RE: Debian on Sparc 4 - woody to sarge - more progress
Hi Jurij et all, I couldn't seem to see why I could type the full path and it would work, yet SILO couldn't get the info from silo.conf. My symlinks are there and so are the files. In the end I went OTT in silo.conf but it has fixed the problem. Now I need to see why I had it in the first place. Below is the info you requested, even though I can now reboot with no issues really. I have a few aesthetics to look at, why it changed my display to white and why I cannot see the correct Debian install dialogs (the blue ones). Being able to upgrade the kernel and access the debian site was the main issue here and that appears to be solved. Thanks a lot for your time Jurij et al, Steve dns:/# cat /etc/silo.conf root=/dev/sda1 partition=1 default=linux timeout=100 read-only image=1/boot/vmlinuz-2.4.27-2-sparc32 initrd=1/boot/initrd.img-2.4.27-2-sparc32 root=/dev/sda1 label=linux image=1/vmlinuz.old label=linux dns:/# ls -al total 55 drwxr-xr-x 20 root root 1024 Jul 1 16:35 . drwxr-xr-x 20 root root 1024 Jul 1 16:35 .. drwxr-xr-x2 root root 2048 Jul 1 13:06 bin drwxr-xr-x2 root root 1024 Jul 2 01:57 boot drwxr-xr-x2 root root 1024 Jun 30 21:38 cdrom drwxr-xr-x5 root root19456 Jul 2 11:48 dev drwxr-xr-x 40 root root 2048 Jul 2 11:48 etc drwxr-xr-x2 root root 1024 Jun 30 21:38 floppy drwxrwsr-x4 root staff1024 Jun 30 20:01 home drwxr-xr-x2 root root 1024 Jun 30 21:38 initrd lrwxrwxrwx1 root root 32 Jul 1 16:35 initrd.img -> boot/initrd.img-2.4.27-2-sparc32 drwxr-xr-x5 root root 4096 Jun 30 21:51 lib drwx--2 root root12288 Jun 30 20:31 lost+found drwxr-xr-x2 root root 1024 Feb 8 2002 mnt drwxr-xr-x2 root root 1024 Jun 30 21:38 opt dr-xr-xr-x 35 root root0 Jul 2 2005 proc drwxr-xr-x3 root root 1024 Jul 1 18:00 root drwxr-xr-x2 root root 2048 Jul 1 13:02 sbin drwxrwxrwt3 root root 1024 Jul 2 11:40 tmp drwxr-xr-x 13 root root 1024 Jun 30 21:38 usr drwxr-xr-x 14 root root 1024 Jun 30 21:38 var lrwxrwxrwx1 root root 29 Jul 1 16:35 vmlinuz -> boot/vmlinuz-2.4.27-2-sparc32 lrwxrwxrwx1 root root 27 Jun 30 20:51 vmlinuz.old -> boot/vmlinuz-2.2.20-sun4cdm dns:/# ls -al /boot/ total 3567 drwxr-xr-x2 root root 1024 Jul 2 01:57 . drwxr-xr-x 20 root root 1024 Jul 1 16:35 .. -rw-r--r--1 root root 321829 Jun 30 20:51 System.map-2.2.20-sun4cdm -rw-r--r--1 root root 314634 Feb 15 11:38 System.map-2.4.27-2-sparc32 -rw-r--r--1 root root 1024 Aug 15 2002 cd.b -rw-r--r--1 root root 2982 Jun 30 20:51 config-2.2.20-sun4cdm -rw-r--r--1 root root12700 Feb 15 09:02 config-2.4.27-2-sparc32 -rw-r--r--1 root root 1024 Aug 15 2002 fd.b -rw-r--r--1 root root 512 Aug 15 2002 first.b -rw-r--r--1 root root 1024 Aug 15 2002 generic.b -rw-r--r--1 root root 784 Aug 15 2002 ieee32.b -rw-r--r--1 root root 1101824 Jul 1 16:34 initrd.img-2.4.27-2-sparc32 -rw-r--r--1 root root 7680 Jun 30 21:53 old.b -rw-r--r--1 root root59904 Jul 2 11:44 second.b -rw-r--r--1 root root57512 Aug 15 2002 silotftp.b -rw-r--r--1 root root 512 Aug 15 2002 ultra.b -rw-r--r--1 root root 890942 Jun 30 20:51 vmlinuz-2.2.20-sun4cdm -rw-r--r--1 root root 842884 Feb 15 11:37 vmlinuz-2.4.27-2-sparc32 -Original Message- From: Jurij Smakov [mailto:[EMAIL PROTECTED] Sent: 02 July 2005 02:40 To: Steve Lewis Cc: debian-sparc@lists.debian.org Subject: RE: Debian on Sparc 4 - woody to sarge - more progress On Sat, 2 Jul 2005, Steve Lewis wrote: > Hi All, > > Well, the sunlance addition to the modules.conf worked like a charm > and I now see my interface card and can get on my network. > Unfortunately I still have the boot problem. > > I think this is related to my silo.conf but have tried the following > with no luck: - > > First attempt: - > > root=/dev/sda1 > partition=1 > default=Linux > timeout=100 > read-only > > image=1/vmlinuz > label=linux > initrd=initrd.img Try changing that to image=1/vmlinuz label=linux initrd=1/initrd.img Do you have the /initrd.img -> /boot/initrd.img-2.4.27-sparc32 symbolic link set up? If you still have problems booting, post your final silo.conf file and the output of the following commands: ls -la / ls -la /boot Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/
