Re: Debian 3.0 on Netra X1, installs but won't run
On Tue, Jun 10, 2003 at 02:55:24PM -0700, Patrick Morris wrote: > Nope, it's not a Sparc thing, it's an IDE thing. Maybe it depends on > the motherboard, but just last week I was building a OpenBSD box out of > spare parts that wouldn't boot because the CD on the second IDE was > jumpered as a master. It's not an IDE thing, it's a CDROM thing. Some just don't want to be masters. I don't know why they even have jumpers. Others are OK being masters. I think most recent drives can handle being masters, but I could be wrong. In my experience it's pretty much a toss-up whether a given drive will work when jumpered as master. noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgpjtqdYd24dZ.pgp Description: PGP signature
Re: Got 2.4.20 to compile/run, but get NULL pointer dereference under load.
On Fri, Jun 06, 2003 at 08:02:08AM -0500, Greg Moeller wrote: > With a combo of egcs64 and gcc 2.95 I managed to get a kernel compiled which > had all the hardware/drivers I needed. (differential SCSI, SMP, RAID) > With some simple testing on a pair of RAID0 drives (in a D1000) using cp -a > I get: > Unable to handle kernel NULL pointer dereference Is the target filesystem ext3? There's some bug that causes this behavior. I could very easily reproduce it by copying lots of data (I was trying to move /usr/ to its own disk). I was completely unable to get any help from the ext3 devel list or the sparc kernel list, or the main kernel dev list, so it's probably still there. See if you get the crash when the target fs is ext2. And, as Ben already said, use ksymoops to give some meaning to that code dump you have. noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgppyisFRxDxZ.pgp Description: PGP signature
Re: sparc64 and ipsec
On Wed, May 28, 2003 at 07:33:31PM +0200, Andreas Pettersson wrote: > I've been planning to test ipsec on my Ultra 1. > But it seems that the FreeS/WAN kernel patches for 2.4.x won't compile on > sparc64 and 2.5.69 was horribly unstable. > So my question now is if there is any other way of getting ipsec for the > sparc64 platform? There is a backport of the IPsec implementation from 2.5 to 2.4. See ftp://ftp.kernel.org/pub/linux/kernel/davem/IPSEC/linux-2.4.21-ipsec.patch It's totally unsupported, but might work for you. noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgpZunpR1sKbG.pgp Description: PGP signature
Re: Kernel 2.4.20 and IPv6 ICMP on Sparc32/64 bomb
On Fri, May 09, 2003 at 11:05:15AM +0200, Peter Keel wrote: > ping6 on an IPv6-enabled Sparc32 or Sparc64 on Kernel 2.4.20 > bombs really nasty. Machine halts instantly. No matter what patches. You're using ping6 from iputils-ping? I maintain that package, and use it on a number of sparc systems. I do quite a bit of IPv6 stuff on these machines, using one as a tunnel gateway and another as a dual-stack web server. Never have I seen any crashes related to IPv6 or ICMPv6 or ping6 or any related code. If you want, I can send you an image of one of the kernels that I use on my Ultra1s. I *know* they do IPv6. It would be interesting to see if they have problems on your machines. noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgpaX9vyxdwKy.pgp Description: PGP signature
Re: Ultra 5 won't boot from HD after install
On Fri, May 02, 2003 at 07:47:09PM -0400, Ben Collins wrote: > Are you sure that you have the disk setup as Master, and not cable > select or something else weird. > Yup, entirely. I found another IDE disk (6x the capacity!) and it works fine. I don't know what was wrong before. I know the disk was jumpered properly. It was a weird ancient disk, so who knows. Now, one more question, to which I should know the answer by now: When compiling the kernel in a default woody installation on this system, the build fails because there is no sparc64-linux-gcc on the system. I've worked around that problem in the past by making that a symlink to gcc-3.0. What's the actual right thing to do? noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgpy9Uahj24LO.pgp Description: PGP signature
Re: Ultra 5 won't boot from HD after install
I have set up a tftp environment from which I can boot the installation system on this machine. What do I need to do to tftp a kernel to this machine and boot it? I obviously can't just send a kernel image to the machine (I tried). I can boot the tftpboot.img file with no troubles, but of course that kicks me into the system installer. I just want to boot the actual installed system via tftp. I've found the silotftp.b bootloader and used tftp to get it running, but it can't seem to access the disk, either. So it really seems like I'll need to download the whole kernel. Of course, even if I manage to do this, who knows if the machine will be stable or not. I'm starting to think there might be something wrong with the PCI bus or the IDE controller. noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgpkDyNYncWZM.pgp Description: PGP signature
Re: Ultra 5 won't boot from HD after install
