Re: Re: VT8237 and installation problem
I've tried the last testing, downloaded yesterday evening. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: adm64 and multiarch
On Mon, May 15, 2006 at 10:42:30AM +0200, Goswin von Brederlow wrote: > Alexander Samad <[EMAIL PROTECTED]> writes: > > > Hi > > > > sorry if this email is a bit off topic, but still enough to ask I think. > > > > What is the status of multiarch and debian (specifically amd64 port). > > > > I like lots of other users have migrated to amd64, I am finding it a > > bit of a pain to get some legacy software and even new stuff to work on > > my machines. > > > > Usually it invovles me having to manually install some i386 lib > > somewhere under the program directory setting up a script to change the > > ld_path. > > Dump them in /emul/ia32-linux/ or add the directory to > ld.so.conf. There is no need to write a wraper for every program to > change the library path. did not realise it handle x86_64 & i386 at the same time, thanks will do > > You can also make a wishlist bugreport against ia32-libs for inclusion > of more libs. We (the ia32-libs maintainer) don't know what libraries > are missing. If it isn't an too obscure software we can certainly add > libs. > > > This works, but is a pain to maintain > > > > I don't like the idea of chroot'ing > > > > My big hurdle to jump next is OO 2.x > > > > I have had a look around debian.org for multiarch and there seem like > > there was some work done back in 04 talk about it has been vary sparce. > > Don't hold your breath. The etch release is coming fast and there has > been pretty little progress in the direction of multiarch. Most > notably nobody steped up and did realy dig into dpkg and apt sources > to make it happen. What was done in the past is both incomplete and > in parts was only a hack for testing purposes. Nothing that would be > ready for a stable release. > > > Also looked around for OO on x86_64, this is also a bit here and a bit > > there, I reliase there are only a small amount of poeple working on it. > > But I would have thought this is a majour show stopper. > > The long term plan is to make it 64bit clean. The short term plan, > meaning at least etch, is to extend ia32-libs or equivalent mechanism > to get a 32bit OOo for amd64. any idea where I can look for these's deb's ? Most of the places I have look talk about it, but then the trail goes cold > > > Alex > > MfG > Goswin > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > signature.asc Description: Digital signature
Re: software raid jfs
Giacomo Mulas wrote: On Mon, 15 May 2006, Karl Schmidt wrote: I'm now thinking it might be the NForce4 chipset - I need to find out where I might snag a newer kernel for sarge - and I am concerned that there is a change about the initrd that might make this a problem? yes, but of course this is only a problem if you use an initrd to boot. I use a backported version of yaird for this purpose. Also, if you use them you will need backported versions of some other packages, e.g. udev, module-init-tools, a handful more packages which I don't remember. All in all, I think I needed some 5 backported packages, and it was not a very big problem. I routinely compile kernels from the linux-source package in unstable on these sarge machines using make-kpkg, and install them without problems. I found some backports at: http://www.backports.org/debian/pool/main/l/linux-2.6/ I think these will work? Any way to apt-get from the installer shell so it brings in the dependencies? ,.,. Can anyone tell me what kernel version the AMD64 kernels changed from kernel-image to linux-image? Karl Schmidt EMail [EMAIL PROTECTED] Transtronics, Inc. WEB http://xtronics.com 3209 West 9th StreetPh (785) 841-3089 Lawrence, KS 66049 FAX (785) 841-0434 When angry count four; when very angry, swear. --Mark Twain -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
VT8237 and installation problem
Hello, I'm trying to install debian sarge (I've also tried testing) in a pc with MSI mainbord with VIA VT8237 southbridge. The problems come at the beginning, just after the boot option. The final result is a kernel panic. I've tried all avaible boot option: safe, linux26, acpi=off, noapic, nolapic but without success!!! What can I do? Anyone have success with that chipset? Thanks, Matteo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: software raid jfs
On Mon, 15 May 2006, Lennart Sorensen wrote: backports.org might have one along with the required tool updates. At least they do for i386. I think they have amd64 packages too but I haven't checked. They do. bye Giacomo -- _ Giacomo Mulas <[EMAIL PROTECTED]> _ OSSERVATORIO ASTRONOMICO DI CAGLIARI Str. 54, Loc. Poggio dei Pini * 09012 Capoterra (CA) Tel. (OAC): +39 070 71180 248 Fax : +39 070 71180 222 Tel. (UNICA): +39 070 675 4916 _ "When the storms are raging around you, stay right where you are" (Freddy Mercury) _ -- Il messaggio e' stato analizzato alla ricerca di virus o contenuti pericolosi da MailScanner, ed e' risultato non infetto. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: software raid jfs
