{"id":1980,"date":"2018-12-14T16:31:52","date_gmt":"2018-12-14T16:31:52","guid":{"rendered":"http:\/\/www.max-sperling.bplaced.net\/?p=1980"},"modified":"2025-12-05T14:42:11","modified_gmt":"2025-12-05T14:42:11","slug":"coredump-analysing-with-gdb","status":"publish","type":"post","link":"http:\/\/www.max-sperling.bplaced.net\/?p=1980","title":{"rendered":"Coredump analysis (GDB)"},"content":{"rendered":"

Setup<\/h2>\n
\r\n$ cat \/proc\/sys\/kernel\/core_pattern ... get coredumps location\r\n$ ulimit -c                         ... check max coredump size\r\n$ gdb <executable> <coredump>       ... load coredump with GDB\r\n<\/pre>\n
\nDetails \u25bc<\/summary>\n
  Activate cordumps for the current session<\/strong><\/p>\n
    \n
  1. Set coredump size to unlimited\n
    \r\n$ ulimit -c unlimited\r\n<\/pre>\n<\/li>\n
  2. Set pattern to store coredump\n
    \r\n$ sudo sysctl -w kernel.core_pattern=\/tmp\/core.%e.%p.%t\r\n<\/pre>\n<\/li>\n<\/ol>\n<\/details>\n
    \n
    Threads & Stack<\/h2>\n
    \r\ninfo threads        ... show running threads\r\nthread apply all bt ... show the bt of all running threads\r\nthread <idx> bt     ... show the bt of specific thread\r\nthread <idx>        ... switch to other thread\r\n\r\nbt [<lvl's>]        ... show the bt of the current thread\r\nup                  ... move up in bt\r\ndown                ... move down in bt\r\nf <idx>             ... select frame <idx>\r\ninfo f              ... shows selected frame\r\n<\/pre>\n
    \nDetails \u25bc<\/summary>\n
    \r\nset debug frame <0\/1> ... De\/activate debug infos for frames\r\n                          (Shows for example:\r\n                          - types (INLINE_FRAME, NORMAL_FRAME, ...)\r\n                          - unwinder (inline, dwarf2, amd64 prologue, ...))\r\n<\/pre>\n<\/details>\n
    \n
    Functions & Variables<\/h2>\n
    \r\nlist [function]            ... show source code\r\ndisas [\/<m>] [function]    ... show disassembly\r\n\r\n<m> (mode)\r\n  - b ... prints bytes and instruction (in memory order)\r\n  - m ... prints assembly and sources (not useful with optimized code)\r\n  - r ... prints bytes and instruction (in instruction order)\r\n  - s ... prints assembly and sources\r\n\r\ninfo register              ... show registers\r\ninfo args                  ... show parameters\r\ninfo locals                ... show local variables\r\n\r\nptype [variable]           ... show type of variable\r\np[\/<f>] $<register>        ... print register\r\np[\/<f>] <variable>         ... print variable\r\np[\/<f>] &<variable>        ... print address of variable\r\np[\/<f>] *<variable>        ... print dereferenced variable\r\np[\/<f>] *this              ... print member variables\r\np[\/<f>] sizeof(<var|type>) ... print size of variable\/type\r\n\r\n<f> (format)\r\n  - t ... binary\r\n  - x ... hexadecimal\r\n  - d ... signed decimal\r\n  - u ... unsigned decimal\r\n  - o ... octal\r\n  - a ... address\r\n  - c ... character\r\n  - f ... float point\r\n  - s ... string\r\n<\/pre>\n
    \nDetails \u25bc<\/summary>\n
      Convenience Variables<\/strong><\/p>\n
    \r\np $_siginfo ... show signal details\r\n<\/pre>\n
      Containers<\/strong><\/p>\n
    \r\np <vec>._M_impl._M_start+<idx> ... get item of vector\r\n<\/pre>\n<\/details>\n
    \n
    Memory<\/h2>\n
    \r\ninfo proc mappings ... show memory mapping\r\n\r\nx\/<n><f><u> <addr> ... print memory\r\n\r\n<n> (number)\r\n\r\n<f> (format)\r\n  - Basic types see <f> for print\r\n  - i ... instruction\r\n\r\n<u> (unit)\r\n  - b ... byte (8-bit integer)\r\n  - h ... half word (16-bit integer)\r\n  - w ... word (32-bit integer)\r\n  - g ... giant word (64-bit integer)\r\n<\/pre>\n
    \nDetails \u25bc<\/summary>\n