[2].type = &info->type_uint32_vector;
} else {
info->vector_fields[1].next = NULL;
}
if (info->vlenb[hartid] >= 8) {
info->vector_fields[2].next = info->vector_fields + 3;
info->vector_fields[3].name = "l";
info->vector_fields[3].type = &info->type_uint64_vector;
} else {
info->vector_fields[2].next = NULL;
}
if (info->vlenb[hartid] >= 16) {
info->vector_fields[3].next = info->vector_fields + 4;
info->vector_fields[4].name = "q";
info->vector_fields[4].type = &info->type_uint128_vector;
} else {
info->vector_fields[3].next = NULL;
}
info->vector_fields[4].next = NULL;
info->vector_union.fields = info->vector_fields;
info->type_vector.type = REG_TYPE_ARCH_DEFINED;
info->type_vector.id = "riscv_vector";
info->type_vector.type_class = REG_TYPE_CLASS_UNION;
info->type_vector.reg_type_union = &info->vector_union;
struct csr_info csr_info[] = {
#define DECLARE_CSR(name, number) { number, #name },
#include "encoding.h"
#undef DECLARE_CSR
};
/* encoding.h does not contain the registers in sorted order. */
qsort(csr_info, DIM(csr_info), sizeof(*csr_info), cmp_csr_info);
unsigned csr_info_index = 0;
unsigned custom_range_index = 0;
int custom_within_range = 0;
riscv_reg_info_t *shared_reg_info = calloc(1, sizeof(riscv_reg_info_t));
if (!shared_reg_info)
return ERROR_FAIL;
shared_reg_info->target = target;
/* When gdb requests register N, gdb_get_register_packet() assumes that this
* is register at index N in reg_list. So if there are certain registers
* that don't exist, we need to leave holes in the list (or renumber, but
* it would be nice not to have yet another set of numbers to translate
* between). */
for (uint32_t number = 0; number < target->reg_cache->num_regs; number++) {
struct reg *r = &target->reg_cache->reg_list[number];
r->dirty = false;
r->valid = false;
r->exist = true;
r->type = &riscv_reg_arch_type;
r->arch_info = shared_reg_info;
r->number = number;
r->size = riscv_xlen(target);
/* r->size is set in riscv_invalidate_register_cache, maybe because the
* target is in theory allowed to change XLEN on us. But I expect a lot
* of other things to break in that case as well. */
if (number <= GDB_REGNO_XPR31) {
r->exist = number <= GDB_REGNO_XPR15 ||
!riscv_supports_extension(target, hartid, 'E');
/* TODO: For now we fake that all GPRs exist because otherwise gdb
* doesn't work. */
r->exist = true;
r->caller_save = true;
switch (number) {
case GDB_REGNO_ZERO:
r->name = "zero";
break;
case GDB_REGNO_RA:
r->name = "ra";
break;
case GDB_REGNO_SP:
r->name = "sp";
break;
case GDB_REGNO_GP:
r->name = "gp";
break;
case GDB_REGNO_TP:
r->name = "tp";
break;
case GDB_REGNO_T0:
r->name = "t0";
break;
case GDB_REGNO_T1:
r->name = "t1";
break;
case GDB_REGNO_T2:
r->name = "t2";
break;
case GDB_REGNO_FP:
r->name = "fp";
break;
case GDB_REGNO_S1:
r->name = "s1";
break;
case GDB_REGNO_A0:
r->name = "a0";
break;
case GDB_REGNO_A1:
r->name = "a1";
break;
case GDB_REGNO_A2:
r->name = "a2";
break;
case GDB_REGNO_A3:
r->name = "a3";
break;
case GDB_REGNO_A4:
r->name = "a4";
break;
case GDB_REGNO_A5:
r->name = "a5";
break;
case GDB_REGNO_A6:
