Re: /boot
On Sat, Feb 26, 2005 at 11:11:50PM -0700, Grant Grundler wrote: Could you replace --update-partition= with --init-partition=? I'm expecting palo to blow everything on f0 away and build a new ext2 file system. THen you can: o mount /dev/sda1 /mnt o copy /boot/* /mnt o edit /etc/fstab so /dev/sda1 is mounted on /boot. If that boots, then: o unmount /boot That should be: o umount /boot o rm /boot/* o mount /boot As you might guess, the order is important. :^) -- Stuart Brady -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: PA-RISC/Linux Boot HOWTO Section 8, Draft 1, 2.4 - 2.6
On Sat, 26 Feb 2005 16:52:58 -0500 Harry Cochran [EMAIL PROTECTED] wrote: Hi, I don't know if this has any value or not, but I am happy to contribute it to PA-RISC/Linux Boot HOWTO with approval of those explicitly copied on this E-mail whose contributions I copied. Cheers, This looks interesting, I'll wrap up something a bit shorter probably (some of what is said is already in my recent changes in other sections), with the comments of Grant and others and will post it here for review, then I'll merge it with my current set of changes to the HOWTO for later publishing on TLDP. Thanks Thibaut VARENE The PA/Linux ESIEE Team http://www.pateam.org/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: PA-RISC/Linux Boot HOWTO Section 8, Draft 1, 2.4 - 2.6
Thanks, I will send an update as for you to use as a starting point as soon as I get through the -e2 business. If I had the money I would buy the newest, sexiest hppa box possible and give it to Dave Sill, author of Life with qmail. That's a guy who knows how to write documentation! Cheers, Harry -Original Message- From: Thibaut VARENE [mailto:[EMAIL PROTECTED] Sent: Sunday, February 27, 2005 6:30 AM To: Harry Cochran Cc: Grant Grundler; Kyle McMartin; debian-hppa@lists.debian.org Subject: Re: PA-RISC/Linux Boot HOWTO Section 8, Draft 1, 2.4 - 2.6 On Sat, 26 Feb 2005 16:52:58 -0500 Harry Cochran [EMAIL PROTECTED] wrote: Hi, I don't know if this has any value or not, but I am happy to contribute it to PA-RISC/Linux Boot HOWTO with approval of those explicitly copied on this E-mail whose contributions I copied. Cheers, This looks interesting, I'll wrap up something a bit shorter probably (some of what is said is already in my recent changes in other sections), with the comments of Grant and others and will post it here for review, then I'll merge it with my current set of changes to the HOWTO for later publishing on TLDP. Thanks Thibaut VARENE The PA/Linux ESIEE Team http://www.pateam.org/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: /boot
On Sat, Feb 26, 2005 at 11:55:19PM -0500, Harry Cochran wrote: Fourth try to identify error: ... On Sat, Feb 26, 2005 at 1:12AM -0500, Grant Grundler wrote: How about reading palo --help? I have it pasted on my wall :-). SinoHub5:/etc# cat palo.conf --commandline=1/vmlinux initrd=1/initrd.img root=/dev/sda3 HOME=/ --format-as=2 --update-partition=/dev/sda SinoHub5:/etc# palo palo version 1.5 [EMAIL PROTECTED] Tue Sep 21 15:14:17 MDT 2004 OK we're doing a format as ext2 ipl: addr 16384 size 36864 entry 0x0 Warning! Boot loader header version is 3, I only know how to handle version 4. It MIGHT work anyway. Could you replace --update-partition= with --init-partition=? Done. Now palo doesn't give an error. I'm expecting palo to blow everything on f0 away and build a new ext2 file system. THen you can: o mount /dev/sda1 /mnt o copy /boot/* /mnt o edit /etc/fstab so /dev/sda1 is mounted on /boot. Done. And it boots 2.6.8-2-32-smp! If that boots, then: o unmount /boot No ... umount /boot or umount -f /boot comes back with: device is busy o rm /boot/* o mount /boot Is there a problem leaving the redundant /boot on sda3? As you might guess, the order is important. :^) This, plus 10 tries at booting, all tells me that: 1. You can't specify a recovery kernel with this method (not very comforting ... but palo finds 2.4.27-32 every time anyway). I was wondering why you didn't specify one in your example. Now I know. One isn't needed. palo can list the contents of an ext2 directory. So any vmlinux/initrd can be used if palo can list it. 2. initrd has nothing to do with the v3 vs. v4 problem right 3. I can't find the magic in front