i386 -current make release available for testing
Hi, I have built a i386 specific install for -current. It is available as: ftp://releng4.freebsd.org/pub/FreeBSD/snapshots/i386/5.0-20030228-CURRENT Floppies are in the typical place. It was built using the patch below passed in as LOCAL_PATCH to the make release process. I'll leave it there for a few days for folks to test against. Enjoy! -John Index: sys/i386/conf/GENERIC === RCS file: /home/ncvs/src/sys/i386/conf/GENERIC,v retrieving revision 1.376 diff -u -r1.376 GENERIC --- sys/i386/conf/GENERIC 13 Feb 2003 22:24:43 - 1.376 +++ sys/i386/conf/GENERIC 28 Feb 2003 03:11:52 - @@ -19,9 +19,10 @@ # $FreeBSD: src/sys/i386/conf/GENERIC,v 1.376 2003/02/13 22:24:43 obrien Exp $ machinei386 -cpuI486_CPU -cpuI586_CPU -cpuI686_CPU +cpuI386_CPU +#cpu I486_CPU +#cpu I586_CPU +#cpu I686_CPU ident GENERIC maxusers 0 @@ -38,9 +39,9 @@ optionsUFS_ACL #Support for access control lists optionsUFS_DIRHASH #Improve performance on big directories optionsMD_ROOT #MD is a potential root device -optionsNFSCLIENT #Network Filesystem Client -optionsNFSSERVER #Network Filesystem Server -optionsNFS_ROOT#NFS usable as root device, requires NFSCLIENT +#options NFSCLIENT #Network Filesystem Client +#options NFSSERVER #Network Filesystem Server +#options NFS_ROOT#NFS usable as root device, requires NFSCLIENT optionsMSDOSFS #MSDOS Filesystem optionsCD9660 #ISO 9660 Filesystem optionsPROCFS #Process filesystem (requires PSEUDOFS) To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: config files and includes.
- Julian Elischer's Original Message -
I have just gone through the process of upgrading or installing several
hundred machines, and Thst includes altering or editing many config
files in /etc. I like the way that rc.conf
is handled, in that defaults/rc.comf can be updated and only the
local changes live in r.conf. I wish that more files had this
capability.
This is not exactly what you are asking for, but this is from
a petty much a been-there/done-that many years ago. Typing in
the logic from memory:
rcfiles=/etc/inetd.conf /etc/syslog.conf /etc/newsyslog.conf
for rcf in $rcfiles; do
if [ -f ${rcf}.local ]; then
if [ ! -f ${rcf}.base ]; then
if diff ${rcf} ${rcf}.base /dev/null; then
cp -p ${rcf} ${rcf}.base
fi
fi
cat ${rcf}.base ${rcf}.local ${rcf}
fi
done
I think you can get the idea.
-John
For example syslogd.conf or newsyslog.conf are updated between releases
but they are also prime candidates for local additions.
What would be really cool is if more config files could
do 'includes' so that you could have a syslogd.local.conf
wher eall your local entries could be. In addition you could make it
look in /usr/local/etc/syslogd.conf for loging requirments for
packages.
To do this for every config file would be a lot of work. I do wonder
however whether there couldn't be some config-file reader library
routine that could be used to pre-pass files and do inclusions..
if the interface was very similar to what is usually used
(people tend to either use open/read/close or
fopen/fscanf (etc).
equivalent calls could be made that use a stream of data that is
pre-processed to do inclusions etc. That would making retrofitting
relatively easy. Programs that use yacc/lex ar emore difficult,
but I haven't looked into it..
anyhow, that was just a thought.
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message
--
--
As said by Napolean Bonaparte:
Never ascribe to malice, that which is adequately explained by incompetence
After being embraced by MS:
When accused of malice, always hide behind incompetence.
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message
UFS2 regression tests?
Hi Folks, I've just put together a 1.7TB filesystem and was looking for some regression tests to run against it. Looking through the mailing lists doesn't turn up anything, nor does a websearch (at least for the keywords I tried). So, does anyone have any comments/ideas on a good way to test the new system? Thanks, John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: UFS2 regression tests?
- Brad Knowles's Original Message - Yeah, especially if it's UFS2, you're doing softupdates with background fsck, and you've used vinum to build the large logical volume. ;-) Actually, I didn't say how the volume was put together :-) It is currently a Raid50 consisting of 3 Raid5 sets. It is hanging off of an Adaptec 5400S controller. I realize I can probably break the system doing abnormal things to it, but I'm more interested (at least for now) in regression style testing. -John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
src/sbin/mount/mount.c argv[0] cleanup
Hi,
I've built and installed a complete system with the patch
below. Amoung other things, this allows the individual mount_$vfstype
executables to not have to differentiate how they were
invoked. They can then have their XXX sections removed along
with a small amount of code (and this will help our poor
boot floppies :-)
Does anyone know any reason why this should not be done?
