SIGTRAP recieved while trying to call C function from GDB

SIGTRAP recieved while trying to call C function from GDB

[Avi Gozlan]

Hello,

We get SIGTRAP when trying to call C functions in GDB prompt. Following is an example program:

#include <stdio.h>

class A{
public:
  A() {}
};

int main(int argc, char **argv)
{
  A *a = new A;

  printf("hello, world\n");
}

Scenario:
1. Run gdb a.out 
2. b main 
3. r 
4. n (GDB stops prior to the printf() call) 
5. p atoi(“1”) 

(gdb) p atoi(“1”)

Program received signal SIGTRAP, Trace/breakpoint trap.
0x004b7411 in malloc () from /lib/ld-linux.so.2
The program being debugged was signaled while in a function called from GDB.
GDB remains in the frame where the signal was received.
To change this behavior use "set unwindonsignal on"
Evaluation of the expression containing the function (malloc) will be abandoned.
(gdb)

This does not happen if calling ‘p atoi(“1”)’ immediately when running the program or following the printf() function call.

GDB version: 6.8
Linux Release: Red Hat Enterprise Linux Server release 5.1 (Tikanga)
Kernel version: 2.6.18-53.1.13.el5PAE
Compiler: either gcc 4.1.2 20070626 or gcc 3.2.3 20030502
Compilation command: gcc –g –l stdc++   myprog.cc

Thanks,

Avi

Re: SIGTRAP recieved while trying to call C function from GDB

[Hui Zhu]

I cannot reproduce this issue in gdb-cvs-head.  AMD64 UBUNTU

(gdb) start
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
Temporary breakpoint 1 at 0x40065c: file 2.c, line 10.
Starting program: /home/teawater/gdb/rec/bgdbno/gdb/a.out

Temporary breakpoint 1, main (argc=
During symbol reading, incomplete CFI data; unspecified registers
(e.g., rax) at 0x400655.
1, argv=0x7fffffffe3b8) at 2.c:10
10	  A *a = new A;
Current language:  auto
The current source language is "auto; currently c++".
(gdb) n
12	  printf("hello, world\n");
(gdb) p atoi("1")
$1 = 1

Thanks,
Hui


2009/9/6 Avi Gozlan <avi@checkpoint.com>:
> Hello,
>
> We get SIGTRAP when trying to call C functions in GDB prompt. Following is an example program:
>
> #include <stdio.h>
>
> class A{
> public:
>   A() {}
> };
>
> int main(int argc, char **argv)
> {
>   A *a = new A;
>
>   printf("hello, world\n");
> }
>
> Scenario:
> 1. Run gdb a.out
> 2. b main
> 3. r
> 4. n (GDB stops prior to the printf() call)
> 5. p atoi(“1”)
>
> (gdb) p atoi(“1”)
>
> Program received signal SIGTRAP, Trace/breakpoint trap.
> 0x004b7411 in malloc () from /lib/ld-linux.so.2
> The program being debugged was signaled while in a function called from GDB.
> GDB remains in the frame where the signal was received.
> To change this behavior use "set unwindonsignal on"
> Evaluation of the expression containing the function (malloc) will be abandoned.
> (gdb)
>
> This does not happen if calling ‘p atoi(“1”)’ immediately when running the program or following the printf() function call.
>
> GDB version: 6.8
> Linux Release: Red Hat Enterprise Linux Server release 5.1 (Tikanga)
> Kernel version: 2.6.18-53.1.13.el5PAE
> Compiler: either gcc 4.1.2 20070626 or gcc 3.2.3 20030502
> Compilation command: gcc –g –l stdc++   myprog.cc
>
> Thanks,
>
> Avi
>
>

Re: SIGTRAP recieved while trying to call C function from GDB

[Daniel Jacobowitz]

On Sun, Sep 06, 2009 at 06:03:37PM +0300, Avi Gozlan wrote:
> GDB version: 6.8
> Linux Release: Red Hat Enterprise Linux Server release 5.1 (Tikanga)
> Kernel version: 2.6.18-53.1.13.el5PAE

Is this the latest RHEL kernel?  I believe this was a known kernel
fault around the 2.6.18 era.

-- 
Daniel Jacobowitz
CodeSourcery

RE: SIGTRAP recieved while trying to call C function from GDB

[Avi Gozlan]

Indeed this is not the latest RHEL kernel. It happens on RHEL 5.1

Thanks,

Avi

-----Original Message-----
From: Daniel Jacobowitz [mailto:drow@false.org] 
Sent: Sunday, September 06, 2009 6:26 PM
To: Avi Gozlan
Cc: gdb@sourceware.org
Subject: Re: SIGTRAP recieved while trying to call C function from GDB

On Sun, Sep 06, 2009 at 06:03:37PM +0300, Avi Gozlan wrote:
> GDB version: 6.8
> Linux Release: Red Hat Enterprise Linux Server release 5.1 (Tikanga)
> Kernel version: 2.6.18-53.1.13.el5PAE

Is this the latest RHEL kernel?  I believe this was a known kernel
fault around the 2.6.18 era.

-- 
Daniel Jacobowitz
CodeSourcery

Scanned by Check Point Total Security Gateway.

Re: SIGTRAP recieved while trying to call C function from GDB

[Jan Kratochvil]

> From: Daniel Jacobowitz [mailto:drow@false.org] 
> On Sun, Sep 06, 2009 at 06:03:37PM +0300, Avi Gozlan wrote:
> > Linux Release: Red Hat Enterprise Linux Server release 5.1 (Tikanga)
> > Kernel version: 2.6.18-53.1.13.el5PAE
> 
> Is this the latest RHEL kernel?  I believe this was a known kernel
> fault around the 2.6.18 era.

On Tue, 08 Sep 2009 12:31:11 +0200, Avi Gozlan wrote:
> Indeed this is not the latest RHEL kernel. It happens on RHEL 5.1

Just a confirmation RHEL-5.1 ptrace in kernel is known it had issues primarily
visible with GDB.  It was fixed in general by the RHEL-5.2 kernel.

Sure I would like to recommend the latest RHEL-5.4 instead now having
backported also various ptrace fixes being done upstream.


Thanks,
Jan