r->name = "a6";
break;
case GDB_REGNO_A7:
r->name = "a7";
break;
case GDB_REGNO_S2:
r->name = "s2";
break;
case GDB_REGNO_S3:
r->name = "s3";
break;
case GDB_REGNO_S4:
r->name = "s4";
break;
case GDB_REGNO_S5:
r->name = "s5";
break;
case GDB_REGNO_S6:
r->name = "s6";
break;
case GDB_REGNO_S7:
r->name = "s7";
break;
case GDB_REGNO_S8:
r->name = "s8";
break;
case GDB_REGNO_S9:
r->name = "s9";
break;
case GDB_REGNO_S10:
r->name = "s10";
break;
case GDB_REGNO_S11:
r->name = "s11";
break;
case GDB_REGNO_T3:
r->name = "t3";
break;
case GDB_REGNO_T4:
r->name = "t4";
break;
case GDB_REGNO_T5:
r->name = "t5";
break;
case GDB_REGNO_T6:
r->name = "t6";
break;
}
r->group = "general";
r->feature = &feature_cpu;
} else if (number == GDB_REGNO_PC) {
r->caller_save = true;
sprintf(reg_name, "pc");
r->group = "general";
r->feature = &feature_cpu;
} else if (number >= GDB_REGNO_FPR0 && number <= GDB_REGNO_FPR31) {
r->caller_save = true;
if (riscv_supports_extension(target, hartid, 'D')) {
r->size = 64;
if (riscv_supports_extension(target, hartid, 'F'))
r->reg_data_type = &type_ieee_single_double;
else
r->reg_data_type = &type_ieee_double;
} else if (riscv_supports_extension(target, hartid, 'F')) {
r->reg_data_type = &type_ieee_single;
r->size = 32;
} else {
r->exist = false;
}
switch (number) {
case GDB_REGNO_FT0:
r->name = "ft0";
break;
case GDB_REGNO_FT1:
r->name = "ft1";
break;
case GDB_REGNO_FT2:
r->name = "ft2";
break;
case GDB_REGNO_FT3:
r->name = "ft3";
break;
case GDB_REGNO_FT4:
r->name = "ft4";
break;
case GDB_REGNO_FT5:
r->name = "ft5";
break;
case GDB_REGNO_FT6:
r->name = "ft6";
break;
case GDB_REGNO_FT7:
r->name = "ft7";
break;
case GDB_REGNO_FS0:
r->name = "fs0";
break;
case GDB_REGNO_FS1:
r->name = "fs1";
break;
case GDB_REGNO_FA0:
r->name = "fa0";
break;
case GDB_REGNO_FA1:
r->name = "fa1";
break;
case GDB_REGNO_FA2:
r->name = "fa2";
break;
case GDB_REGNO_FA3:
r->name = "fa3";
break;
case GDB_REGNO_FA4:
r->name = "fa4";
break;
case GDB_REGNO_FA5:
r->name = "fa5";
break;
case GDB_REGNO_FA6:
r->name = "fa6";
break;
case GDB_REGNO_FA7:
r->name = "fa7";
break;
case GDB_REGNO_FS2:
r->name = "fs2";
break;
case GDB_REGNO_FS3:
r->name = "fs3";
break;
case GDB_REGNO_FS4:
r->name = "fs4";
break;
case GDB_REGNO_FS5:
r->name = "fs5";
break;
case GDB_REGNO_FS6:
r->name = "fs6";
break;
case GDB_REGNO_FS7:
r->name = "fs7";
break;
case GDB_REGNO_FS8:
r->name = "fs8";
break;
case GDB_REGNO_FS9:
r->name = "fs9";
break;
case GDB_REGNO_FS10:
r->name = "fs10";
break;
case GDB_REGNO_FS11:
r->name = "fs11";
break;
case GDB_REGNO_FT8:
r->name = "ft8";
break;
case GDB_REGNO_FT9:
r->name = "ft9";
break;
case GDB_REGNO_FT10:
r->name = "ft10";
break;
case GDB_REGNO_FT11:
r->name = "ft11";
break;
}
r->group = "float";
r->feature = &feature_fpu;
} else if (number >= GDB_REGNO_CSR0 && number <= GDB_REGNO_CSR4095) {
r->group = "csr";
r->feature = &feature_csr;
unsigned csr_number = number - GDB_REGNO_CSR0;
while (csr_info[csr_info_index].number < csr_number &&
csr_info_index < DIM(csr_info) - 1) {
csr_info_index++;
}
if (csr_info[csr_info_index].number == csr_number) {
