[Cooker] Helium SCSI install problem?
Hello fellow Cookers, Under Helium 7.1, I'm trying to set up a new SCSI-based server. However, my 4g boot drive is setting itself to LUN #6 on the bus. Unfortunately, Helium's SCSI module during install seems to only be scanning up to probably LUN #4 (happens to be the tape drive). Therefore, it never even gets around to seeing the 6th LUN. Is there *ANY* way to send a LILO switch, or a command or anything on the F2 install console, to get it to scan all the LUNs? I understand this is a compile-time thing, possibly. What would I need to do to recompile the kernel on the install disk, as that is most likely a stripped version of some other kernel. Please reply to [EMAIL PROTECTED], as I had to take down the mailserver that my subscription to this list was being processed by. ;) Thanks! --Scott Balmos
Re: [Cooker] Wheel mouse and XF4
Under the section Mouse, as usual... Option "ZAxisMapping" "4 5" -- Scott Balmos [EMAIL PROTECTED] On Fri, 28 Jul 2000, Eugenio Diaz wrote: > What is the equivalent of "ZAxisMapping 4 5" in the new XF86Config for > XF4? > > -- > Eugenio Diaz, BSEE/BSCE > Linux Engineer > [EMAIL PROTECTED] > > >
Re: [Cooker] SMP & IO-APIC IRQs?
No, it wasn't bundled as-is, I will admit. The machine came with the SL22V. I added in the second processor. I requested model V, but I got model T. Still within the same stepping, I thought they could work. I'm trying to contact the dealer now, as the sale was only a week ago. I'll also try the vanilla kernel attempt when I get home. Thanks. ;) Anyonen else? LOL On Wed, 26 Jul 2000, Francis Galiegue wrote: > On Wed, 26 Jul 2000, Scott Balmos wrote: > > > Hi again fellow cookers, > > > > I am trying to install Helium onto an HP Vectra XU 6/200 series server, > > equipped with dual Pentium Pro-200/256k cache chips. Both of the > > processors are Stepping 9, but the main processor is a model SL22V, and > > the secondary is model SL22T. > > > > Was the machine bundled as is? Looks strange. > > > When the system boots, it looks like all is going well, until it hits the > > synchronization part for the two procs. Without trying to hand type what's > > on the screen, I can basically say that it looks like the kernel > > calibrates the first processor fine, detects the second processor and > > calibrates it fine, is able to add together the two BogoMIPS (like we > > care), and then craps out on this setting the IO-APIC IRQ line. > > > > I've had a large amount of experience with computers, but not really with > > multiproc ones. What the heck is IO-APIC? I'm thinking something like > > Automatic InterProcessor Communication? > > > > Advanced Programmable Interrupt Controller - an evolution of PCs' PICs. Allows > interrupts to be distributed among processors. No APICs, no SMP. > > > Anyway, here are the last few lines before the kernel locks: > > > > enabling symmetric IO mode... ...done. > > ENABLING IO-APIC IRQs > > ...changing IO-APIC physical APIC ID to 16 > > Kernel anic: could not set ID > > in swapper task - not syncing > > > > It goes to note, obviously, that I can easily boot into the uniproc > > version of the multiproc kernel, which is 2.2.15-mdk4secure, if it > > requires a recompilation. > > > > Please let me know if this is a kernel problem, a BIOS thing maybe, or the > > fact that the two procs have to be both the same stepping and model? Maybe > > I have to flip them to get the "newer" SL22V in the secondary position? > > > > I'd say that yes, the two processors have to be the same. I've no idea about > the differences between the two models. > > You can try and compile a vanilla SMP kernel (ie, one from ftp.kernel.org) and > see if you can reproduce the problem. > >
Re: [Cooker] SMP & IO-APIC IRQs?
