inline void
insque(void *a, void *b)
{
register struct quehead *element = (struct quehead *) a;
register struct quehead *head = (struct quehead *) b;
element->qh_link = head->qh_link;
(line 27 is the last line)
(gdb) p *element
Cannot access memory at address 0x0
(gdb) p a
$3 = (void *) 0x0
This is called from here in slirp/if.c:
/* If there are more packets for this session, re-queue them */
if (ifm->ifs_next != /* ifm->ifs_prev != */ ifm) {
insque(ifm->ifs_next, ifqt);
ifs_remque(ifm);
}
It sounds like ifm expects its last element to have ifm->ifs_next = ifm,
but it's actually == NULL.
I don't see any changes to this file likely to have introduced the
regression, looking further up the stack.
** Also affects: qemu
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/932539
Title:
qemu exits with -11 when connecting to a port redirect before the
service starts listening
Status in QEMU:
New
Status in “qemu-kvm” package in Ubuntu:
Confirmed
Bug description:
This was detected initially as a crash in the auto upgrade tester.
The code of the upgrade tester basically spawns a kvm instance in the
background with a port redirect from localhost:54322 to tcp:22 in the VM, then
wait for that port to allow for a ssh connection before continuing the upgrade
testing.
In the past (Oneiric), all worked well but since Precise, we now get
qemu exitting with -11 at every single test :(
A quick reproducer is:
- start a VM that has openssh-server installed with: -net
user,hostfwd=tcp::54322-:22
- immediately start "ssh -p 54322 127.0.0.1" before the VM starts booting
(BIOS/GRUB state)
Then wait for sshd to start in the VM and qemu will exit with -11.
To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/932539/+subscriptions