On Mon, 15 May 2006, Karl Schmidt wrote: I'm now thinking it might be the NForce4 chipset - I need to find out where I might snag a newer kernel for sarge - and I am concerned that there is a change about the initrd that might make this a problem? yes, but of course this is only a problem if you use an initrd to boot. I use a backported version of yaird for this purpose. Also, if you use them you will need backported versions of some other packages, e.g. udev, module-init-tools, a handful more packages which I don't remember. All in all, I think I needed some 5 backported packages, and it was not a very big problem. I routinely compile kernels from the linux-source package in unstable on these sarge machines using make-kpkg, and install them without problems. Bye Giacomo -- _ Giacomo Mulas <[EMAIL PROTECTED]> _ OSSERVATORIO ASTRONOMICO DI CAGLIARI Str. 54, Loc. Poggio dei Pini * 09012 Capoterra (CA) Tel. (OAC): +39 070 71180 248 Fax : +39 070 71180 222 Tel. (UNICA): +39 070 675 4916 _ "When the storms are raging around you, stay right where you are" (Freddy Mercury) _ -- Il messaggio e' stato analizzato alla ricerca di virus o contenuti pericolosi da MailScanner, ed e' risultato non infetto. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: software raid jfs
On Mon, May 15, 2006 at 02:14:23PM -0500, Karl Schmidt wrote: > I'm now thinking it might be the NForce4 chipset - I need to find out where > I might snag a newer kernel for sarge - and I am concerned that there is a > change about the initrd that might make this a problem? > > I also think they changed kernel-image to linux-kernel ? > > My plan is to drop to shell and wget a newer kernel if I can find one. backports.org might have one along with the required tool updates. At least they do for i386. I think they have amd64 packages too but I haven't checked. Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: software raid jfs
Lennart Sorensen wrote: On Mon, May 15, 2006 at 12:52:04PM -0500, Karl Schmidt wrote: Is anyone using software raid (mdraid) and/or jfs with success on AMD64? I finally finished an install on a Tyan S2865 and was installing some extra packages - I suddenly had massive file corruption and an unbootable system. I have done raid1 no problem on amd64. The sarge installer had no problem setting that up. LVM and ext3 works fine with it too. Never even looked at jfs. Len Sorensen I'm now thinking it might be the NForce4 chipset - I need to find out where I might snag a newer kernel for sarge - and I am concerned that there is a change about the initrd that might make this a problem? I also think they changed kernel-image to linux-kernel ? My plan is to drop to shell and wget a newer kernel if I can find one. Karl Schmidt EMail [EMAIL PROTECTED] Transtronics, Inc. WEB http://xtronics.com 3209 West 9th StreetPh (785) 841-3089 Lawrence, KS 66049 FAX (785) 841-0434 Time wounds all heels. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: software raid jfs
On Mon, May 15, 2006 at 12:52:04PM -0500, Karl Schmidt wrote: > Is anyone using software raid (mdraid) and/or jfs with success on AMD64? > > I finally finished an install on a Tyan S2865 and was installing some extra > packages - I suddenly had massive file corruption and an unbootable system. I have done raid1 no problem on amd64. The sarge installer had no problem setting that up. LVM and ext3 works fine with it too. Never even looked at jfs. Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Re: Problem booting Debian amd64 3.1r0a
> Maybe, the scsi-driver is not seen properly. So it will not find the hard
> drives, and cannot load needed modules. my suggestion to fix this:
> Make a kernel with all needed drivers statically compiled in and no initrd
on
> another machine using make-kpkg. Then start with the installation system and
> install as normal. Do not reboot !!! Now install your static-kernel with no
> initrd and make it known in grub or lilo.
Unfortunately, I didn't have time to pursue this matter and ended up using a
"server" install of Ubuntu 5.10. However, I had an issue with that which
seems like it may have been related, which I mention here in case it comes in
useful for people in the future.
My aim in setting up the Debian amd64 system originally was as a
"host" ("dom0") system for Xen. Part of setting up the Xen system involved
creating an initrd from the Ubuntu system for the Xen kernel, using Ubuntu's
mkinitramfs. I used the same procedure as I would have done on Debian 3.1,
only using "mkinitramfs" instead of "mkinitrd".
