Re: svn commit: r315662 - in head: contrib/bsnmp/snmp_mibII contrib/ipfilter/ipsend lib/libprocstat sys/netinet sys/sys usr.bin/netstat usr.bin/sockstat usr.bin/systat usr.sbin/tcpdrop usr.sbin/trpt
On Tue, Mar 21, 2017 at 08:22:19AM +0100, Baptiste Daroussin wrote: B> On Mon, Mar 20, 2017 at 11:54:33PM -0700, Gleb Smirnoff wrote: B> > On Tue, Mar 21, 2017 at 07:50:32AM +0100, Antoine Brodin wrote: B> > A> > This change is known to break a ton of ports. More than 100 if B> > A> > counting depends. I'm sorry for that and I already started to fix B> > A> > them. B> > A> > B> > A> > Please send all new breakages to me. B> > A> B> > A> Hi, B> > A> B> > A> Exp-runs should happen before breakage happens, not after. B> > A> If you already know that it breaks hundreds of ports, please revert B> > A> and request an exp-run. B> > B> > The exp-run has already been made: B> > B> > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=210673 B> > B> > Wasn't very helpful. B> > B> B> I disagree it was useful it showed you some ports that were breaking given there B> was some high level one you were expected to provide patches to fix them. I'm working on them right now. -- Totus tuus, Glebius. ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: svn commit: r315662 - in head: contrib/bsnmp/snmp_mibII contrib/ipfilter/ipsend lib/libprocstat sys/netinet sys/sys usr.bin/netstat usr.bin/sockstat usr.bin/systat usr.sbin/tcpdrop usr.sbin/trpt
On Tue, Mar 21, 2017 at 07:50:32AM +0100, Antoine Brodin wrote: A> > This change is known to break a ton of ports. More than 100 if A> > counting depends. I'm sorry for that and I already started to fix A> > them. A> > A> > Please send all new breakages to me. A> A> Hi, A> A> Exp-runs should happen before breakage happens, not after. A> If you already know that it breaks hundreds of ports, please revert A> and request an exp-run. The exp-run has already been made: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=210673 Wasn't very helpful. -- Totus tuus, Glebius. ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: svn commit: r315662 - in head: contrib/bsnmp/snmp_mibII contrib/ipfilter/ipsend lib/libprocstat sys/netinet sys/sys usr.bin/netstat usr.bin/sockstat usr.bin/systat usr.sbin/tcpdrop usr.sbin/trpt
Hi!
This change is known to break a ton of ports. More than 100 if
counting depends. I'm sorry for that and I already started to fix
them.
Please send all new breakages to me.
On Tue, Mar 21, 2017 at 06:39:49AM +, Gleb Smirnoff wrote:
T> Author: glebius
T> Date: Tue Mar 21 06:39:49 2017
T> New Revision: 315662
T> URL: https://svnweb.freebsd.org/changeset/base/315662
T>
T> Log:
T> Hide struct inpcb, struct tcpcb from the userland.
T>
T> This is a painful change, but it is needed. On the one hand, we avoid
T> modifying them, and this slows down some ideas, on the other hand we still
T> eventually modify them and tools like netstat(1) never work on next
version of
T> FreeBSD. We maintain a ton of spares in them, and we already got some
ifdef
T> hell at the end of tcpcb.
T>
T> Details:
T> - Hide struct inpcb, struct tcpcb under _KERNEL || _WANT_FOO.
T> - Make struct xinpcb, struct xtcpcb pure API structures, not including
T> kernel structures inpcb and tcpcb inside. Export into these structures
T> the fields from inpcb and tcpcb that are known to be used, and put there
T> a ton of spare space.
T> - Make kernel and userland utilities compilable after these changes.
T> - Bump __FreeBSD_version.
