From owner-freebsd-current@FreeBSD.ORG Thu Jun 30 17:02:58 2011 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id E28DD106564A for ; Thu, 30 Jun 2011 17:02:57 +0000 (UTC) (envelope-from yanegomi@gmail.com) Received: from mail-vx0-f182.google.com (mail-vx0-f182.google.com [209.85.220.182]) by mx1.freebsd.org (Postfix) with ESMTP id A122E8FC18 for ; Thu, 30 Jun 2011 17:02:57 +0000 (UTC) Received: by vxg33 with SMTP id 33so2470275vxg.13 for ; Thu, 30 Jun 2011 10:02:56 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; bh=+ud8KdR7+/go6JzJ6avVBJU6J3X+PzBdO14HsQ63rfg=; b=BVxXP0mrW8oleHJTtclRkh9qbkJ9IEO8U2UHhNVuFk/Kzk1vt4T47aieGYzSqAyyES 0etjKXwql5RCm3QZrLI6/S5iRJZ6nBnHJRU+U5bikZhVvB5rNH7iYj3M8XNLFUOq4SZL fByj+rMntxrSqNzZVlo3kmI9ZsYVsMhJupe3A= MIME-Version: 1.0 Received: by 10.220.59.193 with SMTP id m1mr76640vch.38.1309453376800; Thu, 30 Jun 2011 10:02:56 -0700 (PDT) Received: by 10.220.92.201 with HTTP; Thu, 30 Jun 2011 10:02:56 -0700 (PDT) In-Reply-To: <841737947.159814.1309447573370.JavaMail.root@mail-01.cse.ucsc.edu> References: <576562817.159786.1309446593826.JavaMail.root@mail-01.cse.ucsc.edu> <841737947.159814.1309447573370.JavaMail.root@mail-01.cse.ucsc.edu> Date: Thu, 30 Jun 2011 10:02:56 -0700 Message-ID: From: Garrett Cooper To: Tim Gustafson Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Cc: freebsd-current@freebsd.org Subject: Re: FreeBSD 9 X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 30 Jun 2011 17:02:58 -0000 On Thu, Jun 30, 2011 at 8:26 AM, Tim Gustafson wrote: > Hi, > > I've installed FreeBSD 9 on a new server because 8.2 doesn't have support= for the LSI SAS2008 controller. =A0I've also built the system as a ZFS-roo= t box, and I have to say that I'm quite happy with the disk performance: we= 're getting about 500MB/s write and 675MB/s read. =A0All in all, I'm very h= appy with FreeBSD 9. > > I have noticed two snafus that I thought I'd send to the group just as fe= edback: > > -------------------------------------------------------------------------= ------- > 1. net-snmp fails to compile with the following error: > -------------------------------------------------------------------------= ------- > > /bin/sh ../../libtool --mode=3Dcompile cc -I../../include -I. -I../../age= nt -I../../agent/mibgroup =A0-I../../snmplib -I/usr/include =A0 -O2 -pipe -= fno-strict-aliasing -Ufreebsd9 -Dfreebsd9=3Dfreebsd9 -c -o mibII/tcpTable.l= o mibII/tcpTable.c > libtool: compile: =A0cc -I../../include -I. -I../../agent -I../../agent/m= ibgroup -I../../snmplib -I/usr/include -O2 -pipe -fno-strict-aliasing -Ufre= ebsd9 -Dfreebsd9=3Dfreebsd9 -c mibII/tcpTable.c =A0-fPIC -DPIC -o mibII/.li= bs/tcpTable.o > mibII/tcpTable.c:94: error: field 'pcb' has incomplete type > mibII/tcpTable.c: In function 'tcpTable_load': > mibII/tcpTable.c:866: error: dereferencing pointer to incomplete type > mibII/tcpTable.c:868: error: dereferencing pointer to incomplete type > mibII/tcpTable.c:868: error: invalid application of 'sizeof' to incomplet= e type 'struct xinpgen' > mibII/tcpTable.c:872: error: dereferencing pointer to incomplete type > mibII/tcpTable.c:876: error: dereferencing pointer to incomplete type > mibII/tcpTable.c:877: error: invalid application of 'sizeof' to incomplet= e type 'struct inpcb' > mibII/tcpTable.c:881: error: dereferencing pointer to incomplete type Someone already filed a PR for this ( http://www.freebsd.org/cgi/query-pr.cgi?pr=3Dports/158266 ) and I'm working on cleaning up the autoconf tests to work properly for $WORK so we can upgrade to 5.6.1.1. The problem is caused by the recent netinet / net content shuffling and the fact that the autoconf tests for net-snmp are broken (and a number of includes on files). Unfortunately the upstream maintainers used a sledgehammer approach for all of the BSDs to detect how headers were supposed to be #include'd, and there's a lot of namespace pollution involved. Thanks, -Garrett