Thanks!
John
Index: mount.c
===
RCS file: /mnt/cvs/src/sbin/mount/mount.c,v
retrieving revision 1.50
diff -c -r1.50 mount.c
*** mount.c 14 Oct 2002 19:40:00 - 1.50
--- mount.c 26 Jan 2003 20:06:36 -
***
*** 469,474
--- 469,475
for (edir = edirs; *edir; edir++) {
(void)snprintf(execname,
sizeof(execname), %s/mount_%s, *edir, vfstype);
+ argv[0] = execname;
execv(execname, (char * const *)argv);
}
if (errno == ENOENT) {
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message
cron/crontab openlog() cleanup
Hi,
While debugging a problem recently I had all.log enabled
in syslog.conf. One of the things I noted was the messages
coming from cron where the ident string was the fully qualified
pathname:
Dec 29 21:00:00 xxx4 /usr/sbin/cron[7409]: (root) CMD (newsyslog)
Dec 29 21:00:00 xxx4 /usr/sbin/cron[7410]: (root) CMD (/usr/libexec/atrun)
Any comments on one of the two patch sets below? Either crontab.c
and cron.c need a patch, or the common routine they call.
cvs server: Diffing .
cvs server: Diffing cron
Index: cron/cron.c
===
RCS file: /home/ncvs/src/usr.sbin/cron/cron/cron.c,v
retrieving revision 1.14
diff -u -r1.14 cron.c
--- cron/cron.c 9 Mar 2001 03:14:09 - 1.14
+++ cron/cron.c 30 Dec 2002 03:24:23 -
@@ -70,7 +70,10 @@
{
cron_db database;
- ProgramName = argv[0];
+ if (ProgramName = strrchr(argv[0],'/')
+ ++ProgramName;
+ else
+ ProgramName = argv[0];
#if defined(BSD)
setlinebuf(stdout);
cvs server: Diffing crontab
Index: crontab/crontab.c
===
RCS file: /home/ncvs/src/usr.sbin/cron/crontab/crontab.c,v
retrieving revision 1.19
diff -u -r1.19 crontab.c
--- crontab/crontab.c 1 Oct 2002 22:59:11 - 1.19
+++ crontab/crontab.c 30 Dec 2002 03:24:23 -
@@ -91,7 +91,10 @@
int exitstatus;
Pid = getpid();
- ProgramName = argv[0];
+ if (ProgramName = strrchr(argv[0],'/')
+ ++ProgramName;
+ else
+ ProgramName = argv[0];
#if defined(POSIX)
setlocale(LC_ALL, );
Or, the single patch to the common openlog interface:
Index: lib/misc.c
===
RCS file: /home/ncvs/src/usr.sbin/cron/lib/misc.c,v
retrieving revision 1.11
diff -u -r1.11 misc.c
--- lib/misc.c 4 Aug 2002 04:32:27 - 1.11
+++ lib/misc.c 30 Dec 2002 03:24:23 -
@@ -508,14 +508,12 @@
#if defined(SYSLOG)
if (!syslog_open) {
- /* we don't use LOG_PID since the pid passed to us by
-* our client may not be our own. therefore we want to
-* print the pid ourselves.
-*/
+ char *Name = strrchr(ProgramName,'/');
+ Name = Name ? Name+1 : ProgramName;
# ifdef LOG_DAEMON
- openlog(ProgramName, LOG_PID, LOG_CRON);
+ openlog(Name, LOG_PID, LOG_CRON);
# else
- openlog(ProgramName, LOG_PID);
+ openlog(Name, LOG_PID);
# endif
syslog_open = TRUE; /* assume openlog success */
}
I removed the comment since it's out-of-date. Maybe I should
leave it (to reduce diffs for historical purposes). I tend to
prefer this single patch.
It's amazing how much this code hasn't changed... I haven't
looked at it since '88/'89 when I ported it to a ps/2 Mod80
running ISC Unix... memories... :-)
Thanks,
John
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message
Re: Run two copies of named from rc.conf?
- Terry Lambert's Original Message - Brad Knowles wrote: Sorry, I wasn't think of transient networks. Indeed, that does make things a lot uglier. I'll have to think some more about all the various implications, however. One of the draft RFC's in the FTP directory I referenced is a Best Current Practices document. This an interesting thread, but it seems to be getting a bit off target. I need to kick off 2 name servers. The first is authoritive for the domain as seen externally and the 2nd which is authoritive for the internal network. The internal forwards to the external when appropriate. These networks are not transient. -John -- Terry To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
ccd disklabel (what should work?)