of the symlinks vmlinux and initrd.img to get 2.6 to load. I get failed to load ram disk every time with 10 tries at initrd=x, where x is 0/initrd.img, 1/initrd.img, 1/boot/initrd.img, /boot.initrd.img, etc. Or maybe it's something else entirely. Can you dump the following info (cut/paste to a shell): fdisk -l /dev/sda SinoHub5:/# fdisk -l /dev/sda Disk /dev/sda: 64 heads, 32 sectors, 17366 cylinders Units = cylinders of 2048 * 512 bytes Device BootStart EndBlocks Id System /dev/sda1 * 195 97264 f0 Linux/PA-RISC boot /dev/sda296 1072 1000448 82 Linux swap /dev/sda3 1073 17366 16685056 83 Linux mount SinoHub5:/# mount /dev/sda3 on / type ext2 (rw,errors=remount-ro) proc on /proc type proc (rw) /dev/sda1 on /boot type ext2 (rw) /dev/sdb1 on /var type ext3 (rw) cat /etc/palo.conf SinoHub5:/# cat /etc/palo.conf --commandline=1/vmlinux initrd=1/initrd.img root=/dev/sda3 HOME=/ --update-partition=/dev/sda ls -l /boot SinoHub5:/# ls -l /boot total 0 The ls command assumes the f0 partition is mounted on /boot. Well I was feeling pretty good until I saw ls -l /boot come back with total 0. I assume that's bad news. You didn't say to change init-partitioned to update-partitioned, so maybe I screwed up there. Should I have just gotten rid of that line? Having an f0 partition and an e2 /boot is looking pretty good right now! /boot on f0 partition definitely works. I'm certainly not the only one using it. hth, grant Thanks again Grant. This is very close now. I just need one more push to complete. I hope you have time to reply today. Best regards, Harry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: PA-RISC/Linux Boot HOWTO Section 8, Draft 1, 2.4 - 2.6
Hi Thibaut, Here is an update, but it's not final. Grant's step 8.3 i. doesn't work (device is busy). I'm waiting to hear back from him. Cheers, Harry -Original Message- From: Thibaut VARENE [mailto:[EMAIL PROTECTED] Sent: Sunday, February 27, 2005 6:30 AM To: Harry Cochran Cc: Grant Grundler; Kyle McMartin; debian-hppa@lists.debian.org Subject: Re: PA-RISC/Linux Boot HOWTO Section 8, Draft 1, 2.4 - 2.6 On Sat, 26 Feb 2005 16:52:58 -0500 Harry Cochran [EMAIL PROTECTED] wrote: Hi, I don't know if this has any value or not, but I am happy to contribute it to PA-RISC/Linux Boot HOWTO with approval of those explicitly copied on this E-mail whose contributions I copied. Cheers, This looks interesting, I'll wrap up something a bit shorter probably (some of what is said is already in my recent changes in other sections), with the comments of Grant and others and will post it here for review, then I'll merge it with my current set of changes to the HOWTO for later publishing on TLDP. Thanks Thibaut VARENE The PA/Linux ESIEE Team http://www.pateam.org/ 8. Upgrading from a Debian hppa 2.4 Kernel to a 2.6 kernel 8.1 Check your 2.4 partition scheme 8.2 First Steps 8.3 Set up f0 as /boot if necessary 8.4 Edit /etc/palo.conf 8.5 Before you reboot 8.6 Conclusion 8.1 Check your 2.4 partition scheme It is relatively easy to set up a partition scheme that will work for a while, but then stop working when root starts to fill up. For example: palo ipl 1.5 [EMAIL PROTECTED] Tue Sep 21 15:14:05 MDT 2004 Partition Start(MB) End(MB) Id Type 1 1 31 f0 Palo 2 321008 82 swap 31009 17366 83 ext2 root Don't despair. There is a way to tell palo to format your f0 partition, copy everything from /boot over too it and the modify fstab to make the f0 partition mount as /boot (See 8.3). Here's an example that works: Disk /dev/sda: 73.4 GB, 73407865856 bytes 255 heads, 63 sectors/track, 8924 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System /dev/sda1 *1 24 192748+ f0 Linux/PA-RISC boot /dev/sda2 25 148 996030 82 Linux swap /dev/sda3 149 214016000740 83 Linux /dev/sda4 2141 892454492480 83 Linux and here's the palo.conf that goes with it: --commandline=0/vmlinux-2.6.11-rc2-pa3-UP root=/dev/sda3 panic=5 console=ttyS1 --update-partitioned=/dev/sda 8.2 First steps I'm the conservative type, so I don't load from testing. At the time of this writing, the 2.6 kernels in unstable are 2.6.8-2-32, 2.6.8-2-32-smp, 2.6.8-2-64 and 2.6.8-2-64-smp. Step 1 For example, apt-get install kernel-image-2.6.7-2-32-smp. When you get near the end of the install, you will get messages telling you