Re: starting X interferes with NFS
On Tue, Apr 18, 2006 at 12:42:12PM -0400, [EMAIL PROTECTED] wrote:
> On Tue, Apr 18, 2006 at 09:52:23AM -0400, Lennart Sorensen wrote:
> > On Tue, Apr 18, 2006 at 08:29:53AM -0400, [EMAIL PROTECTED] wrote:
> >
> > > Not that I have any particular aversion to xorg 7.0, but if that's
> > > needed it *might* be simplest to wait for it to stabilize into testing.
> > > I suspect that bug reports from mixing xorg7.0 into an etch system at
> > > this point would only confuse things, since it doesn't look as if it
> > > was quite ready for sid yet.
> >
> > Actually it is lookking pretty good already. It has actually gone
> > rather quick to get rid of the worst bugs in the packaging. Given it
> > went to entirely modular, I am impressed.
>
> Unfortunately, aptitude tells me:
>
> * nvidia-glx [nonfree] depends on x11-common (>= 1:7.0.0) [UNAVAILABLE]
>
> Evidently, it wants xorg 7.0 from sid, and not xorg 6.9 from etch.
> I hope the bugs ar out and it migrates down soon.
Managed to get
nvidia-glx_1.0.8756-1_amd64.deb
nvidia-glx-dev_1.0.8756-1_amd64.deb
nvidia-kernel-common_20051028+1_all.deb
nvidia-kernel-source_1.0.8756-1_amd64.deb
nvidia-settings_1.0+20051122-1_amd64.deb
using a private repository. The modules compiled properly, providing
nvidia-kernel-2.6.12-1-amd64-generic_1.0.8756-1_amd64.deb
The packages installed smoothly, and the X, gdm, and icewm come up
beautifully. Yes, beautifully. Except for a few analog artifacts, I've
never seen that old screen look that good.
Now for the problem.
I used gdm to start up idcmp (not sure I have the name right:
that's the thing that allows you to log into another machine from gdm)
on april, my AMD64 machine, which runs Debian AMD64 etch.
Used it to connect to lovesong (a machine running sarge, and named
after a T.S Eliot poem), got lovesong's login screen, and logged in.
icewm came up nicely. But when I started up emacs, everything froze.
That is, the window was created, with a frame and everything, but its
contents remained blank. The mouse pointer would no longer respong to
mouse movements, control-alt-backspace and conteol-alt-F1 would no
longer work. lovesong (where the emacs would actually have run) was
still up and unperturbed, but I was now unable to ssh into april from
lovesong ("no route to host").
I had to hard-reset and reboot april.
Is this a known problem?
Is it even likely to have been caused by xorg 6.9/nvidia8756?
Is it worth investigating right now? After all, I'm using old
software.
Should I just meekly wait until xorg 7.0 reaches etch and
then investigate only if the problem recurs?
By the way, I have the same home directory on the two systems. april
exports it to lovesong via NFS.