In trying to setup a ccd on a -current system I
seem to have run into a few issues:
/etc/ccd.conf:
# ccd ileave flags component devices
ccd0 64 none/dev/ad1s1a /dev/ad2s1a
# ccdconfig -g
ccd064 0 /dev/ad1s1a /dev/ad2s1a
# ls -al /dev/ccd*
crw-r- 1 root operator 74, 2 Nov 17 22:09 /dev/ccd0c
So far, so good...
# disklabel -rw ccd0c auto
disklabel: auto: unknown disk type
# disklabel -rw ccd0 auto
disklabel: cannot open /dev/ccd0: No such file or directory
disklabel: auto: unknown disk type
#disklabel -e ccd0c
disklabel: ioctl DIOCWDINFO: no disk label on disk;
add -r to install initial label
re-edit the label? [y]:
Ok, let's just newfs the 'c' partition
# newfs -U /dev/ccd0c
/dev/ccd0c: 76345.1MB (156354688 sectors) block size 16384, fragment size 2048
using 416 cylinder groups of 183.69MB, 11756 blks, 23552 inodes.
with soft updates
super-block backups (for fsck -b #) at:
...
154614944, 154991136, 155367328, 155743520, 156119712
newfs: ioctl (WDINFO): /dev/ccd0c: can't rewrite disk label: No such process
Should we be able to disklabel 'auto' a ccd?
Should we be able to create multiple partitions? ('a','c','d', etc)
If not, the ccdconfig man page and handbook probably need
a bit of updating. They seem to indicate that the above should work.
Comments?
Thanks,
John
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message
Run two copies of named from rc.conf?
It would be nice if rc.conf could start a 2nd copy
of named (split dns). Comments on the following simplistic
patch?
-John
cvs diff: Diffing src/etc
Index: src/etc/rc.network
===
RCS file: /home/ncvs/src/etc/rc.network,v
retrieving revision 1.138
diff -u -r1.138 rc.network
--- src/etc/rc.network 2 Nov 2002 08:21:26 - 1.138
+++ src/etc/rc.network 18 Nov 2002 04:22:00 -
@@ -24,7 +24,7 @@
# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
# SUCH DAMAGE.
#
-# $FreeBSD$
+# $FreeBSD: src/etc/rc.network,v 1.138 2002/11/02 08:21:26 ume Exp $
# From: @(#)netstart 5.9 (Berkeley) 3/30/91
#
@@ -561,6 +561,11 @@
case ${named_enable} in
[Yy][Ee][Ss])
echo -n ' named'; ${named_program:-named} ${named_flags}
+ ;;
+ esac
+ case ${named_enable2} in
+ [Yy][Ee][Ss])
+ echo -n ' named'; ${named_program2:-named}
${named_flags2}
;;
esac
cvs diff: Diffing src/etc/defaults
Index: src/etc/defaults/rc.conf
===
RCS file: /home/ncvs/src/etc/defaults/rc.conf,v
retrieving revision 1.163
diff -u -r1.163 rc.conf
--- src/etc/defaults/rc.conf15 Nov 2002 08:26:36 - 1.163
+++ src/etc/defaults/rc.conf18 Nov 2002 04:25:01 -
@@ -13,7 +13,7 @@
#
# All arguments must be in double or single quotes.
#
-# $FreeBSD$
+# $FreeBSD: src/etc/defaults/rc.conf,v 1.163 2002/11/15 08:26:36 ru Exp $
##
### Important initial Boot-time options
@@ -140,6 +140,10 @@
# components of named. See /etc/rc.d/named.
named_symlink_enable=YES # Symlink /var/run/named.pid and /var/run/ndc
# to their chrooted counterparts.
+
+named_enable2=NO # Run a 2nd copy of named?