I cannot determine the BIOS or chipset information from here at work, at least from the Vectra's PDF manuals. I will try and look later when I get home, along with trying that config option. I will say that the docs say that the motherboard is APIC compatible. And if you had really read my message, I said that the procs are not identical. They are the same stepping, but different models. However, for all intents and purposes, the BIOS correctly sees both procs, and I thought Linux did also up until the point of the APIC problem. If needed I will try and look for another identical processor to one of them. But I need definite assurance that the fact that they're different models *IS* the problem. Like I said, I'll try that config switch when I get home. Thanks for any more answers! --Scott Balmos ([EMAIL PROTECTED]) On 26 Jul 2000, Thierry Vignaud wrote: > The following message is a courtesy copy of an article > that has been posted to liste.cooker as well. > > [EMAIL PROTECTED] (Scott Balmos) writes: > > I am trying to install Helium onto an HP Vectra XU 6/200 series server, > > equipped with dual Pentium Pro-200/256k cache chips. Both of the > > processors are Stepping 9, but the main processor is a model SL22V, and > > the secondary is model SL22T. > > > > When the system boots, it looks like all is going well, until it hits the > > synchronization part for the two procs. Without trying to hand type what's > > on the screen, I can basically say that it looks like the kernel > > calibrates the first processor fine, detects the second processor and > > calibrates it fine, is able to add together the two BogoMIPS (like we > > care), and then craps out on this setting the IO-APIC IRQ line. > > > > I've had a large amount of experience with computers, but not really with > > multiproc ones. What the heck is IO-APIC? I'm thinking something like > > Automatic InterProcessor Communication? > > > > Anyway, here are the last few lines before the kernel locks: > > > > enabling symmetric IO mode... ...done. > > ENABLING IO-APIC IRQs > > ...changing IO-APIC physical APIC ID to 16 > > Kernel anic: could not set ID > > in swapper task - not syncing > > > > It goes to note, obviously, that I can easily boot into the uniproc > > version of the multiproc kernel, which is 2.2.15-mdk4secure, if it > > requires a recompilation. > > > > Please let me know if this is a kernel problem, a BIOS thing maybe, or the > > fact that the two procs have to be both the same stepping and model? Maybe > > I have to flip them to get the "newer" SL22V in the secondary position? > > in which mode your bios is : MP1.1 or 1.4 ?? > > does options "noapic" and/or "reboot=bios" works ?? > > have you two identical CPU ?? (same step, family, ...) ?? > > what chipset (440FX is bogus) >
[Cooker] SMP & IO-APIC IRQs?
Hi again fellow cookers, I am trying to install Helium onto an HP Vectra XU 6/200 series server, equipped with dual Pentium Pro-200/256k cache chips. Both of the processors are Stepping 9, but the main processor is a model SL22V, and the secondary is model SL22T. When the system boots, it looks like all is going well, until it hits the synchronization part for the two procs. Without trying to hand type what's on the screen, I can basically say that it looks like the kernel calibrates the first processor fine, detects the second processor and calibrates it fine, is able to add together the two BogoMIPS (like we care), and then craps out on this setting the IO-APIC IRQ line. I've had a large amount of experience with computers, but not really with multiproc ones. What the heck is IO-APIC? I'm thinking something like Automatic InterProcessor Communication? Anyway, here are the last few lines before the kernel locks: enabling symmetric IO mode... ...done. ENABLING IO-APIC IRQs ...changing IO-APIC physical APIC ID to 16 Kernel anic: could not set ID in swapper task - not syncing It goes to note, obviously, that I can easily boot into the uniproc version of the multiproc kernel, which is 2.2.15-mdk4secure, if it requires a recompilation. Please let me know if this is a kernel problem, a BIOS thing maybe, or the fact that the two procs have to be both the same stepping and model? Maybe I have to flip them to get the "newer" SL22V in the secondary position? Thanks! Scott Balmos ([EMAIL PROTECTED])
[Cooker] GDM Broadcast?
Hi everyone, This may amount to nothing more than a config line change in my system that I have yet to find, as usual, but I am wondering whether or not it would be possible to run GDM's computer chooser at login time? I can easily have it run, obviously, if I'm running a remote X session to my computer from the outside with the XDM broadcast protocol. But what if I'm not doing anything remote. What if I'm just sitting at my normal Helium GDM login screen? Is there a way to select and remotely log into another XDMCP-capable box on the local LAN from just normal Helium? I will admit, the only way that I can really think at the moment is to drop down to the console in runlevel 3 and then manually type out a startx command with the remote computer's address in the DISPLAY variable. But since GDM has a computer chooser available, like all good XDM clones, this seems an easy fix. Anyone? :) --Scott Balmos ([EMAIL PROTECTED])
Re: [Cooker] Erratic Intellimouse Optical Behavior
