Re: r248583 Kernel panic: negative refcount 0xfffffe0031b59168
I'm seeing what I believe is related panic, reliably being generated by the Python regression test suite on a newly created FreeBSD 10-CURRENT buildbot. Symptoms first seen in an freebsd.org FTP snapshot dated "Thu May 30 20:01:46 UTC 2013" and also reproducible on a freshly updated r252400 It is additionally reproducible after checking out pure upstream python sources, using the following steps: hg clone http://hg.python.org/cpython cd cpython && configure && make buildbottest An interesting possible correlation is that it seems to drop out during/around "test_socket" Backtrace below: > koobs@10-CURRENT-amd64:~ % sudo kgdb /boot/kernel/kernel /var/crash/vmcore.1 > GNU gdb 6.1.1 [FreeBSD] > Copyright 2004 Free Software Foundation, Inc. > GDB is free software, covered by the GNU General Public License, and you are > welcome to change it and/or distribute copies of it under certain conditions. > Type "show copying" to see the conditions. > There is absolutely no warranty for GDB. Type "show warranty" for details. > This GDB was configured as "amd64-marcel-freebsd"... > > Unread portion of the kernel message buffer: > panic: negative refcount 0xfe0009316de8 > cpuid = 0 > KDB: enter: panic > > Reading symbols from /boot/kernel/zfs.ko.symbols...done. > Loaded symbols for /boot/kernel/zfs.ko.symbols > Reading symbols from /boot/kernel/opensolaris.ko.symbols...done. > Loaded symbols for /boot/kernel/opensolaris.ko.symbols > #0 doadump (textdump=114425856) at pcpu.h:236 > 236 pcpu.h: No such file or directory. > in pcpu.h > (kgdb) bt > #0 doadump (textdump=114425856) at pcpu.h:236 > #1 0x80338a75 in db_fncall (dummy1=, > dummy2=, dummy3=, dummy4= optimized out>) > at /usr/src/sys/ddb/db_command.c:578 > #2 0x8033875d in db_command (cmd_table=) at > /usr/src/sys/ddb/db_command.c:449 > #3 0x803384d4 in db_command_loop () at > /usr/src/sys/ddb/db_command.c:502 > #4 0x8033ae80 in db_trap (type=, code=0) at > /usr/src/sys/ddb/db_main.c:231 > #5 0x808ec1b3 in kdb_trap (type=3, code=0, tf=) > at /usr/src/sys/kern/subr_kdb.c:654 > #6 0x80c4cadb in trap (frame=0xff80978a7880) at > /usr/src/sys/amd64/amd64/trap.c:579 > #7 0x80c35ca2 in calltrap () at > /usr/src/sys/amd64/amd64/exception.S:232 > #8 0x808eb98e in kdb_enter (why=0x80f10487 "panic", > msg=) at cpufunc.h:63 > #9 0x808b5a76 in vpanic (fmt=, ap= optimized out>) at /usr/src/sys/kern/kern_shutdown.c:747 > #10 0x808b5926 in kassert_panic (fmt=) at > /usr/src/sys/kern/kern_shutdown.c:642 > #11 0x80875b4f in closef (fp=, td= optimized out>) at refcount.h:66 > #12 0x80873890 in closefp (fdp=0xfe0009652000, fd= optimized out>, fp=0xfe0009316dc0, td=0xfe0009337490, > holdleaders=) > at /usr/src/sys/kern/kern_descrip.c:1140 > #13 0x80c4d725 in amd64_syscall (td=0xfe0009337490, traced=0) at > subr_syscall.c:134 > #14 0x80c35f8b in Xfast_syscall () at > /usr/src/sys/amd64/amd64/exception.S:391 > #15 0x00080119b00a in ?? () > Previous frame inner to this frame (corrupt stack?) > Current language: auto; currently minimal > (kgdb) -- koobs ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
panic: Lock filedesc structure not share locked