RE: Debian on Sparc 4 - woody to sarge - more progress
On Sat, 2 Jul 2005, Steve Lewis wrote: Hi All, Well, the sunlance addition to the modules.conf worked like a charm and I now see my interface card and can get on my network. Unfortunately I still have the boot problem. I think this is related to my silo.conf but have tried the following with no luck: - First attempt: - root=/dev/sda1 partition=1 default=Linux timeout=100 read-only image=1/vmlinuz label=linux initrd=initrd.img Try changing that to image=1/vmlinuz label=linux initrd=1/initrd.img Do you have the /initrd.img -> /boot/initrd.img-2.4.27-sparc32 symbolic link set up? If you still have problems booting, post your final silo.conf file and the output of the following commands: ls -la / ls -la /boot Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
RE: Debian on Sparc 4 - woody to sarge - more progress
Title: Message Hi All, Well, the sunlance addition to the modules.conf worked like a charm and I now see my interface card and can get on my network. Unfortunately I still have the boot problem. I think this is related to my silo.conf but have tried the following with no luck: - First attempt: - root=/dev/sda1partition=1default=Linuxtimeout=100read-only image=1/vmlinuz label=linux initrd=initrd.img image=1/vmlinuz.old label=linux -- Second attempt: - root=/dev/sda1partition=1default=Linuxtimeout=100read-only image=/vmlinuz label=linux initrd=initrd.img image=/vmlinuz.old label=linux The reason for trying without the 1 in the image line was because I thought the 'partition=1' was perhaps making it read image=1/1/vmlinuz I also tried moving the file to the boot directory and doing a silo -C /boot/silo.conf, but that didn't work either. This _could_ be my last hurdle to finishing off the kernel upgrade and move to sarge on my SPARC4. I should really do this on my P3 as well, but may take a bit of time off :-) Thanks for your help all, Steve
RE: Debian on Sparc 4 - woody to sarge - some progress
I was too eager ... thanks Jurij, I tried modprobe sunlance.o instead of just sunlance. I was kinda getting into the right general idea but led myself astray a bit it seems. modprobe sunlance worked just fine. I shall have a good look around to see if there are any other modules which should be listed in /etc/modules as I certainly missed that. Now the question is this, as the interface issue will be cleared up on reboot, the error I experienced when I rebooted the new kernel: - VFS: Disk quotas vdquot_6.5.1 devfs: v1.12c (20020818) Richard Gooch ([EMAIL PROTECTED]) devfs: boot_options 0x0 Console: switching to mono PROM 80x34 pty: 256 Unix98 ptys configured Floppy drive(s): fd0 is 1.44M FDC 0 is a post-1991 82077 RAMDISK driver initialized: 16 RAM disks of 812K size 1024 blocksize Initializing Cryptographic API NET4: Linux TCP/IP 1.0 for NET4.0 IP: routing cache hash table of 512 buckets, 4Kbytes TCP: Hash tables configured (established 4096 bind 8192) NET4: Unix domian sockets 1.0/SMP for Linux NET4.0. VFS: Cannot open root device "sda1" or 08:01 Please append a correct "root=" boot option Kernel panic: VFS: Unable to mount root fs on 08:01 Press L1-A to return to the boot prom Was that a user issue because I have symlinks in / pointing to /boot and _also_ have the following silo.conf file? I still had the Please append a correct "root=" boot option even though it is clearly in my silo.conf $ cat /etc/silo.conf root=/dev/sda1 partition=1 timeout=100 read-only image=1/vmlinuz label=linux initrd=/initrd.img image=1/vmlinuz.old label=linuxOLD Getting there slowly but surely. I am still reluctant to reboot just yet as I would like to minimize