--- Begin Message ---
- To: submit@bugs.debian.org
- Subject: My own recompilation of 2.6.16 or -17 kernel on my AMD64 fails. Kernel panics, says "tried to kill init!!"
- From: Willy Gommel <wg3w@spamcop.net>
- Date: Fri, 10 Nov 2006 15:53:11 -0800
- Message-id: <200611101553.11818.wg3w@spamcop.net>
Package: linux-source
Version: 2.6.17-9; testing "Etch" version, AMD64
Whenever I try to compile the kernel without any initrd support (the only way
I know how to do so), and then install the resulting kernel, the system
becomes unbootable using it.
I use the LILO loader. What happens is this:
When I turn the machine on, or reboot it, LILO loads the kernel, which then
begins processing. Less than a second later, the kernel panics, saying
something about "tried to kill init!!" At this point, the screen is locked up
tight, which means I cannot review more than ~60 lines of preceding
procedure. Since the kernel, in panicking, fills most of the screen with what
looks like register hex-dumps, I must say that this is pretty useless to me.
This only happens on my AMD64 machine. I currently have (and am using) an AMD
1.33-GHz 32-bit machine which, using the same package (but with appropriately
different headers), handles the booting task just fine. This actually
suggests to me that the problem is in the 2.6.16&17 header packages, not
linux-source itself. Mentioning this is all I can do, however.
Both of the amchines are home-built, no-name-brand machines. The AMD64 is
running at 2.2 GHz, and is described as a "3700+". It has identical twin
250GB HDDs from Western Digital. The only way I can boot the AMD64 at present
is to use your precompiled kernels, loaded with heaps of useless junk --
including INITRD.
I have yet to try regressing to your unofficial "stable" version for AMD64. If
necessary for testing purposes, I will be happy to do so.
--
.......................................................... Willy Gommel
http://www.wg3.net
--- End Message ---