Hi, with head as of r252381 on amd64, I got the following panic after starting tmux and creating a 2nd terminal window inside tmux (ctrl- + c): ---snip--- panic: Lock filedesc structure not share locked @ /space/system/usr_src/sys/kern/kern_descrip.c:3448 cpuid = 2 KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xff839ee566d0 kdb_backtrace() at kdb_backtrace+0x39/frame 0xff839ee56780 vpanic() at vpanic+0x126/frame 0xff839ee567c0 panic() at panic+0x43/frame 0xff839ee56820 _sx_assert() at _sx_assert+0x134/frame 0xff839ee56830 _sx_sunlock() at _sx_sunlock+0x46/frame 0xff839ee56860 kern_proc_filedesc_out() at kern_proc_filedesc_out+0x420/frame 0xff839ee568e0 sysctl_kern_proc_filedesc() at sysctl_kern_proc_filedesc+0x66/frame 0xff839ee56950 sysctl_root() at sysctl_root+0x1bd/frame 0xff839ee569a0 userland_sysctl() at userland_sysctl+0x192/frame 0xff839ee56a40 sys___sysctl() at sys___sysctl+0x74/frame 0xff839ee56af0 amd64_syscall() at amd64_syscall+0x23c/frame 0xff839ee56bf0 Xfast_syscall() at Xfast_syscall+0xfb/frame 0xff839ee56bf0 ---snip--- Full textdump output available at http://www.leidinger.net/test/textdump.tar.2 Bye, Alexander. -- http://www.Leidinger.netAlexander @ Leidinger.net: PGP ID = B0063FE7 http://www.FreeBSD.org netchild @ FreeBSD.org : PGP ID = 72077137 ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
network.subr (r252360) changes break ifconfig_aliasX
Hi I can't figure out how to use rc.conf to configure my interfaces after these recent charges. My use case is that I have interfaces to configure but I don't need to put an IP address on them. I do need to change the MAC address though. What I have been doing is probably wrong, but it worked up until r252360: lagg0: link state changed to DOWN /etc/rc: WARNING: $ifconfig_lagg0_alias0 needs "inet" keyword for an IPv4 address. ifconfig: ether: bad value But, there is no ip address for this interface. The config looks like: cloned_interfaces="lagg0 \ vlan2 vlan3 vlan4 vlan5 vlan6 vlan7 vlan8 vlan9 vlan14 vlan15 vlan19 \ vlan20 vlan21 vlan22 vlan23 vlan24 vlan25 vlan26 vlan27 vlan30 \ vlan33 vlan37 vlan39 vlan40 vlan41 vlan44 vlan45 vlan999 \ " ifconfig_igb0="up" ifconfig_igb1="up" ifconfig_igb2="up" ifconfig_igb3="up" ifconfig_lagg0="up laggproto lacp laggport igb0 laggport igb1 laggport igb2 laggport igb3" ifconfig_lagg0_alias0="ether 00:1e:c9:53:3e:15" # Neotel ifconfig_vlan2="vlandev lagg0 vlan 2" #should probably be create_args ifconfig_vlan2_alias0="inet 41.161.56.4/29" ifconfig_vlan2_alias1="inet 41.154.2.106/29" ifconfig_vlan2_alias2="inet 196.46.25.156/25" ifconfig_vlan2_alias3="inet 41.161.56.2/29 vhid 2 pass XXX advskew 20" ifconfig_vlan2_alias4="inet 41.154.2.105/29 vhid 2 pass XXX advskew 20" ifconfig_vlan2_alias5="inet 196.46.25.155/25 vhid 2 pass XXX advskew 20" etc. I've tried create_args but I still get the following error: ifconfig: ether: bad value and, I cannot use ifconfig_lagg0 to set the MAC address because the rc system just seems to ignore the configured string. I can however 'ifconfig lagg0 ether 00:1e:c9:53:3e:15' from the cammand line, so there's something in the rc scripts that's mangling the argument. Ian -- Ian Freislich ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: r248583 Kernel panic: negative refcount 0xfffffe0031b59168
On Sun, Jun 30, 2013 at 05:21:42PM +1000, Kubilay Kocak wrote: > I'm seeing what I believe is related panic, reliably being generated by > the Python regression test suite on a newly created FreeBSD 10-CURRENT > buildbot. > > Symptoms first seen in an freebsd.org FTP snapshot dated "Thu May 30 > 20:01:46 UTC 2013" and also reproducible on a freshly updated r252400 > > It is additionally reproducible after checking out pure upstream python > sources, using the following steps: > > hg clone http://hg.python.org/cpython > cd cpython && configure && make buildbottest > > An interesting possible correlation is that it seems to drop out > during/around "test_socket" Can you extract the autonomous test case, in C ? pgpCR_7rbQtav.pgp Description: PGP signature