T>
T> Reviewed by: rrs, gnn
T> Differential Revision: D10018
T>
T> Modified:
T> head/contrib/bsnmp/snmp_mibII/mibII_tcp.c
T> head/contrib/bsnmp/snmp_mibII/mibII_udp.c
T> head/contrib/ipfilter/ipsend/sock.c
T> head/lib/libprocstat/libprocstat.c
T> head/sys/netinet/in_pcb.c
T> head/sys/netinet/in_pcb.h
T> head/sys/netinet/ip_divert.c
T> head/sys/netinet/raw_ip.c
T> head/sys/netinet/tcp_subr.c
T> head/sys/netinet/tcp_syncache.c
T> head/sys/netinet/tcp_timer.c
T> head/sys/netinet/tcp_timer.h
T> head/sys/netinet/tcp_var.h
T> head/sys/netinet/udp_usrreq.c
T> head/sys/sys/param.h
T> head/usr.bin/netstat/inet.c
T> head/usr.bin/sockstat/sockstat.c
T> head/usr.bin/systat/extern.h
T> head/usr.bin/systat/netcmds.c
T> head/usr.bin/systat/netstat.c
T> head/usr.sbin/tcpdrop/tcpdrop.c
T> head/usr.sbin/trpt/trpt.c
T>
T> Modified: head/contrib/bsnmp/snmp_mibII/mibII_tcp.c
T>
==
T> --- head/contrib/bsnmp/snmp_mibII/mibII_tcp.cTue Mar 21 05:15:10
2017(r315661)
T> +++ head/contrib/bsnmp/snmp_mibII/mibII_tcp.cTue Mar 21 06:39:49
2017(r315662)
T> @@ -310,7 +310,7 @@ op_tcpconn(struct snmp_context *ctx __un
T> switch (value->var.subs[sub - 1]) {
T>
T>case LEAF_tcpConnState:
T> -switch (tcpoids[i].tp->xt_tp.t_state) {
T> +switch (tcpoids[i].tp->t_state) {
T>
T>case TCPS_CLOSED:
T> value->v.integer = 1;
T>
T> Modified: head/contrib/bsnmp/snmp_mibII/mibII_udp.c
T>
==
T> --- head/contrib/bsnmp/snmp_mibII/mibII_udp.cTue Mar 21 05:15:10
2017(r315661)
T> +++ head/contrib/bsnmp/snmp_mibII/mibII_udp.cTue Mar 21 06:39:49
2017(r315662)
T> @@ -105,8 +105,8 @@ fetch_udp(void)
T> ptr->xig_len > sizeof(struct xinpgen);
T> ptr = (struct xinpgen *)(void *)((char *)ptr + ptr->xig_len)) {
T> inp = (struct xinpcb *)ptr;
T> -if (inp->xi_inp.inp_gencnt > xinpgen->xig_gen ||
T> -(inp->xi_inp.inp_vflag & INP_IPV4) == 0)
T> +if (inp->inp_gencnt > xinpgen->xig_gen ||
T> +(inp->inp_vflag & INP_IPV4) == 0)
T> continue;
T>
T> udp_total++;
T> @@ -128,17 +128,17 @@ fetch_udp(void)
T> ptr->xig_len > sizeof(struct xinpgen);
T> ptr = (struct xinpgen *)(void *)((char *)ptr + ptr->xig_len)) {
T> inp = (struct xinpcb *)ptr;
T> -if (inp->xi_inp.inp_gencnt > xinpgen->xig_gen ||
T> -(inp->xi_inp.inp_vflag & INP_IPV4) == 0)
T> +if (inp->inp_gencnt > xinpgen->xig_gen ||
T> +(inp->inp_vflag & INP_IPV4) == 0)
T> continue;
T> oid->inp = inp;
T> oid->index.len = 5;
T> -inaddr = ntohl(inp->xi_inp.inp_laddr.s_addr);
T> +inaddr = ntohl(inp->inp_laddr.s_addr);
T> oid->index.subs[0] = (inaddr >> 24) & 0xff;
T> oid->index.subs[1] = (inaddr >> 16) & 0xff;
T> oid->index.subs[2] = (inaddr >> 8) & 0xff;
T> oid->index.
Re: removing _WANT_IFADDR breaks net-snmp
On Thu, Feb 26, 2015 at 10:08:40PM -0500, Michael Butler wrote: M The recent changes which served to hide struct ifaddr have broken M net-snmp: I know and slowly working on that: https://lists.freebsd.org/pipermail/svn-src-head/2015-February/068674.html -- Totus tuus, Glebius. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: r263478 - /stable/10/sys/net/route.h breaks net-snmp
On Sun, Apr 06, 2014 at 10:11:31AM +0200, Oliver Lehmann wrote: O Hi Gleb, O O your change to route.h breaks the compilation of O net-mgmt/net-snmp. Can you please fix it? O O Here is a log from CURRENT, the error message on 10-STABLE is the same. O With 10-RELEASE the port compiled fine: O O http://beefy1.isc.freebsd.org/bulk/head-i386-default/latest/logs/errors/net-snmp-5.7.2_4.log Strange. This is how it builds on my CURRENT: checking for struct rtentry.rt_dst... no checking for struct rtentry.rt_hash... no checking for struct rtentry.rt_next... no checking for struct rtentry.rt_refcnt... no checking for struct rtentry.rt_unit... no checking for struct rtentry.rt_use... no checking type of rtentry structure... unknown checking for struct rtentry... rtentry After that, it compiles fine. In your log it detects 'struct ortentry' and surprisingly prefers it. -- Totus tuus, Glebius. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: r263478 - /stable/10/sys/net/route.h breaks net-snmp
