If a non-NBD client connects to qemu-nbd, we would end up with
a SIGSEGV in nbd_cilent_put() because we were trying to
unregister the client's association to the export, even though
we skipped inserting the client into that list.  Easy trigger
in two terminals:

$ qemu-nbd -p 30001 --format=raw file
$ nmap 127.0.0.1 -p 30001

nmap claims that it thinks it connected to a pago-services1
server (which probably means nmap could be updated to learn the
NBD protocol and give a more accurate diagnosis of the open
port - but that's not our problem), then terminates immediately,
so our call to nbd_negotiate() fails.  The fix is to reorder
nbd_co_client_start() to ensure that all initialization occurs
before we ever try talking to a client in nbd_negotiate(), so
that the teardown sequence on negotiation failure doesn't fault
while dereferencing a half-initialized object.

While debugging this, I also noticed that nbd_update_server_watch()
called by nbd_client_closed() was still adding a channel to accept
the next client, even when the state was no longer RUNNING.  That
is fixed by making nbd_can_accept() pay attention to the current
state.

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1451614

Signed-off-by: Eric Blake <eblake@redhat.com>
---

I'm planning to run a bisect to see which patch actually introduced
the problem, but wanted to post the patch first to get review started.

 nbd/server.c | 14 ++++++--------
 qemu-nbd.c   |  2 +-
 2 files changed, 7 insertions(+), 9 deletions(-)

diff --git a/nbd/server.c b/nbd/server.c
index 0c4f456..d8dfac8 100644
--- a/nbd/server.c
+++ b/nbd/server.c
@@ -1601,16 +1601,14 @@ static coroutine_fn void nbd_co_client_start(void *opaque)

     if (exp) {
         nbd_export_get(exp);
-    }
-    if (nbd_negotiate(data)) {
-        client_close(client);
-        goto out;
-    }
-    qemu_co_mutex_init(&client->send_lock);
-
-    if (exp) {
         QTAILQ_INSERT_TAIL(&exp->clients, client, next);
     }
+    qemu_co_mutex_init(&client->send_lock);
+
+    if (nbd_negotiate(data)) {
+        client_close(client);
+        goto out;
+    }

     nbd_client_receive_next_request(client);

diff --git a/qemu-nbd.c b/qemu-nbd.c
index 27a4e3a..5410854 100644
--- a/qemu-nbd.c
+++ b/qemu-nbd.c
@@ -322,7 +322,7 @@ out:

 static int nbd_can_accept(void)
 {
-    return nb_fds < shared;
+    return state == RUNNING && nb_fds < shared;
 }

 static void nbd_export_closed(NBDExport *exp)
-- 
2.9.4

On 05/26/2017 10:04 PM, Eric Blake wrote:
> If a non-NBD client connects to qemu-nbd, we would end up with
> a SIGSEGV in nbd_cilent_put() because we were trying to
> unregister the client's association to the export, even though
> we skipped inserting the client into that list.  Easy trigger
> in two terminals:
> 
> $ qemu-nbd -p 30001 --format=raw file
> $ nmap 127.0.0.1 -p 30001
> 

> Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1451614
> 
> Signed-off-by: Eric Blake <eblake@redhat.com>
> ---
> 
> I'm planning to run a bisect to see which patch actually introduced
> the problem, but wanted to post the patch first to get review started.

Looks like the problem of split initialization has existed since at
least 2.6; commit 1a6245a split up the QTAILQ_INSERT_TAIL and
nbd_export_get as part of refactoring to create nbd_co_client_start.
But even trying the commit before that, I still got a different
assertion failure:
qemu-nbd: nbd/server.c:521: nbd_client_put: Assertion `client->closing'
failed.

Going all the way back to 2.5 worked, but 2.6 was far enough back that I
didn't want to bisect further which commit first broke probes from nmap.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org

On 05/26/2017 10:04 PM, Eric Blake wrote:
> If a non-NBD client connects to qemu-nbd, we would end up with
> a SIGSEGV in nbd_cilent_put() because we were trying to
> unregister the client's association to the export, even though
> we skipped inserting the client into that list.  Easy trigger
> in two terminals:
> 
> $ qemu-nbd -p 30001 --format=raw file
> $ nmap 127.0.0.1 -p 30001

Since this is now part of a CVE fix, I'm adding qemu-stable in cc.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org