Re: network.subr (r252360) changes break ifconfig_aliasX
"Ian FREISLICH" wrote: > What I have been doing is probably wrong, but it worked up until > r252360: I see from the commit log that it was actually 252015 that broke my router. -- Ian Freislich ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: panic: Lock filedesc structure not share locked
On Sun, Jun 30, 2013 at 09:41:50AM +0200, Alexander Leidinger wrote:
> Hi,
>
> with head as of r252381 on amd64, I got the following panic after
> starting tmux and creating a 2nd terminal window inside tmux
> (ctrl- + c):
> ---snip---
> panic: Lock filedesc structure not share locked @
> /space/system/usr_src/sys/kern/kern_descrip.c:3448
>
> cpuid = 2
> KDB: stack backtrace:
> db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xff839ee566d0
> kdb_backtrace() at kdb_backtrace+0x39/frame 0xff839ee56780
> vpanic() at vpanic+0x126/frame 0xff839ee567c0
> panic() at panic+0x43/frame 0xff839ee56820
> _sx_assert() at _sx_assert+0x134/frame 0xff839ee56830
> _sx_sunlock() at _sx_sunlock+0x46/frame 0xff839ee56860
> kern_proc_filedesc_out() at kern_proc_filedesc_out+0x420/frame
> 0xff839ee568e0
> sysctl_kern_proc_filedesc() at sysctl_kern_proc_filedesc+0x66/frame
> 0xff839ee56950
> sysctl_root() at sysctl_root+0x1bd/frame 0xff839ee569a0
> userland_sysctl() at userland_sysctl+0x192/frame 0xff839ee56a40
> sys___sysctl() at sys___sysctl+0x74/frame 0xff839ee56af0
> amd64_syscall() at amd64_syscall+0x23c/frame 0xff839ee56bf0
> Xfast_syscall() at Xfast_syscall+0xfb/frame 0xff839ee56bf0
> ---snip---
>
Can you try this (only compile-tested):
diff --git a/sys/kern/kern_descrip.c b/sys/kern/kern_descrip.c
index e760fe5..7aa17cd 100644
--- a/sys/kern/kern_descrip.c
+++ b/sys/kern/kern_descrip.c
@@ -3272,6 +3272,8 @@ export_fd_to_sb(void *data, int type, int fd, int fflags,
int refcnt,
if (efbuf->remainder < kif->kf_structsize) {
/* Terminate export. */
efbuf->remainder = 0;
+ if (!locked && efbuf->fdp != NULL)
+ FILEDESC_SLOCK(efbuf->fdp);
return (0);
}
efbuf->remainder -= kif->kf_structsize;
--
Mateusz Guzik
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: r248583 Kernel panic: negative refcount 0xfffffe0031b59168
On Sun, Jun 30, 2013 at 05:21:42PM +1000, Kubilay Kocak wrote:
> I'm seeing what I believe is related panic, reliably being generated by
> the Python regression test suite on a newly created FreeBSD 10-CURRENT
> buildbot.
>
> Symptoms first seen in an freebsd.org FTP snapshot dated "Thu May 30
> 20:01:46 UTC 2013" and also reproducible on a freshly updated r252400
>
> It is additionally reproducible after checking out pure upstream python
> sources, using the following steps:
>
> hg clone http://hg.python.org/cpython
> cd cpython && configure && make buildbottest
>
> An interesting possible correlation is that it seems to drop out
> during/around "test_socket"
>
Turns out the bug is quite funny ;)
Try this:
diff --git a/sys/kern/uipc_usrreq.c b/sys/kern/uipc_usrreq.c
index 5d8e814..7a4db04 100644
--- a/sys/kern/uipc_usrreq.c
+++ b/sys/kern/uipc_usrreq.c
@@ -1764,8 +1764,8 @@ unp_externalize(struct mbuf *control, struct mbuf
**controlp, int flags)
}
for (i = 0; i < newfds; i++, fdp++) {
fde = &fdesc->fd_ofiles[*fdp];
- fde->fde_file = fdep[0]->fde_file;
- filecaps_move(&fdep[0]->fde_caps,
+ fde->fde_file = fdep[i]->fde_file;
+ filecaps_move(&fdep[i]->fde_caps,
&fde->fde_caps);
if ((flags & MSG_CMSG_CLOEXEC) != 0)
fde->fde_flags |= UF_EXCLOSE;
--
Mateusz Guzik
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: another -Wunsequenced topic
There are more.