+named_program2=/usr/sbin/named
+named_flags2=-u bind -g bind
#
# kerberos. Do not run the admin daemons on slave servers
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message
Re: Kernel not booting....Immediate crash
- Nate Lawson's Original Message - On Thu, 7 Nov 2002, David Rhodu wrote: Sidcarter wrote: Hi Folks, I just did a cvsup and installed a kernel. I have been trying this from the past few days with the same error. I am copying this by hand, since it crashes immediately after loading the modules. The error message is here Booting [/boot/kernel/kernel]... /boot/kernel/acpi.ko text=0x30dbc data=0x1a58+0xb48 syms=[0x4+0x5770+0x4+0x73b9/]\ Try disabling acpi: echo hint.acpi.0.disable=1 /boot/device.hints I appear to be having the exact same problem with a new dual Xeon system. It boots 4.7 correctly. The header from a 4.7 dmesg shows: Copyright (c) 1992-2002 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved. FreeBSD RELENG_4_7_0_RELEASE #0: Fri Nov 8 12:08:15 GMT 2002 root@:/usr/src/sys/compile/GENERIC Timecounter i8254 frequency 1193182 Hz CPU: Pentium 4 (2791.01-MHz 686-class CPU) Origin = GenuineIntel Id = 0xf27 Stepping = 7 Features=0xbfebfbffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,b28,ACC,b31 real memory = 1073676288 (1048512K bytes) avail memory = 1039872000 (1015500K bytes) Preloaded elf kernel kernel at 0xc050e000. Pentium Pro MTRR support enabled md0: Malloc disk Using $PIR table, 19 entries at 0xc00f30a0 Trying to boot from a 5.0 iso fails immediately after loading the acpi.ko module. It panics with no traceback available. Breaking into the boot sequence and issuing: set hint.acpi.0.disable=1 has no effect. Any other ideas on how to boot and get 5.0 installed onto this hardware? Thanks, John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
5.0-20021101-CURRENT snap iso
For all you weekend install warriors out there... There is a new 5.0 snap iso available via anonymous ftp at: usw2.FreeBSD.Org /pub/FreeBSD/snapshots/i386/5.0-20021101-CURRENT and the iso: /pub/FreeBSD/snapshots/i386/5.0-20021101-CURRENT.iso I have verified that this iso boots and can reference itself for a CD/DVD install. The only (non-critical) problem I've seen so far is refresh problems within sysinstall. Many thanks to those who downloaded and tested the previous (not exactly working) iso and provided feedback. -John ps: This iso does not contain the ports tree. Cvsup it at your leisure. Pkg_add is your friend. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Poor 5.0/nfs performance
Hi, I have a 5.0 system from 10/27. In an attempt to improve performance I commented out the INVARIANTS/WITNESS options: #optionsINVARIANTS #Enable calls of extra sanity checking #optionsINVARIANT_SUPPORT #Extra sanity checks of internal structures, required by INVARIANTS #optionsWITNESS #Enable checks to detect deadlocks and cycles #optionsWITNESS_SKIPSPIN#Don't run witness on spinlocks for speed and then started a make release. Since doing this, the machine has become almost totally unresponsive. Command execution is measured in hours. A page from top which finally came up shows some very high load averages: last pid: 1892; load averages: 7.14, 6.00, 5.67up 1+10:49:19 23:50:26 34 processes: 1 running, 33 sleeping CPU states: 0.0% user, 0.0% nice, 78.5% system, 0.8% interrupt, 20.7% idle Mem: 69M Active, 909M Inact, 214M Wired, 51M Cache, 112M Buf, 255M Free Swap: 4096M Total, 36K Used, 4096M Free PID USERNAME PRI NICE SIZERES STATETIME WCPUCPU COMMAND 99689 root -85 2928K 2236K biowr0:59 0.00% 0.00% cvs The cvs is being executed by 'make release' updating the chroot area. The repo lives in /home/ncvs which is an nfs mount of a 4.7 system. A kernel with the above options does not exibit this behaviour. When I killed the cvs process, the machine returns to normal. I guess my basic question is: Are the INVARIANTS and WITNESS options required at this point? Thanks, John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
5.0-20021027-CURRENT.iso cdrom will not mount
Hi, The last three 5.x install iso cdroms that I have created will not mount. The iso(s) boot correctly and sysinstall works fine. The disks are formated and newfs'd correctly. However, when sysinstall tries to mount the cdrom, the following error is received: Error mounting /dev/acd0c on /dist: Operation not supported by device (19). The only thing I think I could be doing wrong/incorrectly is the iso is currently created on a 4.7 system. I'll try creating the iso on a 5.x system later this evenning. The system I am trying to install onto works fine running 4.7, and the iso cd mounts ok on a 4.7 system. Could this be a GEOM issue? Something else? Has anyone else run into this problem? Thanks! John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Large 'label'ing defaults for sysinstall
Hi,
In this day of larger disk drives, I've modified
the code in sysinstall to automatically create a /home
partition and increase the rest of the sizes if the
size of the disk (or slice) exceeds a given size (currently
58gig in my patch). For example, using A(uto in the label
editor on a 100gig drive now gives:
Disk: ad6 Partition name: ad6s1 Free: 0 blocks (0MB)
Part Mount Size Newfs Part Mount Size Newfs
- - - -
ad6s1a/1024MB UFS Y
ad6s1bswap 4096MB SWAP
ad6s1e/var 4096MB UFS+S Y
ad6s1f/tmp 8192MB UFS+S Y
ad6s1g/usr15360MB UFS+S Y
ad6s1h/home 62625MB UFS+S Y
The patch is below. I left the one 'if (large)' condition to
the very left to show indention while attempting to reduce whitespace
diffs. A whitespace only diff can be committed posthumously.
