| /* |
| * Copyright (c) 2020 Intel Corporation. |
| * |
| * SPDX-License-Identifier: Apache-2.0 |
| */ |
| |
| #include <zephyr/device.h> |
| #include <zephyr/kernel.h> |
| |
| #include <zephyr/logging/log.h> |
| LOG_MODULE_REGISTER(gdbstub); |
| |
| #include <zephyr/sys/util.h> |
| |
| #include <ctype.h> |
| #include <stdbool.h> |
| #include <stdio.h> |
| #include <stdlib.h> |
| #include <string.h> |
| #include <zephyr/toolchain.h> |
| #include <sys/types.h> |
| #include <zephyr/sys/util.h> |
| |
| #include <zephyr/debug/gdbstub.h> |
| #include "gdbstub_backend.h" |
| |
| /* +1 is for the NULL character added during receive */ |
| #define GDB_PACKET_SIZE (CONFIG_GDBSTUB_BUF_SZ + 1) |
| |
| /* GDB remote serial protocol does not define errors value properly |
| * and handle all error packets as the same the code error is not |
| * used. There are informal values used by others gdbstub |
| * implementation, like qemu. Lets use the same here. |
| */ |
| #define GDB_ERROR_GENERAL "E01" |
| #define GDB_ERROR_MEMORY "E14" |
| #define GDB_ERROR_OVERFLOW "E22" |
| |
| static bool not_first_start; |
| |
| /* Empty memory region array */ |
| __weak const struct gdb_mem_region gdb_mem_region_array[0]; |
| |
| /* Number of memory regions, default to 0 */ |
| __weak const size_t gdb_mem_num_regions; |
| |
| /** |
| * Given a starting address and length of a memory block, find a memory |
| * region descriptor from the memory region array where the memory block |
| * fits inside the memory region. |
| * |
| * @param addr Starting address of the memory block |
| * @param len Length of the memory block |
| * |
| * @return Pointer to the memory region description if found. |
| * NULL if not found. |
| */ |
| static inline const |
| struct gdb_mem_region *find_memory_region(const uintptr_t addr, const size_t len) |
| { |
| const struct gdb_mem_region *r, *ret = NULL; |
| unsigned int idx; |
| |
| for (idx = 0; idx < gdb_mem_num_regions; idx++) { |
| r = &gdb_mem_region_array[idx]; |
| |
| if ((addr >= r->start) && |
| (addr < r->end) && |
| ((addr + len) >= r->start) && |
| ((addr + len) < r->end)) { |
| ret = r; |
| break; |
| } |
| } |
| |
| return ret; |
| } |
| |
| bool gdb_mem_can_read(const uintptr_t addr, const size_t len, uint8_t *align) |
| { |
| bool ret = false; |
| const struct gdb_mem_region *r; |
| |
| if (gdb_mem_num_regions == 0) { |
| /* |
| * No region is defined. |
| * Assume memory access is not restricted, and there is |
| * no alignment requirement. |
| */ |
| *align = 1; |
| ret = true; |
| } else { |
| r = find_memory_region(addr, len); |
| if (r != NULL) { |
| if ((r->attributes & GDB_MEM_REGION_READ) == |
| GDB_MEM_REGION_READ) { |
| if (r->alignment > 0) { |
| *align = r->alignment; |
| } else { |
| *align = 1; |
| } |
| ret = true; |
| } |
| } |
| } |
| |
| return ret; |
| } |
| |
| bool gdb_mem_can_write(const uintptr_t addr, const size_t len, uint8_t *align) |
| { |
| bool ret = false; |
| const struct gdb_mem_region *r; |
| |
| if (gdb_mem_num_regions == 0) { |
| /* |
| * No region is defined. |