Take the first hunk with caution.
Index: contrib/gdb/gdb/dwarf2-frame.c
===
--- contrib/gdb/gdb/dwarf2-frame.c (revision 252384)
+++ contrib/gdb/gdb/dwarf2-frame.c (working copy)
@@ -1361,7 +1361,7 @@
else if (*augmentation == 'P')
{
/* Skip. */
- buf += size_of_encoded_value (*buf++);
+ buf += size_of_encoded_value (*buf) + 1;
augmentation++;
}
Index: usr.sbin/moused/moused.c
===
--- usr.sbin/moused/moused.c(revision 252384)
+++ usr.sbin/moused/moused.c(working copy)
@@ -2455,7 +2455,7 @@
return (FALSE);
lbutton = atoi(s);
- arg = skipspace(++arg);
+ arg = skipspace(arg + 1);
s = arg;
while (isdigit(*arg))
++arg;
Index: lib/libstand/nfs.c
===
--- lib/libstand/nfs.c (revision 252384)
+++ lib/libstand/nfs.c (working copy)
@@ -1465,8 +1465,9 @@
d->d_name[d->d_namlen] = '\0';
pos = roundup(d->d_namlen, sizeof(uint32_t)) / sizeof(uint32_t);
- fp->off = cookie = ((uint64_t)ntohl(rent->nameplus[pos++]) << 32) |
- ntohl(rent->nameplus[pos++]);
+ fp->off = cookie = ((uint64_t)ntohl(rent->nameplus[pos]) << 32) |
+ ntohl(rent->nameplus[pos + 1]);
+ pos += 2;
buf = (u_char *)&rent->nameplus[pos];
return (0);
}
Index: contrib/sendmail/src/recipient.c
===
--- contrib/sendmail/src/recipient.c(revision 252384)
+++ contrib/sendmail/src/recipient.c(working copy)
@@ -1834,7 +1834,7 @@
/* #@# introduces a comment anywhere */
/* for Japanese character sets */
- for (p = buf; (p = strchr(++p, '#')) != NULL; )
+ for (p = buf; (p = strchr(p + 1, '#')) != NULL; )
{
if (p[1] == '@' && p[2] == '#' &&
isascii(p[-1]) && isspace(p[-1]) &&
Index: usr.sbin/pkg_install/create/perform.c
===
--- usr.sbin/pkg_install/create/perform.c (revision 252384)
+++ usr.sbin/pkg_install/create/perform.c (working copy)
@@ -149,7 +149,7 @@
/* Count number of dependencies */
for (cp = Pkgdeps; cp != NULL && *cp != '\0';
- cp = strpbrk(++cp, " \t\n")) {
+ cp = strpbrk(cp + 1, " \t\n")) {
ndeps++;
}
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
"make -j4 buildwolrd" success when here are problems (files was not found in depend / all targets)
Hello, Freebsd-current. I'm trying to build world with patch from Rui Paulo, which imports hostapd/wpa_supplicant 2.0 I've used "worng" patch, which didn't add some files, but "make -j4 buildworld" succeed! Here are quotes from "buildworld" log: ===> usr.sbin/wpa/wpa_supplicant (depend) make: make: don't know how to make driver_common.c. Stop make: stopped in /data/src/usr.sbin/wpa/wpa_supplicant ===> usr.sbin/wpa/wpa_cli (depend) make: make: don't know how to make edit.c. Stop make: stopped in /data/src/usr.sbin/wpa/wpa_cli ===> usr.sbin/wpa/wpa_passphrase (depend) --- .depend --- rm -f .depend CC='/usr/bin/cc --sysroot=/data/obj.nano/gateway.v2/data/src/tmp -B/data/obj.nano/gateway.v2/data/src/tmp/usr/bin' mkdep -f .depend -a -I/data/src/usr.sbin/wpa/wpa_passphrase -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//hostapd -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/common -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/crypto -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/drivers -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/l2_packet -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/utils -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/wps -DCONFIG_CTRL_IFACE -DCONFIG_CTRL_IFACE_UNIX -DINTERNAL_SHA1 -DINTERNAL_MD5 -I/data/src/usr.sbin/wpa/wpa_passphrase -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//hostapd -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../con trib/wpa//src -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/common -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/crypto -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/drivers -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/l2_packet -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/utils -I/data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/wps -DCONFIG_CTRL_IFACE -DCONFIG_CTRL_IFACE_UNIX -std=gnu99 /data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/utils/common.c /data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/crypto/md5-internal.c /data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/crypto/md5.c /data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/utils/os_unix.c /data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/crypto/sha1-internal.c /data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/crypto /sha1-pbkdf2.c /data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//src/crypto/sha1.c /data/src/usr.sbin/wpa/wpa_passphrase/../../../contrib/wpa//wpa_supplicant/wpa_passphrase.c echo wpa_passphrase: /data/obj.nano/gateway.v2/data/src/tmp/usr/lib/libc.a >> .depend ===> usr.sbin/wpa/hostapd (depend) make: make: don't know how to make driver_common.c. Stop make: stopped in /data/src/usr.sbin/wpa/hostapd ===> usr.sbin/wpa/hostapd_cli (depend) make: make: don't know how to make edit.c. Stop make: stopped in /data/src/usr.sbin/wpa/hostapd_cli And, later ===> usr.sbin/wpa/wpa_supplicant (all) make: make: don't know how to make driver_common.c. Stop make: stopped in /data/src/usr.sbin/wpa/wpa_supplicant ===> usr.sbin/wpa/wpa_cli (all) make: make: don't know how to make edit.c. Stop make: stopped in /data/src/usr.sbin/wpa/wpa_cli But -- >>> World build completed on Sun Jun 30 21:13:10 MSK 2013 -- Later, "installworld" failed due to absence of "wpa_ssupplicant" program! I'm using r252420 as base. -- // Black Lion AKA Lev Serebryakov ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: network.subr (r252360) changes break ifconfig_aliasX
"Ian FREISLICH" wrote in : ia> Hi ia> ia> I can't figure out how to use rc.conf to configure my interfaces ia> after these recent charges. My use case is that I have interfaces ia> to configure but I don't need to put an IP address on them. I do ia> need to change the MAC address though. Please try r252426. -- Hiroki pgpPp9RCG9oqY.pgp Description: PGP signature
Zoneminder build failure
Running -CURRENT, I'm getting this: c++ -O2 -pipe -march=core2 -fno-strict-aliasing -L/usr/local/lib -L/usr/local/lib/mysql -lpthread -o zmc zmc.o zm_box.o zm_buffer.o zm_camera.o zm_comms.o zm_config.o zm_coord.o zm.o zm_db.o zm_logger.o zm_event.o zm_exception.o zm_file_camera.o zm_ffmpeg_camera.o zm_image.o zm_jpeg.o zm_local_camera.o zm_monitor.o zm_ffmpeg.o zm_mpeg.o zm_poly.o zm_regexp.o zm_remote_camera.o zm_remote_camera_http.o zm_remote_camera_rtsp.o zm_rtp.o zm_rtp_ctrl.o zm_rtp_data.o zm_rtp_source.o zm_rtsp.o zm_sdp.o zm_signal.o zm_stream.o zm_thread.o zm_time.o zm_timer.o zm_user.o zm_utils.o zm_zone.o -lz -lbz2 -lswscale -lavdevice -lavformat -lavcodec -lavutil -lz -lpcre -lcrypto -lc -lpthread -ljpeg -lmysqlclient zm_timer.o: In function `Timer::TimerThread::run()': zm_timer.cpp:(.text+0x428): undefined reference to `ThreadData::setValue(bool)' c++: error: linker command failed with exit code 1 (use -v to see invocation) *** Error code 1 Is there a fix for it? -- Cheers, Saul ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: panic: Lock filedesc structure not share locked
On Sun, Jun 30, 2013 at 10:27:57AM +0200, Mateusz Guzik wrote:
> On Sun, Jun 30, 2013 at 09:41:50AM +0200, Alexander Leidinger wrote:
> > Hi,
> >
> > with head as of r252381 on amd64, I got the following panic after
> > starting tmux and creating a 2nd terminal window inside tmux
> > (ctrl- + c):
> > ---snip---
> > panic: Lock filedesc structure not share locked @
> > /space/system/usr_src/sys/kern/kern_descrip.c:3448
> >
> > cpuid = 2
> > KDB: stack backtrace:
> > db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame
> > 0xff839ee566d0
> > kdb_backtrace() at kdb_backtrace+0x39/frame 0xff839ee56780