On Fri, May 02, 2003 at 04:13:56PM -0400, Ben Collins wrote: > Boot back to the install and switch to a the shell console (ALT+F2) and > use "dmesg | more" to see if you can find some info about your IDE. I > have an Ultra5 here working perfectly with a non-stock IDE drive, so I > know it works. Here's the boottime output when initializing the IDE busses: Uniform Multi-Platform E-IDE driver Revision: 6.31 ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx CMD646: IDE controller on PCI bus 01 dev 18 CMD646: chipset revision 3 CMD646: chipset revision 0x03, MultiWord DMA Force Limited CMD646: 100% native mode on irq 4,7e0 ide0: BM-DMA at 0x1fe02c00020-0x1fe02c00027, BIOS settings: hda:pio, hdb:pio ide1: BM-DMA at 0x1fe02c00028-0x1fe02c0002f, BIOS settings: hdc:pio, hdd:pio hda: Conner Peripherals 1275MB - CFS1276A, ATA DISK drive ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx hdc: CRD-8322B, ATAPI CD/DVD-ROM drive ide0 at 0x1fe02c0-0x1fe02c7,0x1fe02ca on irq 4,7e0 ide1 at 0x1fe02c00010-0x1fe02c00017,0x1fe02c0001a on irq 4,7e0 (shared with ide) hda: 2502308 sectors (1281 MB), CHS=2482/16/63, DMA hdc: ATAPI 32X CD-ROM drive, 128kB Cache Uniform CD-ROM driver Revision: 3.12 Partition check: hda: hda1 hda2 hda3 The partition table looks like: Disk /dev/hda (Sun disk label): 16 heads, 63 sectors, 2480 cylinders Units = cylinders of 1008 * 512 bytes Device FlagStart EndBlocks Id System /dev/hda1 0 2379 1199016 83 Linux native /dev/hda2 u 2379 2480 50904 82 Linux swap /dev/hda3 0 2480 12499205 Whole disk noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgp5fZaEbU0jD.pgp Description: PGP signature
Re: Ultra 5 won't boot from HD after install
On Fri, May 02, 2003 at 03:09:04PM -0400, Ben Collins wrote: > I'm only guessing but OBP doesn't seem to detect the device, atleast not > at the OBP address that the "disk" alias points to. > > Try doing "boot disk1" or "boot disk2", "boot disk3"... That didn't work. show-devs reveals, among a bunch of other non-disk things, a disk at /[EMAIL PROTECTED],0/[EMAIL PROTECTED],1/[EMAIL PROTECTED]/disk, which is the device that disk seems to be pointing to (how do I view device aliases?). I don't suppose there's some IDE equivalent to probe-scsi-all that I need to run, or something like that, is there? noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgphyYnflY4uL.pgp Description: PGP signature
Ultra 5 won't boot from HD after install
Hi all. I just installed potato on an Ultra 5 (will upgrade once it's working, but that's all I had on CD) but the machine doesn't seem to want to boot from the IDE hard drive. It had no problem booting the install CD and the installation seemed to go fine. No errors were reported when installing SILO. Here's the output of my interaction with OpenBoot: ok boot disk Boot device: /[EMAIL PROTECTED],0/[EMAIL PROTECTED],1/[EMAIL PROTECTED]/[EMAIL PROTECTED],0 File and args: Can't open boot device The partition table has a Linux partition first, occupying most of the disk, followed by swap. The 3rd partition is the "entire disk" in the Sun disklable. The disk in the machine is not the original Sun-supplied disk but is an old 1.2 gig IDE disk. The "manual with hardware detected defaults" option when creating a sun disklabel in fdisk seemed to pick the right values, according to the label on the disk, and as I said, there was no sign of problems during the installation of the system or of SILO. Any suggestions? noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgpgsd6UTA0jh.pgp Description: PGP signature
Re: Mail Server?
On Mon, Jun 17, 2002 at 06:43:33PM +0200, Ottavio Campana wrote: > I don't think exim's good for an ISP. It could be useful for a home > computer but nothing more. I don't like qmail. > I'm surprised to hear you say that about exim. You realize that it was written to handle mail for a rather large site, right? See http://www.exim.org/FAQ.html#TOC188 for some testimonials from other large sites that use it. I won't comment on qmail or those involved in its development. I suspect I don't need to comment on things like sendmail. noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgpQrFANOXU6i.pgp Description: PGP signature
Re: kernel 2.4.x compile failure on Ultra 1
On Wed, Sep 12, 2001 at 03:47:20AM +0200, Ben Castricum wrote: > I think a came across this problem as well. Try compiling the kernel with > SMP, i think the atomic_dec_and_lock only gets exported when SMP is being > used. Unfortunately I run accross more problems later on when trying to > compile 2.4.9, hopefully you have more luck. Yup, that did it. It also turned out that I needed to enable PCI support, despite the fact that the Ultra1 is an SBus machine. But all seems to be well now. Thanks. noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgpZMOfwVw4Nw.pgp Description: PGP signature
Re: SPARC 20 firmware error at boot
On Sat, Aug 25, 2001 at 12:14:52PM +0800, Jean-Paul Blaquiere wrote: > what's probably happened is the internal battery has died. It is a trivial > matter to attach a new lithium/NiCad battery to the top of your CMOS chip > after disconnecting the old one. All you need is a soldering iron and a > small hacksaw :) I ended up pulling an NVRAM chip from an unused SPARC 10. According to the SPARC FAQ the SS10 and 20 used the same chip. Lo and behold, my SS20 is now happily working again. Yay! noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgpOOYjueOjVa.pgp Description: PGP signature
Re: SPARC 20 firmware error at boot
On Fri, Aug 24, 2001 at 12:57:59PM +0800, Jean-Paul Blaquiere wrote: > google is your friend. > or the copy I keep at home ;) > http://japester.ucc.asn.au/faqsunnvram.txt Thanks! Now I only need to determine whether or not it's worth it to me to buy a replacement NVRAM chip... ;^) noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html
SPARC 20 firmware error at boot
I've got a SPARC 20 running sid that will no longer boot. It fails before running the kernel with the following error: PROMLIB: obio_ranges 5 IDPROM: Checksum failure (nvram=5d, calc=d0)! Does anybody know how to fix this? It's clearly not a Debian, or even Linux specific problem, but I don't know of any other SPARC experts to consult. The problem showed up after a power outtage; the machine worked fine prior to the outtage. Thanks. noah -- ___ | Web: http://web.morgul.net/~frodo/ | PGP Public Key: http://web.morgul.net/~frodo/mail.html pgp8jMnfTjP8g.pgp Description: PGP signature