I'd like to commit this if there are no major objections. The
values of the LARGE defines, will I'm sure, be refined with use :-)
-John
The patch can also be found at:
http://people.freebsd.org/~jwd/sysinstall/label.diff
--- label.c.origMon Oct 28 05:34:00 2002
+++ label.c Mon Oct 28 06:36:44 2002
@@ -80,11 +80,12 @@
* for this configuration we scale things relative to the NOM vs DEFAULT
* sizes. If the disk is larger then /home will get any remaining space.
*/
-#define ROOT_DEFAULT_SIZE 128
-#define USR_DEFAULT_SIZE 3072
-#define VAR_DEFAULT_SIZE 256
-#define TMP_DEFAULT_SIZE 256
-#define HOME_DEFAULT_SIZE USR_DEFAULT_SIZE
+#define ROOT_DEFAULT_SIZE (large ? ROOT_LARGE_SIZE : 128)
+#define SWAP_DEFAULT_SIZE (large ? SWAP_LARGE_SIZE :0)
+#define USR_DEFAULT_SIZE (large ? USR_LARGE_SIZE : 3072)
+#define VAR_DEFAULT_SIZE (large ? VAR_LARGE_SIZE : 256)
+#define TMP_DEFAULT_SIZE (large ? TMP_LARGE_SIZE : 256)
+#define HOME_DEFAULT_SIZE (large ? HOME_LARGE_SIZE : USR_DEFAULT_SIZE)
/*
* Nominal partition sizes. These are used to scale the default sizes down
@@ -97,6 +98,23 @@
#define TMP_NOMINAL_SIZE 64
#define HOME_NOMINAL_SIZE USR_NOMINAL_SIZE
+/*
+ * Large partition sizes. If we have a 'large' disk (60gig?) then we
+ * try to scale things up a bit.. In this case, we create the /home
+ * mount point.
+ */
+#define ROOT_LARGE_SIZE( 1024)
+#define SWAP_LARGE_SIZE( 4*1024)
+#define USR_LARGE_SIZE (15*1024)
+#define VAR_LARGE_SIZE ( 4*1024)
+#define TMP_LARGE_SIZE ( 8*1024)
+#define HOME_LARGE_SIZE(17*1024)
+
+/*
+ * Let's define a what we think a large disk is..
+ */
+#define LARGE_DISK_SIZE(58*1024*1024)
+
/* The bottom-most row we're allowed to scribble on */
#define CHUNK_ROW_MAX 16
@@ -1178,6 +1196,7 @@
try_auto_label(Device **devs, Device *dev, int perc, int *req)
{
int sz;
+int large = 0;
struct chunk *root_chunk = NULL;
struct chunk *swap_chunk = NULL;
struct chunk *usr_chunk = NULL;
@@ -1192,6 +1211,11 @@
char *msg = NULL;
sz = space_free(label_chunk_info[here].c);
+if (sz LARGE_DISK_SIZE)
+ large = 1;
+else
+ large = 0; /* Just in case */
+
if (sz = FS_MIN_SIZE)
return(Not enough free space to create a new partition in the slice);
@@ -1214,7 +1238,7 @@
record_label_chunks(devs, dev);
}
if (!swapdev) {
- sz = requested_part_size(VAR_SWAP_SIZE, 0, 0, perc);
+ sz = requested_part_size(VAR_SWAP_SIZE, 0, SWAP_DEFAULT_SIZE, perc);
if (sz == 0) {
int nom;
int def;
@@ -1279,9 +1303,9 @@
}
if (!usrdev !variable_get(VAR_NO_USR)) {
sz = requested_part_size(VAR_USR_SIZE, USR_NOMINAL_SIZE, USR_DEFAULT_SIZE,
perc);
-#if AUTO_HOME == 0
+if (!large)
sz = space_free(label_chunk_info[here].c);
-#endif
+
if (sz) {
if (sz (USR_MIN_SIZE * ONE_MEG)) {
*req = 1;
@@ -1303,7 +1327,7 @@
record_label_chunks(devs, dev);
}
}
-#if AUTO_HOME == 1
+if (large)
if (!homedev !variable_get(VAR_NO_HOME)) {
sz = requested_part_size(VAR_HOME_SIZE, HOME_NOMINAL_SIZE, HOME_DEFAULT_SIZE,
perc);
if (sz space_free(label_chunk_info[here].c))
@@ -1330,7 +1354,7 @@
record_label_chunks(devs, dev);
}
}
-#endif
+
/* At this point, we're reasonably labelled */
if (variable_cmp(DISK_LABELLED, written))
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message
Re: 5.0-20021025-CURRENT snapshot
- Daniel Flickinger's Original Message - Sent: Sun, 27 Oct 2002 01:03:43 +0900 by MAR + jwd A new 5.0-20021025-CURRENT snapshot is available + jwd via anonymous ftp at usw2.freebsd.org: + + Wonderful! Is it a time to switch back 'current.FreeBSD.org' name to + that machine? NO! it is a fraction of what you have on snapshot.jp.freebsd.org usw2.freebsd.org has a serious lack of disk space. If/when more disk space is added usw2 can become a standard 5.x distribution site again. Until then, it is limited to a small number of ftp'able install areas and an ISO (or two). -John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Local DNS lookup by sshd?