-- hendrik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: starting X interferes with NFS
On Tue, Apr 18, 2006 at 09:52:23AM -0400, Lennart Sorensen wrote: > On Tue, Apr 18, 2006 at 08:29:53AM -0400, [EMAIL PROTECTED] wrote: > > > Not that I have any particular aversion to xorg 7.0, but if that's > > needed it *might* be simplest to wait for it to stabilize into testing. > > I suspect that bug reports from mixing xorg7.0 into an etch system at > > this point would only confuse things, since it doesn't look as if it > > was quite ready for sid yet. > > Actually it is lookking pretty good already. It has actually gone > rather quick to get rid of the worst bugs in the packaging. Given it > went to entirely modular, I am impressed. Unfortunately, aptitude tells me: * nvidia-glx [nonfree] depends on x11-common (>= 1:7.0.0) [UNAVAILABLE] Evidently, it wants xorg 7.0 from sid, and not xorg 6.9 from etch. I hope the bugs ar out and it migrates down soon. -- hendrik > > > But picking up the new nvidia-glx package and seeing if it works in etch > > might still be useful as long as I can remember how to boot without X if > > it crashes hopelessly. > > > > And it probably wants a compatibly-numbered nvidia-kernel-source. I'll > > have to get that too. > > > > nvidia-glx_1.0.8756-4_amd64.deb > > nvidia-kernel-source_1.0.8756-4_amd64.deb > > Those would be the versions to use for that chipset. Anything older > won't work as far as I know. > > Len Sorensen > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: starting X interferes with NFS
On Tue, Apr 18, 2006 at 08:29:53AM -0400, [EMAIL PROTECTED] wrote: > Which, looking it up on packages.debian.org, is in stable and unstable, > but has somehow skipped by testing. That's why aptitude didn't find it. testing is VERY VERY incomplete for amd64 right now, since everything in unstable got rebuilt from scratch and is now moving in to testing when they can. testing has NOT been rebuilt from scratch since that is not how the buildds normally do things. libc6 has been holding back probably the majority of stuff from testing so far, but hopefully that will clear soon. > Not that I have any particular aversion to xorg 7.0, but if that's > needed it *might* be simplest to wait for it to stabilize into testing. > I suspect that bug reports from mixing xorg7.0 into an etch system at > this point would only confuse things, since it doesn't look as if it > was quite ready for sid yet. Actually it is lookking pretty good already. It has actually gone rather quick to get rid of the worst bugs in the packaging. Given it went to entirely modular, I am impressed. > But picking up the new nvidia-glx package and seeing if it works in etch > might still be useful as long as I can remember how to boot without X if > it crashes hopelessly. > > And it probably wants a compatibly-numbered nvidia-kernel-source. I'll > have to get that too. > > nvidia-glx_1.0.8756-4_amd64.deb > nvidia-kernel-source_1.0.8756-4_amd64.deb Those would be the versions to use for that chipset. Anything older won't work as far as I know. Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: starting X interferes with NFS
On Mon, Apr 17, 2006 at 10:01:34AM -0400, Lennart Sorensen wrote: > On Thu, Apr 13, 2006 at 11:15:20PM -0400, [EMAIL PROTECTED] wrote: > > Yeah. I've been trying to set up X on my ASUS A8N-VM-UAYGZ system. > > Its on-board graphics chip is, I believe, an GFORCE4 6300. This > > machine is successfully in production use as an NFS server, serving an > > 80G reiser filesystem on LVM on RAID-1 to several other machines on the > > LAN. > > > > I'm using AMD64 etch with Len Sorenson's -12 kernel, and csail at MIT is > > my package repository. I've managed to compile and install the > > so-called nvidia-kernel, but nvidia is not one of the options I get when > > running dpkg-reconfigure xserver-xorg. > > > > Selecting nv gets nowhere; when I boot, X tries to start > > up and fails, finding no screens. But more than that, when it starts > > and fails on boot, the NFS server doesn't seem to work. > > remote-mounting the big partition on another machine times out. > > > > Of the options I've tried, only vga works at all. X still doesn't > > start (apparently I've specified 24-bit graphics and VGA doesn't like > > it), but at least the NFS server comes up properly, and my remote users > > are