[Bug binutils/29278] New: gdb crashes on interrupt (non-stop mode, target extended-remote)

[Manfred.Steiner at gmx dot at]

https://sourceware.org/bugzilla/show_bug.cgi?id=29278

            Bug ID: 29278
           Summary: gdb crashes on interrupt (non-stop mode, target
                    extended-remote)
           Product: binutils
           Version: unspecified
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: binutils
          Assignee: unassigned at sourceware dot org
          Reporter: Manfred.Steiner at gmx dot at
  Target Milestone: ---

binutils-gdb Version: 13.0.50.20220615-git

When working with a AVR Stub in non-stop mode via target extended-remote, it is
possible to start without program running (stub returns an empty thread list on
qfThreadInfo request), or to detach from program. 

If you execute interrupt in this situation on gdb command line, gdb crashes
with backtrace output.

inferior.c:304: internal-error: find_inferior_pid: Assertion `pid != 0' failed.
A problem internal to GDB has been detected,
...
.../binutils-gdb/gdb/inferior.c:299
...


Looking inside source code shows the cause:

file gdb/inferior.c:298:

struct inferior *
find_inferior_pid (process_stratum_target *targ, int pid)
{
  /* Looking for inferior pid == 0 is always wrong, and indicative of
     a bug somewhere else.  There may be more than one with pid == 0,
     for instance.  */
  gdb_assert (pid != 0);
  ...

I think the comment is outdated and it would be better to replace the assert
with:
if (pid == 0) {
  return NULL;
}

because pid == 0 is a possible case in non-stop mode when using target
extended-remote.

Then the interrupt command would be ignored (or result in an error message when
the stub returns an error).

-- 
You are receiving this mail because:
You are on the CC list for the bug.