Hi, When logging into a current 5.0 system via ssh, I see the following written to the system console (the 'xxx's are my whiteout): ... kernel: Connection attempt to UDP xxx.58.184.35:53 from xxx.58.184.35:49253 ... kernel: Connection attempt to UDP xxx.58.184.35:53 from xxx.58.184.35:49254 ... kernel: Connection attempt to UDP xxx.58.184.35:53 from xxx.58.184.35:49255 ... kernel: Connection attempt to UDP xxx.58.184.35:53 from xxx.58.184.35:49256 Basically, it looks like it is trying to talk to a DNS on the localhost. However, I do not have DNS running. I do not have localhost listed in /etc/resolv.conf. /etc/nsswitch.conf lists 'hosts: files dns' and putting my ssh origination id in /etc/hosts has no effect. It appears to be related to code in canohost.c. Before I start debugging, I thought I'd ask if anyone knew if there is a reason for this behaviour, or where it might be coming from (specifically). Comments Welcome. -John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
5.0 disklabel warnings against 4.7 disk
Hi, I've (re)scanned my -current folder for issues related to the following but didn't see a good match. Pointing out my blindness is allowed if this was discussed... I have a system onto which I installed a 4.7-RC a couple of weeks ago. I then upgraded that newly installed system to -current. After the upgrade, disklabel now complains about every disk in the system, for example: # disklabel -r ad6s1 # /dev/ad6s1c: type: unknown disk: amnesiac label: flags: bytes/sector: 512 sectors/track: 63 tracks/cylinder: 16 sectors/cylinder: 1008 cylinders: 193820 sectors/unit: 195371505 rpm: 3600 interleave: 1 trackskew: 0 cylinderskew: 0 headswitch: 0 # milliseconds track-to-track seek: 0 # milliseconds drivedata: 0 8 partitions: #size offsetfstype [fsize bsize bps/cpg] a: 195371505 634.2BSD 2048 16384 28512 # (Cyl.0*- 193820*) c: 195371505 63unused0 0# (Cyl.0*- 193820*) partition a: partition extends past end of unit partition c: partition extends past end of unit Warning, partition c doesn't start at 0! Warning, An incorrect partition c may cause problems for standard system utilities # This is an fdisk'd drive, with the entire disk allocated to the 1st slice. An identical disk (ad4): ad4: 95396MB WDC WD1000JB-32CWE0 [193821/16/63] at ata2-master UDMA100 ad6: 95396MB WDC WD1000JB-32CWE0 [193821/16/63] at ata3-master UDMA100 which I then fdisk and disklabel under -current shows up as: 8 partitions: #size offsetfstype [fsize bsize bps/cpg] c: 1953715050unused0 0# (Cyl.0 - 193820*) The only difference being the offset of 0 (and lack of warning msgs). Both systems come up with the same disk size, but the offsets not matching seems to be a problem. An older 4.7 pre-release system from Sep 16 creates the 'c' slice with an offset of 0. I don't think this is something I did wrong. Any thoughts, comments, or pointy hats are welcome. Thanks, John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
5.0-20021025-CURRENT snapshot
This will be the last post on this topic. A new 5.0-20021025-CURRENT snapshot is available via anonymous ftp at usw2.freebsd.org: /pub/FreeBSD/snapshots/i386/5.0-20021025-CURRENT New snaps will appear each day if the build completes without errors. Enjoy! John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