fallout. Cheers, Steve -Original Message- From: Jurij Smakov [mailto:[EMAIL PROTECTED] Sent: 01 July 2005 23:58 To: Steve Lewis Cc: debian-sparc@lists.debian.org Subject: Re: Debian on Sparc 4 - woody to sarge - some progress On Fri, 1 Jul 2005, Patrick Morris wrote: > Have you tried "modprobe sunhme" ? > > On Fri, 01 Jul 2005, Steve Lewis wrote: > >> I see under the /lib/module/2.2.x directory there is no sunhme.o, yet >> there is under /lib/modules/2.4.x/kernel/drivers/net ... doesn't >> sunhme.o take care of my eth0 interface? ... or not as the case maybe Yes, the network card drivers used to be compiled into the kernel in 2.2.20, but are available as modules in 2.4. From previous postings it is clear that you have a Sun Lance network card, so the command to activate it should be 'modprobe sunlance', not sunhme. You can also add the name of the module (sunlance, that is) to /etc/modules, so it is automatically loaded on every boot. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- 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: Debian on Sparc 4 - woody to sarge - some progress
Hi Patrick et al, I hadn't, just did and ... "No such device" Also dmesg just shows the same as the kern.log. Ethernet address I used to (before kernel upgrade) have the following line in in kern.log on boot: - sunlance.c:v1.12 11/Mar/99 Miguel de Icaza ([EMAIL PROTECTED]) But I just noticed that it is missing since, also not present on dmesg. I thought perhaps that sunhme.o may find it, but perhaps that is just for the later ULTRA SPARC stations. I cannot see a lance module under the old /lib/modules/2.2.20 directory but as I mentioned in an earlier post, I thought only 2.4.x was modular. Interesting enough is that I have found a sunlance.o module under the following: - /lib/modules/2.4.27-2-sparc32/kernel/drivers/net/sunlance.o However, when I try and use it with modprobe I get the following error: - Modprobe: Can't locate module /lib/modules/2.4.27-2-sparc32/kernel/drivers/net/sunlance.o Strange too, modprobe -c shows that my kernel module path is /lib/module/kernel but the directory doesn't exist. I also have another top level path /lib/modules/2.4 which doesn't exist. Ah, in actual fact, I have just one toplevel directory which is correct, pointing to /lib/modules/2.4.27-2-sparc32 all the other path statements are pointing to /lib/modules/ and the directory doesn't exist. Perhaps that is why it cannot find the sunlance module. I suppose I could actually direct the output to a file, edit the paths and use the -C option to use the newly created config file. I wonder what I would break doing that :-) Cheers, Steve -Original Message- From: Patrick Morris [mailto:[EMAIL PROTECTED] Sent: 01 July 2005 23:31 To: Steve Lewis Cc: debian-sparc@lists.debian.org Subject: Re: Debian on Sparc 4 - woody to sarge - some progress Have you tried "modprobe sunhme" ? On Fri, 01 Jul 2005, Steve Lewis wrote: > I see under the /lib/module/2.2.x directory there is no sunhme.o, yet > there is under /lib/modules/2.4.x/kernel/drivers/net ... doesn't > sunhme.o take care of my eth0 interface? ... or not as the case maybe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Debian on Sparc 4 - woody to sarge - some progress
On Fri, 1 Jul 2005, Patrick Morris wrote: Have you tried "modprobe sunhme" ? On Fri, 01 Jul 2005, Steve Lewis wrote: I see under the /lib/module/2.2.x directory there is no sunhme.o, yet there is under /lib/modules/2.4.x/kernel/drivers/net ... doesn't sunhme.o take care of my eth0 interface? ... or not as the case maybe Yes, the network card drivers used to be compiled into the kernel in 2.2.20, but are available as modules in 2.4. From previous postings it is clear that you have a Sun Lance network card, so the command to activate it should be 'modprobe sunlance', not sunhme. You can also add the name of the module (sunlance, that is) to /etc/modules, so it is automatically loaded on every boot. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Debian on Sparc 4 - woody to sarge - some progress