| * Assume memory access is not restricted, and there is |
| * no alignment requirement. |
| */ |
| *align = 1; |
| ret = true; |
| } else { |
| r = find_memory_region(addr, len); |
| if (r != NULL) { |
| if ((r->attributes & GDB_MEM_REGION_WRITE) == |
| GDB_MEM_REGION_WRITE) { |
| if (r->alignment > 0) { |
| *align = r->alignment; |
| } else { |
| *align = 1; |
| } |
| |
| ret = true; |
| } |
| } |
| } |
| |
| return ret; |
| } |
| |
| size_t gdb_bin2hex(const uint8_t *buf, size_t buflen, char *hex, size_t hexlen) |
| { |
| if ((hexlen + 1) < buflen * 2) { |
| return 0; |
| } |
| |
| for (size_t i = 0; i < buflen; i++) { |
| if (hex2char(buf[i] >> 4, &hex[2 * i]) < 0) { |
| return 0; |
| } |
| if (hex2char(buf[i] & 0xf, &hex[2 * i + 1]) < 0) { |
| return 0; |
| } |
| } |
| |
| return 2 * buflen; |
| } |
| |
| __weak |
| int arch_gdb_add_breakpoint(struct gdb_ctx *ctx, uint8_t type, |
| uintptr_t addr, uint32_t kind) |
| { |
| return -2; |
| } |
| |
| __weak |
| int arch_gdb_remove_breakpoint(struct gdb_ctx *ctx, uint8_t type, |
| uintptr_t addr, uint32_t kind) |
| { |
| return -2; |
| } |
| |
| |
| /** |
| * Add preamble and termination to the given data. |
| * |
| * It returns 0 if the packet was acknowledge, -1 otherwise. |
| */ |
| static int gdb_send_packet(const uint8_t *data, size_t len) |
| { |
| uint8_t buf[2]; |
| uint8_t checksum = 0; |
| |
| /* Send packet start */ |
| z_gdb_putchar('$'); |
| |
| /* Send packet data and calculate checksum */ |
| while (len-- > 0) { |
| checksum += *data; |
| z_gdb_putchar(*data++); |
| } |
| |
| /* Send the checksum */ |
| z_gdb_putchar('#'); |
| |
| if (gdb_bin2hex(&checksum, 1, buf, sizeof(buf)) == 0) { |
| return -1; |
| } |
| |
| z_gdb_putchar(buf[0]); |
| z_gdb_putchar(buf[1]); |
| |
| if (z_gdb_getchar() == '+') { |
| return 0; |
| } |
| |
| /* Just got an invalid response */ |
| return -1; |
| } |
| |
| /** |
| * Receives one whole GDB packet. |
| * |
| * @retval 0 Success |
| * @retval -1 Checksum error |
| * @retval -2 Incoming packet too large |
| */ |
| static int gdb_get_packet(uint8_t *buf, size_t buf_len, size_t *len) |
| { |
| uint8_t ch = '0'; |
| uint8_t expected_checksum, checksum = 0; |
| uint8_t checksum_buf[2]; |
| |
| /* Wait for packet start */ |
| checksum = 0; |
| |
| /* wait for the start character, ignore the rest */ |
| while (ch != '$') { |
| ch = z_gdb_getchar(); |
| } |
| |
| *len = 0; |
| /* Read until receive '#' */ |
| while (true) { |
| ch = z_gdb_getchar(); |
| |
| if (ch == '#') { |
| break; |
| } |
| |
| /* Only put into buffer if not full */ |
| if (*len < (buf_len - 1)) { |
| buf[*len] = ch; |
| } |
| |
| checksum += ch; |
| (*len)++; |
| } |
| |
| buf[*len] = '\0'; |
| |
| /* Get checksum now */ |
| checksum_buf[0] = z_gdb_getchar(); |
| checksum_buf[1] = z_gdb_getchar(); |
| |
| if (hex2bin(checksum_buf, 2, &expected_checksum, 1) == 0) { |
| return -1; |
| } |
| |
| /* Verify checksum */ |
| if (checksum != expected_checksum) { |
| LOG_DBG("Bad checksum. Got 0x%x but was expecting: 0x%x", |
| checksum, expected_checksum); |