that this is an initrd kernel and asking you if you want to abort. At this point say, Yes. Now it's time for: Step 2 you need to add sym53c8xx to /etc/mkinitrd/modules. modules is created from /etc/mkinitrd/modules.dpkg-new. This is how it looks after editing: % cat /etc/mkinitrd/modules.dpkg-new # /etc/mkinitrd/modules: Kernel modules to load for initrd. # # This file should contain the names of kernel modules and their arguments # (if any) that are needed to mount the root file system, one per line. # Comments begin with a `#', and everything on the line after them are ignored. # # You must run mkinitrd(8) to effect this change. # # Examples: # # ext2 # wd io=0x300 tulip sym53c8xx Okay. Now we are ready for: Step 3 apt-get install kernel-image-2.6.8-2-32-smp again, but this time when asked if you want to abort, say, No. Step 4 Run apt-get upgrade Step 5 Run palo -? and make sure your palo has -format-as=type 8.3 Set up f0 as /boot if necessary One of the weaknesses of the current hppa installer is that it won't let you mount the f0 partition as /boot. If you created a small (50MB-200MB) ext2 partition and mounted it as /boot during the install process, you can skip this section and go to 8.4. If not, and you have a / that goes out beyond 2GB from the beginning of your disk, here is your salvation. Of course, if you're starting from scratch, you can set up an f0 and an ext2 partition that you mount as /boot, but the preferred partitioning scheme is to just set up one f0 partition during installation (at least 50MB for a production system and say 200MB for a development system) and then use palo to format it, copy everything from /boot over too it and the modify fstab to make the f0 partition mount as /boot. Here's what to do: a. Edit /etc/palo.conf, on the --commandline change your pointer to vmlinux to 1/vmlinux and insert initrd=1/initrd.img. Add a line that says --format-as=2 Leave the line that says --init-partitioned=/dev/sda (or similar) Remove the line that says --recoverykernel=/boot/vmlinux.old (or similar) b. Now run palo. This
Re: /boot
On Sun, Feb 27, 2005 at 08:55:41AM -0500, Harry Cochran wrote: On Sat, Feb 26, 2005 at 1:12AM -0500, Grant Grundler wrote: I'm expecting palo to blow everything on f0 away and build a new ext2 file system. THen you can: o mount /dev/sda1 /mnt o copy /boot/* /mnt o edit /etc/fstab so /dev/sda1 is mounted on /boot. Done. And it boots 2.6.8-2-32-smp! Great! If that boots, then: o unmount /boot No ... umount /boot or umount -f /boot comes back with: device is busy Did you do a cd /boot, by any chance? If so, you need to go out of /boot before you can unmount it. Is there a problem leaving the redundant /boot on sda3? No. In fact, you need it -- it's the mount point. The redundant /boot (on sda3) would ideally be empty, but do not delete the contents of the mounted /boot (on sda1)! If there's anything inside /boot on sda3, you simply won't see it while sda1 is mounted there, which could be confusing, but shouldn't cause any harm. -- Stuart Brady -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: (SPAM: 50) Mail Delivery (failure jobs@google.com) (VIRUS REMOVED)
We want to thank you for your interest in joining the Google team. We received your email inquiry and look forward to the opportunity to review your background and experience. Unfortunately, we are unable to give a personal reply to every applicant. However, please know that we do review all resumes by hand so it takes us just a little bit longer to get back to applicants we feel might be a fit for one of our positions. If you do not hear from one of us, we may not have a position available for you at this time. We thank you for your patience and want to thank you again for your interest in Google. Google Staffing http://www.google.com/jobs.html -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /boot