happy. > > > > Is this a known problem? > > > > Actually, I'm not completely certain whether it was nv or vesa (or > > both) I was using when the NFS server failed to operate -- I'd have to > > try it again to be sure, but it's getting late and my users are > > complaining about yet another shutdown. I will test it again and give > > more complete details, probably tomorrow. > > > > In the meantime, if someone could tell me just what data to gather to > > help in diagnosis, I'd appreciate it. > > > > -- hendrik > > > > P.S. Does anyone know which NFS demon does proper locking *and* handles > > files larger than 4G? > > You only get the nvidia choice after you install nvidia-glx as well. Which, looking it up on packages.debian.org, is in stable and unstable, but has somehow skipped by testing. That's why aptitude didn't find it. > nvidia-kernel is the kernel module part. > > I suspect for your particular board you require the new driver that just > entered unstable, although you should be able to grab it and rebuild it > for amd64 if it hasn't already been built for amd64. Not sure if the > x.org 7.0 transition stuff will make it not work on older versions. Not that I have any particular aversion to xorg 7.0, but if that's needed it *might* be simplest to wait for it to stabilize into testing. I suspect that bug reports from mixing xorg7.0 into an etch system at this point would only confuse things, since it doesn't look as if it was quite ready for sid yet. But picking up the new nvidia-glx package and seeing if it works in etch might still be useful as long as I can remember how to boot without X if it crashes hopelessly. And it probably wants a compatibly-numbered nvidia-kernel-source. I'll have to get that too. nvidia-glx_1.0.8756-4_amd64.deb nvidia-kernel-source_1.0.8756-4_amd64.deb -- hendrik > > Len Sorensen > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: starting X interferes with NFS
On Thu, Apr 13, 2006 at 11:15:20PM -0400, [EMAIL PROTECTED] wrote: > Yeah. I've been trying to set up X on my ASUS A8N-VM-UAYGZ system. > Its on-board graphics chip is, I believe, an GFORCE4 6300. This > machine is successfully in production use as an NFS server, serving an > 80G reiser filesystem on LVM on RAID-1 to several other machines on the > LAN. > > I'm using AMD64 etch with Len Sorenson's -12 kernel, and csail at MIT is > my package repository. I've managed to compile and install the > so-called nvidia-kernel, but nvidia is not one of the options I get when > running dpkg-reconfigure xserver-xorg. > > Selecting nv gets nowhere; when I boot, X tries to start > up and fails, finding no screens. But more than that, when it starts > and fails on boot, the NFS server doesn't seem to work. > remote-mounting the big partition on another machine times out. > > Of the options I've tried, only vga works at all. X still doesn't > start (apparently I've specified 24-bit graphics and VGA doesn't like > it), but at least the NFS server comes up properly, and my remote users > are happy. > > Is this a known problem? > > Actually, I'm not completely certain whether it was nv or vesa (or > both) I was using when the NFS server failed to operate -- I'd have to > try it again to be sure, but it's getting late and my users are > complaining about yet another shutdown. I will test it again and give > more complete details, probably tomorrow. > > In the meantime, if someone could tell me just what data to gather to > help in diagnosis, I'd appreciate it. > > -- hendrik > > P.S. Does anyone know which NFS demon does proper locking *and* handles > files larger than 4G? You only get the nvidia choice after you install nvidia-glx as well. nvidia-kernel is the kernel module part. I suspect for your particular board you require the new driver that just entered unstable, although you should be able to grab it and rebuild it for amd64 if it hasn't already been built for amd64. Not sure if the x.org 7.0 transition stuff will make it not work on older versions. Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: starting X interferes with NFS