| /* NACK packet */ |
| z_gdb_putchar('-'); |
| return -1; |
| } |
| |
| /* ACK packet */ |
| z_gdb_putchar('+'); |
| |
| if (*len >= (buf_len - 1)) { |
| return -2; |
| } else { |
| return 0; |
| } |
| } |
| |
| /* Read memory byte-by-byte */ |
| static inline int gdb_mem_read_unaligned(uint8_t *buf, size_t buf_len, |
| uintptr_t addr, size_t len) |
| { |
| uint8_t data; |
| size_t pos, count = 0; |
| |
| /* Read from system memory */ |
| for (pos = 0; pos < len; pos++) { |
| data = *(uint8_t *)(addr + pos); |
| count += gdb_bin2hex(&data, 1, buf + count, buf_len - count); |
| } |
| |
| return count; |
| } |
| |
| /* Read memory with alignment constraint */ |
| static inline int gdb_mem_read_aligned(uint8_t *buf, size_t buf_len, |
| uintptr_t addr, size_t len, |
| uint8_t align) |
| { |
| /* |
| * Memory bus cannot do byte-by-byte access and |
| * each access must be aligned. |
| */ |
| size_t read_sz, pos; |
| size_t remaining = len; |
| uint8_t *mem_ptr; |
| size_t count = 0; |
| int ret; |
| |
| union { |
| uint32_t u32; |
| uint8_t b8[4]; |
| } data; |
| |
| /* Max alignment */ |
| if (align > 4) { |
| ret = -1; |
| goto out; |
| } |
| |
| /* Round down according to alignment. */ |
| mem_ptr = UINT_TO_POINTER(ROUND_DOWN(addr, align)); |
| |
| /* |
| * Figure out how many bytes to skip (pos) and how many |
| * bytes to read at the beginning of aligned memory access. |
| */ |
| pos = addr & (align - 1); |
| read_sz = MIN(len, align - pos); |
| |
| /* Loop till there is nothing more to read. */ |
| while (remaining > 0) { |
| data.u32 = *(uint32_t *)mem_ptr; |
| |
| /* |
| * Read read_sz bytes from memory and |
| * convert the binary data into hexadecimal. |
| */ |
| count += gdb_bin2hex(&data.b8[pos], read_sz, |
| buf + count, buf_len - count); |
| |
| remaining -= read_sz; |
| if (remaining > align) { |
| read_sz = align; |
| } else { |
| read_sz = remaining; |
| } |
| |
| /* Read the next aligned datum. */ |
| mem_ptr += align; |
| |
| /* |
| * Any memory accesses after the first one are |
| * aligned by design. So there is no need to skip |
| * any bytes. |
| */ |
| pos = 0; |
| }; |
| |
| ret = count; |
| |
| out: |
| return ret; |
| } |
| |
| /** |
| * Read data from a given memory address and length. |
| * |
| * @return Number of bytes read from memory, or -1 if error |
| */ |
| static int gdb_mem_read(uint8_t *buf, size_t buf_len, |
| uintptr_t addr, size_t len) |
| { |
| uint8_t align; |
| int ret; |
| |
| /* |
| * Make sure there is enough space in the output |
| * buffer for hexadecimal representation. |
| */ |
| if ((len * 2) > buf_len) { |
| ret = -1; |
| goto out; |
| } |
| |
| if (!gdb_mem_can_read(addr, len, &align)) { |
| ret = -1; |
| goto out; |
| } |
| |
| if (align > 1) { |
| ret = gdb_mem_read_aligned(buf, buf_len, |
| addr, len, |
| align); |
| } else { |
| ret = gdb_mem_read_unaligned(buf, buf_len, |
| addr, len); |
| } |
| |
| out: |
| return ret; |
| } |
| |
| /* Write memory byte-by-byte */ |
| static int gdb_mem_write_unaligned(const uint8_t *buf, uintptr_t addr, |