On Sun, Feb 27, 2005 at 08:55:41AM -0500, Harry Cochran wrote: On Sat, Feb 26, 2005 at 1:12AM -0500, Grant Grundler wrote: ls -l /boot SinoHub5:/# ls -l /boot total 0 The ls command assumes the f0 partition is mounted on /boot. Well I was feeling pretty good until I saw ls -l /boot come back with total 0. I assume that's bad news. You didn't say to change init-partitioned to update-partitioned, so maybe I screwed up there. Should I have just gotten rid of that line? If sda1 is mounted on /boot and it is empty, then yes, that's bad. If I understand correctly, --init-partition formats the partition, and as a result, clears any files that were it previously held... If that's right, it's better to pass --init-partition to palo, than it is to add --init-partition in palo.conf, since you won't have to worry about changing it back afterwards. You'll have to copy the files from /boot again (using /mnt for sda1, making sure you've unmounted /boot first). -- Stuart Brady -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: /boot
On Sun, Feb 27, 2005 at 08:55:41AM -0500, Harry Cochran wrote: On Sat, Feb 26, 2005 at 1:12AM -0500, Grant Grundler wrote: ls -l /boot SinoHub5:/# ls -l /boot total 0 The ls command assumes the f0 partition is mounted on /boot. Well I was feeling pretty good until I saw ls -l /boot come back with total 0. I assume that's bad news. You didn't say to change init-partitioned to update-partitioned, so maybe I screwed up there. Should I have just gotten rid of that line? On Sun, Feb 27, 2005 at 11:46:41AM -0500, Stuart Brady wrote: If sda1 is mounted on /boot and it is empty, then yes, that's bad. If I understand correctly, --init-partition formats the partition, and as a result, clears any files that were it previously held... Am I right that with /dev/sda1 mounted as /boot, if I am at say /var a cd /boot should go to the /dev/sda1 partition? If so, that /boot is empty as shown above. Now for the bad news ... cd / and then cd boot which should bring me to the boot directory under the root is also empty. Arrrg. Will an apt-get install kernel-image-2.6.8-2-32-smp reload /dev/sda1 with the files I need to boot? If that's right, it's better to pass --init-partition to palo, than it is to add --init-partition in palo.conf, since you won't have to worry about changing it back afterwards. Assuming I can get a loadable kernel and initrd.img back on the /boot on /dev/sda1/, I guess I don't need --update-partitioned /dev/sda in palo.conf. Can you please confirm that? You'll have to copy the files from /boot again (using /mnt for sda1, making sure you've unmounted /boot first). -- Stuart Brady Also I did cd /var and then tried umount /boot and got device is busy still. What am I doing wrong? Thanks for your help Stuart! Harry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /boot
On Sun, Feb 27, 2005 at 08:55:41AM -0500, Harry Cochran wrote: Done. And it boots 2.6.8-2-32-smp! good. Now you know what to write up :^) If that boots, then: o unmount /boot No ... umount /boot or umount -f /boot comes back with: device is busy Sounds like sda1 did not get mounted as /boot Does mount list /boot? Is there a problem leaving the redundant /boot on sda3? Only that it might cause confusion later. SinoHub5:/# fdisk -l /dev/sda ... Good ... SinoHub5:/# mount /dev/sda3 on / type ext2 (rw,errors=remount-ro) proc on /proc type proc (rw) /dev/sda1 on /boot type ext2 (rw) /dev/sdb1 on /var type ext3 (rw) Good SinoHub5:/# cat /etc/palo.conf Good SinoHub5:/# ls -l /boot total 0 Not Good. vmlinux/initrd symlinks and *-2.6.8.1-2-* files should be here. The ls command assumes the f0 partition is mounted on /boot. Well I was feeling pretty good until I saw ls -l /boot come back with total 0. I assume that's bad news. Right. You didn't say to change init-partitioned to update-partitioned, so maybe I screwed up there. Should I have just gotten rid of that line? No - you've already run palo --init-partition once and that's sufficient. The vmlinux/initrd didn't get copied to sda1 when it was mounted as /mnt. umount sda1 by hand and mount /dev/sda1 /mnt so you can clone the original contents of /boot (sda3) to /mnt (sda1). Thanks again Grant. This is very close now. I just need one more push to complete. I hope you have time to reply today. For some definition of today :^) welcome, grant -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /boot