> > vpanic() at vpanic+0x126/frame 0xff839ee567c0
> > panic() at panic+0x43/frame 0xff839ee56820
> > _sx_assert() at _sx_assert+0x134/frame 0xff839ee56830
> > _sx_sunlock() at _sx_sunlock+0x46/frame 0xff839ee56860
> > kern_proc_filedesc_out() at kern_proc_filedesc_out+0x420/frame
> > 0xff839ee568e0
> > sysctl_kern_proc_filedesc() at sysctl_kern_proc_filedesc+0x66/frame
> > 0xff839ee56950
> > sysctl_root() at sysctl_root+0x1bd/frame 0xff839ee569a0
> > userland_sysctl() at userland_sysctl+0x192/frame 0xff839ee56a40
> > sys___sysctl() at sys___sysctl+0x74/frame 0xff839ee56af0
> > amd64_syscall() at amd64_syscall+0x23c/frame 0xff839ee56bf0
> > Xfast_syscall() at Xfast_syscall+0xfb/frame 0xff839ee56bf0
> > ---snip---
> >
>
> Can you try this (only compile-tested):
> diff --git a/sys/kern/kern_descrip.c b/sys/kern/kern_descrip.c
> index e760fe5..7aa17cd 100644
> --- a/sys/kern/kern_descrip.c
> +++ b/sys/kern/kern_descrip.c
> @@ -3272,6 +3272,8 @@ export_fd_to_sb(void *data, int type, int fd, int
> fflags, int refcnt,
> if (efbuf->remainder < kif->kf_structsize) {
> /* Terminate export. */
> efbuf->remainder = 0;
> + if (!locked && efbuf->fdp != NULL)
> + FILEDESC_SLOCK(efbuf->fdp);
> return (0);
> }
> efbuf->remainder -= kif->kf_structsize;
>
Mateusz, thank you for spotting this lock leakage. Regardless if this
is the root cause of the reported panic (it looks like it is), this
fix should be definetly committed. Will you do this?
--
Mikolaj Golub
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: panic: Lock filedesc structure not share locked
On Mon, Jul 01, 2013 at 12:23:35AM +0300, Mikolaj Golub wrote:
> On Sun, Jun 30, 2013 at 10:27:57AM +0200, Mateusz Guzik wrote:
> > On Sun, Jun 30, 2013 at 09:41:50AM +0200, Alexander Leidinger wrote:
> > > Hi,
> > >
> > > with head as of r252381 on amd64, I got the following panic after
> > > starting tmux and creating a 2nd terminal window inside tmux
> > > (ctrl- + c):
> > > ---snip---
> > > panic: Lock filedesc structure not share locked @
> > > /space/system/usr_src/sys/kern/kern_descrip.c:3448
> > >
> > > cpuid = 2
> > > KDB: stack backtrace:
> > > db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame
> > > 0xff839ee566d0
> > > kdb_backtrace() at kdb_backtrace+0x39/frame 0xff839ee56780
> > > vpanic() at vpanic+0x126/frame 0xff839ee567c0
> > > panic() at panic+0x43/frame 0xff839ee56820
> > > _sx_assert() at _sx_assert+0x134/frame 0xff839ee56830
> > > _sx_sunlock() at _sx_sunlock+0x46/frame 0xff839ee56860
> > > kern_proc_filedesc_out() at kern_proc_filedesc_out+0x420/frame
> > > 0xff839ee568e0
> > > sysctl_kern_proc_filedesc() at sysctl_kern_proc_filedesc+0x66/frame
> > > 0xff839ee56950
> > > sysctl_root() at sysctl_root+0x1bd/frame 0xff839ee569a0
> > > userland_sysctl() at userland_sysctl+0x192/frame 0xff839ee56a40
> > > sys___sysctl() at sys___sysctl+0x74/frame 0xff839ee56af0
> > > amd64_syscall() at amd64_syscall+0x23c/frame 0xff839ee56bf0
> > > Xfast_syscall() at Xfast_syscall+0xfb/frame 0xff839ee56bf0
> > > ---snip---
> > >
> >
> > Can you try this (only compile-tested):
> > diff --git a/sys/kern/kern_descrip.c b/sys/kern/kern_descrip.c
> > index e760fe5..7aa17cd 100644
> > --- a/sys/kern/kern_descrip.c
> > +++ b/sys/kern/kern_descrip.c
> > @@ -3272,6 +3272,8 @@ export_fd_to_sb(void *data, int type, int fd, int
> > fflags, int refcnt,
> > if (efbuf->remainder < kif->kf_structsize) {
> > /* Terminate export. */
> > efbuf->remainder = 0;
> > + if (!locked && efbuf->fdp != NULL)
> > + FILEDESC_SLOCK(efbuf->fdp);
> > return (0);
> > }
> > efbuf->remainder -= kif->kf_structsize;
> >
>
> Mateusz, thank you for spotting this lock leakage. Regardless if this
> is the root cause of the reported panic (it looks like it is), this
> fix should be definetly committed. Will you do this?