# depmod 2.6.16-xen
# mkinitramfs -o /boot/initrd-2.6-xen.img 2.6.16-xen
When I came to boot this Xen kernel (having configured Grub appropriately), I
had the same symptoms as I'd had with the Debian kernel. It seems that
mkinitramfs had failed to include the mptspi driver for my LSI Logic /
Symbios Logic 53c1030 PCI-X Fusion-MPT Dual Ultra320 SCSI controller, despite
this driver being included in /lib/modules/2.6.16-xen
The solution to this problem ultimately was to
change /etc/mkinitramfs/initramfs.conf from "MODULES = most" to "MODULES =
dep" and rerun mkinitramfs. This included mptspi in the initramfs allowing my
kernel to boot.
I suspect that had I perservered with the Debian amd64 install, I may have
found the same solution, regenerating the initrd for the default
kernel-image. I post this information in the hope that others will find it
useful.
Alex
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
software raid jfs
Is anyone using software raid (mdraid) and/or jfs with success on AMD64? I finally finished an install on a Tyan S2865 and was installing some extra packages - I suddenly had massive file corruption and an unbootable system. Karl Schmidt EMail [EMAIL PROTECTED] Transtronics, Inc. WEB http://xtronics.com 3209 West 9th StreetPh (785) 841-3089 Lawrence, KS 66049 FAX (785) 841-0434 When everything is coming my way, I'm probably in the wrong lane. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: APC (American Power Conversion) APC BACK-UPS RS 800VA
On Mon, May 15, 2006 at 04:29:09PM +0200, Francesco Pietra wrote: > I forgot two possible problems: it deals of a workstation for very heavy > computation (quantum mechanical), not a server. Although I'll benefit of full > property (not having to share cpu time with colleagues, such as in the > institution) it will be at the very low limit to accomplish part of my tasks. > Therefore, what about the usb connection to the APC apparatus? usb makes use > of the cpu and because of this it has a bad repution in our quarters (read > scsi by far preferred): how much it will detract from the performance? Basicly nothing. There is so little communications between the UPS and the system. I think you are looking at maybe 30 or 40 bytes of data per minute typically. Unless the power fails the UPS really doesn't have anything new to say so it says nothing for the most part. > Second problem: voltage surge from lightnings, which may occur because the ws > will also run unattended. As far as i know, back-ups do not provide any > reliable barrier. I remember vaguely that any computer should by default be > provided of varistors to block voltage surge from lightnings. Is it correct? > In particular do you know if Enermax case has bu default real varistors and > efficiently connected (there were once upon a time controversial views how > many varistors, how coccected, and if they are real or not). Well APC claims to insure your connected equipment against damage caused by lightning. I hope that means they have good enough protection that they don't have to pay that very often. What the power supply does I don't know. > I hope these questions are of general interest. They concern the safety of > expensive machines and extensive work by the user. Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: APC (American Power Conversion) APC BACK-UPS RS 800VA
Hi Len: Thanks for all information. I forgot two possible problems: it deals of a workstation for very heavy computation (quantum mechanical), not a server. Although I'll benefit of full property (not having to share cpu time with colleagues, such as in the institution) it will be at the very low limit to accomplish part of my tasks. Therefore, what about the usb connection to the APC apparatus? usb makes use of the cpu and because of this it has a bad repution in our quarters (read scsi by far preferred): how much it will detract from the performance? Second problem: voltage surge from lightnings, which may occur because the ws will also run unattended. As far as i know, back-ups do not provide any reliable barrier. I remember vaguely that any computer should by default be provided of varistors to block voltage surge from lightnings. Is it correct? In particular do you know if Enermax case has bu default real varistors and efficiently connected (there were once upon a time controversial views how many varistors, how coccected, and if they are real or not). I hope these questions are of general interest. They concern the safety of expensive machines and extensive work by the user. Regards francesco On Monday 15 May 2006 15:00, Lennart Sorensen wrote: > On Sat, May 13, 2006 at 06:14:59PM +0200, Francesco Pietra wrote: > > I forgot to mention that for my two amd64 dual opterons set the > > transformer is 600W. Would the 540W of APC BACK-UPS RS 800VA be enough? > > The 1000W model is priced 470 euros in Italy. > > Well a system is amost never using the full capacity of the power > supply. So it really depends what is in the system. The 800VA might be > a bit small for a server, but it depends. I run a desktop system of one > of the 800VAs, althought it might be a bit overkill for it (but it was > cheap so I got it). > > Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Problem booting Debian amd64 3.1r0a