Have you tried "modprobe sunhme" ? On Fri, 01 Jul 2005, Steve Lewis wrote: > I see under the /lib/module/2.2.x directory there is no sunhme.o, yet > there is under /lib/modules/2.4.x/kernel/drivers/net ... doesn't > sunhme.o take care of my eth0 interface? ... or not as the case maybe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
RE: Debian on Sparc 4 - woody to sarge - some progress
Title: Message Hi again All, After looking around a bit more it seems like this could be some sort of module issue. I have the correct settings in /etc/network/interfaces I noticed a post on the net regarding a similar issue which said that there was no discover1 seen in the output and that modules were being loaded specifically. Looking at the kern.log I cannot find a line with discover1 and I see 5 modules being loaded and lsmod shows them to be: - ext3 jbd sd_mod esp scsi_mod /etc/modules is empty but it says it should have a list of modules being loaded at boot time. under /etc/modutils/arch there isn't anything which looks like a sparc module, just alpha, atari, i386, various m68k files, various powerpc files, s390 and s390x Although in the modules.conf there is no processing of the arch files, so it seems that it isn't needed there. This could be tantamount to chasing a gazelle and about as much use as a chocolate fireguard but I see the file paths under modutils is empty but should specify a path to where the modules reside. Perhaps that is taken care of in the code somewhere as I see under /lib/modules there is a directory for both kernels. One thing to note here is that there are more modules under the old kernel, which as it is 2.2.x I wasn't sure if it was a modular kernel, I thought that was with 2.4.x only. I see under the /lib/module/2.2.x directory there is no sunhme.o, yet there is under /lib/modules/2.4.x/kernel/drivers/net ... doesn't sunhme.o take care of my eth0 interface? ... or not as the case maybe :-) I have managed to find the 5 modules under various subdirectories of /lib/modules/2.4.x why are there just 5 loaded? I cannot find a reference to the 5 modules loaded in either the /etc/modules.conf or the associated /etc/modutils files. Seems I will definitely need some help here as I cannot manually configure the interface. Even though there is a reference to it in the kern.log, I get a "No such device" when I try to configure it. Here's hoping that someone out there knows the answer as I am very reluctant to just start editing files and rebooting, which I still have not done by the way ;-) Regards all, Steve -Original Message-From: Steve Lewis [mailto:[EMAIL PROTECTED] Sent: 01 July 2005 20:20To: debian-sparc@lists.debian.orgSubject: RE: Debian on Sparc 4 - woody to sarge - some progress Hi All, Well, I gave it another bash, but I still experienced a problem with the reboot. Applicable output seen are: - VFS: Disk quotas vdquot_6.5.1 devfs: v1.12c (20020818) Richard Gooch ([EMAIL PROTECTED]) devfs: boot_options 0x0 Console: switching to mono PROM 80x34 pty: 256 Unix98 ptys configured Floppy drive(s): fd0 is 1.44M FDC 0 is a post-1991 82077 RAMDISK driver initialized: 16 RAM disks of 812K size 1024 blocksize Initializing Cryptographic API NET4: Linux TCP/IP 1.0 for NET4.0 IP: routing cache hash table of 512 buckets, 4Kbytes TCP: Hash tables configured (established 4096 bind 8192) NET4: Unix domian sockets 1.0/SMP for Linux NET4.0. VFS: Cannot open root device "sda1" or 08:01 Please append a correct "root=" boot option Kernel panic: VFS: Unable to mount root fs on 08:01 Press L1-A to return to the boot prom I have no idea what L1-A is and cannot seem to get a prompt, so I rebooted and used 'stop-a' to get the boot prompt. I used the following command: - linux