On Fri, Apr 14, 2006 at 08:51:34AM +0100, Matthew Robinson wrote, but forgot to reply to the list: > [EMAIL PROTECTED] wrote: > > >Yeah. I've been trying to set up X on my ASUS A8N-VM-UAYGZ system. > >Its on-board graphics chip is, I believe, an GFORCE4 6300. This > >machine is successfully in production use as an NFS server, serving an > >80G reiser filesystem on LVM on RAID-1 to several other machines on the > >LAN. > > > >I'm using AMD64 etch with Len Sorenson's -12 kernel, and csail at MIT is > >my package repository. I've managed to compile and install the > >so-called nvidia-kernel, but nvidia is not one of the options I get when > >running dpkg-reconfigure xserver-xorg. > > > >Selecting nv gets nowhere; when I boot, X tries to start > >up and fails, finding no screens. But more than that, when it starts > >and fails on boot, the NFS server doesn't seem to work. > >remote-mounting the big partition on another machine times out. > > > >Of the options I've tried, only vga works at all. X still doesn't > >start (apparently I've specified 24-bit graphics and VGA doesn't like > >it), but at least the NFS server comes up properly, and my remote users > >are happy. > > > >Is this a known problem? > > > >Actually, I'm not completely certain whether it was nv or vesa (or > >both) I was using when the NFS server failed to operate -- I'd have to > >try it again to be sure, but it's getting late and my users are > >complaining about yet another shutdown. I will test it again and give > >more complete details, probably tomorrow. > > > >In the meantime, if someone could tell me just what data to gather to > >help in diagnosis, I'd appreciate it. > > > >-- hendrik > > > >P.S. Does anyone know which NFS demon does proper locking *and* handles > >files larger than 4G? > > > > > > > > > The name of the binary nvidia driver is 'nvidia'. 'nv' is the name of > the open source driver (you could use that?). (1) 'nvidia' does not appear on the list of drivers when I do dpkg-reconfigure xserver-xorg. Presumably something still needs to be installed. (2) 'vga' does, and works. Unfortunately, it has a lot of limitations. I use icewm, and the tiny icons in the quarter-inch icon bar at the bottom show up slightly more than one inch high, so only a few fit. I can reach the 'logout' munu item, the dialog for confirming the logout is larger than the screen, but fortunately I can reach the 'OK' button. Seems to be (at the very least) a screen resolution problem. (3) But 'nv', the one I'm currently trying for, still doesn't come up. However, it no longer blocks the NFS server, which is a good thing. The log file is appended below. The video chip I have on the motherboard is described on the box as 'GeForce6100+nForce410'. As far as I can tell, it's not mentioned in the huge list of devices mentioned in the log. The key lines seem to be (EE) No devices detected. Fatal server error: no screens found > > have a look in /var/log/Xorg.0.log (or something like that) for lines > that match 'EE' Here's the /var/log/Xorg.0.log, cut-and-pasted from an emacs window. (emacs is running on a different machine from both the AMD64 with problems and the mail software. Love networks!) -- hendrik cut here --- X Window System Version 6.9.0 (Debian 6.9.0.dfsg.1-6 20060405004539 Frederik Sch??ler <[EMAIL PROTECTED]>) Release Date: 21 December 2005 X Protocol Version 11, Revision 0, Release 6.9 Build Operating System: Linux 2.6.16-1-vserver-amd64-k8 x86_64 [ELF] Current Operating System: Linux april 2.6.12-1-amd64-generic #1 Wed Sep 28 02:05:15 CEST 2005 x86_64 Build Date: 05 April 2006 Before reporting problems, check http://wiki.X.Org to make sure that you have the latest version. Module Loader present OS Kernel: Linux version 2.6.12-1-amd64-generic ([EMAIL PROTECTED]) (gcc version 4.0.2 (Debian 4.0.1-9)) #1 Wed Sep 28 02:05:15 CEST 2005 Markers: (--) probed, (**) from config file, (==) default setting, (++) from command line, (!!) notice, (II) informational, (WW) warning, (EE) error, (NI) not implemented, (??) unknown. (==) Log file: "/var/log/Xorg.0.log", Time: Fri Apr 14 09:54:20 2006 (==) Using config file: "/etc/X11/xorg.conf" (==) ServerLayout "Default Layout" (**) |-->Screen "Default Screen" (0) (**) | |-->Monitor "Generic Monitor" (**) | |-->Device "Generic Video Card" (**) |-->Input Device "Generic Keyboard" (**) Option "XkbRules" "xorg" (**) XKB: rules: "xorg" (**) Option "XkbModel" "pc104" (**) XKB: model: "pc104" (**) Option "XkbLayout" "us" (**) XKB: layout: "us" (==) Keyboard: CustomKeycode disabled (**) |-->Input Device "Configured Mouse" (WW) The directory "/usr/lib/X11/fonts/cyrillic" does not exist. Entry deleted from font path. (WW) The directory "/usr/lib/X11/fonts/CID" does not exist. Entry deleted from font path. (**) FontPath set to "unix/:7100,/usr/lib/X11/fonts/m
