my expectation is that cntrl-C from the ddd console will have the same=20
as "kill -INT" to the kernel thread. It doesn't seem to work that way.
Here's a transcript from the ddd console:
GNU DDD 3.3.1 (i386-suse-linux), by Dorothea L=FCtkehaus and Andreas Zell=
Copyright =A9 1995-1999 Technische Universit=E4t Braunschweig, Germany.
Copyright =A9 1999-2001 Universit=E4t Passau, Germany.
Breakpoint 1 at 0xa00101e3: file panic.c, line 52.
Breakpoint 2 at 0xa00dbacb: file user_util.c, line 51.
Program received signal SIGINT, Interrupt.
0xa019c0c5 in __sigprocmask (how=3D1, set=3D0xa023b7a4, oset=3D0xa023b724=
When I do the "continue" at the (gdb) prompt, the UML is stuck. I'm not s=
which thread is getting the SIGINT, but it doesn't look like the kernel=20
thanks for your help,
Jeff Dike wrote:
>>This console handles cntrl-C correctly (sending SIGINT to the running
>>user task). When I try to cntrl-C from the gdb console (in ddd), it
>>appears SIGINT is not handled correctly. SIGINT works fine when sent
>>directly to the kernel thread with "kill -INT".
>Why do you think this behavior is incorrect? That sounds exactly right