January 2005 - openSUSE AMD64 - openSUSE Mailing Lists

strange nfsd message
by Jose Luis Ricardo Chavez 11 Jan '05

11 Jan '05

Hello all, does anybody know what the following message means? nfsd: doing automatic ADDCLIENT for <IP address here> It is strange since the host with the listed IP address does not have permission to mount the nfs partitions. Thanks, - Jose Luis

1 0

ESOTERIC QUESTION FOR THE GURUS
by Constantine 'Gus' Fantanas 11 Jan '05

11 Jan '05

Hi everybody. Since quite a few gurus frequent this board, I decided to ask a question that has been nagging me: How does the kernel calculate the processor speed? I am running SuSE 9.2 Professional on a Compaq Presario 3000Z laptop (2.2 GHz AMD64, 512 MB RAM, 60 GB HD, nVidia 440 toGo). For the record, I am also running 32-bit SuSE 9.2 Professional on a 32-bit Sony Vaio PCG-K35 so, I got two great distributions for the price of one and a great manual (so much for the British 'Linux Format' Magazine, which claims that SuSE is great but expensive!). The reason I ask this question is the following: My Presario laptop had a series of hardware problems, one of them being overheating. I had to send it to HP for repairs three times. After I got it back the third time, everything seemed to run fine except that the evaluation version of the burnin software I downloaded from www.passmark.com reports 1795 MHz (under Windows XP Home; the "speedometer" in the system tray of Windows has also disappeared). Under Linux, the KDE power manager and 'powersave -r' do show 2200 MHz occasionally. Does the kernel use the real time clock to come up with these values? The CPU does show up as a 2.2 GHz one. Does the kernel assume the maximum processor speed is what the processor reports at powerup (2.2 GHz in this case) and calculate the speed of the CPU based on that number (accounting for throttling)? If so, could a 2.2 GHz processor be actually run at 1.8 GHz and the kernel still report 2.2 GHz? I would appreciate your responses. CF

3 2

Tyan s4882
by Joe Landman 11 Jan '05

11 Jan '05

Folks: We are still having significant problems getting these quad boards with 16 GB ram to work with SuSE 9.2 pro. We have verified that they work using other distributions (RHEL, and a few others). Symptoms are kernel panic immediately after boot menu when more than 1 processor is installed. Symptoms are consistent across multiple units (Tyan transport tx46's). Disabling USB had no effect. Has anyone had a similar experience that they have successfully dealt with? If someone has this board, with 4 CPUs and 16 GB ram, could you tell me what your bios settings/versions are for some key items? Thanks. Joe

2 1

RE: [suse-amd64] System memory missing - AGP card taking it?
by Langsdorf, Mark 10 Jan '05

10 Jan '05

> Sorry if this is not specifically an X86-64 question. > > I have a Dual Opteron 250 with 4 GB of RAM with an NVIDIA > QuadroFX 1100 > card, with Suse 9.1 X86-64 installed with updates (kernel version > 2.5.5-7.111.19-smp). The BIOS recognizes all 4 GB of RAM, but the OS > only sees 3.26GB. I've been told this is because the video card is > taking up system memory. But could it really be taking up this much? > > Changing the AGP aperture size in the BIOS does affect how > much RAM is > occupied by the video card. Currently it is set to 32 MB. > If I set it > to the maximum of 512 MB, a measly 2.4GB are available. > > I also thought this was strange - if I set the AGP aperture to 32 MB, > and then try 1, 2, 3, and 4 GB, I see different percentages of system > memory being unavailable. Here's what I see: > > 1 GB -> 91.8% > 2 GB -> 95.2% > 3 GB -> 96.4% > 4 GB -> 81.5% > > Is their anything I can do in Suse to give me back more of my system > RAM? Seems like that last GB was a waste of money... > > Thanks for any help you can offer. The OS needs to allocate memory addresses for the framebuffer, AGP, and memory-mapped PCI accesses. These address regions must be below 4GB. Because the same physical address cannot be used to address a memory cell and a non-memory device at the same time, you lose access to the underlying memory cells. There is nothing the OS can do about this limitation. All machines with more than 3.5 GB of RAM will see some of the RAM missing to these various memory regions. There are some things that the BIOS can do to hoist one or more of the memory sticks above 4 GB, thus creating a hole in the memory map that can be filled by the framebuffer, PCI space, etc. Writing a BIOS to support memory hoisting is tricky and I don't expect that most BIOS vendors will support it until a larger part of the market buys machines with 4 or more GB of memory. -Mark Langsdorf Linux Drivers Team Lead AMD, Inc.