| size_t len) |
| { |
| uint8_t data; |
| int ret; |
| size_t count = 0; |
| |
| while (len > 0) { |
| size_t cnt = hex2bin(buf, 2, &data, sizeof(data)); |
| |
| if (cnt == 0) { |
| ret = -1; |
| goto out; |
| } |
| |
| *(uint8_t *)addr = data; |
| |
| count += cnt; |
| addr++; |
| buf += 2; |
| len--; |
| } |
| |
| ret = count; |
| |
| out: |
| return ret; |
| } |
| |
| /* Write memory with alignment constraint */ |
| static int gdb_mem_write_aligned(const uint8_t *buf, uintptr_t addr, |
| size_t len, uint8_t align) |
| { |
| size_t pos, write_sz; |
| uint8_t *mem_ptr; |
| size_t count = 0; |
| int ret; |
| |
| /* |
| * Incoming buf is of hexadecimal characters, |
| * so binary data size is half of that. |
| */ |
| size_t remaining = len; |
| |
| union { |
| uint32_t u32; |
| uint8_t b8[4]; |
| } data; |
| |
| /* Max alignment */ |
| if (align > 4) { |
| ret = -1; |
| goto out; |
| } |
| |
| /* |
| * Round down according to alignment. |
| * Read the data (of aligned size) first |
| * as we need to do read-modify-write. |
| */ |
| mem_ptr = UINT_TO_POINTER(ROUND_DOWN(addr, align)); |
| data.u32 = *(uint32_t *)mem_ptr; |
| |
| /* |
| * Figure out how many bytes to skip (pos) and how many |
| * bytes to write at the beginning of aligned memory access. |
| */ |
| pos = addr & (align - 1); |
| write_sz = MIN(len, align - pos); |
| |
| /* Loop till there is nothing more to write. */ |
| while (remaining > 0) { |
| /* |
| * Write write_sz bytes from memory and |
| * convert the binary data into hexadecimal. |
| */ |
| size_t cnt = hex2bin(buf, write_sz * 2, |
| &data.b8[pos], write_sz); |
| |
| if (cnt == 0) { |
| ret = -1; |
| goto out; |
| } |
| |
| count += cnt; |
| buf += write_sz * 2; |
| |
| remaining -= write_sz; |
| if (remaining > align) { |
| write_sz = align; |
| } else { |
| write_sz = remaining; |
| } |
| |
| /* Write data to memory */ |
| *(uint32_t *)mem_ptr = data.u32; |
| |
| /* Point to the next aligned datum. */ |
| mem_ptr += align; |
| |
| if (write_sz != align) { |
| /* |
| * Since we are not writing a full aligned datum, |
| * we need to do read-modify-write. Hence reading |
| * it here before the next hex2bin() call. |
| */ |
| data.u32 = *(uint32_t *)mem_ptr; |
| } |
| |
| /* |
| * Any memory accesses after the first one are |
| * aligned by design. So there is no need to skip |
| * any bytes. |
| */ |
| pos = 0; |
| }; |
| |
| ret = count; |
| |
| out: |
| return ret; |
| } |
| |
| /** |
| * Write data to a given memory address and length. |
| * |
| * @return Number of bytes written to memory, or -1 if error |
| */ |
| static int gdb_mem_write(const uint8_t *buf, uintptr_t addr, |
| size_t len) |
| { |
| uint8_t align; |
| int ret; |
| |
| if (!gdb_mem_can_write(addr, len, &align)) { |
| ret = -1; |
| goto out; |
| } |
| |
| if (align > 1) { |
| ret = gdb_mem_write_aligned(buf, addr, len, align); |
| } else { |
| ret = gdb_mem_write_unaligned(buf, addr, len); |
| } |
| |
| out: |
| return ret; |
| } |
| |
| /** |
| * Send a exception packet "T <value>" |
| */ |
