samba_spnupdate invoked oom-killer ( samba BUG )

[Jeremy Allison]

On Sat, Oct 04, 2014 at 10:13:31PM +0200, Volker Lendecke wrote:
> On Sat, Oct 04, 2014 at 11:50:21AM +0200, Volker Lendecke wrote:
> > You might want to take a look at the attached patchset. This
> > is what I've dug out of my huge messaging mess. It compiles,
> > but right now I don't have the time to even do some simple
> > tests. At least that's the direction where I want to go
> > eventually.
> 
> ... and with
> 
> 
> source3/lib/messages.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/source3/lib/messages.c b/source3/lib/messages.c
> index 94a738f..f153727 100644
> --- a/source3/lib/messages.c
> +++ b/source3/lib/messages.c
> @@ -206,7 +206,6 @@ static void messaging_recv_cb(const uint8_t *msg, size_t msg_len,
>  {
>  	struct messaging_context *msg_ctx = talloc_get_type_abort(
>  		private_data, struct messaging_context);
> -	uint8_t hdr[MESSAGE_HDR_LENGTH];
>  	uint32_t msg_type;
>  	struct server_id src, dst;
>  	struct server_id_buf idbuf;
> @@ -214,7 +213,7 @@ static void messaging_recv_cb(const uint8_t *msg, size_t msg_len,
>  	int64_t fds64[MIN(num_fds, INT8_MAX)];
>  	size_t i;
>  
> -	if (msg_len < sizeof(*hdr)) {
> +	if (msg_len < MESSAGE_HDR_LENGTH) {
>  		for (i=0; i < num_fds; i++) {
>  			close(fds[i]);
>  		}
> @@ -239,7 +238,7 @@ static void messaging_recv_cb(const uint8_t *msg, size_t msg_len,
>  		fds[i] = -1;
>  	}
>  
> -	message_hdr_get(&msg_type, &src, &dst, hdr);
> +	message_hdr_get(&msg_type, &src, &dst, msg);
>  
>  	DEBUG(10, ("%s: Received message 0x%x len %u (num_fds:%u) from %s\n",
>  		   __func__, (unsigned)msg_type,
> 
> on top it even does survive a few simple tests.
> 
> smbcontrol <samba-pid> ping
> 
> works both in standard as well as in single process model. That's what you
> get if you dig up stuff from big heap of stuff... Now to run autobuild.

Did the autobuild work ?

I'm really interested in getting this in and
unifying the messaging infrastructures.

Let me know when you think this code is ready
for review and possible push.

This is significant progress !!!

Cheers,

	Jeremy.