r->name = csr_info[csr_info_index].name;
} else {
sprintf(reg_name, "csr%d", csr_number);
/* Assume unnamed registers don't exist, unless we have some
* configuration that tells us otherwise. That's important
* because eg. Eclipse crashes if a target has too many
* registers, and apparently has no way of only showing a
* subset of registers in any case. */
r->exist = false;
}
switch (csr_number) {
case CSR_FFLAGS:
case CSR_FRM:
case CSR_FCSR:
r->exist = riscv_supports_extension(target, hartid, 'F');
r->group = "float";
r->feature = &feature_fpu;
break;
case CSR_SSTATUS:
case CSR_STVEC:
case CSR_SIP:
case CSR_SIE:
case CSR_SCOUNTEREN:
case CSR_SSCRATCH:
case CSR_SEPC:
case CSR_SCAUSE:
case CSR_STVAL:
case CSR_SATP:
r->exist = riscv_supports_extension(target, hartid, 'S');
break;
case CSR_MEDELEG:
case CSR_MIDELEG:
/* "In systems with only M-mode, or with both M-mode and
* U-mode but without U-mode trap support, the medeleg and
* mideleg registers should not exist." */
r->exist = riscv_supports_extension(target, hartid, 'S') ||
riscv_supports_extension(target, hartid, 'N');
break;
case CSR_PMPCFG1:
case CSR_PMPCFG3:
case CSR_CYCLEH:
case CSR_TIMEH:
case CSR_INSTRETH:
case CSR_HPMCOUNTER3H:
case CSR_HPMCOUNTER4H:
case CSR_HPMCOUNTER5H:
case CSR_HPMCOUNTER6H:
case CSR_HPMCOUNTER7H:
case CSR_HPMCOUNTER8H:
case CSR_HPMCOUNTER9H:
case CSR_HPMCOUNTER10H:
case CSR_HPMCOUNTER11H:
case CSR_HPMCOUNTER12H:
case CSR_HPMCOUNTER13H:
case CSR_HPMCOUNTER14H:
case CSR_HPMCOUNTER15H:
case CSR_HPMCOUNTER16H:
case CSR_HPMCOUNTER17H:
case CSR_HPMCOUNTER18H:
case CSR_HPMCOUNTER19H:
case CSR_HPMCOUNTER20H:
case CSR_HPMCOUNTER21H:
case CSR_HPMCOUNTER22H:
case CSR_HPMCOUNTER23H:
case CSR_HPMCOUNTER24H:
case CSR_HPMCOUNTER25H:
case CSR_HPMCOUNTER26H:
case CSR_HPMCOUNTER27H:
case CSR_HPMCOUNTER28H:
case CSR_HPMCOUNTER29H:
case CSR_HPMCOUNTER30H:
case CSR_HPMCOUNTER31H:
case CSR_MCYCLEH:
case CSR_MINSTRETH:
case CSR_MHPMCOUNTER3H:
case CSR_MHPMCOUNTER4H:
case CSR_MHPMCOUNTER5H:
case CSR_MHPMCOUNTER6H:
case CSR_MHPMCOUNTER7H:
case CSR_MHPMCOUNTER8H:
case CSR_MHPMCOUNTER9H:
case CSR_MHPMCOUNTER10H:
case CSR_MHPMCOUNTER11H:
case CSR_MHPMCOUNTER12H:
case CSR_MHPMCOUNTER13H:
case CSR_MHPMCOUNTER14H:
case CSR_MHPMCOUNTER15H:
case CSR_MHPMCOUNTER16H:
case CSR_MHPMCOUNTER17H:
case CSR_MHPMCOUNTER18H:
case CSR_MHPMCOUNTER19H:
case CSR_MHPMCOUNTER20H:
case CSR_MHPMCOUNTER21H:
case CSR_MHPMCOUNTER22H:
case CSR_MHPMCOUNTER23H:
case CSR_MHPMCOUNTER24H:
case CSR_MHPMCOUNTER25H:
case CSR_MHPMCOUNTER26H:
case CSR_MHPMCOUNTER27H:
case CSR_MHPMCOUNTER28H:
case CSR_MHPMCOUNTER29H:
case CSR_MHPMCOUNTER30H:
case CSR_MHPMCOUNTER31H:
r->exist = riscv_xlen(target) == 32;
break;
case CSR_VSTART:
case CSR_VXSAT:
case CSR_VXRM:
case CSR_VL:
case CSR_VTYPE:
case CSR_VLENB:
r->exist = riscv_supports_extension(target, hartid, 'V');
break;
}
if (!r->exist && expose_csr) {
for (unsigned i = 0; expose_csr[i].low <= expose_csr[i].high; i++) {
if (csr_number >= expose_csr[i].low && csr_number <= expose_csr[i].high) {
LOG_INFO("Exposing additional CSR %d", csr_number);