image=1/boot/vmlinuz-2.4.27-2-sparc32 initrd=1/boot/initrd.img-2.4.2.27-2-sparc32 root=/dev/sda1 It appeared to be working, decompressing the image, but then I just got what appeared to be the prom prompt back: - Uncompressing image... Loading initial ramdisk PROMLIB: obio_ranges 1 bootmem_init: Scan sp_banks, init_bootmem(spfn[20a],bpfn[20a],mlpfn[3faf]) free_bootmem: base[0] size[3faf000] reserve_bootmem: base[100] size[10d000] reserve_bootmem: base[0] size[20a000] reserve_bootmem: base[20a000] size[7f8] Booting Linux Type 'go' to resume Type help for more information Typing 'go' booted into Linux with the upgraded kernel surprisingly, I thought I was back at the beginning again ... shows you how often I have done that before ... never :-) So, before doing anything else, I thought it would be wise to check a few things and also understand why I got the problem booting into the new kernel and what to do to repair everything. I found out my ethernet interface is missing, an ifconfig -a just shows my loopback, which is possibly unfortunate. There is no associated hostname file in /etc anymore. I think this means I will have to configure the kernel wh
RE: Debian on Sparc 4 - woody to sarge - some progress
Title: Message Hi All, Well, I gave it another bash, but I still experienced a problem with the reboot. Applicable output seen are: - VFS: Disk quotas vdquot_6.5.1 devfs: v1.12c (20020818) Richard Gooch ([EMAIL PROTECTED]) devfs: boot_options 0x0 Console: switching to mono PROM 80x34 pty: 256 Unix98 ptys configured Floppy drive(s): fd0 is 1.44M FDC 0 is a post-1991 82077 RAMDISK driver initialized: 16 RAM disks of 812K size 1024 blocksize Initializing Cryptographic API NET4: Linux TCP/IP 1.0 for NET4.0 IP: routing cache hash table of 512 buckets, 4Kbytes TCP: Hash tables configured (established 4096 bind 8192) NET4: Unix domian sockets 1.0/SMP for Linux NET4.0. VFS: Cannot open root device "sda1" or 08:01 Please append a correct "root=" boot option Kernel panic: VFS: Unable to mount root fs on 08:01 Press L1-A to return to the boot prom I have no idea what L1-A is and cannot seem to get a prompt, so I rebooted and used 'stop-a' to get the boot prompt. I used the following command: - linux image=1/boot/vmlinuz-2.4.27-2-sparc32 initrd=1/boot/initrd.img-2.4.2.27-2-sparc32 root=/dev/sda1 It appeared to be working, decompressing the image, but then I just got what appeared to be the prom prompt back: - Uncompressing image... Loading initial ramdisk PROMLIB: obio_ranges 1 bootmem_init: Scan sp_banks, init_bootmem(spfn[20a],bpfn[20a],mlpfn[3faf]) free_bootmem: base[0] size[3faf000] reserve_bootmem: base[100] size[10d000] reserve_bootmem: base[0] size[20a000] reserve_bootmem: base[20a000] size[7f8] Booting Linux Type 'go' to resume Type help for more information Typing 'go' booted into Linux with the upgraded kernel surprisingly, I thought I was back at the beginning again ... shows you how often I have done that before ... never :-) So, before doing anything else, I thought it would be wise to check a few things and also understand why I got the problem booting into the new kernel and what to do to repair everything. I found out my ethernet interface is missing, an ifconfig -a just shows my loopback, which is possibly unfortunate. There is no associated hostname file in /etc anymore. I think this means I will have to configure the kernel which I have also not done before, so it could be interesting, which is why I say it is _possibly_ unfortunate :-) Checking the kern.log I see this is indeed a SPARC4 sun4m which clears up an earlier question: - ARCH: SUN4M TYPE: SPARCstation 4 :-) I also see that it has seen my ethernet interface as it states the MAC in the log. Checking the syslog I see only the following: - Ethernet address: Whereas I used to see the below entries when it worked ok: - Ethernet address: eth0: LANCE eth0: using auto-carrier-detection. eth0: Carrier Lost, trying TPE As this was a completely basic system with hardly any software installed (wanted to upgrade it first) there is not too much to check, I can create, edit and remove files. It would be good to get to a system which needs no interaction on a reboot and can communicate with other devices though :-) Can anyone help out? Thanks again for your time all. Cheers, Steve