3 2

Adobe pdf viewer
by Janis Klava 10 Jan '05

10 Jan '05

After the latest YOU upgrade which included the Adobe pdf viewer, the pdf viewer does not display pages correctly. Opening a pdf file now defaults to +235 magnification and about 1/4 of the page is lost on the right hand margin. There is no vertical or horizontal scroll bar. Playing excessively with the zoom results in a frozen Konqueror. Janis

1 0

9.2 (FTP version) kernel compile error:
by Darren Lissimore 10 Jan '05

10 Jan '05

Hey all, Did a fresh install of the 9.2 FTP version and came across a kernel recompile error. ld: BFD 2.15.91.0.2 20040727 (SuSE Linux) assertion fail ../../bfd/linker.c:619 Apparently the same sort of error popped up with Fedora Core x86-64, an update of bin-utils took care of it for Fedora Core. Has anyone found the same error in their 9.2 kernel recompiles? I'm compiling an more recent bin-utils from src to see if that will correct the assetion problem. Darren ============================================================== linux:/usr/src/linux # make install CHK include/linux/version.h make[1]: `arch/x86_64/kernel/asm-offsets.s' is up to date. CHK include/linux/compile.h CC arch/x86_64/kernel/process.o LD arch/x86_64/kernel/built-in.o GEN .version CHK include/linux/compile.h UPD include/linux/compile.h CC init/version.o LD init/built-in.o LD .tmp_vmlinux1 ld: BFD 2.15.91.0.2 20040727 (SuSE Linux) assertion fail ../../bfd/linker.c:619 KSYM .tmp_kallsyms1.S AS .tmp_kallsyms1.o LD .tmp_vmlinux2 ld: BFD 2.15.91.0.2 20040727 (SuSE Linux) assertion fail ../../bfd/linker.c:619 KSYM .tmp_kallsyms2.S AS .tmp_kallsyms2.o LD vmlinux ld: BFD 2.15.91.0.2 20040727 (SuSE Linux) assertion fail ../../bfd/linker.c:619 SYSMAP System.map SYSMAP .tmp_System.map AS arch/x86_64/boot/setup.o LD arch/x86_64/boot/setup OBJCOPY arch/x86_64/boot/compressed/vmlinux.bin GZIP arch/x86_64/boot/compressed/vmlinux.bin.gz LD arch/x86_64/boot/compressed/piggy.o LD arch/x86_64/boot/compressed/vmlinux OBJCOPY arch/x86_64/boot/vmlinux.bin BUILD arch/x86_64/boot/bzImage Root device is (3, 5) Boot sector 512 bytes. Setup is 4770 bytes. System is 1566 kB Kernel: arch/x86_64/boot/bzImage is ready sh /usr/src/linux-2.6.8-24.10/arch/x86_64/boot/install.sh 2.6.8-24.10-default arch/x86_64/boot/bzImage System.map "" Root device: /dev/hda5 (mounted on / as ext3) Module list: sata_promise sata_via jbd ext3 Kernel image: /boot/vmlinuz-2.6.8-24.10-default Initrd image: /boot/initrd-2.6.8-24.10-default Shared libs: lib/ld-2.3.3.so lib/libc.so.6 lib64/ld-2.3.3.so lib64/libc.so.6 lib64/libselinux.so.1 Modules: kernel/drivers/scsi/scsi_mod.ko kernel/drivers/scsi/sd_mod.ko kernel/drivers/scsi/libata.ko kernel/drivers/scsi/sata_promise.ko kernel/drivers/scsi/sata_via.ko kernel/fs/jbd/jbd.ko kernel/fs/ext3/ext3.ko Including: udev klibc Using grub, re-install of bootloader not required. linux:/usr/src/linux #

1 0

Compiling flags for Mozilla Firefox
by Alexandre Moutinho Santos 09 Jan '05

09 Jan '05

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, I installed the rpm from ftp.suse.com of Mozilla Firefox 1.0 for SuSE 9.2 x86_64. Writing about:buildconfig in the URL box, I noticed that it was compiled without the -march=k8 or the -m64 flags. I then downloaded the src.rpm and compiled it with this flags - no problems at all. I would like to know why it wasn't compiled at first place with these flags, as in the man page of gcc it says that these flags turn on a set of options specific for the AMD64 processors. Best regards, Alex -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.5 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFB4Dv1PjVgz/mpp6YRAgXvAKCfemoheI46FQljhQg3STjjJyrkGACdFbKo mcCmn2AcBVSzwbRFMVyaB0I= =MAfO -----END PGP SIGNATURE-----

4 5

Konqueror message
by Mike Roy 08 Jan '05

08 Jan '05

Hello everyone: When I open Konqueror as a web browser, I get the following error message: Listing folders is not supported for protocol about How do I correct this or shut it off? Thanks for any ideas, suggestions, etc. I should mention that Konqueror seems to work in every aspect and once I close the error message, nothing else happens. Cheers, Mike

2 1

Re: [suse-amd64] Acer Aspire 1523 WLMI und 64bit-Linux Suse-9.2
by patheve2 08 Jan '05