On Sun, Feb 27, 2005 at 04:45:37PM +, Stuart Brady wrote: If sda1 is mounted on /boot and it is empty, then yes, that's bad. If I understand correctly, --init-partition formats the partition, and as a result, clears any files that were it previously held... correct If that's right, it's better to pass --init-partition to palo, than it is to add --init-partition in palo.conf, since you won't have to worry about changing it back afterwards. Yes - palo.conf should have --update-partition in it by default. (any Debian-installer folks noting this?) You'll have to copy the files from /boot again (using /mnt for sda1, making sure you've unmounted /boot first). *nod* grant -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /boot
On Sun, Feb 27, 2005 at 12:00:25PM -0500, Harry Cochran wrote: Am I right that with /dev/sda1 mounted as /boot, if I am at say /var a cd /boot should go to the /dev/sda1 partition? If so, that /boot is empty as shown above. Now for the bad news ... cd / and then cd boot which should bring me to the boot directory under the root is also empty. No - mount /dev/sda1 /boot will overlay and hide the original contents on /dev/sda3. That's why we keep telling you to mount /dev/sda1 /mnt and then copy the /boot contents to /mnt. Arrrg. Will an apt-get install kernel-image-2.6.8-2-32-smp reload /dev/sda1 with the files I need to boot? Probably not. But dpkg -i kernel-image-2.6.8-2-32-smp.deb will. (just get the name of the .deb file right) Assuming I can get a loadable kernel and initrd.img back on the /boot on /dev/sda1/, I guess I don't need --update-partitioned /dev/sda in palo.conf. Can you please confirm that? You do need --update-partitioned /dev/sda in palo.conf. Also I did cd /var and then tried umount /boot and got device is busy still. What am I doing wrong? It sounds like /boot is the original directory on /dev/sda3. No one uses /boot for anything *except* palo to load vmlinux/initrd. grant -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /boot
Harry Cochran wrote: Also I did cd /var and then tried umount /boot and got device is busy still. What am I doing wrong? I encounter the same pb a time ago: if /boot is a fs mounted at boot time iirc you would have to stop klog daemon (iirc too fuser -c /boot would have to show you the process(es) using your fs). hth, Joel -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: /boot
On Sun, Feb 27, 2005 at 12:00:25PM -0500, Harry Cochran wrote: Am I right that with /dev/sda1 mounted as /boot, if I am at say /var a cd /boot should go to the /dev/sda1 partition? If so, that /boot is empty as shown above. Now for the bad news ... cd / and then cd boot which should bring me to the boot directory under the root is also empty. On Sun, Feb 27, 2005 at 12:37:25PM -0500, Grant Grundler wrote: No - mount /dev/sda1 /boot will overlay and hide the original contents on /dev/sda3. So you are saying that unless you can umount /boot, you can't get to /boot on sda3, right? The problem that started this thread is that I can't umount /boot. I just get device is busy. I have cd'd to /var which is on sdb1, but I still can't umount /boot. That's why we keep telling you to mount /dev/sda1 /mnt and then copy the /boot contents to /mnt. Arrrg. Will an apt-get install kernel-image-2.6.8-2-32-smp reload /dev/sda1 with the files I need to boot? Probably not. But dpkg -i kernel-image-2.6.8-2-32-smp.deb will. (just get the name of the .deb file right) Actually apt-get remove kernel-image-2.6.8-2-32-smp and then apt-get install kernel-image-2.6.8-2-32-smp did work with a lot of complaining :-) (I did it before I got your email). Assuming I can get a loadable kernel and initrd.img back on the /boot on /dev/sda1/, I guess I don't need --update-partitioned /dev/sda in palo.conf. Can you please confirm that? You do need --update-partitioned /dev/sda in palo.conf. Thanks ... glad I got at least one thing right :-). Also I did cd /var and then tried umount /boot and got device is busy still. What am I doing wrong? It sounds like /boot is the original directory on /dev/sda3. No one uses /boot for anything *except* palo to load vmlinux/initrd. Bottom line is I can't umount /boot or umount /dev/sda1 ... I always get device is busy. I know I'm missing something here, but I don't know what it is. Thanks, Harry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /boot