RE: Debian on Sparc 4 - woody to sarge
On Fri, 1 Jul 2005, Steve Lewis wrote: [..] root=/dev/sda1 partition=1 timeout=100 read-only image=1/vmlinuz label=linux initrd=/initrd.img image=1/vmlinuz.old label=linuxOLD initrd=/initrd.img.old I should then remove the existing symlink and create a new one called vmlinuz.old pointing to my 2.2.20 kernel and create a new symlink called vmlinuz pointing to the new 2.4.27 kernel Yes, except that initrd= option for 2.2.20 (woody) kernel is not necessary, since it does not need/support initrd. And you will also need to create a symlink: /initrd.img -> /boot/initrd.img-2.4.27-2-sparc32 [..] And failing that to be able to use the old kernel by doing: - linux image=1/boot/vmlinux-2.2.20-sun4cdm initrd=1/boot/initrd.img-2.2.20-sun4cdm root=/dev/sda1 Again, initrd option is not necessary for booting 2.2.20 kernel. Do these need to be proceeded with the command 'boot'? You'll need to type these at the 'boot:' prompt given by SILO, not at the PROM 'ok' prompt. Architecture-specific modutils configuration not found, using defaults depmod: *** Unresolved symbols in /lib/modules/2.2.20/hfs.o This can be safely ignored. Setting up ash (0.3.8-37) ... scsi0: MEDIUM ERROR on channel 0, id 3, lun 0, CDB: 0x03 00 00 00 10 00 Info fld=0x1e7ec2, Current sd08:07: sns = f0 3 ASC=11 ASCQ=43 Raw sense data:0xf0 0x00 0x03 0x00 0x1e 0x7e 0xc2 0x0c 0x0d 0x32 0x04 0x2a 0x11 0x15 0x80 scsidisk I/O error: dev 08:07, sector 590354 That looks like a hardware problem. It might be just a controller quirk, or the disk might actually have physical problems on it. I have not edited the silo.conf or symlinks just yet, I thought it wiser to see if there any tips from this message. I remember that when the kernel package is installed that there is a question about symlinking but cannot quite remember it, something about doing it from scratch. Yeah, after kernel package installation it will try to set up the symbolic links /boot/vmlinuz and /boot/initrd.img, pointing to the latest installed kernel and initrd image respectively. You can either use those in your silo.conf, or create symlinks manually in / and use silo.conf mentioned above. Either way should work. Good luck, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
RE: Debian on Sparc 4 - woody to sarge
Hi Jurij et all, Firstly, thanks for the help, it's appreciated. This is making more sense now ... finally ... sort of :) I have not as yet ever tried to upgrade a kernel and perhaps it would have been more straight forward if the releases had not just changed. I do not have a /boot partition. I have created a / partition on sda1 which is where my /boot directory is. Perhaps it would have been simpler if I had. I have vmlinux in / (root) which is symlinked to vmlinuz-2.2.20-sun4cdm in the /boot directory: - lrwxrwxrwx1 root root 27 Apr 2 16:52 vmlinuz -> boot/vmlinuz-2.2.20-sun4cdm My silo.conf reads: - partition=1 root=/dev/sda1 timeout=100 image=1/vmlinuz label=linux read-only So, it seems (as I can boot with no problems from the current kernel and silo.conf) that once I have gone through the downloading of packages and editing the /etc/apt/sources.list again, the silo.conf should be edited to read: - root=/dev/sda1 partition=1 timeout=100 read-only image=1/vmlinuz label=linux initrd=/initrd.img image=1/vmlinuz.old label=linuxOLD initrd=/initrd.img.old I should then remove the existing symlink and create a new one called vmlinuz.old pointing to my 2.2.20 kernel and