gdb除錯cordump檔案

2021-10-01 07:32:56 字數 3830 閱讀 6304

自己在除錯的時候，簡單的使用方法，目前只是菜鳥水平，後面繼續學習

資料使應用崩潰可以生成core檔案

ulimit -c unlimited
gdb 應用名稱 ***.core
程式崩潰，bt檢視堆疊資訊

switching to thread 1 (lwp 687)]


#0 kill (pid=, sig=6) at src/signal/kill.c:7


7 src/signal/kill.c: no such file or directory.


(gdb) bt


#0 kill (pid=, sig=6) at src/signal/kill.c:7


#1 #2 congruent_aligned32 () at src/string/arm/memcpy_le.s:149


#3 0xb6dddb7c in fb_fliphwsu***ce () from /usr/lib/libminigui_ths-3.2.so.0


#4 0x0012c000 in ?? ()


backtrace stopped: previous frame identical to this frame (corrupt stack?)


info threads 檢視幾個執行緒前面的數字是執行緒id
檢視有幾個執行緒,並且資訊會標識出在哪個執行緒死掉

(gdb) info thread


id target id frame 


29 lwp 1005 __munmap (start=0xb66d3000, len=2457600)


at src/mman/munmap.c:12


28 lwp 757 __cp_end () at src/thread/arm/syscall_cp.s:25


27 lwp 1269 __cp_begin () at src/thread/arm/syscall_cp.s:23


26 lwp 1267 __cp_begin () at src/thread/arm/syscall_cp.s:23


25 lwp 1265 __cp_begin () at src/thread/arm/syscall_cp.s:23


24 lwp 1049 __cp_begin () at src/thread/arm/syscall_cp.s:23


23 lwp 1047 __cp_begin () at src/thread/arm/syscall_cp.s:23


22 lwp 1045 __cp_begin () at src/thread/arm/syscall_cp.s:23


21 lwp 1043 __cp_begin () at src/thread/arm/syscall_cp.s:23


20 lwp 1041 __cp_begin () at src/thread/arm/syscall_cp.s:23


19 lwp 1039 __cp_begin () at src/thread/arm/syscall_cp.s:23


18 lwp 1037 __cp_begin () at src/thread/arm/syscall_cp.s:23


17 lwp 1035 __cp_begin () at src/thread/arm/syscall_cp.s:23


16 lwp 1033 __cp_begin () at src/thread/arm/syscall_cp.s:23


15 lwp 1031 __cp_begin () at src/thread/arm/syscall_cp.s:23


14 lwp 1029 __cp_begin () at src/thread/arm/syscall_cp.s:23


13 lwp 1027 __cp_begin () at src/thread/arm/syscall_cp.s:23


12 lwp 1025 __cp_begin () at src/thread/arm/syscall_cp.s:23


11 lwp 1023 __cp_begin () at src/thread/arm/syscall_cp.s:23


10 lwp 1021 __cp_begin () at src/thread/arm/syscall_cp.s:23


9 lwp 1019 __cp_begin () at src/thread/arm/syscall_cp.s:23


---type to continue, or q to quit---


8 lwp 1017 __cp_begin () at src/thread/arm/syscall_cp.s:23


7 lwp 1015 __cp_begin () at src/thread/arm/syscall_cp.s:23


6 lwp 1013 __cp_begin () at src/thread/arm/syscall_cp.s:23


5 lwp 1011 __cp_begin () at src/thread/arm/syscall_cp.s:23


4 lwp 755 __cp_begin () at src/thread/arm/syscall_cp.s:23


3 lwp 754 __cp_begin () at src/thread/arm/syscall_cp.s:23


2 lwp 753 __cp_begin () at src/thread/arm/syscall_cp.s:23


1 lwp 687 kill (pid=, sig=6)


at src/signal/kill.c:7
可以看到我們bt看到的崩潰資訊是執行緒1的資訊

跳轉到29執行緒

bt 檢視29執行緒的堆疊

(gdb) thread 29


[switching to thread 29 (lwp 1005)]


#0 __munmap (start=0xb66d3000, len=2457600) at src/mman/munmap.c:12


12 src/mman/munmap.c: no such file or directory.


(gdb) bt


#0 __munmap (start=0xb66d3000, len=2457600) at src/mman/munmap.c:12


#1 0xb6dddc48 in fb_doublebufferenable ()


from /usr/lib/libminigui_ths-3.2.so.0


#2 0x000283b8 in cvtedoublebufferenable ()


#3 0x000174a4 in customswitchproc ()


#4 0xb6d8a724 in updatewindow () from /usr/lib/libminigui_ths-3.2.so.0


#5 0x00031f04 in cvtesettingswifiswitchproc ()


#6 0x00017414 in customswitchproc ()


#7 0xb6da5a94 in dispatchmessage () from /usr/lib/libminigui_ths-3.2.so.0


#8 0x00029710 in cvtesettingsthread ()


#9 0xb6f5ffdc in start (p=0xb5ce3d2c) at src/thread/pthread_create.c:145
從上面的資訊可以看出，29執行緒正常執行，然後呼叫到fb_doublebufferenable->fb_fliphwsu***ce觸發主線程/也就是1號執行緒死掉，整個應用死掉

GDB除錯core檔案

1.什麼是core dump？core dump說的是作業系統執行的乙個動作，當某個程序因為一些原因意外終止 crash 的時候，作業系統會將這個程序當時的記憶體資訊轉儲 dump 到磁碟上，產生的檔案就是core檔案了，一般會以core.形式命名。2.那些訊號能夠引起core dump？經常用到的...

GDB除錯coredump檔案

linux上程式崩潰起來挺煩人，不過linux 比較好的是有gdb.echo ulimit c unlimited etc profile 然後記得敲入命令 source etc profile然後敲入命令 ulimit c效果如下 確認能否生成coredump檔案，使用如下命令 使用時注意，我在測...

gdb除錯coredump檔案

linux上程式崩潰起來挺煩人，不過linux 比較好的是有gdb.echo ulimit c unlimited etc profile 然後記得敲入命令 source etc profile然後敲入命令 ulimit c效果如下 確認能否生成coredump檔案，使用如下命令 使用時注意，我在測...