Re: [PATCH 1/3] Windows gdb: Show CONTEXT_EXCEPTION_REQUEST info in "info threads"

[Eli Zaretskii <eliz@gnu.org>]

> From: Pedro Alves <pedro@palves.net>
> Date: Mon,  4 May 2026 18:33:40 +0100
> 
> This patch makes the Windows native target expose
> CONTEXT_EXCEPTION_REQUEST information in "info threads" extra info,
> like so:
> 
>  (gdb) info threads
>    Id   Target Id                                     Frame
>    1    Thread 15648.0x4338 "sleeper"                 main () at sleeper.c:5
>    2    Thread 15648.0x3d58 (in syscall)              0x00007ffde37057b4 in ntdll!ZwWaitForWorkViaWorkerFactory () from C:/WINDOWS/SYSTEM32/ntdll.dll
>    3    Thread 15648.0x27e0 "sig" (in syscall)        0x00007ffde3701bc4 in ntdll!ZwReadFile () from C:/WINDOWS/SYSTEM32/ntdll.dll
>    4    Thread 15648.0x26f0 (in syscall)              0x00007ffde37057b4 in ntdll!ZwWaitForWorkViaWorkerFactory () from C:/WINDOWS/SYSTEM32/ntdll.dll
>  * 5    Thread 15648.0xff8 (in exception)             0x00007ffde00aa464 in KERNELBASE!CtrlRoutine () from C:/WINDOWS/System32/KERNELBASE.dll
>  (gdb)
> 
> Above, we can see that thread 1 is running user space code, threads 2
> to 4 are in some system call, and thread 5 raised an exception (a
> Ctrl-C).
> 
> This is useful information to see, as system calls are not
> interruptible on Windows.  E.g. an infcall on a thread that is blocked
> in a system call will appear to hang, until the system call returns on
> its own.
> 
> Change-Id: I04221f123eef81d59b5cc1c9fbb298f7a33fa001
> commit-id:d93544e4
> ---
>  gdb/nat/windows-nat.h |  3 ++-
>  gdb/windows-nat.c     | 22 ++++++++++++++++++++++
>  2 files changed, 24 insertions(+), 1 deletion(-)
> 
> diff --git a/gdb/nat/windows-nat.h b/gdb/nat/windows-nat.h
> index 52378765438..d66d5ec0ed3 100644
> --- a/gdb/nat/windows-nat.h
> +++ b/gdb/nat/windows-nat.h
> @@ -527,7 +527,8 @@ struct WindowsContext<CONTEXT *>
>  				     | CONTEXT_SEGMENTS
>  #endif
>  				     | CONTEXT_DEBUG_REGISTERS
> -				     | CONTEXT_EXTENDED_REGISTERS);
> +				     | CONTEXT_EXTENDED_REGISTERS
> +				     | CONTEXT_EXCEPTION_REQUEST);
>  };
>  
>  #ifdef __x86_64__
> diff --git a/gdb/windows-nat.c b/gdb/windows-nat.c
> index a9647e90bb8..d01fbc8e4ba 100644
> --- a/gdb/windows-nat.c
> +++ b/gdb/windows-nat.c
> @@ -3363,6 +3363,28 @@ windows_nat_target::extra_thread_info (thread_info *info)
>  	   || th->last_event.dwDebugEventCode == EXIT_PROCESS_DEBUG_EVENT)
>      return "exiting process";
>  
> +  /* Specifying CONTEXT_EXCEPTION_REQUEST in ContextFlags as input
> +     (which we do), asks Windows to report back why the thread entered
> +     kernel mode.  If GetThreadContext returns a context with
> +     CONTEXT_EXCEPTION_REPORTING set, it means that it understood the
> +     request.  */
> +  windows_process->fill_thread_context (th);
> +  DWORD context_flags = *windows_process->context_flags_ptr (th);
> +  if ((context_flags & CONTEXT_EXCEPTION_REPORTING) != 0)
> +    {
> +      /* The thread was running user space code which raised an
> +	 exception, which we intercepted.  */
> +      if ((context_flags & CONTEXT_EXCEPTION_ACTIVE) != 0)
> +	return "in exception";
> +
> +      /* The thread was running a system call.  */
> +      if ((context_flags & CONTEXT_SERVICE_ACTIVE) != 0)
> +	return "in syscall";
> +
> +      /* Otherwise, the thread was simply suspended while running
> +	 user space code.  */
> +    }
> +
>    return nullptr;
>  }

The CONTEXT_* constants you are adding aren't defined in mingw.org's
MinGW headers.  I'm guessing they were introduced for Vista or
something.  So I think we will need to have their explicit definitions
in nat/windows-nat.h, guarded with #ifndef.

Also, this page:

  https://zachsaw.blogspot.com/2010/11/wow64-bug-getthreadcontext-may-return.html

seems to say (near the end) that XP doesn't support
CONTEXT_EXCEPTION_REQUEST, so maybe this feature should be guarded by
a later Windows version, say 7 or 8.1?