[ENet-discuss] [PATCH] init ENetHost struct memory in enet_host_create

[Lee Salzman]

Committed to CVS. It's just complaining about some of the ENetBuffer
structs being uninitialized even though they are outside the count passed
to sendmsg.

On Thu, Dec 8, 2011 at 3:48 AM, Lucas Beyer <pompei2 at gmail.com> wrote:

> Dear list,
>
> I have come to the habit of running my small programs through valgrind.
> One error being reported comes from enet.
>
> This is the valgrind output:
>
> ==18627== Syscall param socketcall.sendmsg(msg.msg_iov[i]) points to
> uninitialised byte(s)
> ==18627==    at 0x56816B0: __sendmsg_nocancel (in /lib64/libc-2.14.1.so)
> ==18627==    by 0x40D5C5: enet_socket_send (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==    by 0x40CBFF: enet_protocol_send_outgoing_commands (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==    by 0x40CEB6: enet_host_service (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==    by 0x4045E4: main (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==  Address 0x593614c is 156 bytes inside a block of size 10,976
> alloc'd
> ==18627==    at 0x4C297CD: malloc (in
> /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
> ==18627==    by 0x40D8B7: enet_malloc (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==    by 0x405C2D: enet_host_create (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==    by 0x405190: enet::Client::Client(std::string const&,
> unsigned short, double, unsigned long, unsigned int, unsigned int) (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==    by 0x40455E: main (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==  Uninitialised value was created by a heap allocation
> ==18627==    at 0x4C297CD: malloc (in
> /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
> ==18627==    by 0x40D8B7: enet_malloc (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==    by 0x405C2D: enet_host_create (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==    by 0x405190: enet::Client::Client(std::string const&,
> unsigned short, double, unsigned long, unsigned int, unsigned int) (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
> ==18627==    by 0x40455E: main (in
> /windows/Xchange/fts/scm/git/fts/gamedbg/build-d/fts-gamedbg)
>
> I have looked into it for a little time and am now pretty confident it's
> not because of an usage error I made. My code is pretty simple and based on
> the tutorial. Applying the following patch fixes it for me:
>
> --- host.c.orig    2011-12-08 12:25:44.187125000 +0100
> +++ host.c    2011-12-08 12:23:43.661363000 +0100
> @@ -38,6 +38,7 @@
>      host = (ENetHost *) enet_malloc (sizeof (ENetHost));
>      if (host == NULL)
>        return NULL;
> +    memset (host, 0, sizeof (ENetHost));
>
>      host -> peers = (ENetPeer *) enet_malloc (peerCount * sizeof
> (ENetPeer));
>      if (host -> peers == NULL)
>
> I also attached the patch as a file to this email.
>
> --
> Kind regards, Lucas
> http://arkana-fts.org
>
> _______________________________________________
> ENet-discuss mailing list
> ENet-discuss at cubik.org
> http://lists.cubik.org/mailman/listinfo/enet-discuss
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cubik.org/pipermail/enet-discuss/attachments/20111208/ea686374/attachment.html>