r->exist = true;
break;
}
}
}
} else if (number == GDB_REGNO_PRIV) {
sprintf(reg_name, "priv");
r->group = "general";
r->feature = &feature_virtual;
r->size = 8;
} else if (number >= GDB_REGNO_V0 && number <= GDB_REGNO_V31) {
r->caller_save = false;
r->exist = riscv_supports_extension(target, hartid, 'V') && info->vlenb[hartid];
r->size = info->vlenb[hartid] * 8;
sprintf(reg_name, "v%d", number - GDB_REGNO_V0);
r->group = "vector";
r->feature = &feature_vector;
r->reg_data_type = &info->type_vector;
} else if (number >= GDB_REGNO_COUNT) {
/* Custom registers. */
assert(expose_custom);
range_t *range = &expose_custom[custom_range_index];
assert(range->low <= range->high);
unsigned custom_number = range->low + custom_within_range;
r->group = "custom";
r->feature = &feature_custom;
r->arch_info = calloc(1, sizeof(riscv_reg_info_t));
if (!r->arch_info)
return ERROR_FAIL;
((riscv_reg_info_t *) r->arch_info)->target = target;
((riscv_reg_info_t *) r->arch_info)->custom_number = custom_number;
sprintf(reg_name, "custom%d", custom_number);
custom_within_range++;
if (custom_within_range > range->high - range->low) {
custom_within_range = 0;
custom_range_index++;
}
}
if (reg_name[0])
r->name = reg_name;
reg_name += strlen(reg_name) + 1;
assert(reg_name < info->reg_names + target->reg_cache->num_regs *
max_reg_name_len);
r->value = info->reg_cache_values[number];
}
return ERROR_OK;
}
void riscv_add_bscan_tunneled_scan(struct target *target, struct scan_field *field,
riscv_bscan_tunneled_scan_context_t *ctxt)
{
jtag_add_ir_scan(target->tap, &select_user4, TAP_IDLE);
memset(ctxt->tunneled_dr, 0, sizeof(ctxt->tunneled_dr));
if (bscan_tunnel_type == BSCAN_TUNNEL_DATA_REGISTER) {
ctxt->tunneled_dr[3].num_bits = 1;
ctxt->tunneled_dr[3].out_value = bscan_one;
ctxt->tunneled_dr[2].num_bits = 7;
ctxt->tunneled_dr_width = field->num_bits;
ctxt->tunneled_dr[2].out_value = &ctxt->tunneled_dr_width;
/* for BSCAN tunnel, there is a one-TCK skew between shift in and shift out, so
scanning num_bits + 1, and then will right shift the input field after executing the queues */
ctxt->tunneled_dr[1].num_bits = field->num_bits + 1;
ctxt->tunneled_dr[1].out_value = field->out_value;
ctxt->tunneled_dr[1].in_value = field->in_value;
ctxt->tunneled_dr[0].num_bits = 3;
ctxt->tunneled_dr[0].out_value = bscan_zero;
} else {
/* BSCAN_TUNNEL_NESTED_TAP */
ctxt->tunneled_dr[0].num_bits = 1;
ctxt->tunneled_dr[0].out_value = bscan_one;
ctxt->tunneled_dr[1].num_bits = 7;
ctxt->tunneled_dr_width = field->num_bits;
ctxt->tunneled_dr[1].out_value = &ctxt->tunneled_dr_width;
/* for BSCAN tunnel, there is a one-TCK skew between shift in and shift out, so
scanning num_bits + 1, and then will right shift the input field after executing the queues */
ctxt->tunneled_dr[2].num_bits = field->num_bits + 1;
ctxt->tunneled_dr[2].out_value = field->out_value;
ctxt->tunneled_dr[2].in_value = field->in_value;
ctxt->tunneled_dr[3].num_bits = 3;
ctxt->tunneled_dr[3].out_value = bscan_zero;
}
jtag_add_dr_scan(target->tap, ARRAY_SIZE(ctxt->tunneled_dr), ctxt->tunneled_dr, TAP_IDLE);
}