"The problem is that (a) the kernel's postinst uses debconf, which uses
stdin and stdout for communication and forbids other junk from appearing
there, and (b) the kernel's postinst calls update-initramfs, which
outputs status information to stdout which debconf thinks is a command
in its protocol and thus confuses any future real debconf commands.
Since the kernel's postinst is the thing with special requirements due
to using debconf, it should be the thing responsible for redirecting
file descriptors around as necessary." -cjwatson
svn path=/dists/sid/linux-2.6/; revision=13347
fixes:
rmdir: failed to remove `/lib/modules/2.6.29-rc8-amd64': Directory not empty
dpkg - warning: while removing linux-image-2.6.29-rc8-amd64, directory
`/lib/modules/2.6.29-rc8-amd64' not empty so not removed.
ls /lib/modules/2.6.29-rc8-amd64
modules.alias.bin modules.dep.bin modules.symbols.bin
should also land in sid branch.
svn path=/dists/trunk/linux-2.6/; revision=13314
/usr/include/drm/drm_mode.h is now also in linux-libc-dev,
so it's also in libdrm-dev 2.4.5.
Fixed for 2.4.5-2 without that file.
Reported-by: Julien Cristau <jcristau@debian.org>
svn path=/dists/sid/linux-2.6/; revision=13258
don't use mkinitramfs-kpkg vestiges.
preinst and postinst needed a bit of surge to not call
the old mkinitramfs-kpkg interfaces. just check which initramfs
binary is around and call them.
they support all relevant linux-2.6 versions.
svn path=/dists/trunk/linux-2.6/; revision=12828
user proably don't even cares about the symbolic link,
nuke 2 messages as they don't provide info.
nor is there any action for the admin.
svn path=/dists/trunk/linux-2.6/; revision=12163
currently is printed:
Running depmod.
Using mkinitramfs-kpkg to build the ramdisk.
just print:
Running depmod.
Running mkinitramfs-kpkg.
(which reminds me that this wrapper should be nuked).
svn path=/dists/trunk/linux-2.6/; revision=11743
-- relevant irc discussion:
00:43 <waldi> maks: sarge->lenny upgrades are not supported
00:44 <maks> waldi i know but people like to install latest linux-26 in etch
00:44 <waldi> not our problem
00:45 <waldi> this is not supported
00:45 <waldi> and even if, only on top of etch, where it is already moved in l-k-h
---
thus revert and close away.
This reverts commit 72cd97dcf534b764c137b62ff2b7fbc6fd362d2b.
svn path=/dists/trunk/linux-2.6/; revision=11442
> Both linux-kernel-headers in etch and amd64-libs-dev in sarge
> provide /usr/include/asm/bootsetup.h (among others), leading to a hiccup
> during sarge->etch upgrade if they are not upgraded in the right order.
a bit late not but better safe than sorry,
can be kicked after Lenny release.
svn path=/dists/trunk/linux-2.6/; revision=11441
later we might want to add amixer output,
but that is very verbose so kept out for now.
fix changelog for removed atl1 drivers too.
svn path=/dists/trunk/linux-2.6/; revision=11427
* debian/rules.real: Install patch.unpatch from templates.
* debian/templates/patch.unpatch
- Move from debian/templates/patch.unpatch.in.
- Only use static informations.
svn path=/dists/trunk/linux-2.6/; revision=11102
- Simplify reverse check.
- Add series list.
* debian/templates/patch.apply.in
- Fix extra matching.
- Use full version check.
svn path=/dists/trunk/linux-2.6/; revision=11097
project smith hit the linux-2.6 templates.
linux-2.6 should generate now much less lintian warnings.
thanks to bubulle!!!
svn path=/dists/trunk/linux-2.6/; revision=11084