On Sun, Feb 27, 2005 at 01:06:13PM -0500, Harry Cochran wrote: On Sun, Feb 27, 2005 at 12:37:25PM -0500, Grant Grundler wrote: No - mount /dev/sda1 /boot will overlay and hide the original contents on /dev/sda3. So you are saying that unless you can umount /boot, you can't get to /boot on sda3, right? Correct. The problem that started this thread is that I can't umount /boot. I just get device is busy. I have cd'd to /var which is on sdb1, but I still can't umount /boot. Run this before you try to unmount /boot: /etc/init.d/klogd stop (Thanks to Joel for pointing this out.) Actually apt-get remove kernel-image-2.6.8-2-32-smp and then apt-get install kernel-image-2.6.8-2-32-smp did work with a lot of complaining :-) (I did it before I got your email). Great, but remember that sda1 should be mounted as /boot when you do the install -- otherwise, you'll have to move the files across afterwards. -- Stuart Brady -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: /boot
Harry Cochran wrote: Also I did cd /var and then tried umount /boot and got device is busy still. What am I doing wrong? Joel Soete wrote: I encounter the same pb a time ago: if /boot is a fs mounted at boot time iirc you would have to stop klog daemon (iirc too fuser -c /boot would have to show you the process(es) using your fs). That was it. Thanks so much Joel!! Harry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: PA-RISC/Linux Boot HOWTO Section 8, Final Draft, 2.4 - 2.6
Hi Thibaut, Okay. As far as I'm concerned, it's soup. I will be interested to see what you come up with out of this. Thanks to all for your help and patience! Cheers, Harry 8. Upgrading from a Debian hppa 2.4 Kernel to a 2.6 kernel 8.1 Check your 2.4 partition scheme 8.2 First Steps 8.3 Set up f0 as /boot if necessary 8.4 Edit /etc/palo.conf 8.5 Before you reboot 8.6 Conclusion 8.1 Check your 2.4 partition scheme It is relatively easy to set up a partition scheme that will work for a while, but then stop working when root starts to fill up. For example: palo ipl 1.5 [EMAIL PROTECTED] Tue Sep 21 15:14:05 MDT 2004 Partition Start(MB) End(MB) Id Type 1 1 31 f0 Palo 2 321008 82 swap 31009 17366 83 ext2 root Don't despair. There is a way to tell palo to format your f0 partition, copy everything from /boot over too it and the modify fstab to make the f0 partition mount as /boot (See 8.3). Here's an example that works: Disk /dev/sda: 73.4 GB, 73407865856 bytes 255 heads, 63 sectors/track, 8924 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System /dev/sda1 *1 24 192748+ f0 Linux/PA-RISC boot /dev/sda2 25 148 996030 82 Linux swap /dev/sda3 149 214016000740 83 Linux /dev/sda4 2141 892454492480 83 Linux and here's the palo.conf that goes with it: --commandline=0/vmlinux-2.6.11-rc2-pa3-UP root=/dev/sda3 panic=5 console=ttyS1 --update-partitioned=/dev/sda 8.2 First steps I'm the conservative type, so I don't load from testing. At the time of this writing, the 2.6 kernels in unstable are 2.6.8-2-32, 2.6.8-2-32-smp, 2.6.8-2-64 and 2.6.8-2-64-smp. Step 1 For example, apt-get install kernel-image-2.6.7-2-32-smp. When you get near the end of the install, you will get messages telling you that this is an initrd kernel and asking you if you want to abort. At this point say, Yes. Now it's time for: Step 2 you need to add sym53c8xx to /etc/mkinitrd/modules. modules is created from /etc/mkinitrd/modules.dpkg-new. This is how it looks after editing: % cat /etc/mkinitrd/modules.dpkg-new # /etc/mkinitrd/modules: Kernel modules to load for initrd. # # This file should contain the names of kernel modules and their arguments # (if any) that are needed to mount the root file system, one per line. # Comments begin with a `#', and everything on the line after them are ignored. # # You must run mkinitrd(8) to effect this change. # # Examples: # # ext2 # wd io=0x300 tulip sym53c8xx Okay. Now we are ready for: Step 3 apt-get install kernel-image-2.6.8-2-32-smp again, but this time when asked if you want to abort, say, No. Step 4 Run apt-get upgrade Step 5 Run palo -? and make sure your palo has -format-as=type 8.3 Set up f0 as /boot if necessary One of the weaknesses of the current hppa installer is that it won't let you mount the f0 partition as /boot. If you created a small (50MB-200MB) ext2 partition and mounted it as /boot during the install process, you can skip this section and go to 8.4. If not, and you have a / that goes out