On Friday 12 May 2006 12:41, Hans wrote: > Am Freitag, 12. Mai 2006 13:08 schrieb Alex Page: > > I'm trying to install Debian amd64 3.1r0a on a dual em64t Xeon box, with > > SCSI hard drives and an IDE CD-ROM. I've previously had Debian i386 > > running on the same hardware. > > > > Once I've completed the base install from CD, I'm prompted to reboot > > into Debian. When I boot my Debian amd64 kernel, the boot process hangs > > with the following error message: > > Hi all, I had a similar problem: installation from cd (3.1r0a netinst) worked fine until reboot, where the new system hung up. I found two errors in the new system, which caused the system-hang: 1) the installer did not create a symlink to the new kernel-image in / 2) "lilo.conf" didn't contain the necessary statement for the initial ramdisk. So, I had to do the following: After the installation, just before the reboot, I had to switch to a second tty, entered a "chroot /target" and then a "ln -s /boot/vmlinuz". After that I edited the lilo.conf with "nano /etc/lilo.conf" and entered something like "initrd=/boot/initrd.img". Then save and run "/sbin/lilo". After these steps my new debian rebooted without a problem... :) HTH, Best regards, Marcus -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: APC (American Power Conversion) APC BACK-UPS RS 800VA
On Sat, May 13, 2006 at 06:14:59PM +0200, Francesco Pietra wrote: > I forgot to mention that for my two amd64 dual opterons set the transformer > is > 600W. Would the 540W of APC BACK-UPS RS 800VA be enough? The 1000W model is > priced 470 euros in Italy. Well a system is amost never using the full capacity of the power supply. So it really depends what is in the system. The 800VA might be a bit small for a server, but it depends. I run a desktop system of one of the 800VAs, althought it might be a bit overkill for it (but it was cheap so I got it). Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: APC (American Power Conversion) APC BACK-UPS RS 800VA
On Sat, May 13, 2006 at 01:55:19PM +0200, Francesco Pietra wrote: > I would like to protect from sparks and temporary power failures my > (continuous) computational work on a workstation (still to be built because > of the difficulty of getting mainboard Tyan K8WE S2895SA3NRF). We already > have two dual amd6a opterons, driven by amd64 debian. > > I have been offered in my country (Italy) APC (American Power Conversion) APC > BACK-UPS RS 800VA (at a price 219 euros, more than twice the price in USA > 120$). The price difference is standard here. It does not reflect the > difference of taxation in the two countries, importation duties, and shipping > expenses. And, from the long delay in having the mainboard delivered from the > official dealer here, could it be that someone is waiting for the importation > price to go down, after having fixed the price with the client? At any event, > I am reiterating that with debian we have not solved our "dependencies" in > toto. > > My main concern is about functionality with the OS i intend to use. I came > across > to /www.pricegrabber.com/rating_getprodrev.php/product_id=1369992/id_type=M > which is not very encouraging (particularly as it probably refers to 32 not > 64), as follows: > > Has anyone experience with this (or other equipment to the same purpose) for > the type on installation of my project? I just use apcupsd which works great for me. I can't imagine powerchute would be an improvement. Works with all versions of linux and most other unixes as far as I know. Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: adm64 and multiarch
Alexander Samad <[EMAIL PROTECTED]> writes: > Hi > > sorry if this email is a bit off topic, but still enough to ask I think. > > What is the status of multiarch and debian (specifically amd64 port). > > I like lots of other users have migrated to amd64, I am finding it a > bit of a pain to get some legacy software and even new stuff to work on > my machines. > > Usually it invovles me having to manually install some i386 lib > somewhere under the program directory setting up a script to change the > ld_path. Dump them in /emul/ia32-linux/ or add the directory to ld.so.conf. There is no need to write a wraper for every program to change the library path. You can also make a wishlist bugreport against ia32-libs for inclusion of more libs. We (the ia32-libs maintainer) don't know what libraries are missing. If it isn't an too obscure software we can certainly add libs. > This works, but is a pain to maintain > > I don't