Re: BSDStats project, interesting returns from OpenBSD
On Thu, Sep 14, 2006 at 08:40:21AM -0500, Doug Poland wrote: I just glanced at the latest statistics on www.bsdstats.org. In the last week or so OpenBSD has overtaken FreeBSD in the USA. Should one conclude that OpenBSD admins have enthusiastically embraced this project and FreeBSD admins have not; or, is OpenBSD really more widely deployed than FreeBSD? One should not conclude anything until the numbers are much larger than they are now, because small fluctuations from e.g. regional promotion of bsdstats in one country but not another, or one large company deploying it on all machine, will dramatically change your conclusions. Kris pgp3vAekpFTu7.pgp Description: PGP signature
Re: BSDStats project, interesting returns from OpenBSD
On Thu, Sep 14, 2006 at 12:53:04PM -0400, Gerard Seibert wrote: On Thursday 14 September 2006 12:09, Kris Kennaway wrote: One should not conclude anything until the numbers are much larger than they are now, because small fluctuations from e.g. regional promotion of bsdstats in one country but not another, or one large company deploying it on all machine, will dramatically change your conclusions. I was just wondering if there is any consensus on adding BSDStats to the base system? If would appear to be a logical step to take so as to insure that all users of FBSD would be counted. An end user could always disable the sending of data by disabling it in the /etc/rc.file. I feel that unless it is part of the base system and turned on by default, too many users will never take part in the reporting process. I highly doubt that it would be enabled by default in FreeBSD, since many of our users (or their employers) would consider it a privacy breach to have their systems reporting back automatically. Kris pgphwA3uRDlUy.pgp Description: PGP signature
Re: BSDStats project, interesting returns from OpenBSD
On Thu, Sep 14, 2006 at 01:29:38PM -0400, Andy Greenwood wrote: What about making it a sysinstall option? Not in the base install, but the option is presented when setting up a new box. That's not ruled out, if someone does the work. Kris pgpIrJBp4JQSq.pgp Description: PGP signature
Re: FreeBSD 6.1 shutting down.
On Thu, Sep 14, 2006 at 07:29:13PM +, Marwan Sultan wrote: hello Lowell, thank you for your reply, i wish you could find some solution for me i tried to google the net, and found many results for atapci1: failed to enable memory mapping! but most with no solutions. I doubt it's the cause of the problems; most likely you have some other failing hardware component (bad memory, power supply, etc). Check the archives for extensive discussion. Kris P.S. Don't top-post, it spoils the logical flow of the thread. pgpvoBONOcdau.pgp Description: PGP signature
Re: mount_ext2fs returning ENODEV on 6.1
On Thu, Sep 14, 2006 at 02:07:56PM -0700, Perry Hutchison wrote: # ll /dev/ad0s7 crw-r- 1 root operator0, 93 Sep 4 02:30 /dev/ad0s7 # file -s /dev/ad0s7 /dev/ad0s7: Linux rev 1.0 ext2 filesystem data # grep -w ad0s7 /etc/fstab /dev/ad0s7 /linux ext2fs ro 0 0 # ll -d /linux drwxr-xr-x 2 root wheel 512 Aug 24 12:09 /linux # mount /linux mount_ext2fs: /dev/ad0s7: Operation not supported by device No ext2fs support in your kernel? I had not thought that was the problem, since according to something in the docs or manpages -- which I now cannot locate -- missing kernel support should have resulted in a different message. How would I check, to be sure? I am using the kernel from the installation CD, not one I have built: Then you don't have kernel support, since it's not enabled by default (the ext2 code is under the GPL and cannot be distributed in a BSD-licensed kernel). Recompile your kernel or load the module per the handbook. Kris pgppw5OdBtg6u.pgp Description: PGP signature
Re: libm.so.3 on FreeBSD 6.1-RELEASE
On Tue, Sep 12, 2006 at 09:23:46PM -0500, Jonathan Horne wrote: cvsuping to -p6 wil not fix the problem. i had the exact same issue (except it was a list of 6 different .so files) getting the NetBackup 5.1 agent for UNIX to run on freebsd. my solution was as simple as: ln -s /lib/libm.so.4 /lib/libm.so.2 (NetBackup agent was looking for so.2) so in your case, just symlink the existing .4 to a .3, and you should be good to go. backwards compatibility should not be an issue. That's a bogus hack; the libraries are not compatible or they'd have the same version! Just install the relevant compat package (compat4x/compat5x). Kris pgp1ELXWqSUsz.pgp Description: PGP signature
Re: Problem installing Mathematica on FreeBSD 6.1
On Tue, Sep 12, 2006 at 10:09:27PM +0400, Alexey Mikhailov wrote: Hello! Sometime ago I posted this message to freebsd-questions@, but had no answer. So I'll try my luck here. - Hello! I installed Mathematica 5.1 on my FreeBSD 6.1 system. And I can't run it.. That's very strange behaviour: ~/Mathematica/SystemFiles/FrontEnd/Binaries/Linux % ./Mathematica ./Mathematica: relocation error: /usr/X11R6/lib/libX11.so.6: undefined symbol: __stderrp But: ~/Mathematica/SystemFiles/FrontEnd/Binaries/Linux % ldd Mathematica Mathematica: libm.so.6 = /lib/libm.so.6 (0x284e5000) ... Since it is a linux binary it should be linked to the linux libraries. Do you have a linux_base package installed? Kris P.S. Redirecting followups to freebsd-questions@ as this is a basic technical support question. pgpXLAppEbfpx.pgp Description: PGP signature
Re: FreeBSD not popular in Asia?
On Fri, Sep 08, 2006 at 10:02:45PM -0300, Marc G. Fournier wrote: Check out http://www.bsdstats.org ... Republic of Korea is about to push the US out of first place, but there are *zero* FreeBSD boxes reporting from there ... DragonFly is first, then NetBSD and then OpenBSD ... Are there *really* no Korean FreeBSD hosts out there ... ? So which korean freebsd mailing lists did you advertise bsdstats on? Kris pgpBxrHKdZtqp.pgp Description: PGP signature
Re: rpc.lockd stalls
On Thu, Sep 07, 2006 at 01:34:08PM -0400, Tom Ierna wrote: Hello, list. For the purposes of ease of software and hardware management, I'm attempting to run a set of PXE-booted Client machines as web/db or mail servers. The NFS/DHCP/YP servers are running on a 5.4-STABLE Server. I mostly followed the PXE guide when building these systems. All of the disk (except for swap) sits on the master Server (which has a bunch of external drive sleds), and all of the Client machines boot via Gig-E. Client machines are running 5.4-STABLE as well, but it is not compiled with the same kernel configuration as the master Server, as the hardware is slightly different. Client machines share userland with the Server. At the moment I have one Client machine running about 40 domains of web and db, with reasonably low traffic (less than 3Mbit/sec total) and one Client machine booted from the master Server, but not doing anything. Resource utilization on the master Server seems pretty low. Sporadically, there appear to be stalls on some locks with rpc.lockd. rpc.lockd is unreliable in all versions of FreeBSD (although it may be worse in 5.x), see the mailing list archives for extensive discussion of this. Try turning it off and using mount_nfs -L instead to fake the lock traffic (See the manpage). Kris pgpYJqMO1v6e5.pgp Description: PGP signature
Re: rpc.lockd stalls
On Thu, Sep 07, 2006 at 02:12:26PM -0400, Tom Ierna wrote: On Sep 7, 2006, at 1:40 PM, Kris Kennaway wrote: On Thu, Sep 07, 2006 at 01:34:08PM -0400, Tom Ierna wrote: Sporadically, there appear to be stalls on some locks with rpc.lockd. rpc.lockd is unreliable in all versions of FreeBSD (although it may be worse in 5.x), see the mailing list archives for extensive discussion of this. Try turning it off and using mount_nfs -L instead to fake the lock traffic (See the manpage). Kris, Is there a way to note -L via fstab? Since these machines are PXE booted, unmounting and re-mounting with -L will be problematic, and I'd like them to inherit this property at reboot. Yes, use the -o format, see the manpage. Kris pgp9m8FlnRVJi.pgp Description: PGP signature
Re: rpc.lockd stalls
On Thu, Sep 07, 2006 at 03:19:51PM -0400, Tom Ierna wrote: On Sep 7, 2006, at 2:39 PM, Kris Kennaway wrote: On Thu, Sep 07, 2006 at 02:12:26PM -0400, Tom Ierna wrote: Is there a way to note -L via fstab? Since these machines are PXE booted, unmounting and re-mounting with -L will be problematic, and I'd like them to inherit this property at reboot. Yes, use the -o format, see the manpage. Under the man page for mount_nfs, I have the following: -o Options are specified with a -o flag followed by a comma sepa- rated string of options. See the mount(8) man page for possible options and their meanings. The following NFS specific options are also available: ... Historic -o Options ... lockd Same as not specifying -L. ... It doesn't have any other reference to -L. Are mounts specified in fstab automatically non-locking, or is the man page incorrect? Prefixing with 'no' negates an option. Kris pgpfmX1pSMpew.pgp Description: PGP signature
Re: DL 380/G5 with 16G of ram
On Thu, Sep 07, 2006 at 01:50:48PM -0700, Josef Grosch wrote: Hello, I've got a DL 380/G5 as an evalu unit. It has 16 gig of ram. I compiled a PAE kernel but I'm finding that it is not very stable. It crashes during heavy disk activity, ie. portupgrade -rav. Does anyone have experience with this sort of machine and would you care to share your kernel config file and/or advice. A good place to start looking would be at the disk driver; is it listed in the PAE kernel config? If not, it's probably known not to work. kris pgpOMYqEa0LtX.pgp Description: PGP signature
Re: whowatch port not working on AMD64 machines
On Thu, Sep 07, 2006 at 09:10:09PM -0400, stan wrote: I'm having probelms with the whowatch port on my AMD64 6.1 machines. Does anyone know how I can make this work? I really use this tool a lot. Try talking to the developers of the software. Kris pgpK2JzktdKSJ.pgp Description: PGP signature
Re: AMD64 Stability with 6.1+?
On Tue, Sep 05, 2006 at 08:09:27AM -0500, Thomas T. Veldhouse wrote: Can anybody on this list indicate what kind of stability they have seen with AMD64 systems using FreeBSD (the 64-bit binaries)? How about the majority of PORTS in the tree? Stability of amd64 is excellent. # uptime 7:45AM up 104 days, 7:39, 2 users, load averages: 8.10, 7.61, 7.06 # uname -r 6.1-RELEASE That's one of the machines that I use for the official package builds, it's been building 8 packages concurrently for the past 104 days (which was the last time I rebooted it to update the kernel). Kris pgpO2J8msw50F.pgp Description: PGP signature
Re: AMD64 Stability with 6.1+?
On Tue, Sep 05, 2006 at 06:54:44PM +0200, Cor van Wandelen wrote: I experience instability with FreeBSD-AMD64 using ext2/ext3 filesystems. Sometimes reading and writing these filesystems work without any problems and the next time the whole system hangs with the only option resetting the pc. Nevertheless I use it because overall it seems to be a lot faster, but I stay away from the mentioned filesystems. Have you submitted a PR? Make sure to follow the directions in the chapter on kernel debugging in the developers handbook. Kris pgpZlfL53MxL9.pgp Description: PGP signature
Re: Various package/ports problems
On Mon, Sep 04, 2006 at 04:52:08AM -0700, Dr. Jennifer Nussbaum wrote: Kris Kennaway [EMAIL PROTECTED] wrote: On Sun, Sep 03, 2006 at 06:31:18PM -0700, Dr. Jennifer Nussbaum wrote: Kris Kennaway wrote: On Sun, Sep 03, 2006 at 05:58:36PM -0700, Dr. Jennifer Nussbaum wrote: Kris Kennaway wrote: On Mon, Aug 28, 2006 at 01:59:14PM -0700, Dr. Jennifer Nussbaum wrote: Thanks, i CAN compile from the ports collection (i never wanted to use packages). Its MANAGING this that i cant do. I can update my ports tree fine, and i can go into ports directory and do a make install, sure. But i cant do portsversion to see what needs updating, and i cant do pkg_info to find out what versions i have. So doing something like Why not (in both cases)?As i posted in the original message - the main question i was asking - both of these are broken in some way. Reposting: --- For example if i try to run portversion i get: undefined method `each' for nil:NilClass /usr/local/sbin/portversion:239: [BUG] Bus Error from Ruby (ruby 1.8.5 is up to date on the system). : pkgdb -f ~ $ sudo pkgdb -f ~ $ portversion undefined method `each' for nil:NilClass /usr/local/sbin/portversion:239: [BUG] Bus Error ruby 1.8.5 (2006-08-25) [i386-freebsd4] Abort trap (core dumped) ~ $ Check the manpage, probably I meant pkgdb -fu If i try to run pkg_info I quickly get: pkg_info: read_plist: bad command '@conflicts apache+ipv6-1.*' : Deinstall the package that was built for a later version of : FreeBSD. KrisHow do i tell which this is? Most of this box seems to be compiled from ports, not installed packages. apache+ipv6 is not installed. ~ $ pkg_info ImageMagick-nox11-6.0.2.7 Image processing tools analog-6.0_1,1 An extremely fast program for analysing WWW logfiles pkg_info: read_plist: bad command '@conflicts apache+ipv6-1.*' ~ I rebuilt analog from ports but that doesnt seem to be it. grep -r @conflicts /var/db/pkg and remove all listed packages. Kris pgpIpriec72os.pgp Description: PGP signature
Re: Various package/ports problems
On Mon, Sep 04, 2006 at 02:00:20PM -0700, Dr. Jennifer Nussbaum wrote: Kris Kennaway [EMAIL PROTECTED] wrote: On Mon, Sep 04, 2006 at 04:52:08AM -0700, Dr. Jennifer Nussbaum wrote: For example if i try to run portversion i get: undefined method `each' for nil:NilClass /usr/local/sbin/portversion:239: [BUG] Bus Error from Ruby (ruby 1.8.5 is up to date on the system). : pkgdb -f ~ $ sudo pkgdb -f ~ $ portversion undefined method `each' for nil:NilClass /usr/local/sbin/portversion:239: [BUG] Bus Error ruby 1.8.5 (2006-08-25) [i386-freebsd4] Abort trap (core dumped) ~ $ ::Check the manpage, probably I meant pkgdb -fu Sorry, i dont understand what im checking hte manpage to have it do. Running pkgdb -fu does rebuild the package database, but i get the same Bus Error thing from Ruby as above. Running which version of portupgrade? Also try 'cd /usr/ports; make fetchindex' first. If i try to run pkg_info I quickly get: pkg_info: read_plist: bad command '@conflicts apache+ipv6-1.*' : Deinstall the package that was built for a later version of : FreeBSD. KrisHow do i tell which this is? Most of this box seems to be compiled from ports, not installed packages. apache+ipv6 is not installed. ~ $ pkg_info ImageMagick-nox11-6.0.2.7 Image processing tools analog-6.0_1,1 An extremely fast program for analysing WWW logfiles pkg_info: read_plist: bad command '@conflicts apache+ipv6-1.*' ~ ::grep -r @conflicts /var/db/pkg and remove all listed packages. Kris Which listed packages? The output of this grep looks like: /var/db/pkg/mysql-server-4.1.20/+CONTENTS:@conflicts mysql-server-3.* /var/db/pkg/mysql-server-4.1.20/+CONTENTS:@conflicts mysql-server-4.[02-9].* /var/db/pkg/mysql-server-4.1.20/+CONTENTS:@conflicts mysql-server-5.* But i dont want to remove mysql-server-4.1.20, because this seems to be the version that is actively running on this machine, and i dont want to disable it. Same for apache+mod_ssl, which accounts for alot of the other conflicts. You need to recognize that you've got the system into a damaged state already, so it's going to take further intrusive work to repair it. Also, when i try to delete one of the ports that seemed less critical, even that didnt work: ~ $ sudo pkg_delete -f jade-1.2.1_9 pkg_delete: package 'jade-1.2.1_9' is required by these other packages and may not be deinstalled (but I'll delete it anyway): sgmlformat-1.7_2 pkg_delete: read_plist: bad command '@conflicts sp-1.*' ~ ...and jade-1.2.1_9 is still there. Thank you again for your continued attention, Kris. What happens if you try to reinstall it from the port, setting FORCE_PKG_REGISTER? Kris pgpNESpPRao4Z.pgp Description: PGP signature
Re: Various package/ports problems
On Mon, Sep 04, 2006 at 08:57:51PM -0700, Dr. Jennifer Nussbaum wrote: Kris Kennaway [EMAIL PROTECTED] wrote: On Mon, Sep 04, 2006 at 02:00:20PM -0700, Dr. Jennifer Nussbaum wrote: Kris Check the manpage, probably I meant pkgdb -fu Sorry, i dont understand what im checking hte manpage to have it do. Running pkgdb -fu does rebuild the package database, but i get the same Bus Error thing from Ruby as above. Kris: Running which version of portupgrade? Also try 'cd /usr/ports; make fetchindex' first. No difference; its portupgrade version 2.0.1 (as reported by portupgrade --version). You'll have to report this to the author. Kris: What happens if you try to reinstall it from the port, setting FORCE_PKG_REGISTER? Oh yes! This does work. After this worked i went and reinstalled all the ports listed, and now pkg_info (and pkg_delete, etc.) is working correctly. Thanks! Something working at last. Glad to hear it :) Kris pgp0BiVQc0oxI.pgp Description: PGP signature
Re: Error in the list of ports www.freebsd.org
On Sun, Sep 03, 2006 at 09:04:21PM +0200, Santos wrote: The http://cvsweb.FreeBSD.org/ports/mail/courier-authlib-vchkpw courier-authlib-ldap-0.58_2 This confused me for a while, because your URL is for a different port. port requires openldap-client-2.2.30, not 2.3.27 You must be looking at an old ports tree; 2.3.27 is used by the current tree (since July). Kris pgpNqPQIUR8Io.pgp Description: PGP signature
Re: portversion is part of what?
On Sun, Sep 03, 2006 at 05:21:49PM -0300, Henry Lenzi wrote: Hi -- Is portversion a part of another big binary? Apparently, I can't just reinstall portversion. sysutils/portupgrade Kris pgpz9bDWiJ0yz.pgp Description: PGP signature
Re: Various package/ports problems
On Sun, Sep 03, 2006 at 05:58:36PM -0700, Dr. Jennifer Nussbaum wrote: Kris Kennaway [EMAIL PROTECTED] wrote: On Mon, Aug 28, 2006 at 01:59:14PM -0700, Dr. Jennifer Nussbaum wrote: I wish i could, but i dont have physical access to the machine - its in a remote location with no serial console. So if anything gets messed up during an upgrade, im in a hell of a lot of trouble. I dont know when i'll be able to fix this problem so until i can upgrade the OS i at least want to keep the rest of it up to date. The packages are compiled for 4.11-stable and do not work with systems much older than this (e.g. 4.8, as you discovered). You might have more luck compiling from the ports collection, although that too only supports 4.11 officially (and modern releases, of course). Thanks, i CAN compile from the ports collection (i never wanted to use packages). Its MANAGING this that i cant do. I can update my ports tree fine, and i can go into ports directory and do a make install, sure. But i cant do portsversion to see what needs updating, and i cant do pkg_info to find out what versions i have. So doing something like Why not (in both cases)? upgrading Perl and all it's dependencies--wow what a nightmare that would be. But for another way of fixing the problem, is there any trustworthy way i can do a clean install of 6.1 on a box i dont have access to? I can backup everything offsite and ship a CD or an external drive to the facility, if there is some way i can be assured of getting back into the box after some install procedure is run by a tech there. You can do an 'upgrade install' from the install media...it's something you should definitely practise on a local system first though (not because it's hard, but so you know what you're doing). Kris pgpj5Q9Ob5X3S.pgp Description: PGP signature
Re: Various package/ports problems
On Sun, Sep 03, 2006 at 06:31:18PM -0700, Dr. Jennifer Nussbaum wrote: Kris Kennaway [EMAIL PROTECTED] wrote: On Sun, Sep 03, 2006 at 05:58:36PM -0700, Dr. Jennifer Nussbaum wrote: Kris Kennaway wrote: On Mon, Aug 28, 2006 at 01:59:14PM -0700, Dr. Jennifer Nussbaum wrote: Thanks, i CAN compile from the ports collection (i never wanted to use packages). Its MANAGING this that i cant do. I can update my ports tree fine, and i can go into ports directory and do a make install, sure. But i cant do portsversion to see what needs updating, and i cant do pkg_info to find out what versions i have. So doing something like Why not (in both cases)?As i posted in the original message - the main question i was asking - both of these are broken in some way. Reposting: --- For example if i try to run portversion i get: undefined method `each' for nil:NilClass /usr/local/sbin/portversion:239: [BUG] Bus Error from Ruby (ruby 1.8.5 is up to date on the system). pkgdb -f If i try to run pkg_info I quickly get: pkg_info: read_plist: bad command '@conflicts apache+ipv6-1.*' (i googled for similar problems and there were a few other reports of this, with other packages, but no suggestion of how to fix them.) Deinstall the package that was built for a later version of FreeBSD. Kris pgpOQzObD7MKY.pgp Description: PGP signature
Re: Snapshot performance
On Fri, Sep 01, 2006 at 04:06:01PM -0700, Skylar Thompson wrote: Kris Kennaway wrote: On Fri, Sep 01, 2006 at 11:08:51AM -0700, Paul Lathrop wrote: Hi all, We're working on deploying a new mail server on FreeBSD 6.1-STABLE. One of the major selling points was the ability to take filesystem snapshots in order to make backups from a consistent filesystem on such a high-traffic system. Unfortunately, when I take a snapshot, performance slows to a crawl - to the point where the system stops responding to network requests (ping, SMTP, etc.). Also, the snapshot takes 10-15 minutes to complete. Is this a typical situation? Will I need to schedule downtime for backups in spite of this nifty new feature? Am I doing something wrong? Time depends on the size of the filesystem - but you are correct that snapshots were not designed with performance in mind (rather, to speed up booting after an unclean shutdown by removing the need to wait for fsck). Kris Are there plans to improve performance of snapshots? Using the freebsd-snapshot port to link FS snapshots to the automounter is pretty nifty, but it does kill I/O performance while that's in progress as the OP mentioned. Unfortunately I don't think anyone is working on it. The closest thing on the horizon is ZFS support which does feature high-performance snapshots. This is still a way off though. Kris pgpQTmIjCScRG.pgp Description: PGP signature
Re: newbie. how to compile gcc-4.1.1
On Sat, Sep 02, 2006 at 10:17:27AM -0400, g wrote: ok, i got a snapshot of the make bootstrap it shows the error. ??? Well, you didn't quite follow my instructions, but you showed enough to see that something is quite odd on your system: srcdir=/usr/ports/gcc-4.1.1/fixincludes /bin/sh /usr/ports/gcc-4.1.1/fixincludes/mkfixinc.sh +i386-unknown-freebsd6.1 sed -e 's/@gcc_version@//'mkheadersT *** Error code 2 Stop in /usr/ports/objdir/build-i386-unknown-freebsd6.1/fixincludes. *** Error code 1 Stop in /usr/ports/objdir. Those directories are completely non-standard. How did you install your ports collection? The gcc41 port is supposed to be in /usr/ports/lang/gcc41, not /usr/ports/gcc-4.1.1. Are you really using the ports collection or did you just download the gcc source code directly? Kris P.S. Don't drop the mailing list from your replies, or no-one else will be able to help you! pgpF3Rv5NRsrz.pgp Description: PGP signature
Re: Snapshot performance
On Fri, Sep 01, 2006 at 11:08:51AM -0700, Paul Lathrop wrote: Hi all, We're working on deploying a new mail server on FreeBSD 6.1-STABLE. One of the major selling points was the ability to take filesystem snapshots in order to make backups from a consistent filesystem on such a high-traffic system. Unfortunately, when I take a snapshot, performance slows to a crawl - to the point where the system stops responding to network requests (ping, SMTP, etc.). Also, the snapshot takes 10-15 minutes to complete. Is this a typical situation? Will I need to schedule downtime for backups in spite of this nifty new feature? Am I doing something wrong? Time depends on the size of the filesystem - but you are correct that snapshots were not designed with performance in mind (rather, to speed up booting after an unclean shutdown by removing the need to wait for fsck). Kris pgpVdsO0XfXbl.pgp Description: PGP signature
Re: newbie. how to compile gcc-4.1.1
On Thu, Aug 31, 2006 at 08:21:04PM -0400, g wrote: hi everyone, i'm trying to compile and install gcc-4.1.1 from the port section. errors: Syntax error: redirection unexpected *** Error code 2 Stop in /usr/home/g/Applications/gcc-4 ... /build-i386-unknown- freebsd6.1/fixincludes. *** Error code 1 i tried making the object code in another location, by mkdir'ing a directory, cd'ing to that directory then running configure, make, after some processing i get the error messages how do i fix this? First, you didn't show us enough of the error. Second, you apparently edited the error so it's hard to tell what is really going on. Can you please post more context? Also tell us what version of FreeBSD you're running. Kris pgpPknz7AOqbo.pgp Description: PGP signature
Re: AMD64 make buildworld failure
On Fri, Aug 25, 2006 at 08:26:54PM -0400, stan wrote: I'm trying to do a buildworld (AMD64) on a Sum Ultra 40, but it's failing like this: cc -O2 -fno-strict-aliasing -pipe -DRRESTORE -DRESCUE -c /usr/src/sbin/restore /restore.c cc -O2 -fno-strict-aliasing -pipe -DRRESTORE -DRESCUE -c /usr/src/sbin/restore /dirs.c /usr/src/sbin/restore/dirs.c: In function `extractdirs': /usr/src/sbin/restore/dirs.c:192: internal compiler error: Segmentation fault: 1 1 Please submit a full bug report, with preprocessed source if appropriate. See URL:http://gcc.gnu.org/bugs.html for instructions. *** Error code 1 Stop in /usr/src/sbin/restore. *** Error code 1 Stop in /usr/obj/usr/src/rescue/rescue. *** Error code 1 This is installed from the 6.1 Release CD, and then cvsup'd Can anyone tell me where I should start looking. That URL, or the FreeBSD FAQ. Kris pgpwSZkdw3gvv.pgp Description: PGP signature
Re: 6.1 crashing, no vm dump
On Tue, Aug 29, 2006 at 05:14:01PM -0500, Laurence Sanford wrote: I'm not much on posting mystery stuff for people to try to solve, but I'm at the end of my rope with this one. I have a box that's been pretty stable until very recently, and I don't know if I may have missed something important that was posted somewhere about a problem, or what. I can reproduce crashes with this system through disk activity though. Example situation: I'm downloading an ISO @ 300+ K/s, and ftp a 3 gig file from this system to another system on my LAN at the same time. It will simply reboot. No error message, nothing in the logs, it just reboots. That typically means marginal or failing hardware. There's lots of discussion of people with similar problems to yours in the archive, but the short answer is: test and/or replace all your hardware components until you find which one it was (start with CPU fan, power supply, cabling, clean out accumulated dust, etc). Kris pgpZKeA5JrSwc.pgp Description: PGP signature
Re: Various package/ports problems
On Mon, Aug 28, 2006 at 01:59:14PM -0700, Dr. Jennifer Nussbaum wrote: I wish i could, but i dont have physical access to the machine - its in a remote location with no serial console. So if anything gets messed up during an upgrade, im in a hell of a lot of trouble. I dont know when i'll be able to fix this problem so until i can upgrade the OS i at least want to keep the rest of it up to date. The packages are compiled for 4.11-stable and do not work with systems much older than this (e.g. 4.8, as you discovered). You might have more luck compiling from the ports collection, although that too only supports 4.11 officially (and modern releases, of course). Kris pgpGbuUs7AIkx.pgp Description: PGP signature
Re: Tuning FreeBSD
On Sat, Aug 19, 2006 at 04:35:29PM -0400, Lisa Casey wrote: Hi Folksm I'm getting these on a FreeBSD 5.3 box: Aug 19 15:46:32 radius sm-mta[12850]: k7JJk1rt012776: SYSERR(UID0): fill_fd: disconnect: cannot open /dev/null: Too many open files in system Aug 19 15:46:40 radius kernel: kern.maxfiles limit exceeded by uid 0, please see tuning(7). Aug 19 15:46:40 radius kernel: kern.maxfiles limit exceeded by uid 0, please see tuning(7). Aug 19 15:46:35 radius sm-mta[12784]: k7JJk7F0012784: SYSERR(UID0): queueup: cannot create queue file ./qfk7JJk 7F0012784, euid=0, fd=-1, fp=0x0: Too many open files in system I checked, and the system shows: radius# sysctl kern.maxfiles kern.maxfiles: 1416 radius# sysctl kern.maxusers kern.maxusers: 43 I googled and read that kern.maxfiles is dependent upon kern.maxusers and that I shou;d raise maxusers which I tried to do via sysctl: radius# sysctl -w kern.maxusers=50 sysctl: oid 'kern.maxusers' is a read only tunable sysctl: Tunable values are set in /boot/loader.conf My /boot/loader.conf is a zero byte length file. How can I fix this too many open files problem? Edit loader.conf and make the specified change, then reboot. See the loader.conf manpage for more. Kris pgpWRgEsbT2Rs.pgp Description: PGP signature
Re: Rebuild Problem
On Fri, Aug 18, 2006 at 06:33:10AM -0700, beno - wrote: Hi; I upgraded my box on the other side of the planet and things went pretty well. I can ssh in and the Web sites serve. However, qmail is down and I tried to nmap and both gave me the same errors from having deleted the old libraries (as someone suggested I do). Here are at least some of the libraries I need to rebuild: libpcap.so.3 ld-elf.so.1 libc.so.5 How do I do this? You can install the compat5x port for now, but beyond the short term you should rebuild those ports so they are 6.x native binaries; after updating to a new major release of FreeBSD you have to rebuild *all* ports to avoid certain problems from a mixed 5.x/6.x system Kris pgpHKdLcQR9nU.pgp Description: PGP signature
Re: GCC - Optimal Optimization
On Fri, Aug 18, 2006 at 09:10:57AM -0700, Sean M. wrote: With GCC 3.4.4, what are the best CFLAGS to use for an AMD Duron ~1000 MHz? By best I mean creating the fastest programs that exploit fully all of the architecture's features, without creating a noticible increase in size. To date I've been using CFLAGS=-O3 -march=athlon-xp -mfp-math=sse -funroll-loops -pipe -ffast-math Don't use random options (e.g. -ffast-math) unless you know what you're doing (yes, it says fast, but that doesn't necessarily mean it will make your application faster). The best options to use are the default ones (you can also set CPUTYPE according to the make.conf documentation) unless you like debugging your system when applications suddenly do not operate correctly. Using nonstandard options like the above might make your system feel sexier, but it's really not a good idea to compile your whole system that way :-) Kris pgpguDEu4MeKU.pgp Description: PGP signature
Re: problem compliling new kernel for SCTP
On Thu, Aug 10, 2006 at 03:16:56PM -0500, [EMAIL PROTECTED] wrote: Iam trying to install SCTP enabled Apache server and Mozilla browser on FreeBSD 6.0 These are things that i have done: 1) I downloaded the kernel patch from sctp.org and applied it. 2) Included the line options SCTP in MY kernel source code at /usr/src/sys after making a copy of GENERIC and called it MYKERNEL 3) ran config to generate kernel source code Problem: this gives error, device mfi and device bce are unknown, so i commented out device mfi and device bce. 3) again ran config to generte kernel source code, which was successful. changed into the /usr/src/compileMYKERNEL. 4) Ran the make depend command. Problem: Now it gives the following error: ERROR(1) ln -s /usr/src/sys/i386/compile/MYKERNEL/opt_bce.h opt_bce.h ln: opt_bce.h: File exists *** Error code 1 Stop in /usr/src/sys/modules/bce. *** Error code 1 Stop in /usr/src/sys/modules. *** Error code 1 Stop in /usr/src/sys/i386/compile/MYKERNEL. 5) Then I copied /usr/src/sys/kern into /usr/include/sys and tried doing make buildworld PROBLEM: This gives the following error: ERROR (2) /usr/src/usr.bin/netstat/mbuf.c: In function `mbpr': /usr/src/usr.bin/netstat/mbuf.c:158: error: `MBUF_JUMBOP_MEM_NAME' undeclared (first use in this function) /usr/src/usr.bin/netstat/mbuf.c:158: error: (Each undeclared identifier is reported only once /usr/src/usr.bin/netstat/mbuf.c:158: error: for each function it appears in.) /usr/src/usr.bin/netstat/mbuf.c:171: error: `MBUF_JUMBO9_MEM_NAME' undeclared (first use in this function) /usr/src/usr.bin/netstat/mbuf.c:184: error: `MBUF_JUMBO16_MEM_NAME' undeclared (first use in this function) *** Error code 1 Stop in /usr/src/usr.bin/netstat. *** Error code 1 Stop in /usr/src/usr.bin. *** Error code 1 Stop in /usr/src. *** Error code 1 Stop in /usr/src. *** Error code 1 Stop in /usr/src. REQUIRE HELP with the following: 1) HOw dO i enable FREEBSD kernel with SCTP? 2) Am i moving in the right direction? Are the steps followed above corect? 3) How do i get rid of errors (1)and (2)? I would really appreciate any kind of assistance. Talk to the author of the SCTP patch. Kris pgpaBMkBHJG56.pgp Description: PGP signature
Re: Once CPU idle at all times
On Thu, Aug 10, 2006 at 06:08:00AM +, Duane Hill wrote: I know this was asked once before. I can't recall if it had received a response or not. If so, tell me to go digging. Otherwise, I have a server that is a dual Zeon 2.8 and I have the hyperthreading enabled within the bios. When I use top, it shows one of the CPU's as apparently not being used. Why is this? 13 root 171 52 0K 8K CPU1 0 0:00 99.85% idle: cpu1 -- This one 11 root 171 52 0K 8K RUN 3 383.1H 61.38% idle: cpu3 14 root 171 52 0K 8K RUN 0 403.4H 61.13% idle: cpu0 12 root 171 52 0K 8K RUN 2 333.4H 49.95% idle: cpu2 Hyperthreading is not enabled in FreeBSD (it's off by default)? On some systems the sysctl for disabling HTT doesn't actually work properly and leaves some of the logical CPUs active. Kris pgpkMCcSjjrz9.pgp Description: PGP signature
Re: pkg_add question
On Wed, Aug 09, 2006 at 01:43:10PM +, V.I.Victor wrote: Generally -- is it OK to do a 'pkg_add' from 'packages-6-stable' to a 5.4 system? Absolutely not. Specifically, I think I need to update 'fetchmail.' What's wrong with using packages-5-stable? :-) What I've read *seems* to indicate it's OK, but... Where did you read this, so we can try to correct the bogus advice? Kris pgpvemMjDxsKZ.pgp Description: PGP signature
Re: Midnight Commander in base distribution set
On Fri, Aug 04, 2006 at 12:41:54PM +0600, Renat S. Nurgaliyev wrote: Hello! Please, please, please, include Midnight Commander into the future releases of FreeBSD! It is extremely time-safing and lightweight tool. It can be installed from ports, but what about disconnected PC's? Thanks a lot. Sorry, this will not happen. Everyone has their favourite tools, and that's what ports are for. Kris pgpB5bjcWZzVC.pgp Description: PGP signature
Re: Possible to go from 5.5 to 5.4?
On Fri, Jun 16, 2006 at 09:39:27PM +0200, K?vesd?n G?bor wrote: Tuc wrote: Hi, I recently upgraded from 5.4 to 5.5, and now if I leave it in X for a few minutes it ends up locking up, usually when the screen blanks on its own. I wanted to go backwards to see if that behaviour goes away. I wasn't sure if there was a risk doing this. Thanks, Tuc Have you tried recompiling X? That might help. Take a look at sysutils/portupgrade, if you haven't used that before. I think going backward is unwise, there must be a better way. Also make sure you recompile any kernel modules you might be using. The most likely one would be the nvidia driver. Kris pgpVlK22lBPGQ.pgp Description: PGP signature
Re: Possible to go from 5.5 to 5.4?
On Fri, Jun 16, 2006 at 04:41:35PM -0400, Scott Tuc Ellentuch at T-B-O-H wrote: On Fri, Jun 16, 2006 at 09:39:27PM +0200, K?vesd?n G?bor wrote: Tuc wrote: Hi, I recently upgraded from 5.4 to 5.5, and now if I leave it in X for= a few minutes it ends up locking up, usually when the screen blanks on its own. I wanted to go backwards to see if that behaviour goes away. I wasn't sure if there was a risk doing this. Thanks, Tuc =20 Have you tried recompiling X? That might help. Take a look at=20 sysutils/portupgrade, if you haven't used that before. I think going=20 backward is unwise, there must be a better way. Also make sure you recompile any kernel modules you might be using. The most likely one would be the nvidia driver. I had upgraded the nvidia one at the same time since I knew it needed to be rebuilt. In what order did you do this? Since you are using the module, it's still the #1 suspect. Kris pgpnZHYfmjWKr.pgp Description: PGP signature
Re: Possible to go from 5.5 to 5.4?
On Fri, Jun 16, 2006 at 05:08:30PM -0400, Scott Tuc Ellentuch at T-B-O-H wrote: 6) portupgrade -Ri of anything that was xorg*, nvidia* If you didn't use -f for nvidia-driver then it may not have rebuilt. Kris pgp09noI5d4Dd.pgp Description: PGP signature
Re: FreeBSD is #1
On Mon, Jun 12, 2006 at 02:36:24PM -0700, Danial Thom wrote: Freebsd 4.x no doubt :) At this point FreeBSD 6.x has had way more intensive stress testing and QA by the project than 4.x did. It is stable. Kris pgp3Ibgo80nS3.pgp Description: PGP signature
Re: accidently deleted /lib/libm.so.4 and /lib/libc.so.6
On Sun, Jun 11, 2006 at 11:21:43AM +0100, Soo-Hyun Choi wrote: Hi, I accidently deleted /lib/libm.so.4 and /lib/libc.so.6 on my FreeBSD 6.1 RELEASE systsem. After this, even 'ls' command is not working. Is there any possibility that I can restore the system? You can use the tools in /rescue to help recover from a backup copy..if you don't have a backup copy accessible then you may have to reinstall. Kris pgp3WtS1oGsZV.pgp Description: PGP signature
Re: portupgrade question (solved)
On Wed, Jun 07, 2006 at 09:29:10PM -0400, Michael S wrote: I did define the to environment variables. For some reason, wget works just fine on the same machine. Different defaults. Thanks a lot. np. Kris pgpbQ76wJJ4Xj.pgp Description: PGP signature
Re: Strange swap behaviour
On Thu, Jun 08, 2006 at 10:07:10AM +0200, Andrea Venturoli wrote: Hello. A server of mine is running 6.1/amd64 with 1GiB of RAM. From time to time swap use will grow enormously and performance will suffer from this; also the nightly clamav run will not complete, as the process will get killed due to lack of swap space. The day after I'll look into top and see this two processes are the largest ones: 672 snort 1 -580 4007M 62344K bpf0 8:53 0.00% snort 669 snort 1 -580 323M 29704K bpf1 1:45 0.00% snort I kill them and start them again; I get: 10295 snort 1 -580 81708K 66432K bpf0 0:01 12.38% snort 10283 snort 1 -580 81704K 66368K bpf0 0:01 5.77% snort However swap utilization will drop from 4095/4096 MiB to 136/4096 MiB. Is this normal? Yes. Does the size figure in top only show physical RAM usage? No, that is what the 'res' field (resident size) shows. Or is only swap space reclaimed after the process gets killed? It's reclaimed when the process releases the memory, which is either when it decides to stop hogging so much VM or when you kill it. Kris pgpxkZ655NdLz.pgp Description: PGP signature
Re: portupgrade question
On Wed, Jun 07, 2006 at 09:38:36AM -0400, Michael S wrote: Good day everyone! I have a portupgrade related question. I am behind a proxy and fetch doesn't appear to cope with it very well. What did you try? fetch works fine with proxies for the rest of us :) Kris pgpQqWxSTgnVY.pgp Description: PGP signature
Re: portupgrade question
On Wed, Jun 07, 2006 at 04:35:12PM -0400, Michael S wrote: wget works like magic. I used the defaults. fetch I guess is the FreeBSD default for fetching files from the web. It was having problems fetching files from FTP sites. No idea what the problem is. Try setting FTP_PASSIVE_MODE. This is set in the default login class, but perhaps you have modified it. Kris On 6/7/06, Kris Kennaway [EMAIL PROTECTED] wrote: On Wed, Jun 07, 2006 at 09:38:36AM -0400, Michael S wrote: Good day everyone! I have a portupgrade related question. I am behind a proxy and fetch doesn't appear to cope with it very well. What did you try? fetch works fine with proxies for the rest of us :) Kris pgpzcb5qHUwRC.pgp Description: PGP signature
Re: PTY's in a FreeBSD Jail
On Tue, Jun 06, 2006 at 12:31:19PM -0400, Kevin Wortman wrote: All, Sometime ago there was a question posed about the error Server refused to allocate pty when trying to ssh into a freebsd jail. It seemed to be answered by someone saying that the command mount_devfs devfs /your/jail/dir/dev needed to enter, which in fact does make the jail start working. However, it was stated in the same posting that this was not considered to be secure. Hence my question, if this is in fact not secure, how can I get my jail properly configured in FreeBSD 6.0 without compromising the security of the box? I ask because I have several boxes currently running in my environment with jails (FreeBSD 4.x) and do not see this command anywhere in the startup scripts yet the jailed environment appears to be working like a champ. I am fairly new to the FreeBSD world so please pardon my ignorance, if I have displayed any. See the jail manpage for how to set up devfs. Kris pgpIFotbZwwNJ.pgp Description: PGP signature
Re: Simultaneous portupgrade in 2 jails - compilation problem?
On Mon, Jun 05, 2006 at 09:21:50AM +0200, Philippe Lang wrote: Hi, I'm running FreeBSD 6.0 / jails as a VPS server. I've just tried doing a portupgrade of mysql-server in 2 jails simultaneously, and I noticed compilations problems. The problem must be due to the fact ports are only available on the base machine, and mounted in every jail, at /usr/ports. Is there a way to avoid that, without having to manage a private /usr/ports in each jail? Set WRKDIRPREFIX to build them outside of /usr/ports. Kris pgpzXjsPfFRwc.pgp Description: PGP signature
Re: Cannot delete stubborn files
On Sun, Jun 04, 2006 at 06:57:27PM -0500, Jack Stone wrote: From: James Long [EMAIL PROTECTED] To: freebsd-questions@freebsd.org, Jack Stone [EMAIL PROTECTED] Subject: Re: Cannot delete stubborn files Date: Sun, 4 Jun 2006 14:57:52 -0700 Date: Sun, 04 Jun 2006 14:42:16 -0500 From: Jack Stone [EMAIL PROTECTED] Subject: Re: Cannot delete stubborn files To: [EMAIL PROTECTED], freebsd-questions@freebsd.org Message-ID: [EMAIL PROTECTED] Content-Type: text/plain; format=flowed From: Ricardo Carvalho [EMAIL PROTECTED] To: Jack Stone [EMAIL PROTECTED] Subject: Re: Cannot delete stubborn files Date: Sun, 4 Jun 2006 15:39:05 -0300 This occurred to me once, I simply booted the system in single user mode and removed the file. Nope, that didn't work either. really strangest undeletes I've ever encounted. THX Jack Have you posted the output of ls -lo (problem filename) ? I have not posted it, but here it is for the 2 files: dr1# ls -lo *.pm -rwxrwxrwx 1 root wheel - 5 Jun 4 13:58 Hostname.pm -rwxrwxrwx 1 root wheel - 0 Jun 6 2004 Syslog.pm HTH Unmount (or drop to single-user mode) and fsck -f the drive to fix the filesystem corruption. Kris pgpiPNNUxs7Uj.pgp Description: PGP signature
Re: ldconfig -m for libs... how about upgrading @INC for perl modules??
On Sun, Jun 04, 2006 at 08:16:41PM -0700, Mark Jayson Alvarez wrote: Hi, After upgrading perl, a lot of applications broke, primarily because the new perl looks into site_perl/5.8.8 however my application modules are still in site_perl/5.8.7 Question: Is there an ldconfig -m sort of thing for this job? My new perl package installation process didn't bother about this one... Any idea? Thanks. /usr/ports/UPDATING Kris pgpixVDllQjEg.pgp Description: PGP signature
Re: pkg_upgrade?
On Wed, May 31, 2006 at 07:51:38AM -0400, Joe wrote: Kris Kennaway wrote: On Tue, May 30, 2006 at 08:12:25PM -0400, Joe wrote: The .ko files have a Nov 3 2005 date, whereas the files in /boot including the kernel directory have a May 6 2006 date. I take it that Nov 3 means 6.0-RELEASE since the announcement was done on Nov 4. So, I guess I'm back to the question of how to do a binary upgrade from 6.0 to 6.1 (and particularly where is this documented). Should I attempt another sysinstall Upgrade? Show me sysctl kern.version kern.version: FreeBSD 6.0-RELEASE #0: Thu Nov 3 09:36:13 UTC 2005 [EMAIL PROTECTED]:/usr/obj/usr/src/sys/GENERIC Yep, it's a 6.0-RELEASE kernel. I noticed that you didn't actually confirm whether /boot/kernel/kernel has the right date - only the files in /boot including the kernel directory. Anyway, retry the binary upgrade as you say. Kris pgp3JWEq4ElCF.pgp Description: PGP signature
Re: pkg_upgrade?
On Wed, May 31, 2006 at 05:05:28PM -0400, Joe wrote: Kris Kennaway wrote: Anyway, retry the binary upgrade as you say. The binary upgrade started from the 6.0-RELEASE CD-ROM didn't work Wait, you were trying to update to 6.1-RELEASE using the 6.0-RELEASE CD-ROM? :) Kris pgpiCq50hC5hV.pgp Description: PGP signature
Re: pkg_upgrade?
On Wed, May 31, 2006 at 10:23:46PM -0400, Joe wrote: Kris Kennaway wrote: On Wed, May 31, 2006 at 05:05:28PM -0400, Joe wrote: Kris Kennaway wrote: Anyway, retry the binary upgrade as you say. The binary upgrade started from the 6.0-RELEASE CD-ROM didn't work Wait, you were trying to update to 6.1-RELEASE using the 6.0-RELEASE CD-ROM? :) Yes, I mentioned that a couple of times before. However, in the options screen, I requested the *6.1-RELEASE* to be fetched, and it did do that--except for the kernel, or actually it seems it fetched that too but it was unable to move it into place. I'm fairly knew at FreeBSD (but definitely not at software) and I don't see why I shouldn't be able to upgrade an OS starting from the earlier release, i.e., without having to bootstrap from the new release, particularly when upgrading within the same major release, from one minor to the next. The necessary steps for upgrading from an old release to the current one sometimes change. The old release doesn't know all of the future upgrading procedures for subsequent releases. That's almost certainly what went on here. Kris pgpdFCUxW8gJd.pgp Description: PGP signature
Re: Illegal instruction while portupgrade
On Thu, Jun 01, 2006 at 10:45:40AM +0800, snnn wrote: Bj?rn K?nig wrote: I don't know exactly who says Illegal instruction in your case, but I would take it as a hint that you try to execute code that doesn't work with your CPU. Did you compiled ruby yourself and used special CFLAGS? What kind of CPU do you use? yes,I've installed every package from ports,so ruby is compiled by myself also. This is the my make.conf WITH_SETUID_LUIT=true SUPHOST=cvsup.FreeBSD.org CPUTYPE=athlon-xp MAKE_IDE=true ENABLE_SUID_SSH=true NO_BLUETOOTH=true DOC_LANG=zh_CN APACHE2=true CJK=true CUPS=false DVD=true A lot of these are invalid, i.e. not the correct name. Is it because of optimizing of AlthonXP ? Probably, are you sure that is the CPU you have? Kris pgpMlO9rfbh3v.pgp Description: PGP signature
Re: pkg_upgrade?
On Tue, May 30, 2006 at 04:51:51PM -0400, Joe wrote: Kris Kennaway wrote: On Mon, May 29, 2006 at 07:42:43PM -0400, Joe wrote: Hi, I've read the documentation and it seems there's no pkg_upgrade or pkg_update, or a way to install an updated/upgraded package. I'd like to determine if that is indeed the case. portupgrade -P or -PP OK, since I had upgraded to 6.1-RELEASE, I used portsnap to get the ports, then I used pkg_add -r to get portupgrade and then, as a test ran portupgrade -PP expat It failed to update expat however, because it kept looking in the 6.0-RELEASE paths. I presume this is because even though the binary upgrade of the base to 6.1 went well, uname, etc., think I still have a 6.0 machine. So where is the *real* version id stored and how can it be (should it be?) safely modified? It's reported by the kernel, so if it's still saying 6.0-RELEASE then that's what you're still running. Kris pgpoB5bPfeXeZ.pgp Description: PGP signature
Re: pkg_upgrade?
On Tue, May 30, 2006 at 06:14:32PM -0400, Joe wrote: Kris Kennaway wrote: On Tue, May 30, 2006 at 04:51:51PM -0400, Joe wrote: [...] It failed to update expat however, because it kept looking in the 6.0-RELEASE paths. I presume this is because even though the binary upgrade of the base to 6.1 went well, uname, etc., think I still have a 6.0 machine. So where is the *real* version id stored and how can it be (should it be?) safely modified? It's reported by the kernel, so if it's still saying 6.0-RELEASE then that's what you're still running. I don't have access to the system right now, but after the sysinstall I looked at some of the files in /bin and saw that most had a date of May 6, 2006 (or thereabouts) so I assumed that meant the binary upgrade had been successful. How can I verify that the actual kernel was upgraded? Also, I haven't been able to locate much information about a binary upgrade in the Handbook. The only reference was in the INSTALL.htm in the release directory. ls -l /boot/kernel Kris pgpBXKG8r2A33.pgp Description: PGP signature
Re: pkg_upgrade?
On Tue, May 30, 2006 at 08:12:25PM -0400, Joe wrote: Kris Kennaway wrote: On Tue, May 30, 2006 at 06:14:32PM -0400, Joe wrote: I don't have access to the system right now, but after the sysinstall I looked at some of the files in /bin and saw that most had a date of May 6, 2006 (or thereabouts) so I assumed that meant the binary upgrade had been successful. How can I verify that the actual kernel was upgraded? Also, I haven't been able to locate much information about a binary upgrade in the Handbook. The only reference was in the INSTALL.htm in the release directory. ls -l /boot/kernel The .ko files have a Nov 3 2005 date, whereas the files in /boot including the kernel directory have a May 6 2006 date. I take it that Nov 3 means 6.0-RELEASE since the announcement was done on Nov 4. So, I guess I'm back to the question of how to do a binary upgrade from 6.0 to 6.1 (and particularly where is this documented). Should I attempt another sysinstall Upgrade? Show me sysctl kern.version and the output of a failed package fetch. Kris pgpGt2JJlb4uC.pgp Description: PGP signature
Re: pkg_upgrade?
On Mon, May 29, 2006 at 07:42:43PM -0400, Joe wrote: Hi, I've read the documentation and it seems there's no pkg_upgrade or pkg_update, or a way to install an updated/upgraded package. I'd like to determine if that is indeed the case. portupgrade -P or -PP Kris pgpGNkBQGy4Ip.pgp Description: PGP signature
Re: 6.1-RELEASE for PPC ?
On Mon, May 29, 2006 at 07:47:32PM -0700, Jason Self wrote: I've read the announcement about FreeBSD 6.1-RELEASE from http://www.freebsd.org/releases/6.1R/announce.html and decided to download it. FreeBSD 6.1-RELEASE supports the i386, pc98, alpha, sparc64, amd64, powerpc, and ia64 architectures... But... 6.1 doesn't seem to exist for PPC??? Ask on the PPC list. Probably the PPC project needs additional developers though. This would be a great opportunity for someone who cares about FreeBSD/PPC to step up and become involved. Kris pgpmPgrSNcuxC.pgp Description: PGP signature
Re: Strange pauses and cpu usage - FreeBSD 6.1
On Sun, May 28, 2006 at 02:18:36AM +0200, Xavier Chantry wrote: On Sat, May 27, 2006 at 01:14:13AM +0100, Shawn wrote: I'm having a strange problem with cpu usage with my Athlon XP. Every so often there seems to be a pause where the system comes under heavy load for no apparent reason. If xmms was playing then I get a buzz sound and the mouse pointer stops just for a split second, the same with mplayer and vlc. Using Opera or firefox also seems to have a strange effect where the mouse pointer freezes while pages load, and if there is an animated gif on the page the cpu usage goes to 100%. The jabber client Gajim also makes the cpu get stuck at 100%. When I used gnome, nautilus did the same thing until i stopped and started the process in the system monitor. I have searched around and found this tip from the handbook but it had no effect. hint.apic.0.disabled=1 in /boot/loader.conf This happens with the generic kernel and the one i compiled for myself. It also lasts for a random amount of time, sometimes it is barely noticeable and sometimes it happens for over a second. Running neverwinter nights makes the system pause for 10-20 seconds seemingly randomly. My system behaves much better than that, I've no mouse freeze and no system pauses, but I did have occasonial sound skips using 4bsd on 6.1. Switching to the ULE scheduler seems to fix this problem, but other things may behave worse. One year ago, it was actually the opposite, sound in ULE was really jerky, and mouse pointer too while eg launching an app. Did you already try both schedulers ? Check for interrupt sharing with vmstat -i. Shared interrupts cause performance loss especially if it's sharing with a giant-locked driver like USB. Kris pgpXJ7A2ZEJuR.pgp Description: PGP signature
Re: LIB_DEPENDS problem
On Thu, May 25, 2006 at 07:39:49AM +, db wrote: Kris Kennaway wrote: #LIB_DEPENDS= cppe:${PORTSDIR}/devel/cppe \ (also a port not in the collection) LIB_DEPENDS=ACE_SSL:${PORTSDIR}/devel/ace \ pqxx:${PORTSDIR}/databases/postgresql-libpqxx Grrr..this is your fault since that's not the library name installed by that port. What? work# more /usr/ports/databases/postgresql-libpqxx/pkg-plist | grep libpqxx.so lib/libpqxx.so That's the symlink; your grep is making the same incorrect assumptions about what to look for: lib/libpqxx.a lib/libpqxx.la lib/libpqxx.so lib/libpqxx-%%PORTVERSION%%.so Kris pgpii3SRRLEvk.pgp Description: PGP signature
Re: Constant segmentation faults with freebsd 6.1-release and -stable
On Thu, May 25, 2006 at 12:50:47AM -0700, Aaron Holmes wrote: Am i the only one who gets this with most port installations? I'm trying to narrow it down to a hardware fault or a 6x fault. I get signal 4's and segmentation faults on nearly every compilation any ideas? Yep, hardware. Kris pgp3vnkH5dbv0.pgp Description: PGP signature
Re: Rebuilding /var/db/pkg
On Thu, May 25, 2006 at 05:55:55PM -0500, Robertsen A. Riehle wrote: Say that the /var/db/pkg directory had been recursively erased off of a workstation that had ~300 packages on it. And, let's hypothetically say that this workstation's ports tree was up to date as of yesterday. Is there any hope of rectifying this or is this workstation is a static ports state forever??? There is not (unless you go through by hand and figure out what was installed), this is why backups are necessary. Kris pgpw2GLSnsJS8.pgp Description: PGP signature
Re: LIB_DEPENDS problem
On Wed, May 24, 2006 at 03:00:10PM +, db wrote: libpqxx.so is not a broken link: work# ls -l /usr/local/lib/*pqxx* -rwxr-xr-x 1 root wheel 332194 May 4 14:56 /usr/local/lib/libpqxx-2.5.5.so -rw-r--r-- 1 root wheel 439800 May 4 14:56 /usr/local/lib/libpqxx.a -rwxr-xr-x 1 root wheel 851 May 4 14:56 /usr/local/lib/libpqxx.la lrwxr-xr-x 1 root wheel 16 May 4 14:56 /usr/local/lib/libpqxx.so - libpqxx-2.5.5.so work# ldconfig -r | grep pqxx work# I don't know why it might not be known to ldconfig, but that is the cause of your problem. The port DTRT to update ldconfig at install time: INSTALLS_SHLIB= yes Try rerunning ldconfig -R (or reboot). Kris pgpKrzF8OgpLf.pgp Description: PGP signature
Re: LIB_DEPENDS problem
On Wed, May 24, 2006 at 08:56:44PM +, db wrote: Kris Kennaway wrote: work# ls -l /usr/local/lib/*pqxx* -rwxr-xr-x 1 root wheel 332194 May 4 14:56 /usr/local/lib/libpqxx-2.5.5.so -rw-r--r-- 1 root wheel 439800 May 4 14:56 /usr/local/lib/libpqxx.a -rwxr-xr-x 1 root wheel 851 May 4 14:56 /usr/local/lib/libpqxx.la lrwxr-xr-x 1 root wheel 16 May 4 14:56 /usr/local/lib/libpqxx.so - libpqxx-2.5.5.so work# ldconfig -r | grep pqxx work# Try rerunning ldconfig -R (or reboot). I tried (both), but no luck :-( Reinstall the port then, something seems to be wrong with it (corrupt library?). Kris pgpjvEMNomiKV.pgp Description: PGP signature
Re: LIB_DEPENDS problem
On Wed, May 24, 2006 at 09:11:58PM +, db wrote: Kris Kennaway wrote: work# ls -l /usr/local/lib/*pqxx* -rwxr-xr-x 1 root wheel 332194 May 4 14:56 /usr/local/lib/libpqxx-2.5.5.so -rw-r--r-- 1 root wheel 439800 May 4 14:56 /usr/local/lib/libpqxx.a -rwxr-xr-x 1 root wheel 851 May 4 14:56 /usr/local/lib/libpqxx.la lrwxr-xr-x 1 root wheel 16 May 4 14:56 /usr/local/lib/libpqxx.so - libpqxx-2.5.5.so work# ldconfig -r | grep pqxx work# Try rerunning ldconfig -R (or reboot). I tried (both), but no luck :-( Reinstall the port then, something seems to be wrong with it (corrupt library?). Ok tried that and still no luck :-( What port are you having problems with again? Kris pgpHt0REyn1Zu.pgp Description: PGP signature
Re: LIB_DEPENDS problem
On Wed, May 24, 2006 at 09:46:18PM +, db wrote: Kris Kennaway wrote: What port are you having problems with again? It's not in the port collection yet: PORTNAME= esad PORTVERSION=0.1 CATEGORIES= net MASTER_SITES= http://esad.trunet.dk/ #no you can't download anything until it is ready MAINTAINER= [EMAIL PROTECTED] COMMENT=Multi-service framework INSTALLS_SHLIB= yes #LIB_DEPENDS= cppe:${PORTSDIR}/devel/cppe \ (also a port not in the collection) LIB_DEPENDS=ACE_SSL:${PORTSDIR}/devel/ace \ pqxx:${PORTSDIR}/databases/postgresql-libpqxx Grrr..this is your fault since that's not the library name installed by that port. Kris pgpKKqy9p2Bt7.pgp Description: PGP signature
Re: LIB_DEPENDS problem
On Tue, May 23, 2006 at 02:37:12PM +, db wrote: Notice that it didn't build this port, but skipped straight to installing it. This means that you have a stale build of the port, and need to run 'make clean' before and/or after your builds as a matter of habit. work# ls /usr/ports/databases/postgresql-libpqxx Makefiledistinfofiles pkg-descr pkg-plist work# make -DNO_CHECKSUM === WARNING: Vulnerability database out of date, checking anyway === Extracting for esad-0.1 === Patching for esad-0.1 === esad-0.1 depends on shared library: ACE_SSL - found === esad-0.1 depends on shared library: pqxx - not found ===Verifying install for pqxx in /usr/ports/databases/postgresql-libpqxx === WARNING: Vulnerability database out of date, checking anyway === Extracting for postgresql-libpqxx-2.5.5 === Patching for postgresql-libpqxx-2.5.5 === Applying FreeBSD patches for postgresql-libpqxx-2.5.5 === postgresql-libpqxx-2.5.5 depends on executable: gmake - found === postgresql-libpqxx-2.5.5 depends on shared library: pq.3 - found === Configuring for postgresql-libpqxx-2.5.5 ^C libpqxx is installed, but it can't find it and try to build and install it again. Why do you say it is installed? You certainly didn't show this above; that is the port skeleton used for building the software, not an installed copy of the software. Kris pgp9t7Ex79Bbw.pgp Description: PGP signature
Re: X11R7 through ports?
On Tue, May 23, 2006 at 10:05:37PM -0400, Jim Stapleton wrote: Is there any way to get X11R7 through ports? The Xorg and XFree86 packages all seem to be 6.8/6.9 Xorg is 6.9, which is functionally identical to 7.0. See the xorg webpage. Kris pgp0dsJJEFkjE.pgp Description: PGP signature
Re: NO_RESCUE option in make.conf
On Mon, May 22, 2006 at 03:18:59PM +0100, Vittorio wrote: To update the system in my /etc/make.conf among other building(-world) options I put a: NO_RESCUE= true Which actually is not declared neither in the man page of make.conf nor in the /usr/share/examples/etc/make.conf file. Nonetheless it seems to work. Am I right? Why do you want to do this? The rescue tools are sometimes all that stands between being able to repair your system and having to reinstall it from scratch. Kris pgplP3d0f7SkO.pgp Description: PGP signature
Re: LIB_DEPENDS problem
On Mon, May 22, 2006 at 02:57:28PM +, db wrote: Hi all I'm trying to make a new port that depends on ACE and pqxx, but it can't find lib pqxx even though it is installed. work# make -DNO_CHECKSUM === WARNING: Vulnerability database out of date, checking anyway === Extracting for esad-0.1 === Patching for esad-0.1 === esad-0.1 depends on shared library: ACE_SSL - found === esad-0.1 depends on shared library: pqxx - not found ===Verifying install for pqxx in /usr/ports/databases/postgresql-libpqxx Notice that it didn't build this port, but skipped straight to installing it. This means that you have a stale build of the port, and need to run 'make clean' before and/or after your builds as a matter of habit. Kris pgplbEOVywjuP.pgp Description: PGP signature
Re: Locale problem after upgrade 5.4 to 6.1
On Mon, May 22, 2006 at 07:52:56PM -0200, Rafael Aquino wrote: Hello ALL, I use to set all my systems to pt_BR.ISO8859-1, and everytime works perfect. The ports that are sensible for locale always uses the correct language (Apache, Postgresql, etc). After upgrate from 5.4 to 6.1, I realize that postgresql wasn?t starting up with rc.d script. Trying to mannualy start it, the message was that my database couldn?t support LATIN-1 (witch was, before upgrade). Also the characters that used to work in console (?, ?, ...) don?t work anymore. Something changed with locale support between 5.x and 6.x; check the release notes, UPDATING, etc. There's also a compatibility package. Kris pgpmpFE1WzHbm.pgp Description: PGP signature
Re: print-cdrom-packages.sh??
On Mon, May 15, 2006 at 01:01:23PM +0100, Howard Jones wrote: Moving on with my automated installs, I want to build some additional packages into the FTP area of my install server, which was seeded from the 6.1-RELEASE ISOs. The FreeBSD Release Engineering for Third Party Software Packages document on the FreeBSD website says that the release-building process uses a script called print-cdrom-packages.sh to produce the list of packages to build. However, my source tree doesn't have this script, and a look in CVS suggests that it's no longer in use, and hasn't been since March 2005 (Revision 1.71, Mon Mar 21 19:19:24 2005 UTC - Retire print-cdrom-packages.sh as it has been replaced.). Is there any current documentation for how the package building process is done? Or am I looking at the wrong URL? http://www.freebsd.org/doc/en_US.ISO8859-1/articles/releng-packages/article.html Ultimately, I'll want to rebuild all the packages, so just getting the few extras I need right now doesn't really help. I intend to use 'make release' to keep a 6_RELENG install with fairly current packages available for new servers, so I don't need to do so much rebuilding immediately after a fresh install. Thanks in advance for any pointers, or even any don't do that, do this instead-type advice. There's now a python script somewhere in release/ that replaced it and does a better job. It should be easy to find, and once you have it can you submit a PR requesting the article be updated? Kris pgpYGsAj7Dase.pgp Description: PGP signature
Re: foomatic-db portupgrade problem
On Sun, May 21, 2006 at 06:57:43PM -0500, Denny White wrote: Searched through the cvs stable mailing lists, as well as in /usr/ports/UPDATING, couldn't find any instance of anyone having problems updating the foomatic-db port. After the initial failure to update, I looked at the Makefile that seems to be the problem, where it's trying to download from. Below is the output of portversion -l along with the output when trying to update, portupgrade -arR. And, the pertinent section of the Makefile is appended also: Yes, it's broken (see http://pointyhat.freebsd.org) Talk to the authors about fixing their CVS server. Kris pgphj7bewxsdc.pgp Description: PGP signature
Re: kernel panic on kldload in 6.1
On Thu, May 18, 2006 at 10:54:19PM -0400, Joe Auty wrote: On May 18, 2006, at 12:36 AM, Kris Kennaway wrote: On Wed, May 17, 2006 at 07:06:59PM -0400, Joe Auty wrote: Hello, A problem I've had since 6.0 remains in 6.1 where my machine kernel panics at boot trying to kldload some module (it doesn't say which one). It also does not log this panic. I suppose this isn't terribly helpful information, but if there is anything you can suggest I do (including pass this on to an interested party), please let me know how I can help =) It sounds like you have an old (6.0) module you're still trying to load. Things like the nvidia driver and other ports are prime candidates. Hi, I tried commenting out all of the kernel modules in /boot/ loader.conf, and I'm still panicing. I wrote down some of the panic message this time. The code is superviser read, page not present the panic output is: panic: page fault Unless you guys have any ideas of things I can try, I'll just file this. Any ideas? So you claimed originally that it was due to kldloading some module, but now you're still getting the same panic when you're not kldloading anything? Can you please clarify exactly what is the situation? I suspect you're still actually kldloading something (modules are loaded on demand for some things): did you boot into single-user mode and confirm that all modules are up-to-date? Kris pgpm1YMEwUxXq.pgp Description: PGP signature
Re: How to install freebsd6 on alpha server ?
On Fri, May 19, 2006 at 06:32:48PM +0300, Halid Faith wrote: Hello I could not install Freebsd6 to a compaq alpha server. in fact I had installed freebsd4.4 before the same machine. But this time I got an error as below at the beginning ; fatal kernel trap memory management fault. after above error message, the server goes reboot. Does anyone has any advice ? Start with the freebsd-alpha mailing list. Kris pgpaqPZYyAwzd.pgp Description: PGP signature
Re: broken port with no makefile?
On Fri, May 19, 2006 at 02:00:01PM -0400, [EMAIL PROTECTED] wrote: I've been upgrading to the latest Apache PHP versions, but seem to be unable to get PHP to work again. The crux of the problem is that mod_php5 has no makefile and seems to be eternally stuck building, even after updating the whole collection: $ su [prompt edited out] $ cvsup /root/ports-supfile Connected to cvsup18.us.FreeBSD.org Updating collection ports-all/cvs Finished successfully $ cd /usr/ports/www/mod_php5; ls -aFl total 30 drwxr-xr-x 3 root wheel512 May 18 05:07 ./ drwxr-xr-x 1056 root wheel 25600 May 19 01:20 ../ drwxr-xr-x 3 root wheel512 Apr 16 13:28 work/ The port was removed but you still had a stale work directory there. Just delete it. Kris pgpDBtqBisEfO.pgp Description: PGP signature
Re: kernel panic on kldload in 6.1
On Fri, May 19, 2006 at 04:59:44PM -0400, Joe Auty wrote: On May 19, 2006, at 2:38 PM, Kris Kennaway wrote: On Thu, May 18, 2006 at 10:54:19PM -0400, Joe Auty wrote: On May 18, 2006, at 12:36 AM, Kris Kennaway wrote: On Wed, May 17, 2006 at 07:06:59PM -0400, Joe Auty wrote: Hello, A problem I've had since 6.0 remains in 6.1 where my machine kernel panics at boot trying to kldload some module (it doesn't say which one). It also does not log this panic. I suppose this isn't terribly helpful information, but if there is anything you can suggest I do (including pass this on to an interested party), please let me know how I can help =) It sounds like you have an old (6.0) module you're still trying to load. Things like the nvidia driver and other ports are prime candidates. Hi, I tried commenting out all of the kernel modules in /boot/ loader.conf, and I'm still panicing. I wrote down some of the panic message this time. The code is superviser read, page not present the panic output is: panic: page fault Unless you guys have any ideas of things I can try, I'll just file this. Any ideas? So you claimed originally that it was due to kldloading some module, but now you're still getting the same panic when you're not kldloading anything? Can you please clarify exactly what is the situation? I suspect you're still actually kldloading something (modules are loaded on demand for some things): did you boot into single-user mode and confirm that all modules are up-to-date? I'm getting this error message during boot time, not during manual loading of modules. As I said, some modules are loaded on demand. Not when booting single-user though. I'm getting this error message despite disabling all my third-party modules listed in /boot/loader.conf. I'm assuming that all modules stored in /boot/kernel are system modules, and it would be expected that they would all be able to run without kernel panicing during boot? Sorry for my lack of clarity! Where I became unclear was probably in saying some module. During boot time, once this panic is spewed onto my screen, it doesn't indicate which (system) module it is choking on. The rest of the panic message would have given more details though. This panic occurred in single user mode under 6.0 too, I haven't tried single user mode in 6.1 yet Kris pgpXjY9PVxWY4.pgp Description: PGP signature
Re: kernel panic on kldload in 6.1
On Thu, May 18, 2006 at 10:57:51AM +0200, Frank Steinborn wrote: Kris Kennaway wrote: It sounds like you have an old (6.0) module you're still trying to load. Things like the nvidia driver and other ports are prime candidates. Kris Does that mean that all the panic-issues with kld(un)load are fixed in 6.1? I crashed an important 6.0-STABLE box several times by trying to unload procfs for example and I don't want to run in such nightmares again by fiddling with klds unless I'm sure it's safe now. No, it means that modules are not cross-compatible between releases. Please retest your problem with 6.1 and file a PR if you encounter it again; it's the only way it will get fixed. Kris pgpZEQzA6e4Uv.pgp Description: PGP signature
Re: ELF interpreter /libexec/ld-elf.so.1 not found
On Thu, May 18, 2006 at 04:02:45PM -0400, David Banning wrote: I am getting this error when attempting to start jboss4. Google doesn't seem to bring up anything on this. My system is 4.11-STABLE. I noticed that ld-elf.so.1 -does- exist in /usr/libexec/ld-elf.so.1 so one thought I had was just symlinking to it. Any direction would be helpful. You are trying to run a non-FreeBSD 4.x binary. Kris pgpUMTVJg2EzI.pgp Description: PGP signature
Re: libm.so.4 not found
On Wed, May 17, 2006 at 08:22:03AM -0500, Derek Ragona wrote: Sounds like you recently upgraded but don't have the compatible libraries in /etc/make.conf make sure you have: COMPAT4X= yes And then rebuild the world. You might also want to have: options COMPAT_FREEBSD4 In your kernel config file. The real question is why do you have binaries depending on old system libraries. The answer is probably that you upgraded from a previous branch of FreeBSD (e.g. 4.x - 6.x) but neglected to rebuild all your installed ports, so they now have an inconsistent mix of old and new libraries. portupgrade -af or similar to repair the damage. Kris pgpa3skdUyw0p.pgp Description: PGP signature
Re: Port marked as IGNORE
On Wed, May 17, 2006 at 09:38:51AM -0500, Cody Holland wrote: I was updating the ports on one of my servers and received this error: [Updating the pkgdb format:bdb1_btree in /var/db/pkg ... - 68 packages found (-1 +1) (...). done] ** Port marked as IGNORE: graphics/php4-ming: is marked as broken: doesn't work with new version of ming library What actions, if any do I need to take to fix this? Make it work with the new version of ming library, or work with the authors or maintainers to do so. Kris pgpprJH1isSu9.pgp Description: PGP signature
Re: makeworld FAILURE on 5.4-STABLE
On Wed, May 17, 2006 at 07:25:50PM +0200, Kyrre Nygard wrote: At 12:29 17.05.2006, David Stanford wrote: Maybe I'm confused as to what you're looking to do. If you're looking to copy data from (or all of) /home to /var, it obviously won't be able to hold anymore than 248MB; and it seems like you have much more data than that. And avoiding the /usr slice won't help with upgrading as you will need to reinstall a new /usr slice anyway using the 6.1-RELEASE disc. Much of the system is located in /usr... If you have more than 248MB worth of data you need to save, and upgrading is absolutely necessary, I would suggest just ponying up the $40.00 and getting an external hard drive to back up the data. Then do a fresh install. I don't know where /var got into the picture, but what I'm trying to do is to reinstall FreeBSD while keeping /usr/home/awad on ad4s1f intact. All files and folders except /usr/home/awad I will have deleted manually. I wonder if this is possible ... Just do an upgrade install from sysinstall, this has been supported for many years. Kris pgpj6qHDbrRbL.pgp Description: PGP signature
Re: kernel panic on kldload in 6.1
On Wed, May 17, 2006 at 07:06:59PM -0400, Joe Auty wrote: Hello, A problem I've had since 6.0 remains in 6.1 where my machine kernel panics at boot trying to kldload some module (it doesn't say which one). It also does not log this panic. I suppose this isn't terribly helpful information, but if there is anything you can suggest I do (including pass this on to an interested party), please let me know how I can help =) It sounds like you have an old (6.0) module you're still trying to load. Things like the nvidia driver and other ports are prime candidates. Kris pgpYkbmbQlDEV.pgp Description: PGP signature
Re: FreeBSD 6.1 - nullfs not working with PAE
On Wed, May 10, 2006 at 07:10:53PM +0200, tonix (Antonio Nati) wrote: I'm trying to setup a system with 8GB of RAM, so I've enabled PAE option. But when I try using a mount_nullfs, I always have this error: mount_nullsfs: Operation not supported by device No problem arises when I compile kernel without PAE. I cannot find any documentation about nullfs and PAE. Why is a so simple filesystem (nullfs) incompatible with PAE? Are you using nullfs from a module? Modules are incompatible between PAE and !PAE, so the easiest thing to do is compile them all into your kernel. Kris pgp4FPywghTTq.pgp Description: PGP signature
Re: panic: userret: Returning with 1 locks held.
On Sun, May 14, 2006 at 11:35:01PM -0700, Malachi de ?lfweald wrote: Has anyone found a solution to the problem reported here: http://marc2.theaimsgroup.com/?l=freebsd-amd64m=111219607318603w=2 From what I can tell, unionfs is compiled into my kernel (said YES in the kernel config file)... I can consistently cause this to happen within a couple seconds if I use the man pages from within a jail that is on unionfs. Don't do that then (see the manpage). The patches floating around may help you, but maybe not. Kris pgpw38CFrQ3ec.pgp Description: PGP signature
Re: upgrade 5.4 - 6.1
On Sun, May 14, 2006 at 07:45:10PM +0300, Perttu Laine wrote: Yeah. I will of course read UPDATING, but I was wondering if there is some major changes or something so I would not go to 6.1 and go with 5.5 instead. But there is nothing big? Well, a new branch of FreeBSD always brings fairly big changes. The ones you need to know about to take special steps when updating are listed in the file above. Kris P.S. Don't top-post, thanks pgp2GjJzSuEIH.pgp Description: PGP signature
Re: limitations on mount_cd9660?
On Sat, May 13, 2006 at 03:39:07PM -0500, Jonathan Horne wrote: ive been trying to mount up my iso files so i can have access to their contents as quickly as possible. ive used mdconfig to create block devices for 4 .iso files, but when i get to the point where i try to mount the 3rd one, i get: mount_cd9660: /dev/md2: Invalid argument i created my block devices like this: mdconfig -a -t vnode -f /path/filename -u 0 (-u 1, -u 2, -u 3, etc etc) all 4 devices created without any complaint, and they show up when i 'ls -la /dev/md*'. are we not able to mount more than 2 .iso files to folders at a time, or am i missing something here? You missed showing us the exact commands you typed, in sequence. Perhaps you made a mistake. Kris pgptrZUQlsv2t.pgp Description: PGP signature
Re: Swap Performance in 6.1?
On Wed, May 10, 2006 at 04:36:15PM -0700, Timothy Beyer wrote: I am curious if this problem has been resolved in 6.1: (it was previously on the FreeBSD 6.1 Open Issues page, but I don't see it there any more) http://www.tr.freebsd.org/releases/6.1R/todo.html swapping on 6.0 is slower than on 4.x Not done Performance on swap handling is much slower than 4.x and this can make a system essentially unusable when moderate paging activity is going on Has this been resolved, (I did not see anything about it in the 6.1 release notes, or the errata) or has it been deferred to a future release? (The reason why I ask this is that programs like Firefox and Xemacs use up all of my 1 GB of memory and I have noticed problems of this nature in 6.0) The TODO item should have been worded better. There is anecdotal evidence that it is slower, but no-one has actually measured it definitively. Therefore it will be deferred until the future, after someone has actual numbers showing whether/how much slower it is, or not. Kris pgpFcIrV4mnuZ.pgp Description: PGP signature
Re: when 5.5
On Thu, May 11, 2006 at 05:47:07PM +0900, Kouji Ito wrote: Hello, I wait very much,too, but it is not readily released. If there is a detailed person for circumstances in 5.5-RELEASE situation, please teach it. Next week or so, now that 6.1 is out. Kris pgpMvT0nOLpkZ.pgp Description: PGP signature
Re: when 5.5
On Thu, May 11, 2006 at 04:26:04PM +0100, Jeff Rollin wrote: Are maintenance releases of old branches always released in step with new ones, or is it just coincidence that it's happening this way this time? It was deliberate this time to avoid an extra ports freeze cycle, since that has very high cost and the benefit is really low since the intended audience for 5.5 is small and everyone who can do so will be much better off with the enormous benefits of 6. Kris pgpcQSjaTyxBS.pgp Description: PGP signature
Re: Question on NFS performance
On Wed, May 10, 2006 at 02:54:39PM +0200, Valerio daelli wrote: Hi all we have a FreeBSD 5.4 exporting some NFS filesystem to a cluster of gentoo boxes (kernel 2.6.12). Our exported storage disk is an Apple XRaid. We have Gigabit Ethernet both on the client and the server. We would like to improve our read performance. This is our performance: about 10Mb reading a file of 1Gb with dd and iozone confirms this result. We already use the normal optimization flags (we use rpc.lockd and rpc.statd, on the client we have read size 65536 and a read ahead of 4 blocks, the async options). Is our performance the best we can get? Can we improve it? 5.4 and filesystem performance cannot be said together in the same sentence. Upgrade to 6.1. Kris pgpndSMng4P0q.pgp Description: PGP signature
Re: New freeBSD logo on freebsd.org
On Wed, May 10, 2006 at 05:47:44PM +, cpghost wrote: On Wed, May 10, 2006 at 05:28:55PM +0200, Erik Trulsson wrote: Many professional logo design people never even knew about the contest to redesign the logo because it was not posted to the questions list where everyone reads. Not everyone reads the questions list. Heck, not even all the developers do, due to the low signal/noise ratio here. Uh-oh! A mindset like this would explain a lot of things. :-( Fortunately, there *are* developers here who do take users' questions seriously. You missed the point. Signal = questions from users about FreeBSD technical support. Noise = lots of whining about the logo. On the plus side, if you guys keep it up I'll be able to dramatically improve my view of the S/N ratio by adding those with nothing better to contribute than their indignation to my killfile (some are there already) so I *can* focus on the user questions. Kris pgp6mSGETyEC7.pgp Description: PGP signature
Re: details about EOL (of FreeBSD 5.4) ?
On Tue, May 09, 2006 at 11:20:21AM +0200, albi wrote: i'm running 5.4 on a few servers, and i wondered till when security-updates will be available http://www.freebsd.org/security/ says that the EOL is estimated may 31 2006, but does that also mean no more security-updates for 5.4 ? Yes. i have no problem with doing a reinstall with the brand new 6.1 (or is the upgrade-jump from 5.4 to 6.1 no problem at all ?) , but i just like to get an idea when that would be really needed :) 5.4 to 6.x is pretty straightforward. The only thing to watch out for is to recompile all your ports (which you have to do when crossing any major revision boundary). portupgrade -fa or -faPP makes this pretty easy though. Kris pgpLKyvv2CVFD.pgp Description: PGP signature
Re: details about EOL (of FreeBSD 5.4) ?
On Tue, May 09, 2006 at 10:40:06AM -0500, Greg Barniskis wrote: Kris Kennaway wrote: On Tue, May 09, 2006 at 11:20:21AM +0200, albi wrote: i'm running 5.4 on a few servers, and i wondered till when security-updates will be available http://www.freebsd.org/security/ says that the EOL is estimated may 31 2006, but does that also mean no more security-updates for 5.4 ? Yes. [blink blink] I'm not at all opposed to RE@ and security@ making hard decisions and doing whatever is needed to best further the project's goals (so please don't read this as flame bait). But I have to confirm what I just read: 5.4 is EOL before 5.5 is released, and even before 4.11 and 5.3 are EOL? Or is it really the case that the EOL table should list /2007/ for 5.4 (same day as RELENG_5)? If 2006 is accurate, this is registering on me as a significant POLA violation. Very hard to believe this is accurate. If accurate, what list/channel/forum should I have been paying more attention to? security@, and the website where this has been announced for a LONG TIME. The policy and rationale is all there. Kris pgpABiyIRbQPT.pgp Description: PGP signature
Re: ULE Scheduler and overall performance on 6.x - Wow
On Sun, May 07, 2006 at 06:43:14AM -0300, Duane Whitty wrote: Hi, I decided to give the ULE scheduler a try a while ago (April 28). when I last built 6-STABLE Anyhow it seems great. I'm running a 2.4GHz Celeron with 512MB RAM and two 40GB, PATA disks. Right now I'm running both a GNOME and a KDE session, I've got Thunderbird and Evolution open, Firefox is running and running well, and I'm updating the my local copy of the FreeBSD repository. Oh yeah, I'm also running a DNS server, a Sendmail server, and SAMBA I can't believe how responsive everything is on this low-end machine I'm running.Wow! (And this with debugging turned on but no WITNESS or INVARIANTS turned on) FYI, in my testing ULE is faster under light workloads but quite a lot slower under heavy loads. It's not recommended, but YMMV. Kris pgppGBKH0ZlRx.pgp Description: PGP signature
Re: Panic: vrele: negative ref cnt
On Sun, May 07, 2006 at 03:16:46PM -0400, Robert H. Perry wrote: I received a panic message (see Subject) yesterday following a GNOME upgrade. The upgrade was successful but I encountered subsequent problems trying to access Firefox websites and my mail site in Thunderbird. I noticed a Plugin error message in Firefox and prepared to paste it into a message to the mailing list but was unable to get a connection and eventually lost it. I reset my modem several times then decided to reboot. First, I re-enabled my firewall then entered shutdown -r now. Next I received the panic message and a repeat of the following: READ_DMA ad0:Timeout - Read_DMA retrying (2 retries left) LBA = 28162271 READ_DMA ad0: FAILURE -READ_DMA status=51 READY, DSC, ERROR error = 40 Uncorrectable LBA =28162271 Your hard drive is dying. This may have caused the panic as a secondary effect after the read failed. Kris pgppLoM0srFLs.pgp Description: PGP signature
Re: Panic: vrele: negative ref cnt
On Sun, May 07, 2006 at 06:24:52PM -0400, Robert H. Perry wrote: Thanks for taking the time to respond. Not good news obviously, but could be worse. This is my mule machine. I was using it as a gateway, to develop some networking skills, and had installed SAMBA and was also trying to install HylaFAX . Fortunately it has a second hard drive which currently holds the /var slice. Out with the bad and in with the good. BTW, can the data on this drive be copied to the surviving drive? What commands do I need to review? My initial thoughts were to simply install 6.0 and go on from there. Any hints/suggestions are appreciated. dump/restore or just copy over the data you need to keep with tar | tar Kris pgpOVfewKmoOM.pgp Description: PGP signature
Re: pkg_add gnuplot
On Fri, May 05, 2006 at 05:00:46AM +, Jonathan Herriott wrote: Hey, I'm trying to add the package gnuplot, and when I do, it goes and tries to add another necessary package pdflib. The issue is that it cannot find pdflib. I get the following: pkg_add -r gnuplot Fetching ftp://ftp.freebsd.org/pub/FreeBSD/ports/i386/packages-6.0-release/Latest/gnuplot.tbz... Done. Error: FTP Unable to get ftp://ftp.freebsd.org/pub/FreeBSD/ports/i386/packages-6.0-release/All/pdflib-6.0.1_2.tbz: File unavailable (e.g., file not found, no access) pkg_add: can't open dependency file '/var/db/pkg/pdflib-6.0.1_2/+REQUIRED_BY '! dependency registration is incomplete For some reason the file does not exist on the server. Is there another method that I can get pdflib that pkg_add will recognize for when I pkg_add -r gnuplot again? pdflib packages may not be distributed according the license on the software. Sorry, you need to compile it yourself from the port. Kris pgpQd0m8wdRWI.pgp Description: PGP signature
Re: panic: page fault - 6.0-RELEASE-p7
On Fri, May 05, 2006 at 09:14:04AM -0600, Nick Wood wrote: Hello, We have a group of web and mail servers that run under a moderate load. We recently upgraded them from 4/5.x to 6.0. While we thought we had done enough testing, apparently we hadn't and are now experiencing panic's on a number of the servers. Some of our more heavily loaded servers have been fine for days, while others will crash every 6 to 36 hours. Below are some pieces of information that may be helpful. Try 6.1 first in case the bug is already fixed. Kris pgpPzXZAJeLJa.pgp Description: PGP signature
Re: rm: Operation not permitted
On Fri, May 05, 2006 at 07:54:07PM -0700, John Pettitt wrote: I'm having trouble erasing some files - this on a 6.1RC system (built from -STABLE) - I created a system image using his script: #!/bin/sh make buildworld KERNCONF=CLOCK make buildkernel KERNCONF=CLOCK mkdir -p /raid/diskless/clock make installworld DESTDIR=/raid/diskless/clock cd /usr/src/etc; make distribution DESTDIR=/raid/diskless/clock cd /usr/src make installkernel KERNCONF=CLOCK DESTDIR=/raid/diskless/clock and now when I try to delete it this happens ... # rm -rf clock rm: clock/bin/rcp: Operation not permitted rm: clock/bin: Directory not empty rm: clock/lib/libcrypt.so.3: Operation not permitted rm: clock/lib/libc.so.6: Operation not permitted rm: clock/lib: Directory not empty rm: clock/libexec/ld-elf.so.1: Operation not permitted rm: clock/libexec: Directory not empty rm: clock/sbin/init: Operation not permitted rm: clock/sbin: Directory not empty and so on. The system is running at securelevel -1 and the rm fails even in single user mode - what am I missing? chflags -R noschg Kris pgp9Nc8nd5xpU.pgp Description: PGP signature