beyond 2GB from the beginning of your disk, here is your salvation. Of course, if you're starting from scratch, you can set up an f0 and an ext2 partition that you mount as /boot, but the preferred partitioning scheme is to just set up one f0 partition during installation (at least 50MB for a production system and say 200MB for a development system) and then use palo to format it, copy everything from /boot over too it and the modify fstab to make the f0 partition mount as /boot. Here's what to do: a. Edit /etc/palo.conf, on the --commandline change your pointer to vmlinux to 1/vmlinux and insert initrd=1/initrd.img. Add a line that says --format-as=2 Leave the line that says --init-partitioned=/dev/sda (or similar) Remove the line that says --recoverykernel=/boot/vmlinux.old (or similar) b. Now run palo. This will format your f0 partition. c. Assuming your f0 partition is on sda1, mount /dev/sda1 mnt d. cp /boot/* /mnt e. edit /etc/fstab so that /dev/sda1 is mounted as /boot by adding this line: # file system mount point type optionsdump pass /dev/sda1 /boot ext2defaults 0 2 f. Edit /etc/palo.conf and remove the --format-as=2 line and change the --init-partitioned=/dev/sda to --update-partitioned=dev/sda. g. Now run palo. h. Reboot. If this fails, look at 8.4, go back over the steps above and try again. If it works, i. Run /etc/init.d/klogd stop (if you dont stop klogd, youll get device is busy when you try to umount /boot in the next step) j. umount /boot k. rm /boot/* l. mount /boot If you get through step l., you're done. 8.4 Edit /etc/palo.conf Edit /etc/palo.conf and insert on the --commandline initrd=x/boot/initrd.img where x is
Re: /boot
Stuart Brady wrote: Great, but remember that sda1 should be mounted as /boot when you do the install -- otherwise, you'll have to move the files across afterwards. yes Joel -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
C360 with A4071B Graphics STI Display Problem
Hello folks, I bought a C360 a while back and it sat in the cupboard for quite some time until the other day when I installed Debian 3.0r4 on it. Today I managed to get my hands on a monitor that can sync-on-green so I've hooked everything up and it all goes swimmingly until Linux boots, at which point the screen goes blank (not blank as in no-signal blank, blank as in [EMAIL PROTECTED] blank). When I ask boot admin about my I/O modules it tells me this about the graphics adapter: == IODC IODC Path Decimal Type Location HVER SVER Vers Dep --- - -- ... GRAPHICS(2) 8/8 HPA4071B GSC slot 2 0040 7700 0x01 0x00 ... == So I go ahead and boot with the following kernel command line: == 2/vmlinux root=/dev/sda4 HOME=/ console=tty0 sti=8/8 sti_font=VGA8x16 TERM=linux == vmlinux is a symlink to the kernel that comes with Debian 3.0r4 (vmlinx-2.4.17-32) and 8/8 is indeed the path to the graphics adapter according to the boot firmware. When I ssh to the machine and dmesg | grep [sS][tT][iI], I get the following: == Kernel command line: root=/dev/sda4 HOME=/ console=tty0 sti=8/8 sti_font=VGA8x16 TERM=linux palo_kernel=2/vmlinux STI byte mode ROM at f400, hpa=f400 STI byte mode ROM, id 2bcb015a-9a02587, conforms to spec rev. 8.04 STI init_graph failed (ret -1, errno 258) sticonsole_init: searching for STI ROMs BUG: Skipping previously registered driver: sti (native) == And fbset -i gives: == open /dev/fb0: No such device == With kernel image vmlinux-2.6.10-pa11-32 from parisc-linux.org, dmesg | grep [sS][tT][iI] gives: == Kernel command line: root=/dev/sda4 HOME=/ console=tty0 sti=8/8 sti_font=VGA8x16 TERM=linux palo_kernel=2/vmlinux-2.6.10-pa11-32 STI GSC/PCI core graphics driver Version 0.9a STI init_graph failed (ret -1, errno 258) sti: probe of 8:8 failed with error 1 == And fbset -i again gives: == open /dev/fb0: No such device == With kernel image vmlinux-2.6.11-rc5-pa3-32 from parisc-linux.org, dmesg | grep [sS][tT][iI] gives: == Kernel command line: root=/dev/sda4 HOME=/ console=tty0 sti=8/8 sti_font=VGA8x16 TERM=linux palo_kernel=2/vmlinux-2.6.11-rc5-pa3-32 STI GSC/PCI core graphics driver Version 0.9a STI init_graph failed (ret -1, errno 258) sti: probe of 8:8 failed with error 1 == And fbset -i yet again gives: == open /dev/fb0: No such device == I have read the HOWTOs and I have Googled