>
I think it will be better if you do this and then MFC all commits.
--
Mateusz Guzik
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Proposal: multi-instance and self-contained rc.d script
Hi,
I am working on rc.d script improvements in terms of the following
two points. A prototype is attached. This still includes some rough
edges but should be enough to understand the concepts. I would like
your comments about them. (This is posted to -current@ and -rc@, but
please reply to freebsd-rc@ only)
1. Multi-instance support
As you know, an rc.d script invokes ${name}_program with
configuration variables in /etc/rc.conf. It works, but it is
difficult to run multiple instances of a program. A typical example
is named---a sysadmin often wants to run two instances of named as a
caching server and a content DNS server by using different
named.conf. In the current framework, two rc.d scripts are needed
by copying /etc/rc.d/named to /etc/rc.d/named_cache and rewriting
name= and rcvar=.
The attached patch allows the following:
named_enable="YES"
named_instances="cache1 cache2"
named_conf="/etc/namedb/named.conf"
named_cache1_enable="YES"
named_cache1_conf="/etc/namedb/named_cache1.conf"
named_cache2_enable="NO"
named_cache2_conf="/etc/namedb/named_cache2.conf"
A new variable "{name}_instances" defines instances. In this
example, it is named_instances="cache1 cache2". All of the default
values of $named_{instname}_foo are automatically set to the same as
$named_foo.
In the implementation, load_rc_config() reads variables for all
instances and run_rc_command() runs each instance in order. When
doing "rc.d/foo stop", run_rc_command() stops the instances in
reverse order.
In the patch, killing the processes without pid file does not work
well yet. This can be improved.
2. Self-contained rc.d script
rc.d scripts depend on /etc/default/rc.conf for the default
configuration and rc.conf(5) manual page describes the knobs.
However, it is difficult to understand which variable is related to
which script. In addition, /etc/defaults/rc.conf is often out of
sync with the rc.d scripts. So, my proposal is as follows:
a) Define rc.conf variables and the default values in the rc.d
script which uses them. "rc.d/foo rcvar" shows the variables
and the default values.
b) Make rc.d/foo always have rc.d/foo(8) manual page.
The attached patch includes an example of rc.d/routed. The primary
difference is declaration part of rc.conf variables:
set_rcvar enable NO
set_rcvar program /sbin/routed
set_rcvar flags-q
These sets the default value of $routed_{enable,program,flags} at
load_rc_config(). The reason why a simple ": ${routed_enable="NO"}"
does not work is that it does not work with multi-instance support.
This is backward-compatible with the current /etc/defaults/rc.conf.
load_rc_config() sets these values first, and then reads
/etc/defaults/rc.conf and /etc/rc.conf.d/$name.
"rc.d/route rcvar" displays the current configuration and available
variables briefly like the following:
# routed: network RIP and router discovery routing daemon
#
routed_enable="NO" # (default: "NO")
routed_program="/sbin/routed" # (default: "/sbin/routed")
routed_flags="-q" # (default: "-q")
When multi-instance is enabled in rc.conf like this:
routed_enable="YES"
routed_instances="hoge fuga"
routed_hoge_desc="hogehoge"
routed_fuga_enable="NO"
routed_fuga_flags=""
The results of rcvar will be the following:
# routed: network RIP and router discovery routing daemon
#
routed_enable="YES" # (default: "NO")
routed_program="/sbin/routed" # (default: "/sbin/routed")
routed_flags="-q" # (default: "-q")
# routed_hoge: network RIP and router discovery routing daemon: hogehoge
#
routed_hoge_enable="YES" # (default: "NO")
routed_hoge_program="/sbin/routed" # (default: "/sbin/routed")
routed_hoge_flags="-q" # (default: "-q")
# routed_fuga: network RIP and router discovery routing daemon
#
routed_fuga_enable="NO" # (default: "NO")
routed_fuga_program="/sbin/routed" # (default: "/sbin/routed")
routed_fuga_flags=""# (default: "-q")
We can remove or comment out all of lines in /etc/defaults/rc.conf,
and mismatch between /etc/defaults/rc.conf and scripts does not
occur. Running "rc.d/foo rcvar" can be used to generate
/etc/defaults/rc.conf if needed.