create a new symlink called vmlinuz pointing to the new 2.4.27 kernel Any issues on rebooting the new kernel would allow me to boot using the following: - linux image=1/boot/vmlinux-2.4.27-2-sparc32 initrd=1/boot/initrd.img-2.4.27-2-sparc32 root=/dev/sda1 And failing that to be able to use the old kernel by doing: - linux image=1/boot/vmlinux-2.2.20-sun4cdm initrd=1/boot/initrd.img-2.2.20-sun4cdm root=/dev/sda1 Do these need to be proceeded with the command 'boot'? I am not very familiar with where initrd installs into but looking at what you have written it seems like it also installs into the /boot directory. Does that all sound ok? I have used apt-get to download and install some of the packages, to gain some headway. The following may be of interest: - apt-get install modutils Architecture-specific modutils configuration not found, using defaults depmod: *** Unresolved symbols in /lib/modules/2.2.20/hfs.o apt-get install initrd-tools Setting up ash (0.3.8-37) ... scsi0: MEDIUM ERROR on channel 0, id 3, lun 0, CDB: 0x03 00 00 00 10 00 Info fld=0x1e7ec2, Current sd08:07: sns = f0 3 ASC=11 ASCQ=43 Raw sense data:0xf0 0x00 0x03 0x00 0x1e 0x7e 0xc2 0x0c 0x0d 0x32 0x04 0x2a 0x11 0x15 0x80 scsidisk I/O error: dev 08:07, sector 590354 Setting up cramfsprogs (1.1-6.woody1) ... Setting up stat (3.3-2) ... Setting up initrd-tools (0.1.79-0.woody1) ... Seems the errors setting up ash is related to /dev/sda7 which is my /usr partition. I have no idea if any of the errors above will affect the kernel upgrade or not. I have also used wget to download the kernel package but have not used dpkg to install it as yet: - wget http://ftp.debian.org/debian/pool/main/k/kernel-image-2.4.27-sparc/kerne l-image-2.4.27-2-sparc32_2.4.27-2_sparc.deb I have not edited the silo.conf or symlinks just yet, I thought it wiser to see if there any tips from this message. I remember that when the kernel package is installed that there is a question about symlinking but cannot quite remember it, something about doing it from scratch. Cheers for your time all, Steve -Original Message- From: Jurij Smakov [mailto:[EMAIL PROTECTED] Sent: 01 July 2005 01:39 To: Steve Lewis Cc: 'Debian Sparc' Subject: RE: Debian on Sparc 4 - woody to sarge On Thu, 30 Jun 2005, Steve Lewis wrote: > I have since tried to upgrade my kernel to the > kernel-image-2.4.27-2-sparc32_2.4.27-2_sparc.deb package but had a > kernel panic, unable to mount root fs error on the reboot. I think > this could be a silo.conf error. I did add the initrd line to > silo.conf but I didn't include a root=/dev/sda1 line. I had a look > around for a decent (full and complete) example of a silo.conf file > but couldn't find one on the net. The manual page accessible with 'man silo.conf' command contains extensive documentation. Here's my working silo.conf: root=/dev/hda2 partition=1 default=Linux read-only timeout=100 image=/vmlinuz label=Linux initrd=/initrd.img image=/vmlinuz.old label=LinuxOLD initrd=/initrd.img.old Note that in my configuration /boot is a separate partition (/dev/hda1), which is also default (set by partition=1), so paths to the files are given relative to that partition. So setting image=/vmlinuz when the default partition is 1 will actually try to load file vmlinuz from the root of partition 1, which translates to /boot/vmlinuz. Also, if your silo.conf is broken, you still should be able to boot by typing something like this at the boot prompt: linux image=1/boot/vmlinux-2.4.27-2-sparc32 initrd=1/boot/initrd.img-2.4.27-2-sparc32 root=/dev/sda1 If you are still h
RE: Debian on Sparc 4 - woody to sarge
