s390 && user_enable_single_step() (Was: odd utrace testing results on s390x)

[Oleg Nesterov]

On 01/04, Oleg Nesterov wrote:
>
> IOW. I think this problem is minor and probably can be ignored,

Or may be not...

Even if the child is not killed by SIGTRAP, it can get a lot of
unnecessary traps.

To verify, I did another trivial patch (below), and the test
case from 6580807da14c423f0d0a708108e6df6ebc8bc83d does trigger
a lot of "false step" printks.

Hmm. And sometimes there is nothing in dmesg, but the test-case
needs a lot of time to complete. "taskset -c" seems to always
trigger printk's. Magic.

Oleg.

--- arch/s390/kernel/traps.c~	2009-12-22 10:41:52.909174198 -0500
+++ arch/s390/kernel/traps.c	2010-01-04 13:19:51.038187586 -0500
@@ -384,6 +384,8 @@ void __kprobes do_single_step(struct pt_
 	}
 	if (tracehook_consider_fatal_signal(current, SIGTRAP))
 		force_sig(SIGTRAP, current);
+	else
+		printk("false step\n");
 }
 
 static void default_trap_handler(struct pt_regs * regs, long interruption_code)