| static int gdb_send_exception(uint8_t *buf, size_t len, uint8_t exception) |
| { |
| size_t size; |
| |
| *buf = 'T'; |
| size = gdb_bin2hex(&exception, 1, buf + 1, len - 1); |
| if (size == 0) { |
| return -1; |
| } |
| |
| /* Related to 'T' */ |
| size++; |
| |
| return gdb_send_packet(buf, size); |
| } |
| |
| /** |
| * Synchronously communicate with gdb on the host |
| */ |
| int z_gdb_main_loop(struct gdb_ctx *ctx) |
| { |
| /* 'static' modifier is intentional so the buffer |
| * is not declared inside running stack, which may |
| * not have enough space. |
| */ |
| static uint8_t buf[GDB_PACKET_SIZE]; |
| |
| enum loop_state { |
| RECEIVING, |
| CONTINUE, |
| ERROR, |
| } state; |
| |
| state = RECEIVING; |
| |
| /* Only send exception if this is not the first |
| * GDB break. |
| */ |
| if (not_first_start) { |
| gdb_send_exception(buf, sizeof(buf), ctx->exception); |
| } else { |
| not_first_start = true; |
| } |
| |
| #define CHECK_ERROR(condition) \ |
| { \ |
| if ((condition)) { \ |
| state = ERROR; \ |
| break; \ |
| } \ |
| } |
| |
| #define CHECK_SYMBOL(c) \ |
| { \ |
| CHECK_ERROR(ptr == NULL || *ptr != (c)); \ |
| ptr++; \ |
| } |
| |
| #define CHECK_INT(arg) \ |
| { \ |
| arg = strtol((const char *)ptr, (char **)&ptr, 16); \ |
| CHECK_ERROR(ptr == NULL); \ |
| } |
| |
| while (state == RECEIVING) { |
| uint8_t *ptr; |
| size_t data_len, pkt_len; |
| uintptr_t addr; |
| uint32_t type; |
| int ret; |
| |
| ret = gdb_get_packet(buf, sizeof(buf), &pkt_len); |
| if ((ret == -1) || (ret == -2)) { |
| /* |
| * Send error and wait for next packet. |
| * |
| * -1: Checksum error. |
| * -2: Packet too big. |
| */ |
| gdb_send_packet(GDB_ERROR_GENERAL, 3); |
| continue; |
| } |
| |
| if (pkt_len == 0) { |
| continue; |
| } |
| |
| ptr = buf; |
| |
| switch (*ptr++) { |
| |
| /** |
| * Read from the memory |
| * Format: m addr,length |
| */ |
| case 'm': |
| CHECK_INT(addr); |
| CHECK_SYMBOL(','); |
| CHECK_INT(data_len); |
| |
| /* Read Memory */ |
| |
| /* |
| * GDB ask the guest to read parameters when |
| * the user request backtrace. If the |
| * parameter is a NULL pointer this will cause |
| * a fault. Just send a packet informing that |
| * this address is invalid |
| */ |
| if (addr == 0L) { |
| gdb_send_packet(GDB_ERROR_MEMORY, 3); |
| break; |
| } |
| ret = gdb_mem_read(buf, sizeof(buf), addr, data_len); |
| CHECK_ERROR(ret == -1); |
| gdb_send_packet(buf, ret); |
| break; |
| |
| /** |
| * Write to memory |
| * Format: M addr,length:val |
| */ |
| case 'M': |
| CHECK_INT(addr); |
| CHECK_SYMBOL(','); |
| CHECK_INT(data_len); |
| CHECK_SYMBOL(':'); |
| |
| if (addr == 0L) { |
| gdb_send_packet(GDB_ERROR_MEMORY, 3); |
| break; |
| } |
| |
| /* Write Memory */ |
| pkt_len = gdb_mem_write(ptr, addr, data_len); |
| CHECK_ERROR(pkt_len == -1); |
| gdb_send_packet("OK", 2); |
| break; |
| |
| /* |
| * Continue ignoring the optional address |
| * Format: c addr |
| */ |
| case 'c': |
| arch_gdb_continue(); |
| state = CONTINUE; |
| break; |
| |
| /* |
| * Step one instruction ignoring the optional address |