That's all. Both changes are fully backward compatible and I believe
they improve flexibility and manageability of rc.d scripts.
An example of rc.d/routed(8) manual page is also attached. If these
changes are acceptable, I would like to split the current (lengthy)
rc.conf(5) manual page into rc.d/foo(8).
-- Hiroki
RC.D/ROUTED(8) FreeBSD System Manager's Manual RC.D/ROUTED(8)
NAME
routed -- rc.d script for routed(8) daemon
SYNOPSIS
routed [fast|force|one]start|stop|restart|rcvar|status|poll
DESCRIPTION
routed is an rc.d(8) script to control routed(8) daemon.
COMMAND LINE ARGUMENTS
Re: Proposal: multi-instance and self-contained rc.d script
Hello, Hiroki. You wrote 1 июля 2013 г., 1:29:53: HS> The attached patch allows the following: HS> named_enable="YES" HS> named_instances="cache1 cache2" HS> named_conf="/etc/namedb/named.conf" HS> named_cache1_enable="YES" HS> named_cache1_conf="/etc/namedb/named_cache1.conf" HS> named_cache2_enable="NO" HS> named_cache2_conf="/etc/namedb/named_cache2.conf" If it will work with ports rc scripts it will be great! For example, nsd DNS server doesn't support views and often need to be run in multiple instances. -- // Black Lion AKA Lev Serebryakov ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: panic: Lock filedesc structure not share locked
On Sun, Jun 30, 2013 at 11:29:59PM +0200, Mateusz Guzik wrote: > I think it will be better if you do this and then MFC all commits. Committed as r252436. Thanks. -- Mikolaj Golub ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: r248583 Kernel panic: negative refcount 0xfffffe0031b59168
On (30/06/2013 13:18), Mateusz Guzik wrote:
> On Sun, Jun 30, 2013 at 05:21:42PM +1000, Kubilay Kocak wrote:
> > I'm seeing what I believe is related panic, reliably being generated by
> > the Python regression test suite on a newly created FreeBSD 10-CURRENT
> > buildbot.
> >
> > Symptoms first seen in an freebsd.org FTP snapshot dated "Thu May 30
> > 20:01:46 UTC 2013" and also reproducible on a freshly updated r252400
> >
> > It is additionally reproducible after checking out pure upstream python
> > sources, using the following steps:
> >
> > hg clone http://hg.python.org/cpython
> > cd cpython && configure && make buildbottest
> >
> > An interesting possible correlation is that it seems to drop out
> > during/around "test_socket"
> >
>
> Turns out the bug is quite funny ;)
Patch fixes chrome for me.
Thanks!
>
> Try this:
> diff --git a/sys/kern/uipc_usrreq.c b/sys/kern/uipc_usrreq.c
> index 5d8e814..7a4db04 100644
> --- a/sys/kern/uipc_usrreq.c
> +++ b/sys/kern/uipc_usrreq.c
> @@ -1764,8 +1764,8 @@ unp_externalize(struct mbuf *control, struct mbuf
> **controlp, int flags)
> }
> for (i = 0; i < newfds; i++, fdp++) {
> fde = &fdesc->fd_ofiles[*fdp];
> - fde->fde_file = fdep[0]->fde_file;
> - filecaps_move(&fdep[0]->fde_caps,
> + fde->fde_file = fdep[i]->fde_file;
> + filecaps_move(&fdep[i]->fde_caps,
> &fde->fde_caps);
> if ((flags & MSG_CMSG_CLOEXEC) != 0)
> fde->fde_flags |= UF_EXCLOSE;
> --
> Mateusz Guzik
> ___
> freebsd-current@freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: network.subr (r252360) changes break ifconfig_aliasX
Hiroki Sato wrote: > ia> I can't figure out how to use rc.conf to configure my interfaces > ia> after these recent charges. My use case is that I have interfaces > ia> to configure but I don't need to put an IP address on them. I do > ia> need to change the MAC address though. > > Please try r252426. Thanks. That fixes it. BTW, nice new features in network.subr. Ian -- Ian Freislich ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