08 Jan '05

Hello Matt, hello Andreas, hello Wigbert At first happy new year ! And the year of all the PCMCIA cards working with all SuSE 9.x AMD64 distribs :-) ! If the SuSE laptop team needs to perform tests by the way of the SuSE users, no problem. Hereafter the main characteristics which can be usefull for the team : - My laptop : Asus L5DF series, chipset nVidia nForce 150, cardBus bridge: Ricoh Co Ltd RL5c476 II (rev ab) - My OS : SuSE 9.0 AMD64 with kernel 2.4.21-268, SuSE 9.1 AMD64 with kernel 2.6.8 of SuSE 9.2 AMD64 (and WinXP pro SP2 32 bits) - My PCMCIA cards : modem card Hayes Optima288 (v34), modem card Com One Platinium Card MC227 (v90), ethernet card NetGear FA411 (16 bits) Hereafter, the status after several tests and suggestions of Matt : These three cards work fine under WinXP on this laptop and under other 32 bits linux distribs. Only the FA411 works under SuSE 9.0 AMD64, no success with the modem cards. Don't hesitate to contact me for performing tests Pascal =================================================== From: "Matt T." <Matt(a)Boons.net> Date: Wed, 5 Jan 2005 20:12:56 +0700 Message-Id: <200501052012.56905.Matt(a)Boons.net> Subject: Re: [suse-amd64] Acer Aspire 1523 WLMI und 64bit-Linux Suse-9.2 On Wednesday 05January 2005 00:19, Andreas Jaeger wrote: /> wlindenbauer(a)t-online.de (Wigbert Lindenbauer) writes: / /> > Hello Andreas, / /> > / /> > is it possible for you to test the PCMCIA-Slots ? / /> / /> No, I don't have any cards myself. But it should work... / /> / /> Andreas / Well, Andreas, unfortunately it does not! Some cards work, and others do not. I spent many days to get it to work with 9.1 on my HP notebook, as reported here earlier. I had been lucky that I tested with a card which works (a Axen / Sagem GPRS card). After that I tried other cards (Sierra Wireless GPRS card and a modem card) which all work fine under 9.1 and 9.2 on a 32 bit notebook). They do not work, as reported. And this is not a config.opts issue anymore, the memory range and the ports are adjusted to what is used on the 64 bit notebook. Taking Pascal's report, and my experience, there are 4 cards out of 6 *not* working. They apparently do not get a device assigned by serial_cs (such as /dev/ttyS1, or whatever else). May be cardmanager does not even get started when these cards are inserted. Andreas, I know you cannot test without the hardware. But obviously we need some help here to get this to work, and this seems to be for the benefit of SuSE also, as there will be more AMD64 notebooks coming. Isn't there anyone in SuSE owner of the 64 bit PCMCIA code? May be we can help by working together? Can you guide us to do some tests? Or would it be helpful to send you a non-working PCMCIA card? As of now I can only recommend to test if a PCMCIA card on an AMD64 notebooks works, *before* buying. 4 out of 6 cards not working is a bad number. And taking into consideration that the 9.1 PCMCIA code was not working without manually adjusting the memory and port configuration in config.opts and without updating the kernel to a kernel >= 2.6.7 it is obvious that this was never tested when 9.1 was released. SuSE 9.2 *is* the very best Linux distro for notebooks. Let's not add to that "but not for amd64 ones, using PCMCIA", please! A Happy New Year! Matt

1 0

RE: [suse-amd64] Artificially limiting memory
by Calkins, Dennis 08 Jan '05

08 Jan '05

Sorry. my application also has a kernel Driver. LIMIT doesn't work . I need EVERY PROCESS and the KERNEL to think there is only 4 GIG of main memory. The previous post where you can boot with mem=4G Is what I was looking for. Thanks again. -----Original Message----- From: Luc Renambot [mailto:luc@evl.uic.edu] Sent: Friday, January 07, 2005 10:16 PM To: skarsch(a)s1.karsch-net.de Cc: Calkins, Dennis; suse-amd64(a)suse.com Subject: Re: [suse-amd64] Artificially limiting memory Can you use the builtin 'limit' from tcsh ? > man limit Luc. On Jan 7, 2005, at 8:53 PM, Andi Kleen wrote: > On Fri, Jan 07, 2005 at 09:23:39PM -0500, Calkins, Dennis wrote: >> Hi, >> >> If I have a SUSE node with 8 GIG of memory, is there a way to >> trick >> the system into thinking it only has 4 GIG of memory? The Monta >> Vista's 2.4 Kernel had a kernel Command line Parameter. Does SUSE? > > Boot with mem=4G > > -Andi > -- Check the List-Unsubscribe header to unsubscribe For additional commands, email: suse-amd64-help(a)suse.com

1 0