gdb attach won't work about pwndocker HOT 4 CLOSED

It seems that it's not the problem of gdb , because I can attach the process outside the python script (using the gdb shell command).

It looks like the problem lies in the pwntools python lib, since I have found some issue about this problem in their repo:
pwntools issue

I have checked two elf files , one is 32 Bits ELF and another is 64 Bits ELF, and both of them can be debugged in the gdb in bash. The same error occur in both of them. So I believe it's not the problem of ELF file to be debug.

I still can't fix the gdb.attach problem in the py script. Hope an elegant way can be found to solve it .

from pwndocker.

@Vccxx , I test with latest docker environment and it seems that it reports the same problem but the debugger works.
I compile a test program in my docker environment and it suceeds:
test.c

#include <stdio.h>
int main()
{
    puts("test");
   getchar();
   puts("done");
}

and I run docker run -it --privileged skysider/pwndocker/ /bin/bash, and it works.

from pwndocker.

I have test your test program on my docker and it doesn't work , here is how I did:
Firstly I start the docker with command :
sudo docker run -it --privileged skysider/pwndocker /bin/bash
Secondly I compile the test.c file with gcc test.c -o test, which contains the following code:

#include <stdio.h>
int main()
{
    puts("test");
   getchar();
   puts("done");
}

Thirdly I write the exp.py:

from pwn import *
context.log_level = "debug"
context.terminal = ["tmux","splitw","-h"]
io = process("./test")
io.send("1")
gdb.attach(io)
io.interactive()

and then I run the exp.py , get stuck here:

root@d3f8a524e41e:/ctf/work# python3 test_exp.py 
[+] Starting local process './test' argv=[b'./test'] : pid 38
[DEBUG] Sent 0x1 bytes:
    49 * 0x1
[*] running in new terminal: /usr/bin/gdb -q  "./test" 38
[DEBUG] Launching a new terminal: ['/usr/bin/tmux', 'splitw', '-h', '/usr/bin/gdb -q  "./test" 38']
[-] Waiting for debugger: debugger exited! (maybe check /proc/sys/kernel/yama/ptrace_scope)

and after about 1 minute, the program just exit without debugger attached:

root@d3f8a524e41e:/ctf/work# python3 test_exp.py 
[+] Starting local process './test' argv=[b'./test'] : pid 38
[DEBUG] Sent 0x1 bytes:
    49 * 0x1
[*] running in new terminal: /usr/bin/gdb -q  "./test" 38
[DEBUG] Launching a new terminal: ['/usr/bin/tmux', 'splitw', '-h', '/usr/bin/gdb -q  "./test" 38']
[-] Waiting for debugger: debugger exited! (maybe check /proc/sys/kernel/yama/ptrace_scope)
[*] Switching to interactive mode
[*] Process './test' stopped with exit code 0 (pid 38)
[DEBUG] Received 0xa bytes:
    b'test\n'
    b'done\n'
test
done
[*] Got EOF while reading in interactive
$ 
[DEBUG] Sent 0x1 bytes:
    10 * 0x1
[*] Got EOF while sending in interactive

Would you please show me your exp.py? I think there maybe some coding problem.

from pwndocker.

OK, It turns out that I didn't run exp.py in a tmux terminal but in a bash terminal......

the debugger works fine in tmux terminal.

Thank you for your time.

from pwndocker.