Re: Again: UML on s390 (31Bit)

[Bodo Stroesser <bstroesser@fujitsu-siemens.com>]

Martin Schwidefsky wrote:
> > Each time when the kernel is entered again and a signal is pending,
> > do_signal() will be called on return to user with regs->trap setup
> > freshly. So, I still believe the patch doesn't have *any* effect.
>
>
> Oh, the patch does have an effect for the debugger. If the debugger
> stopped on the sys_sig_return system call and does e.g. an inferior
> function call, then the kernel might want to restart a system call
> that isn't there because the debugger did a "jump" but could not
> change regs->trap.

AFAICS, it not even has an effect for the debugger.

do_signal() is the only routine, that examines regs->trap. On each
kernel-entry, regs->trap is set freshly. What will be the effect of
changing it *after* it had been examined?

The only exceptions are sys_(rt_)sigsuspend. Here do_signal() might
be called twice, while *and* after processing the syscall. But even
here the patch has no effect, as regs->gprs[2] contains -EINTR, if
so_signal is called by sys_(rt_)sigreturn.

Regards
	Bodo

> blue skies,
>    Martin
>
> Martin Schwidefsky
> Linux for zSeries Development & Services
> IBM Deutschland Entwicklung GmbH
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/