5.0-20021023-CURRENT available
Hi, I have have uploaded a 5.0-20021023-CURRENT snapshot to usw2.freebsd.org available via anonymous ftp: /pub/FreeBSD/snapshots/i386/5.0-20021023-CURRENT I am currently unable to boot from the kern/mfsroot floppies due to Cannot find /mfsroot. I have not had a chance to try burning boot.flp to a cd to see if it will boot correctly. I will rotate in a new snapshot each day assuming good builds. Enjoy! -John To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
5.0-R make release panic
Hi Folks, I've been running repetitive 'make release's on a box and it died once (no serial console), and I finally got it to crash again after about 2 days. uname: (world and kernel up-to-date as of) 5.0-CURRENT FreeBSD 5.0-CURRENT #0: Thu Oct 17 02:12:59 GMT 2002 df: Filesystem 1K-blocks Used Avail Capacity Mounted on /dev/ad0s1a5160627691039786816%/ devfs 11 0 100%/dev /dev/ad0s1e 4129310 8814 3790152 0%/var /dev/ad0s1f 206430263392 1835766 3%/tmp /dev/ad0s1g 10322414 831488 8665134 9%/usr /dev/ad0s1h 17318166 24 15932690 0%/home procfs 44 0 100%/proc /dev/ad1s1a 96135654 596042 87848760 1%/vol/vol0 /dev/ad2s1a 961382062 88447148 0%/vol/vol1 /dev/ad4s1a 961382062 88447148 0%/vol/vol2 /dev/ad6s1a 961382062 88447148 0%/vol/vol3 rtp1:/vol/vol0/ncvs 192318366 17520608 167105024 9%/home/ncvs Basically, I reload the source tree (inside of make release) from the nfs mounted repo. The box itself is a single processor amd system running a uni-processor kernel. Follows is a backtrace and process dump. I've restarted the process. Any ideas for where to head with this are welcome. I'll be glad to do a deeper debug session on the next crash. -John db t Debugger(c049e8b5,c053ea80,c04af39c,e86aa5e8,1) at Debugger+0x54 panic(c04af39c,1000,11aa9000,e86aa614,e86aa624) at panic+0xab kmem_malloc(c0832078,1000,4,e86aa67c,c040c854) at kmem_malloc+0xf8 page_alloc(c30cfa00,1000,e86aa663,4,c031a133) at page_alloc+0x27 slab_zalloc(c30cfa00,0,c871d5b0,cc99a380,0) at slab_zalloc+0x124 uma_zalloc_internal(c30cfa00,0,0,cebc2400,0) at uma_zalloc_internal+0x166 uma_zalloc_arg(c30cfa00,0,0,e86aa774,e86aa788) at uma_zalloc_arg+0x38b nfs_nget(c8a4fa00,c41f905c,1c,e86aa770,e86aa784) at nfs_nget+0x163 nfs_lookup(e86aa90c,c871d5b0,0,c871d5b0,c871d5b0) at nfs_lookup+0x9f2 lookup(e86aabc4,c8999000,400,e86aabe0,c871d5b0) at lookup+0x302 namei(e86aabc4,e86aaa38,e86aaafc,c03d2223,c871d5b0) at namei+0x20b vn_open_cred(e86aabc4,e86aacc4,1a4,c8ad5e80,e86aacb0) at vn_open_cred+0x258 vn_open(e86aabc4,e86aacc4,1a4,e86aac68,e86aaafc) at vn_open+0x29 kern_open(c871d5b0,80ce000,0,1,1b6) at kern_open+0x1e3 open(c871d5b0,e86aad10,c,e86aad40,3) at open+0x30 syscall(2f,2f,2f,4,281ebb00) at syscall+0x39a Xint0x80_syscall() at Xint0x80_syscall+0x1d --- syscall (5, FreeBSD ELF32, open), eip = 0x28166dd3, esp = 0xbfbfef5c, ebp = 0xbfbfef88 --- db ps pid proc addruid ppid pgrp flag stat wmesgwchan cmd 35707 cd2c8540 ea95b0000 35705 27194 0004002 norm[CPU 0] cvs 35705 ce1f9a80 ea9cc0000 27203 27194 0004002 norm[SLPQwait ce1f9a80][SLP] sh 27203 ce1f9540 ea9bd0000 27194 27194 0004002 norm[SLPQwait ce1f9540][SLP] make 