sorry... Realized that a couple days ago after I intially sent the message, just never got around to writing a response. LOL It's running off of the PS/2 port, is the Intellimouse Explorer (the more expensive one), and I'm running X 4.0... If you think it might possibly work better on the USB port for whatever reason, I can switch it over at my next reboot Saturday night. Thanks for the insight! --Scott On Thu, 6 Jul 2000, Guy T. Rice wrote: > Scott, you didn't mention whether you were using PS/2 or USB. For what > it's worth, my Intellimouse Optical works flawlessly. But (1) it's a > different mouse (your subject line says Intellimouse Optical but your > message said Intellimouse Explorer. What's the difference? About $20. > And the Intellimouse Optical is symetrical, i.e. it works right handed > or left handed equally well, whereas the Intellimouse Explorer is > sculpted asymetrically IIRC), and (2) I'm using XFree86 3.3.6. Also, > I do use it on the USB port. >
[Cooker] Erratic Intellimouse Optical Behavior
Good morning everyone (at least for me), I have lived with this problem for a little while, and now I've finally fed up. First off, I do not believe this problem ever came up for me in 7.0... I have an Intellimouse Explorer (the optical kind with all the whizzo side-buttons and everything... oh BTW, anyone know how to keymap those buttons in Nutscrape and elsewhere for back and forward, like they were in the 'doze world?). It runs great, wheel runs great, and all life is happy. But there are numerous times per day that while I'm working in X (X 4.0, if it makes a difference) that the system seems to lose track of where the mouse is tracking to. It locks the mouse and refuses to move the pointer UNTIL I switch out to a non-X virtual terminal (good time to check the logs on console 12 for anyone deciding to reverse ICMP me to death), and then back into X. The mouse resets itself, and life returns to normal for a few minutes, until I have to do it again. This can get fairly annoying a lot of times, as there is usually a 3 second or so delay between me hitting Ctrl-Alt-F12 and it actually switching to that console. And there have been numerous times where I've been working in Gnome, listening to XMMS, etc etc etc and generally raising heck on my machine that when it switches consoles, the system freezes hard, requiring a reset-button push. Here's the details of my system, just in case: Hydrogen Beta 1 2.2.15 (yeah I know... I have the release discs next to me. Just too lazy... :D) X-4.0 with the nvidia drivers 192 megs of RAM, K6-2-300 Gnome 1.2 from Helix MS Intellimouse Explorer (optical) imwheel loaded kudzu loaded (long explanation, but the mouse seems to "power up" when this loads during the bootsequence) Thanks for the ideas. And if anyone knows those button-bindings for those side butttons, I'd be really thankful. :D --Scott Balmos ([EMAIL PROTECTED])
Re: [Cooker] RpmDrak suddenly loses list of packages to install
Thanks Pixel. FYI, I did have to add in the media tags, so it ended up being: urpmi.addmedia cdrom_1 removable_cdrom_1://mnt/cdrom2/Mandrake/RPMS "/Mandrake/RPMS2 Thanks anyway! On 25 Jun 2000, Pixel wrote: > Scott Balmos <[EMAIL PROTECTED]> writes: > > > Definite problem here. :D > > > > [sbalmos@sbalmos urpmi]$ ls -l > > total 104 > > -rw-r--r--1 root root 101174 May 12 22:44 depslist > > -rw-r--r--1 root root 0 May 19 18:35 list.cdrom1 > > -rw-r--r--1 root root 0 May 16 20:48 list.cdrom2 > > > > Looks like urpmi needs a swift kick in the jaw. So... How to rebuild the > > indexes? > > well, it should be: > > - urpmi.removemedia -a > - put cd1 in drive > - urpmi.addmedia removable_cdrom_1://mnt/cdrom > - put cd2 in drive > - urpmi.addmedia removable_cdrom_2://mnt/cdrom >
Re: [Cooker] RpmDrak suddenly loses list of packages to install
Definite problem here. :D [sbalmos@sbalmos urpmi]$ ls -l total 104 -rw-r--r--1 root root 101174 May 12 22:44 depslist -rw-r--r--1 root root 0 May 19 18:35 list.cdrom1 -rw-r--r--1 root root 0 May 16 20:48 list.cdrom2 Looks like urpmi needs a swift kick in the jaw. So... How to rebuild the indexes? On 24 Jun 2000, Pixel wrote: > Scott Balmos <[EMAIL PROTECTED]> writes: > > > Hello everyone, > > > > Odd problem that I've never seen before in Mandrake... I'm not sure if > > it's just lost track or what, but RpmDrak has suddenly lost its tree of > > what packages are *available to be installed*. It keeps the list of > > currently installed packages just fine. I checked the default installation > > paths of cdrom1 and cdrom2, which I assume lead to the package lists of > > the two installation CD-ROMs. Those are fine, and are found easily. > > > > So... question is... How do I 1) either rebuild the RPM database of what's > > available, or 2) Kick RpmDrak in the rear by readding the original two > > installation paths? > > look at /var/lib/urpmi, there should be hdlist.cz2 and another one. > mail the ls -l of /var/lib/urpmi, it should help >
[Cooker] RpmDrak suddenly loses list of packages to install
Hello everyone, Odd problem that I've never seen before in Mandrake... I'm not sure if it's just lost track or what, but RpmDrak has suddenly lost its tree of what packages are *available to be installed*. It keeps the list of currently installed packages just fine. I checked the default installation paths of cdrom1 and cdrom2, which I assume lead to the package lists of the two installation CD-ROMs. Those are fine, and are found easily. So... question is... How do I 1) either rebuild the RPM database of what's available, or 2) Kick RpmDrak in the rear by readding the original two installation paths? Thanks! Scott Balmos ([EMAIL PROTECTED])