| * s addr..addr |
| */ |
| case 's': |
| arch_gdb_step(); |
| state = CONTINUE; |
| break; |
| |
| /* |
| * Read all registers |
| * Format: g |
| */ |
| case 'g': |
| pkt_len = arch_gdb_reg_readall(ctx, buf, sizeof(buf)); |
| CHECK_ERROR(pkt_len == 0); |
| gdb_send_packet(buf, pkt_len); |
| break; |
| |
| /** |
| * Write the value of the CPU registers |
| * Format: G XX... |
| */ |
| case 'G': |
| pkt_len = arch_gdb_reg_writeall(ctx, ptr, pkt_len - 1); |
| CHECK_ERROR(pkt_len == 0); |
| gdb_send_packet("OK", 2); |
| break; |
| |
| /** |
| * Read the value of a register |
| * Format: p n |
| */ |
| case 'p': |
| CHECK_INT(addr); |
| |
| /* Read Register */ |
| pkt_len = arch_gdb_reg_readone(ctx, buf, sizeof(buf), addr); |
| CHECK_ERROR(pkt_len == 0); |
| gdb_send_packet(buf, pkt_len); |
| break; |
| |
| /** |
| * Write data into a specific register |
| * Format: P register=value |
| */ |
| case 'P': |
| CHECK_INT(addr); |
| CHECK_SYMBOL('='); |
| |
| pkt_len = arch_gdb_reg_writeone(ctx, ptr, strlen(ptr), addr); |
| CHECK_ERROR(pkt_len == 0); |
| gdb_send_packet("OK", 2); |
| break; |
| |
| /* |
| * Breakpoints and Watchpoints |
| */ |
| case 'z': |
| __fallthrough; |
| case 'Z': |
| CHECK_INT(type); |
| CHECK_SYMBOL(','); |
| CHECK_INT(addr); |
| CHECK_SYMBOL(','); |
| CHECK_INT(data_len); |
| |
| if (buf[0] == 'Z') { |
| ret = arch_gdb_add_breakpoint(ctx, type, |
| addr, data_len); |
| } else if (buf[0] == 'z') { |
| ret = arch_gdb_remove_breakpoint(ctx, type, |
| addr, data_len); |
| } |
| |
| if (ret == -2) { |
| /* breakpoint/watchpoint not supported */ |
| gdb_send_packet(NULL, 0); |
| } else if (ret == -1) { |
| state = ERROR; |
| } else { |
| gdb_send_packet("OK", 2); |
| } |
| |
| break; |
| |
| |
| /* What cause the pause */ |
| case '?': |
| gdb_send_exception(buf, sizeof(buf), |
| ctx->exception); |
| break; |
| |
| /* |
| * Not supported action |
| */ |
| default: |
| gdb_send_packet(NULL, 0); |
| break; |
| } |
| |
| /* |
| * If this is an recoverable error, send an error message to |
| * GDB and continue the debugging session. |
| */ |
| if (state == ERROR) { |
| gdb_send_packet(GDB_ERROR_GENERAL, 3); |
| state = RECEIVING; |
| } |
| } |
| |
| #undef CHECK_ERROR |
| #undef CHECK_INT |
| #undef CHECK_SYMBOL |
| |
| return 0; |
| } |
| |
| int gdb_init(const struct device *arg) |
| { |
| ARG_UNUSED(arg); |
| |
| if (z_gdb_backend_init() == -1) { |
| LOG_ERR("Could not initialize gdbstub backend."); |
| return -1; |
| } |
| |
| arch_gdb_init(); |
| return 0; |
| } |
| |
| #ifdef CONFIG_XTENSA |
| /* |
| * Interrupt stacks are being setup during init and are not |
| * available before POST_KERNEL. Xtensa needs to trigger |
| * interrupts to get into GDB stub. So this can only be |
| * initialized in POST_KERNEL, or else the interrupt would not be |
| * using the correct interrupt stack and will result in |
| * double exception. |
| */ |
| SYS_INIT(gdb_init, POST_KERNEL, CONFIG_KERNEL_INIT_PRIORITY_DEFAULT); |
| #else |
| SYS_INIT(gdb_init, PRE_KERNEL_2, CONFIG_KERNEL_INIT_PRIORITY_DEFAULT); |
| #endif |