Alan BRASLAU writes:
> On Thursday 19 November 2009 23:55:43 Manoj Srivastava wrote:
> > Ou, Nov 19 2009, Alan BRASLAU wrote:
> > > kernel-package is now severly broken! Building and installing a[nd]
> > > new kernel image now results in *two* identical vmlinuz nodes with
> > > *identical* names under /boot. grub then rightly finds both images,
> > > *neither one which is bootable*.
> >
> > I can't reproduce this with 12.029.
> >
> Reproduces with kernel-package (12.029):
>
> $ dpkg -i linux-image-2.6.31_custom.1.2_amd64.deb
> Selecting previously deselected package linux-image-2.6.31.
> (Reading database ... 251822 files and directories currently installed.)
> Unpacking linux-image-2.6.31 (from linux-image-2.6.31_custom.1.2_amd64.deb)
> ...
> Done.
> Setting up linux-image-2.6.31 (custom.1.2) ...
> Running depmod.
> Examining /etc/kernel/postinst.d.
> run-parts: executing /etc/kernel/postinst.d/initramfs 2.6.31
> /boot/vmlinuz-2.6.31
> run-parts: executing /etc/kernel/postinst.d/initramfs-tools 2.6.31
> /boot/vmlinuz-2.6.31
> Running postinst hook script update-grub.
> Generating grub.cfg ...
> Found Debian background: moreblue-orbit-grub.png
> Found linux image: /boot/vmlinuz-2.6.31-1-amd64
> Found initrd image: /boot/initrd.img-2.6.31-1-amd64
> Found linux image: /boot/vmlinux-2.6.31
> Found initrd image: /boot/initrd.img-2.6.31
> Found linux image: /boot/vmlinuz-2.6.31
> Found initrd image: /boot/initrd.img-2.6.31
> done
grub2 is picking vmlinux over vmlinuz; it should be configured
not to do that.
> $ ls -li /boot
> total 50212
> 49089 -rw-r--r-- 1 root root 101944 Nov 20 06:51 config-2.6.31
> 49143 -rw-r--r-- 1 root root 101865 Nov 16 05:54 config-2.6.31-1-amd64
> 48871 drwxr-xr-x 2 root root 4096 Nov 20 07:43 grub
> 49093 -rw-r--r-- 1 root root 8220827 Nov 20 07:43 initrd.img-2.6.31
> 49098 -rw-r--r-- 1 root root 8228707 Nov 17 08:32 initrd.img-2.6.31-1-amd64
> 49091 -rw-r--r-- 1 root root 1591417 Nov 20 07:41 System.map-2.6.31
> 49142 -rw-r--r-- 1 root root 1591417 Nov 16 05:54 System.map-2.6.31-1-amd64
> 49092 -rw-r--r-- 1 root root 14498311 Nov 20 07:41 vmlinux-2.6.31
> 49090 -rw-r--r-- 1 root root 14498311 Nov 20 07:41 vmlinuz-2.6.31
> 49041 -rw-r--r-- 1 root root 2480736 Nov 16 05:54 vmlinuz-2.6.31-1-amd64
>
> $ diff /etc/kernel/postinst.d/initramfs /usr/share/doc/kernel-
> package/examples/etc/kernel/postinst.d/
>
> $ cat /etc/kernel/postinst.d/initramfs-tools
>
> #!/bin/sh
>
> # passing the kernel version is required
> [ -z "$1" ] && exit 0
>
> # kernel-package passes an extra arg; hack to not run under kernel-package
> [ -z "$2" ] || exit 0
>
> # we're good - create initramfs. update runs do_bootloader
> update-initramfs -c -t -k "$1"
>
> $
>
>
> > > The bug report should therefore be *reopened* and marked *critical*.
> > >
> > > Not only did the new handling of XEN schemes "caused no end of
> > > confusion", it was broken and is still the source of system damage.
> > > This is rather severe as a less experienced user trying to install a
> > > custom tailored kernel can easily end up with an unbootable
> > > installation.
> >
> > Well, this is _unstable_, after all. But thanks for testing,
> > things should be better now.
> >
>
> So, should _critical_ bugs not be allowed in unstable?
This is not really a critical bug. It is only triggered when
the user asks for a XEN dom0 image, and is using grub2, and has not
told grub2 to ignore vmlinux.
In any case, 12.030 will not ship vmlinux unless asked.
I think there is a mismatch between our expectations; I expect
people compiling kernels to be aware of their configuration, and to
have set the configuration according to their desires. In this case,
a dom0 image is being asked for; and you are hitting a flux in
kernel-package's treatment of XEN images.
This is not encountered with the standard (that is, make
defconfig) configurations.
manoj
--
"The shortest distance between two points is through Hell." --Brian
Clark
Manoj Srivastava <[email protected]>
1024D/BF24424C print 4966 F272 D093 B493 410B 924B 21BA DABB BF24 424C
--
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]
