[Bug c++/90964] New: SJLJ: Backtrace stopped: previous frame inner to this frame (corrupt stack?)

ocroquette at free dot fr gcc-bugzilla@gcc.gnu.org
Sat Jun 22 16:14:00 GMT 2019


https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90964

            Bug ID: 90964
           Summary: SJLJ: Backtrace stopped: previous frame inner to this
                    frame (corrupt stack?)
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: ocroquette at free dot fr
  Target Milestone: ---

On Windows with the SJLJ exception handling, the backtrace doesn't work in GDB
as soon a method enters a try/catch block, instead GDB reports:

Backtrace stopped: previous frame inner to this frame (corrupt stack?)

The problem can be easily reproduced with any SJLJ Mingw-w64 toolchain.
The problem doesn't occur with SEH exception handling.

It is not clear if the problem is in GCC, GDB or MinGW-w64. I don't know how to
find out, hopefully someone competent can tell if it is a GCC bug or not.

main.cpp:

#include <stdexcept>

 void
 dummy()
 {
         volatile int i = 0;
 }

 bool
 expectTrue( bool value )
 {
     dummy();

     if ( !static_cast<bool>( value ) )
     {
         throw std::runtime_error("value failed");
     }

     dummy();

     return value;
 }

 int
 main( int argc, char** argv )
 {
     dummy();
     expectTrue( true );
     dummy();
 }


gcc --version

gcc (x86_64-posix-sjlj-rev0, Built by MinGW-W64 project) 7.2.0
Copyright (C) 2017 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.


g++ -g -O0 main.cpp -o main


gdb ./main

GNU gdb (GDB) 8.0.1
...
(gdb) break dummy
Breakpoint 1 at 0x401578: file main.cpp, line 7.
(gdb) run
...

Thread 1 hit Breakpoint 1, dummy () at main.cpp:7
7               volatile int i = 0;
(gdb) backtrace
#0  dummy () at main.cpp:7
#1  0x0000000000401728 in main (argc=1, argv=0x25d1d20) at main.cpp:28
(gdb) continue
Continuing.

Thread 1 hit Breakpoint 1, dummy () at main.cpp:7
7               volatile int i = 0;
(gdb) backtrace
#0  dummy () at main.cpp:7
#1  0x0000000000401626 in expectTrue (value=true) at main.cpp:13
#2  0x0000000000000000 in ?? ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
(gdb) continue
Continuing.

Thread 1 hit Breakpoint 1, dummy () at main.cpp:7
7               volatile int i = 0;
(gdb) backtrace
#0  dummy () at main.cpp:7
#1  0x000000000040167b in expectTrue (value=true) at main.cpp:20
#2  0x0000000000000000 in ?? ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
(gdb) continue
Continuing.

Thread 1 hit Breakpoint 1, dummy () at main.cpp:7
7               volatile int i = 0;
(gdb) backtrace
#0  dummy () at main.cpp:7
#1  0x0000000000401737 in main (argc=1, argv=0x25d1d20) at main.cpp:30
(gdb) continue
Continuing.
[Thread 5844.0x1750 exited with code 0]
[Thread 5844.0x1320 exited with code 0]
[Thread 5844.0x2334 exited with code 0]
[Inferior 1 (process 5844) exited normally]
(gdb)


More information about the Gcc-bugs mailing list