[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#659075: [rt.cpan.org #61577] ->sockdomain and ->socktype undefined on newly ->accept'ed sockets



On 02/16/2012 04:58 PM, Dominic Hargreaves wrote:
> On Thu, Feb 16, 2012 at 12:18:03AM -0500, Daniel Kahn Gillmor wrote:
>> Debian folks -- is this now a patch that folks feel comfortable
>> applying in anticipation of such a fix upstream?
> 
> Given the way the thread's gone so far, I think I'd prefer to see an
> upstream commit/release first, unless the issue is particularly urgent?

Well, it appears to be related to a couple other bugs, and upstream has
done nothing about it for over a year, which leaves me wondering when it
will get resolved upstream.

The change doesn't introduce any API modifications, and (as far as i can
tell) the only possible hiccup for debian will be dealing with the
unimplemented symbol Socket::SO_PROTOCOL on kfreebsd.

Apparently freebsd doesn't currently implement SO_PROTOCOL, though i
haven't been able to find a reason for that (i did find someone's
proposed patch, though [0]).

(i'm cc'ing the debian-bsd list here in hopes that they know of some
other way to retrieve the protocol of an existing socket from userspace
with that kernel)

	--dkg

[0] http://pastebin.com/5spLFG3V

--- src/sys/sys/socket.h.orig	2011-02-16 16:18:46.000000000 +0200
+++ src/sys/sys/socket.h	2011-11-21 05:02:21.000000000 +0200
@@ -140,6 +140,8 @@
 #define	SO_LISTENINCQLEN	0x1013	/* socket's incomplete queue length */
 #define	SO_USETFIB	0x1014	/* use this FIB to route */
 #endif
+#define SO_PROTO	0x1015
+#define SO_PROTOCOL	SO_PROTO

 /*
  * Structure used for manipulating linger option.
--- src/sys/kern/uipc_socket.c.orig	2011-02-16 16:18:46.000000000 +0200
+++ src/sys/kern/uipc_socket.c	2011-11-21 05:05:09.000000000 +0200
@@ -2734,6 +2734,10 @@
 			optval = so->so_type;
 			goto integer;

+		case SO_PROTO:
+			optval = so->so_proto->pr_protocol;
+			goto integer;
+
 		case SO_LISTENQLEN:
 			optval = so->so_qlen;
 			goto integer;


Reply to: