December 2008 - openSUSE Users - openSUSE Mailing Lists

[opensuse] fglrx 8-12 driver released and builds and Installs Fine - Performance Terrible
by David C. Rankin 08 Jan '09

08 Jan '09

Listmates, ATI, cc:Stefan: I have downloaded, built and installed the 8-12 release (version 8.561) of the fglrx driver on my Toshiba P205D laptop running openSuSE 11.0. The driver packages builds and installs just fine from the ati-installer: ati-driver-installer-8-12-x86.x86_64.run Also, the driver now supports 11.1, so I presume it will build fine there too. Building results in the rpm: fglrx64_7_1_0_SUSE110-8.561-1.x86_64.rpm which builds the fglrx.ko module against the kernel source just fine and installs with no errors. That is where the good news ends. For my laptop with the radeon 1200/1300 mobile card, the performance of the 8-10, 8-11, and the 8-12 drivers is just terrible compared to the 8-8 and 8-9 drivers. Immediately before installing the 8-12 driver I was running the 8-9 driver and dumped glxgears frame rates to compare, again, against the new 8-12 driver. With the 8-9 driver: 23:43 alchemy:~> glxgears 4889 frames in 5.0 seconds = 977.692 FPS 4870 frames in 5.0 seconds = 973.831 FPS With the 8-12 driver: 23:59 alchemy:~> glxgears 3021 frames in 5.0 seconds = 604.191 FPS 2613 frames in 5.0 seconds = 522.544 FPS 2614 frames in 5.0 seconds = 522.795 FPS WTF? Applications are "visibly" slower as well. That is nearly a 50% decrease in performance. I don't know what has happened to the fglrx driver, but it hasn't been right since the 8-9 release. Back to the 8-9 release and all is well. Stefan, if you still have a contact at AMD, would you pass this along? I would bug it with novell, but we both know that would an absolute waste of time. -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

4 4

[opensuse] 11.0 cusrses mode yast printer config fail
by Brian K. White 07 Jan '09

07 Jan '09

Although I've been using 11.0 since it came out on several boxes, I haven't happened to need a print spooler on any of those until yesterday, when I discovered that the yast cups interface is broken, at least for me at least in the curses interface. Starting with an 11.0 install with up to the minute online updates applied, and the cups service running (though never been configurd yet): yast, Printer, Printers & queues, add, (x) Network Printers (x) Print Directly to a Network Printer (x) Direct TCP Port Printing ┌Connection Information────────────────────────────────────┐ │ Hostname of Print Server │ │ mycustomer.dyndns.biz▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒↓[Look Up↓] │ │ TCP Port Number │ │ 9100▒▒▒▒▒▒▒▒▒▒▒ │ │ │ └──────────────────────────────────────────────────────────┘ [Test Remote Socket Access] ┌──────────────────────────────────┐ │ │ │ The print server is accessible. │ │ └───────────│ [OK] │ ──────────┘ └──────────────────────────────────┘ ┌Queue Name and Spooler Settings───────────────────────────────────┐ │Name for Printing │ │printer▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒│ │ │ │Printer Description │ │▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒│ │Printer Location │ │▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒│ │[ ] Share Printer │ │ │ │[ ] Do Local Filtering │ │ │ └──────────────────────────────────────────────────────────────────┘ Now it brings you back to the screen where you can add/edit/delete more queues, or test printing or finish. At this point there is still no /etc/cups/printers.conf When you either finish or if you try to test printing, either way it tries to save the config you just defined and fails. (x) Test Text Printing │ ┌─────────────────────────────────────────────────────┐ │ │ │ Error │ │ │ │ An error occurred while communicating with the │ │ │ │ CUPS server while saving queue y2test. │ │ │ │ │ │ │ │ cups(File)DoRequest error: client-error-bad-request │ │ │ │ [OK] │ ────┤───┘ │ └─────────────────────────────────────────────────────┘ ────────┐ Then this error which is unimportant, it's expected given above. The queue was not actually created, and so of course the lpd client can't use it. │ ┌────────────────────────────────────────────────────┐ │ │ │ Unable to print the test page. │ │ │ │ │ │ │ │ Output from lpr: │ │ │ │ /usr/bin/lpr: The printer or class was not found. │ │ │ │ [OK] │ ────┤───┘ │ └────────────────────────────────────────────────────┘ ────────┐ You have to choose Abort to get out of the printer setup screen since Finish just fails the same way forever. I also tried variations on the specific config options, like, although I very much should be able to say no local filtering (ie: don't use any kind of printer driver or ppd file) maybe that's not so common and some part of the tool just assumes theres always a driver. etc.. So I also tried with local filtering and selecting an appropriate printer model (HP LaserJet 1320, totally common pcl driver), tried enabling sharing, tried lpd as well as raw tcp, to a known good and accessible lpd server. In all cases I got the same error above when it came time to actually save the config at the end. However, when I copied the printers.conf from the working 10.2 box, it worked. just a straight rsync copy, no edits at all. Then rccups stop, rccups start. And printing to those printers immediately worked. printers.conf just has this, one lpd printer and two jetdirect, no sharing, no filters/drivers. The comment shows a recent write because I had gone into yast-printers on the old box to delete a printer. The 3 remaining printers haven't been touched in around a year. The cups version shows it didn't come from the 11.0 box that's using it. The new box has cups 1.3.7. Obviousy I rewrote the customer name and their hostname in the email. ---TOF--- # Printer configuration file for CUPS v1.2.7 # Written by cupsd on 2008-10-14 10:08 <DefaultPrinter mycustomer_checks> Info HP DeskJet 540 Location My Customer, Eileen's Office, winxp psu share DeviceURI lpd://mycustomer.dyndns.biz/checks State Idle StateTime 1229696182 Accepting Yes Shared No JobSheets none none QuotaPeriod 0 PageLimit 0 KLimit 0 OpPolicy default ErrorPolicy retry-job </Printer> <Printer mycustomer_laser> Info HP LaserJet 1320 Location My Customer DeviceURI socket://mycustomer.dyndns.biz:9100 State Idle StateTime 1229727065 Accepting Yes Shared No JobSheets none none QuotaPeriod 0 PageLimit 0 KLimit 0 OpPolicy default ErrorPolicy retry-job </Printer> <Printer mycustomer_matrix> Info Okidata Microline 421 Location My Customer DeviceURI socket://mycustomer.dyndns.biz:9101 State Idle StateTime 1229726894 Accepting Yes Shared No JobSheets none none QuotaPeriod 0 PageLimit 0 KLimit 0 OpPolicy default ErrorPolicy retry-job </Printer> ---EOF--- I looked at cupsd.conf too but I never made any changes to that on either the old or new boxes and they looked about the same on both boxes. Both boxes have the same rules in susefirewall and it's active on the old box and disabling the firewall didn't help the new box, nor did reenabling the firewall break it. I reproduced the problem on 4 boxes, but, they are all practically clones of each other. Installed at the same time onto identical hardware from the same install source using the same install recipe. This all works fine in at least 10.0, 10.1, 10.2, 10.3, not just 10.2. 9.3 and 9.1 the interface is a little different but I don't have any problems there either. I haven't istalled 11.1 anywhere yet. Can anyone else confirm this or know about it and know what the problem is and a work-around other than manually creating & editing printers.conf? Pretend I only have serial console access and that http://localhost:631 can't be operated well enough via lynx. -- Brian K. White brian(a)aljex.com http://www.myspace.com/KEYofR +++++[>+++[>+++++>+++++++<<-]<-]>>+.>.+++++.+++++++.-.[>+<---]>++. filePro BBx Linux SCO FreeBSD #callahans Satriani Filk! -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

4 7

[opensuse] KDE4 performance tip
by Amedee Van Gasse 07 Jan '09

07 Jan '09

I have a performance tip that I would like to share. I noticed that something called nepomuk was eating 80% of my cpu, and it was also eating a lot of ram. Nepomuk is (as far as google can tell me) a backend for the strigi desktop search. If you're a bit like me, you have no use for a frakkin' desktop search like strigi or beagle. All my data is well organized. Read about "Inbox Zero" and "The Hamster Revolution" - everything about organizing email can also be applied to every other kind of user data. The bad thing for me was that strigi is enabled by default. I disabled it in Configure Desktop -> Advanced -> Nepomuk. Then I uninstalled strigi. I'm not going to report this somewhere as a bug, because then I'll have the userfriendly police on my back. Imho the ubiquitous usage of desktop search is a Bad Thing(tm). It's like a junior DBA who has read his first book about database optimization and says, hey let's just index everything! While the *real* DBA knows that a well designed database is a better and faster database. You don't index everything, you just index where you need it and where it's useful. Same thing about desktop search, I don't want it to index my /home. Sorry, I'm getting off my soapbox now :) -- Amedee -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

13 22

[opensuse] 11.1 upgrade on md raid root fails to boot
by Joe Morris 06 Jan '09

06 Jan '09

Well, i have managed to get enough info to know this is bug https://bugzilla.novell.com/show_bug.cgi?id=445490, which says it is resolved, but obviously not in the GM version of packages. Unfortunately, I cannot see what was needed to fix this, as I get an access denied when clicking on https://bugzilla.novell.com/show_bug.cgi?id=435778 at the end (where the fix was supposed to be). Is there an update of mkinitrd, or udev, or something to fix this? I upgraded from 10.3 yesterday, but not being able to boot after it was finished was a real problem. I was finally able to boot installed system (hidden in the repair tools) to be able to boot again and was able to find the above bug report and found out with a few tweaks I could eventually boot up by hitting Ctrl-D to seemingly exit the bootup system. I have been running mdraid 1 since 8.0, and mdraid 1 roots since 9.1 IIRC, but this was the hardest to figure out, and appears to be a problem that is quite old ( http://lists.opensuse.org/opensuse-bugs/2008-08/msg07297.html ). I would really appreciate it if anyone knows how I can repair the boot process in 11.1 so boot.md does not fail, and I do not need to hit Ctrl-D to be able to boot? BTW, I have a mdraid root and home, and yesterday trying to repair it I was seeing it go from md0 & md1 to md126 and md127, trying the repair option was a disaster, as it fscked the individual disk, which failed (and I canceled, but that I eventually figured out had somehow split the 2 partitions to one disk was md125_0 and the other md127_0. It was bizarre. Seems the mdraid part of 11.1 has some real problems this time around (and the server at our office has 4 mdraid devices, not looking forward to that upgrade). Does anyone know if this was worse because of the upgrade path (or in other words, has anyone installed 11.1 on a mdraid root device and did not have problems)? TIA for any help. -- Joe Morris Registered Linux user 231871 running openSUSE 10.3 x86_64 -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

3 2

[opensuse] Controlling zypper's Timedout when it is updating -- REPOST
by Basil Chupin 06 Jan '09

06 Jan '09

When zypper is downloading upgrades in YaST2 (Online Update), a loss ofconnection to the internet can occur or some other hiccup where zypper either cannot connect to the URL or it cannot get the filename to download. When this occurs zypper sits around for seems like an eternity doing nothing and then announces that an error occurred; it expects a click of the mouse on the OK button after which another window comes up giving you the chance to Retry etc. In 99+% of cases simply clicking on Retry get zypper to get on with the job. This sitting around doing nothing is a PITA for 'hours' - and there is nothing that one can do except to wait until zypper decides to advise you that an error occurred. The ABORT button does not work so one just sits and waits. (Of course, there is always the drastic step of CTRL-ALT-ESC to exit zypper.) Somewhere there must be timer setting which controls how long zypper keeps trying to resume its task. Does anyone know where this setting is located so that the user can set the timeout period. Nothing I can see in zypper help and nothing shows up in /etc/zypp. Can anyone provide an answer, please? Ciao. -- Be nice to people on your way up - you'll see the same people on your way down. -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

10 27

[opensuse] Samba Question
by Donald D Henson 05 Jan '09

05 Jan '09

I can't seem to find the answer to the following question, not even in "Samba for Dummies". I guess the question is too simple. Anyway... Is it okay to have more than one samba server in a given local network? Or put another way... Do you need a samba server on each linux machine that wants to serve shares? Responses appreciated. -- Don Henson

7 10

[opensuse] openSUSE 11.1 - of grub,md/lvm root, and related headaches
by Moby 05 Jan '09

05 Jan '09

I have had a system running fine since oss 10.1 (upgraded from there on to the current nightmare that is 11.1). Here is the system disk layout: Two 300 GB SATA hard drives - sda and sdba sda1 and sdb1 are 130 MG, them mirrored using md raid into /dev/md0. /dev/md0 has ext3 mounted as /boot. sda2 and sdb2 are 1 GB, then mirrored using md raid into /dev/md1. /dev/md1 is swap. sda3 and sdb3 are 274 GB, then mirrored using md raid into /dev/md2. /dev/md2 has lvm on it, with / being /dev/system/system. Grub installed in mbr of sda. This system has been running great. I upgraded in-place from 11.0 to 11.1, and everything was still fine. Then I built a custom kernel (as I have done with all other versions - the only thing I change is set the processor type). During the custom kernel install phase, make modules_install worked fine, but I noticed some weird device-mapper "no such file or device" error message during make install. I did not think much of it (probably a BIG mistake). Verified new vmlinuz and initrd were in place and /boot/grub/menu.lst looked kosher. Rebooted the system, and it went to pot. On reboot, the machine displays the word "GRUB" and just hangs there. Only thing I can do at that point is to reboot. Booted from 11.1 install disk, went into "repair installed system", no joy - this "repair installed system" did not see any md devices, nor any lvs. Booted from 11.1 install disk, went into rescue mode. I was able to mount all my partitions successfully, I was able to chroot to my original root, go into yast etc. I reinstalled grub to mbr of sda, no errors. I googled and found a fix for the device mapper errors during mkinitrd (it had to do with a bug in Tools.pm), re-made and installed the kernel etc. Rebooted the system, same behavior, it just hangs after printing GRUB. At this point, I have done about everything I can think of to revive it to no joy. Finally I made a grub boot CD, and have it use the kernel and initrd in place on the hard drives, and the system comes up fine. I am completely lost as to why grub will not boot from the HDD. I have tried reinstalled Grub to theMBR, as well as to the boot sector of the active parition (/dev/sda1) to no avail. During googling for solutions, I read bits and pieces about how oss 11.1 uses a different mechanism for mkinitrd, something about relying on device mapper etc. While I do not fully understand what the changes are, I hope someone on the list does and can readily see what could be ailing my system and suggest a way out. In the meantime, the system continues to run after booting from a CD. Thanks in advance for any help, -- --Moby They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety. -- Benjamin Franklin -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

5 12

[opensuse] Wolfgang - 11.0 Tbird, any way to assign the 'del' key to 'd' for messages view?
by David C. Rankin 05 Jan '09

05 Jan '09

List (mostly for Wolfgang): On 11.0 in thunderbird is there any way to assign 'del' to the 'd' key to make the 'd' key work as delete while in the message view window? Any of your brilliant: gconftool-2 -t string -s /desktop/gnome/url-handlers/http/command blah blah commands up your sleeve? -- David C. Rankin, J.D.,P.E. | openSoftware und SystemEntwicklung Rankin Law Firm, PLLC | Countdown for openSuSE 11.1 www.rankinlawfirm.com | http://counter.opensuse.org/11.1/small -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

2 2

[opensuse] ATI Graphics Cards on openSuse
by Jim Flanagan 05 Jan '09

05 Jan '09

Greetings all, I have moved away from ATi cards for a year or so now, and am considering upgrading the card on one box. Before I blindly go out a get an Nvidia card I wanted to check here to see if the situation with the Ati drivers had improved. Does anyone have any recent experiences with these newer card and openSuse, good or bad. Thanks, Jim F -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

8 15

[opensuse] madwifi RPMS for 11.1 (temporary)
by David C. Rankin 05 Jan '09

05 Jan '09

Listmates, With the trouble getting my Toshiba Atheros wireless working and without madwifi rpms available, I built a set from the latest svn build of madwifi and created a repository for 11.1. I only have 256K of upstream bandwidth, so if anyone else wants to copy them and host them until the Buildservice rpms are done, that would help. You can find the repository and a bit of additional information here: http://www.3111skyline.com/linux/openSuSE111-install.php#madwifi -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

4 8