On Mon, Apr 07, 2014 at 12:28:01PM +0200, Oliver Lehmann wrote: O Strange. This is how it builds on my CURRENT: O O checking for struct rtentry.rt_dst... no O checking for struct rtentry.rt_hash... no O checking for struct rtentry.rt_next... no O checking for struct rtentry.rt_refcnt... no O checking for struct rtentry.rt_unit... no O checking for struct rtentry.rt_use... no O checking type of rtentry structure... unknown O checking for struct rtentry... rtentry O O After that, it compiles fine. O O In your log it detects 'struct ortentry' and surprisingly prefers it. O O Please check 10-STABLE. O - ortentry is visible O - rtentry is not visible O I guess this is the reason why. I didn't checked CURRENT, just grabbed O up the log via google (I have no CURRENT system atm). My issue is on O STABLE where you merged this to. I wonder why this API-breakage got O merged anyway?! The ortentry wasn't part of any API or ABI. It was just defined. I am very surprised that net-snmp picks it and utilizes it during compilation. If it uses the structure anywhere, then it obtains invalid info from it. So that is actually a bug on net-snmp side taking that structure. I guess that in CURRENT, where ortentry is wiped entirely, and net-snmp takes rtentry, we've got the bug accidentially fixed. I'm now running universe build with r263203 merged. If it succeeds, I will merge r263203 removing declaration of ortentry. -- Totus tuus, Glebius. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: astro/gpsd fails to build on CURRENT
Rainer, On Tue, May 07, 2013 at 06:19:16PM +0200, Rainer Hurling wrote: R When I try to build astro/gpsd on 10.0-CURRENT it fails with the R following messages (devel/scons should be up to date): R R /usr/ports/astro/gpsd#make R === Found saved configuration for gpsd-3.9 ... R Checking for C header file sys/endian.h... yes R Checking that xsltproc can make man pages... yes R Altered configuration variables: R mtk3301 = False (default True): MTK-3301 support R nmea2000 = False (default True): NMEA2000/CAN support R bluez = False (default True): BlueZ support for Bluetooth devices R libQgpsmm = False (default True): build QT bindings R chrpath = False (default True): use chrpath to edit library load paths R mandir = man (default share/man): manual pages directory R pkgconfig = libdata/pkgconfig (default lib/pkgconfig): pkgconfig file R directory R TypeError: Tried to lookup Dir '/usr/local/lib' as a File.: R File /usr/ports/astro/gpsd/work/gpsd-3.9/SConstruct, line 955: R parse_flags=gpsdlibs + ncurseslibs + ['-lm']) R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 258: R return MethodWrapper.__call__(self, target, source, *args, **kw) R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 222: R return self.method(*nargs, **kwargs) R File /usr/local/lib/scons-2.1.0/SCons/Builder.py, line 631: R env = env.Override(env_kw) R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 635: R if merges: env.MergeFlags(merges) R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 810: R args = self.ParseFlags(args) R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 796: R do_parse(arg) R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 670: R for t in arg: do_parse(t) R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 726: R dict['LIBS'].append(self.fs.File(arg)) R File /usr/local/lib/scons-2.1.0/SCons/Node/FS.py, line 1339: R return self._lookup(name, directory, File, create) R File /usr/local/lib/scons-2.1.0/SCons/Node/FS.py, line 1318: R return root._lookup_abs(p, fsclass, create) R File /usr/local/lib/scons-2.1.0/SCons/Node/FS.py, line 2223: R result.must_be_same(klass) R File /usr/local/lib/scons-2.1.0/SCons/Node/FS.py, line 626: R (self.__class__.__name__, self.path, klass.__name__)) R *** [do-build] Error code 2 R Stop in /usr/ports/astro/gpsd. R *** [build] Error code 1 R Stop in /usr/ports/astro/gpsd. Sorry, I can't reproduce that on a clean 10.0-CURRENT amd64. -- Totus tuus, Glebius. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: astro/gpsd fails to build on CURRENT
On Sat, May 11, 2013 at 10:05:22PM +0200, Rainer Hurling wrote: R On 11.05.2013 20:53 (UTC+2), Gleb Smirnoff wrote: RRainer, R R On Tue, May 07, 2013 at 06:19:16PM +0200, Rainer Hurling wrote: R R When I try to build astro/gpsd on 10.0-CURRENT it fails with the R R following messages (devel/scons should be up to date): R R R R /usr/ports/astro/gpsd#make R R === Found saved configuration for gpsd-3.9 R R ... R R R Checking for C header file sys/endian.h... yes R R Checking that xsltproc can make man pages... yes R R Altered configuration variables: R R mtk3301 = False (default True): MTK-3301 support R R nmea2000 = False (default True): NMEA2000/CAN support R R bluez = False (default True): BlueZ support for Bluetooth devices R R libQgpsmm = False (default True): build QT bindings R R chrpath = False (default True): use chrpath to edit library load paths R R mandir = man (default share/man): manual pages directory R R pkgconfig = libdata/pkgconfig (default lib/pkgconfig): pkgconfig file R R directory R R TypeError: Tried to lookup Dir '/usr/local/lib' as a File.: R R File /usr/ports/astro/gpsd/work/gpsd-3.9/SConstruct, line 955: R R parse_flags=gpsdlibs + ncurseslibs + ['-lm']) R R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 258: R R return MethodWrapper.