[Cooker] Getting GDM to accept remote connections
Hi everyone, Before I start, yes I know XDMCP is an insecure protocol. But I'm sure that running XDMCP will be much faster than VNC, which runs slower than molasses even over the localhost loopback interface. So the question is... How go I get GDM (Gnome's XDM) to accept incoming remote X sessions? I've already read GDM's docs, set the appropriate Enable=1 line in gdm.conf under [xdmcp], and done everything else. Yet when I try to set up a remote session from my Windows laptop running an X server for Win (3 different kinds, so it can't be a single client-software problem), all of them timeout on the connection and/or never connect. What gives? Any ideas? BTW as far as I know, I've only got X running on Display 0. Do I need to run it on Display 1? It's not like I'm going to be support multi-X sessions, as I'm the only one who accesses the system and I can't be in two different places at the same time. ;) Thanks for your help! --Scott Balmos ([EMAIL PROTECTED])
[Cooker] RPMs of Non-Helix Gnome
Would it be safe to simply grab the standard vanilla Gnome RPMs from the Gnome site, or should I try building RPMs on my own from sources? I'm thinking about getting rid of the Helix extensions, as I've traced GTK program incompatibilities as best as I can to various calls that Helix has dubbed over. If there are 7.x-specific plain Gnome RPMs somewhere, could I be informed? Thanks! --Scott Balmos ([EMAIL PROTECTED])
[Cooker] Various minor problems
'evening all again, Forget my last message about chewed memory. See what happens when you leave the Linux scene for 5 months to beta test Winderz 2x10^-16 Not-So-Advanced Server? You get used to reading its reports. :D Anyway, here's a few problems I've seen so far with 7.1B2, some of which have probably been fixed already and I just have never seen an email about it: 1) Obviously the KDE menubuilder isn't in there... which completely left me clueless as to what I had installed until I actually wrote out a list of the binary names (yes I'm dull). Heh. :D 2) The EsounD daemon and the KDE sound servers are both set to not be executable by anyone but root. I understand how there may be a security breach by allowing normal people to use the mic input to possibly eavesdrop on the server (??? Read that somewhere... Linux.com I think?). But to a normal user who has his/her sound card configured just fine, can play sound in X as root, but not as a normal user, I'd get freaked out. 3) Apache's default content directory, /home/httpd, is set to the wrong user and usergroup. Instead of root/root it should be set to the user that Apache runs under, which is nobody/nobody if I remember correctly. Otherwise Apache will scream a 403 Forbidden. ;) 4) I don't know what the heck happened to WU-FTPd between 7.0 and 7.1, but two odd things happen: 1) The FTP server, by itself, does something strange. Whenever any user (normal or anonymous) logs in, the FTP server, by all that I can tell, tries to access a directory listing of /mnt before it dumps the user into their respective home directory. I can tell this because I have automount installed and whenever I try to log into my FTP server, I get greeted by a nice stream of console messages stating that the kernel can't access my floppy and two CD drives. Afterwards, all is fine and I go on with my business in my home directory. 2) Anonymous FTP is broken. On top of the above problem, the anonymous user is dumped into some empty directory of God-knows-where, not /home/ftp. This is in spite of the fact that the "ftp" user is correctly set to a home directory of /home/ftp. Don't ask me anything more about that, as I'm still tracking the logs... which are another story. 5) Speaking of logs, logrotate.conf has multiple entries for various system logs (messages, secure.log, etc) when you look at the include file logrotate.d and the ones listed in the main conf file. 6) When starting syslog at boot-time, syslog complains about the nonexistence of various news-server related log files, even though I don't have a news server installed. Such files are news.critical news.err, etc. Sorry I can't remember exact filenames at the moment. 7) Finally(!!!), Postfix seems completely broken with procmail. This has happened since MDK-7 release. Apparently Postfix can't correctly run procmail for local mail delivery. Hence such, all mail for any user on the system ends up being dumped into deferred/ for all eternity. Now if you remove Postfix and dump Sendmail in its place, it works like a charm. Just a few tiny bites out of a candybar for everyone to chew on. --Scott Balmos [EMAIL PROTECTED]
[Cooker] Horribly-bloated memory usage?
Good evening everyone, While this may seem a rather odd question... It seems alarming to me that my system (running 7.1 Beta 2) loves to idle around 185 megs when everything's all and done loaded up. That's with MySQL, Apache, blah blah blah... loaded, and X alone chewing 60 megs. Does this seem even *REMOTELY* possible? I imagine it might, with all the binaries probably having debugging symbols in them. On a tangent to that question, what part of the filesystem is it *NOT* safe to use "strip" on? If it is the debugging symbols that are the problem, then I'd might as well strip them out save for KDE 2. :) Thanks for the info! --Scott Balmos ([EMAIL PROTECTED])