for several hours trying to solve this problem and have found only one specific mention of it with this graphics adapter: http://lists.debian.org/debian-hppa/2003/10/msg00073.html That message shows the same card working in a 715 but not a C360, so might it be a firmware thing? Surely if the boot firmware is able to output stuff to the display with this card then the kernel can use it (via STI)? Are there other parameters I can pass to the kernel to help it out? Does anyone out there have the same hardware, but working with no problems? I'll happily provide further info if it might help find a solution. Yours with heavy heart and eyelids more so, David Cosgrove -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
PA-RISC/Linux Boot HOWTO: 2.4 - 2.6
On Sun, Feb 27, 2005 at 02:05:45PM -0500, Harry Cochran wrote: 8.1 Check your 2.4 partition scheme It is relatively easy to set up a partition scheme that will work for a while, but then stop working when root starts to fill up. For example: This was due to using an incorrect partition scheme in the first place. I'm not sure that this belongs in a HOWTO that explains how to upgrade from 2.4 to 2.6. Documentation explaining how to use the palo partition as the /boot partition seems worth having, but it's surely a separate issue. I'm the conservative type, so I don't load from testing. At the time of this writing, the 2.6 kernels in unstable are 2.6.8-2-32, 2.6.8-2-32-smp, 2.6.8-2-64 and 2.6.8-2-64-smp. I don't get this -- testing is safer than unstable. -- Stuart Brady -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: PA-RISC/Linux Boot HOWTO: 2.4 - 2.6
On Mon, Feb 28, 2005 at 12:36:50AM +, Stuart Brady wrote: On Sun, Feb 27, 2005 at 02:05:45PM -0500, Harry Cochran wrote: 8.1 Check your 2.4 partition scheme It is relatively easy to set up a partition scheme that will work for a while, but then stop working when root starts to fill up. For example: This was due to using an incorrect partition scheme in the first place. I'm not sure that this belongs in a HOWTO that explains how to upgrade from 2.4 to 2.6. I agree. Maybe we need a FAQ for I upgraded my kernel and now it doesn't work? Documentation explaining how to use the palo partition as the /boot partition seems worth having, but it's surely a separate issue. In /usr/share/doc/palo/README perhaps? -- Next the statesmen will invent cheap lies, putting the blame upon the nation that is attacked, and every man will be glad of those conscience-soothing falsities, and will diligently study them, and refuse to examine any refutations of them; and thus he will by and by convince himself that the war is just, and will thank God for the better sleep he enjoys after this process of grotesque self-deception. -- Mark Twain -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: PA-RISC/Linux Boot HOWTO: 2.4 - 2.6
On Sun, Feb 27, 2005 at 02:05:45PM -0500, Harry Cochran wrote: 8.1 Check your 2.4 partition scheme It is relatively easy to set up a partition scheme that will work for a while, but then stop working when root starts to fill up. For example: On Sun, Feb 27, 2005 at 07:36:45PM -0500, Stuart Brady wrote: This was due to using an incorrect partition scheme in the first place. I'm not sure that this belongs in a HOWTO that explains how to upgrade from 2.4 to 2.6. Documentation explaining how to use the palo partition as the /boot partition seems worth having, but it's surely a separate issue. From a newbie point of view, I think it's important to know that you might have a partition scheme where you are booting something like 3/boot/vmlinux that might just stop working if root starts to fill up and because of format-as=2 you have a chance to fix it on a 2.6 upgrade. I'm the conservative type, so I don't load from testing. At the time of this writing, the 2.6 kernels in unstable are 2.6.8-2-32, 2.6.8-2-32-smp, 2.6.8-2-64 and 2.6.8-2-64-smp. I don't get this -- testing is safer than unstable. Thanks Stuart, I finally sat down and read the explanation of unstable and testing. Someone else told me that unstable was more reliable that testing which clearly isn't the case. Sorry for putting my misunderstanding in the document. Regards, Harry -- 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]