like the idea of chroot'ing > > My big hurdle to jump next is OO 2.x > > I have had a look around debian.org for multiarch and there seem like > there was some work done back in 04 talk about it has been vary sparce. Don't hold your breath. The etch release is coming fast and there has been pretty little progress in the direction of multiarch. Most notably nobody steped up and did realy dig into dpkg and apt sources to make it happen. What was done in the past is both incomplete and in parts was only a hack for testing purposes. Nothing that would be ready for a stable release. > Also looked around for OO on x86_64, this is also a bit here and a bit > there, I reliase there are only a small amount of poeple working on it. > But I would have thought this is a majour show stopper. The long term plan is to make it 64bit clean. The short term plan, meaning at least etch, is to extend ia32-libs or equivalent mechanism to get a 32bit OOo for amd64. > Alex MfG Goswin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: amd64-archive update
Alexander Samad <[EMAIL PROTECTED]> writes: > On Sun, May 14, 2006 at 08:57:53PM -0700, Alexandru Cardaniuc wrote: >> Matthieu Helleboid <[EMAIL PROTECTED]> writes: >> >> > Hello, >> > >> > I've made an update to amd64-archive made by Goswin von Brederlow. >> > >> > It will make a local repository for amd64 without a chroot for : >> > * firefox, mozilla, opera, flash >> > * mplayer, w32codecs >> >> > * skype >> >> >> Since recently ia32-libs includes all necessary libraries to make skype >> work in am64. Just grab the dynamic or static binary from www.skype.com >> No need for amd64-archive anymore to make skype work. > > are there plans to increase ia32-lib to include more and more lib's ? More the opposite. MfG Goswin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: alsaconf locks up my system
On Sun, May 14, 2006 at 17:09:26 -0500, Russ Cook wrote: > Florian Kulzer wrote: > >On Sun, May 14, 2006 at 13:32:31 -0500, Russ Cook wrote: > > > >>I'm running a A8N32-SLI Deluxe, with dual core amd64. My kernel > >>is 2.6.16-1-amd64-k8-smp. I also run kernel 2.6.14.4. With either > >>kernel, I have no audio. I used to have audio until about a week > >>ago, after running apt-get upgrade. I'm running the 64 bit dist, > >>unstable. When trying to run XMMS or any other audio or video > >>player, I get no audio. XMMS tells me it couldn't open audio, and > >>request I check soundcard configured, correct output plugin selected, > >>and no other program is blocking the soundcard. > >> > >>In the past, I could get past this problem by re-running alsaconf. > >>Now however, alsaconf never completes before my X system freezes > >>up. I can still access my system from another machine using ssh, > >>but all output to the screen is frozen, and the keyboard is > >>unresponsive. The mouse pointer still moves across > >>the screen, but nothing responds. > >> > >>Has anyone else had this problem, and can anyone offer hints or > >>suggestions as to the cause and cure? > >> > > > >I would try to shut down X and run alsaconf from a virtual terminal. If > >this still locks up your machine you can log in with ssh and check which > >process is blocking the CPU with the "top" command. It would then also > >be interesting if killing that process unlocks the computer or not. > >Another thing to try is "/etc/init.d/alsa force-reload". > > > >If alsaconf works when run from the terminal you can use a simple > >application such as "speaker-test" to check your sound. If this is > >successful you can try if it still works when you start X again. This > >should help you to isolate the cause of the problem. > > > > > Another problem - I used to be able to access a virtual terminal by > pressing ctl-alt-f2. This key sequence no longer works. I'm running > gnome. How can I temporarily disable X so I can run alsaconf from > a virtual terminal? Open a Gnome terminal or an xterm, become root and run chvt 2 to go to virtual terminal 2, for example. Log in as your normal user, su to root level and run /etc/init.d/gdm stop This will shut down Gnome and X. The command to start it again is /etc/init.d/gdm start As far as the ctrl-alt-fx problem is concerned, have a look at this thread: http://lists.debian.org/debian-user/2006/04/msg02492.html If that does not help you, start a new thread and post the keyboard section of your /etc/X11/xorg.conf. -- Regards, Florian -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
RE: APC (American Power Conversion) APC BACK-UPS RS 800VA
> > Seconded. acpupsd works fine on an AMD64 system. > true. follow the debian readme. first test how your ups reacts to -killpower. new APC devices have killpower delay, and in that case you should send the killpower command in the "doshutdown" event in apccontrol - you have to add one line as described in the readme. Zrin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