__call__(self, target, source, *args, **kw) R R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 222: R R return self.method(*nargs, **kwargs) R R File /usr/local/lib/scons-2.1.0/SCons/Builder.py, line 631: R R env = env.Override(env_kw) R R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 635: R R if merges: env.MergeFlags(merges) R R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 810: R R args = self.ParseFlags(args) R R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 796: R R do_parse(arg) R R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 670: R R for t in arg: do_parse(t) R R File /usr/local/lib/scons-2.1.0/SCons/Environment.py, line 726: R R dict['LIBS'].append(self.fs.File(arg)) R R File /usr/local/lib/scons-2.1.0/SCons/Node/FS.py, line 1339: R R return self._lookup(name, directory, File, create) R R File /usr/local/lib/scons-2.1.0/SCons/Node/FS.py, line 1318: R R return root._lookup_abs(p, fsclass, create) R R File /usr/local/lib/scons-2.1.0/SCons/Node/FS.py, line 2223: R R result.must_be_same(klass) R R File /usr/local/lib/scons-2.1.0/SCons/Node/FS.py, line 626: R R (self.__class__.__name__, self.path, klass.__name__)) R R *** [do-build] Error code 2 R R Stop in /usr/ports/astro/gpsd. R R *** [build] Error code 1 R R Stop in /usr/ports/astro/gpsd. R R Gleb, R R thanks for answering. R R R Sorry, I can't reproduce that on a clean 10.0-CURRENT amd64. R R For me this happens on three boxes, all 10.0-CURRENT amd64 with more R than 1.500 ports installed (most because of their use for scientific R workplaces, some multimedia ones). R R I think Christoph is right here in suspecting scons interaction with R ncurses (wrong parsing). If I patch /usr/local/bin/ncurses5-config in R the way, Christoph suggested, all went fine. The attached zip contains R two log files, one with the messages of the unpatched build, one with R the patched one. R R My systems are far away from 'clean' systems, because they growed over R the years. On the other hand, even if I try to build astro/gpsd without R /etc/make.conf, it fails. I have no clue, where to look next. May be maintainer of scons has idea. I've added him to Cc. -- Totus tuus, Glebius. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
openssh-portable should depend on perl
Grzegorz, I've noticed that during openssh-portable build it runs a ./fixprogs script. This script is a perl program, so openssh-portable should have perl as build dependency. -- Totus tuus, Glebius. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: CFT: security/openssh-portable 5.8p2
On Sun, Oct 09, 2011 at 06:38:10PM +0200, Grzegorz Blach wrote: G New snapshot is ready to testing: G https://github.com/downloads/Roorback/mgk_ports/openssh-portable-5.8p2-t2.shar G In this version WITH_LPK knob is fixed. G Thanks to Gleb Smirnoff. btw, one more issue with the port is that configure autodetects wtmp/utmp/lastlog stuff, using not only header include files, but also actual logs in /var. So, compiling openssh-portable on a 9.x or 10.x system, that was once upgraded from 8.x or earlier, would lead to incorrect autodetection of logging API. Even if you have run 'make delete-old', since the latter doesn't delete anything from /var. Thus, before compiling port, one needs to: # rm /var/log/wtmp /var/run/utmp /var/log/lastlog -- Totus tuus, Glebius. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: FreeBSD Port: mpd4-4.0b5
On Sat, Nov 11, 2006 at 02:17:47AM +0100, [EMAIL PROTECTED] wrote: l this my problem when i try to make mpd l l === Building for mpd4-4.0b5 l === src l Warning: Object directory not changed from original l /usr/ports/net/mpd/work/mpd4-4.0b5/src l make: don't know how to make mpd4.1. Stop l *** Error code 2 l l Stop in /usr/ports/net/mpd/work/mpd4-4.0b5. l *** Error code 1 l l Stop in /usr/ports/net/mpd. l l there is an error What is your FreeBSD version? -- Totus tuus, Glebius. GLEBIUS-RIPN GLEB-RIPE ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to [EMAIL PROTECTED]