27194 cd2cd700 ea99e0000 509 27194 0004002 norm[SLPQwait cd2cd700][SLP] sh 509 c874fe00 ea8e50000 508 508 0004002 norm[SLPQ ttyin c89c2240][SLP] zsh 508 c8ab4000 ea8e80000 439 508 0004102 norm[SLPQwait c8ab4000][SLP] su 503 c8ab48c0 ea9260000 1 503 0004002 norm[SLPQ ttyin c8505c10][SLP] getty 502 c8ab4a80 ea9270000 1 502 0004002 norm[SLPQ ttyin c8a4fe10][SLP] getty 501 c8ab4c40 ea9280000 1 501 0004002 norm[SLPQ ttyin c8a50210][SLP] getty 500 c8ab4e00 ea9290000 1 500 0004002 norm[SLPQ ttyin c8a50610][SLP] getty 499 c8ab6000 ea92a0000 1 499 0004002 norm[SLPQ ttyin c8993c10][SLP] getty 498 c8ab61c0 ea92b0000 1 498 0004002 norm[SLPQ ttyin c89c2010][SLP] getty 497 c8ab6380 ea92c0000 1 497 0004002 norm[SLPQ ttyin c89c2610][SLP] getty 496 c874f700 ea8e10000 1 496 0004002 norm[SLPQ ttyin c89c2a10][SLP] getty 495 c871b000 e868d0000 1 495 0004002 norm[SLPQ ttyin c84a7a10][SLP] getty 478 c871e1c0 e86d0 1 478 000 norm[SLPQ nanslp c053f394][SLP] cron 440 c874f540 ea8e 25 1 440 2000100 norm[SLPQ pause ea8e][SLP] sendmail 439 c871e380 e86d1000 227 438 439 0004002 norm[SLPQwait c871e380][SLP] sh 438 c871e540 e86d2000 227 432 426 100 norm[CVQ select c04f3ec4][SLP] sshd 435 c871ec40 e86d60000 1 435 100 norm[CVQ select c04f3ec4][SLP] sendmail 432 c874f380 ea8df0000 426 426 100 norm[SLPQ sbwait c89b9c04][SLP] sshd 426 c874f8c0 ea8e20000 1 426 100 norm[CVQ select c04f3ec4][SLP] sshd 357 c871ee00 e86d70000 352 352 000 norm[SLPQnfsd c8993a00][SLP] nfsd 356 c8503e00 e868c0000 352 352 000 norm[SLPQnfsd c8993400][SLP] nfsd 355 c874f000 ea8dd0000 352 352 000 norm[SLPQnfsd c8993600][SLP] nfsd 354 c8503c40 e868b0000 352 352 000 norm[SLPQnfsd c8993800][SLP]
Inetd configuration message logging (or lack thereof)
Hi,
Comments on the following patch. The messages
are invisible with a default install. This patch
gets them into /var/log/messages where they can be
seen.
-John
ps: Not in this patch, but I wonder if syslog
cannot be redirected to the console when run -d.
Something like:
#define syslog InetdSyslog
void InetdSyslog(int,char *, ...);
void
InetdSyslog(int priority,char *message, ...)
{
va_list ap;
va_start(ap, message);
if (debug) {
vfprintf(stderr,message,ap);
fprintf(stderr,\n);
}
else
vsyslog(priority,message,ap);
va_end(message);
return;
}
RCS file: /home/ncvs/src/usr.sbin/inetd/inetd.c,v
retrieving revision 1.96
diff -u -r1.96 inetd.c
--- inetd.c 8 May 2002 17:20:08 - 1.96
+++ inetd.c 19 May 2002 01:04:56 -
@@ -1613,7 +1613,7 @@
}
if (strncmp(sep-se_proto, rpc/, 4) == 0) {
if (no_v4bind != 0) {
- syslog(LOG_INFO, IPv4 bind is ignored for %s,
+ syslog(LOG_NOTICE, IPv4 bind is ignored for %s,
sep-se_service);
freeconfig(sep);
goto more;
@@ -1654,7 +1654,7 @@
#ifdef INET6
if (sep-se_proto[strlen(sep-se_proto) - 1] == '6') {
if (no_v6bind != 0) {
- syslog(LOG_INFO, IPv6 bind is ignored for %s,
+ syslog(LOG_NOTICE, IPv6 bind is ignored for %s,
sep-se_service);
freeconfig(sep);
goto more;
@@ -1686,7 +1686,7 @@
#endif
{ /* default to v4 bind if not v6 bind */
if (no_v4bind != 0) {
- syslog(LOG_INFO, IPv4 bind is ignored for %s,
+ syslog(LOG_NOTICE, IPv4 bind is ignored for %s,
sep-se_service);
freeconfig(sep);
goto more;
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message