On Thu, 30 Jun 2005, Steve Lewis wrote: I have since tried to upgrade my kernel to the kernel-image-2.4.27-2-sparc32_2.4.27-2_sparc.deb package but had a kernel panic, unable to mount root fs error on the reboot. I think this could be a silo.conf error. I did add the initrd line to silo.conf but I didn't include a root=/dev/sda1 line. I had a look around for a decent (full and complete) example of a silo.conf file but couldn't find one on the net. The manual page accessible with 'man silo.conf' command contains extensive documentation. Here's my working silo.conf: root=/dev/hda2 partition=1 default=Linux read-only timeout=100 image=/vmlinuz label=Linux initrd=/initrd.img image=/vmlinuz.old label=LinuxOLD initrd=/initrd.img.old Note that in my configuration /boot is a separate partition (/dev/hda1), which is also default (set by partition=1), so paths to the files are given relative to that partition. So setting image=/vmlinuz when the default partition is 1 will actually try to load file vmlinuz from the root of partition 1, which translates to /boot/vmlinuz. Also, if your silo.conf is broken, you still should be able to boot by typing something like this at the boot prompt: linux image=1/boot/vmlinux-2.4.27-2-sparc32 initrd=1/boot/initrd.img-2.4.27-2-sparc32 root=/dev/sda1 If you are still having trouble, post your silo.conf and we should be able to figure it out. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
RE: Debian on Sparc 4 - woody to sarge
Thanks for the tip Martin, Looking at the page it seems I have the following, which is indeed sun4m: - SPARCstation 4 Processor(s): MicroSPARC II @ 70MHz Bus:SBus, 1 slot Architecture: sun4m Notes: Optional 16-bit audio, onboard framebuffer. All, I have since tried to upgrade my kernel to the kernel-image-2.4.27-2-sparc32_2.4.27-2_sparc.deb package but had a kernel panic, unable to mount root fs error on the reboot. I think this could be a silo.conf error. I did add the initrd line to silo.conf but I didn't include a root=/dev/sda1 line. I had a look around for a decent (full and complete) example of a silo.conf file but couldn't find one on the net. I was hoping that upgrading the kernel would allow me to move to the new stable 'sarge' release. Any good pointers? Once again though, getting my box back to a decent version of woody was not straight forward, I had to edit my source.list before completing the install as the floppy disks always point to stable, which of course is now 'sarge'. I tried getting the floppy rescue and driver files over the net thinking they may have been updated, but alas no. Cheers all, Steve -Original Message- From: Martin [mailto:[EMAIL PROTECTED] Sent: 30 June 2005 21:14 To: Debian Sparc Subject: RE: Debian on Sparc 4 > I think the SPARC 4 is a sun4m architecture although I am no guru on > SUN architecture at all. It is a sun4m machine. For hardware info for this generation of machine, this site: http://www.sunhelp.org/faq/sunref1.html is good (IMHO). > I don't know the differences between c and m but I > think the sun4c is the SPARC 1 and 2. They are a different abstract architecture, more critically sun4m machines use SPARC v8 processors rather then SPARC v7 in sun4c machines (I can't think of an exception - although there might be one). This only matters as SPARC v8 has unsigned int multiply and divides (as well as atomic swap I believe) which has to be emulated on SPARC v7 machines. In short, for a few key apps (such as ssh, GPG and libc), having support for SPARC v7 processors slows down the code on faster machines, thus (and as sun4c machines are getting quite old) there is limited support for them in Debian 3.1. > I shall have to take a closer > look at this although looking at the SUN site I am sure it is sun4m. > I shall see if sarge supports it. Perhaps the sun4m can use a => > 2.4.21 kernel and then be upgraded to sarge. I believe sarge on sun4m is viable, not sure about the status of 2.6 kernels on sun4m machines, know there used to be at least some bugs on SMP, check the archives and http://sparclinux.mit.edu/sparc/ if you have problems with it. HTH Cheers, - Martin -- 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]