Re: [Qemu-devel] Signal management in qemu-user

[Peter Maydell]

On 17 May 2012 14:33, Andreas Färber <address@hidden> wrote:
> Am 17.05.2012 11:23, schrieb Alex Barcelo:
>> Running it in a i386 machine works and gives an output of "0x0d\n0x20".
>> Running it in a qemu-i386 segfaults. Because the self-modifying code
>> raises a SIGSEGV in the qemu (I understand that it is the method used by
>> qemu to handle self-modifying code). But the sigprocmask disables the
>> SIGSEGV and the qemu-user... does nothing to avoid it. So the SIGSEGV is
>> unmanaged and breaks the program.
>
> Alex has the following SIGSEGV workaround queued for our openSUSE package:
> http://repo.or.cz/w/qemu/agraf.git/commit/0760e24b52ff20a328f168ed23b52c9b9c0fd28f
>
> Don't know if it fixes your specific problem.

No, that's a different issue, see the analysis here:
 http://lists.gnu.org/archive/html/qemu-devel/2012-02/msg02868.html
In that case an internal-to-QEMU allocation fails (because the guest
imposes a severe rlimit), we don't handle it very well and die
horribly. The commit you link to above is a sticking plaster on
the problem and definitely not OK for master.

> Peter had some ideas how to refactor signal handling but iirc
> didn't have time to work on it himself.

That was mostly to do with solving the problem of signals occurring
during emulation of a system call and the consequent race conditions:
 http://lists.gnu.org/archive/html/qemu-devel/2011-12/msg00384.html

I think Alex Barcelo's problem here is unrelated. I'm not very
surprised by it, though -- linux-user is not very robust in
complex edge cases (and in particular is really poor for
threaded programs).

-- PMM