Re: starting X interferes with NFS
On Thu, Apr 13, 2006 at 11:15:20PM -0400, [EMAIL PROTECTED] wrote: > Yeah. I've been trying to set up X on my ASUS A8N-VM-UAYGZ system. > Its on-board graphics chip is, I believe, an GFORCE4 6300. This > machine is successfully in production use as an NFS server, serving an > 80G reiser filesystem on LVM on RAID-1 to several other machines on the > LAN. > > I'm using AMD64 etch with Len Sorenson's -12 kernel, and csail at MIT is > my package repository. I've managed to compile and install the > so-called nvidia-kernel, but nvidia is not one of the options I get when > running dpkg-reconfigure xserver-xorg. > > Selecting nv gets nowhere; when I boot, X tries to start > up and fails, finding no screens. But more than that, when it starts > and fails on boot, the NFS server doesn't seem to work. > remote-mounting the big partition on another machine times out. > > Of the options I've tried, only vga works at all. X still doesn't > start (apparently I've specified 24-bit graphics and VGA doesn't like > it), but at least the NFS server comes up properly, and my remote users > are happy. > > Is this a known problem? > > Actually, I'm not completely certain whether it was nv or vesa (or > both) I was using when the NFS server failed to operate -- I'd have to > try it again to be sure, but it's getting late and my users are > complaining about yet another shutdown. I will test it again and give > more complete details, probably tomorrow. > > In the meantime, if someone could tell me just what data to gather to > help in diagnosis, I'd appreciate it. > > -- hendrik > There are two issues here. One, which I expect will be fixed by time, effort, routine upgrades, very much appreciated advice from this list, and a lot of my virtual elbow grease, is tht I'm having touble getting X up. Th other, which I think is more serious, although it will probably affect me less, is that when C fails to start up properly at boot, it appears to disable the NFS server. That seems to suggest * a serious lack of compartmentalisation in the kernel. * X's failure to behave properly when failing to start up. -- hendrik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Correction:Re: starting X interferes with NFS
On Thu, Apr 13, 2006 at 11:15:20PM -0400, [EMAIL PROTECTED] wrote: > Yeah. I've been trying to set up X on my ASUS A8N-VM-UAYGZ system. > Its on-board graphics chip is, I believe, an GFORCE4 6300. Actually GeForce 6100. It says "GeForce6100+nForce410" on the box. -- hendrik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
starting X interferes with NFS
Yeah. I've been trying to set up X on my ASUS A8N-VM-UAYGZ system. Its on-board graphics chip is, I believe, an GFORCE4 6300. This machine is successfully in production use as an NFS server, serving an 80G reiser filesystem on LVM on RAID-1 to several other machines on the LAN. I'm using AMD64 etch with Len Sorenson's -12 kernel, and csail at MIT is my package repository. I've managed to compile and install the so-called nvidia-kernel, but nvidia is not one of the options I get when running dpkg-reconfigure xserver-xorg. Selecting nv gets nowhere; when I boot, X tries to start up and fails, finding no screens. But more than that, when it starts and fails on boot, the NFS server doesn't seem to work. remote-mounting the big partition on another machine times out. Of the options I've tried, only vga works at all. X still doesn't start (apparently I've specified 24-bit graphics and VGA doesn't like it), but at least the NFS server comes up properly, and my remote users are happy. Is this a known problem? Actually, I'm not completely certain whether it was nv or vesa (or both) I was using when the NFS server failed to operate -- I'd have to try it again to be sure, but it's getting late and my users are complaining about yet another shutdown. I will test it again and give more complete details, probably tomorrow. In the meantime, if someone could tell me just what data to gather to help in diagnosis, I'd appreciate it. -- hendrik P.S. Does anyone know which NFS demon does proper locking *and* handles files larger than 4G? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
