Paul Sokolovsky [Mon, 26 Nov 2007 16:09:20 +0000 (16:09 +0000)]
gpe-login 0.90: Lock screen on suspend, not on resume.
* When using gpe-login --autolock, lock screen before going to suspend,
not after coming out of resume, as otherwise there's noticeable delay with
screen unprotected.
* Closes #3370.
* Upstream bug http://bugs.linuxtogo.org/show_bug.cgi?id=125
Paul Sokolovsky [Mon, 26 Nov 2007 13:28:46 +0000 (13:28 +0000)]
gpe-login 0.90: Die, die, die gpe-xcalibrate.sh!
* Unbreaks "Record" button press for recalibration during initial login screen
(autolock screen has been fixed by previously applied and mainstreamed patch).
Paul Sokolovsky [Mon, 26 Nov 2007 13:25:29 +0000 (13:25 +0000)]
keylaunch 2.0.10: Apply patch to unbreak key event capture.
* We of course still want to receive key events even after subscribing to
property events. This fixes when keylaunch doesn't receive forwarded
power buttons events from gpe-login --autolock in its locked state.
In plain words: now power button works with autolock screen.
Paul Sokolovsky [Sun, 25 Nov 2007 23:04:30 +0000 (23:04 +0000)]
keylaunch-conf: Never-ever use "Pressed" events in public configs.
* It is possible to unambigously distingish events only of two types:
<default> (short press) and "Held" (long press). Using any other event
type is ambigous. In particular, with the old default config there was
ambiguity with power off and backlight toggle bindings - most of the time,
backlight toggling also caused power off. As freedom of using any key in any
way should be left to user, ambigouos bindings should be avoided.
Mike Westerhof [Sat, 24 Nov 2007 17:10:21 +0000 (17:10 +0000)]
ixp4xx-kernel: defconfig for 2.6.21.6 - add many crypt modules, add requested cdc
modules, and add various USB video device modules. (note: no built-in devices or
crypto drivers have been changed to modules in this commit)
Paul Sokolovsky [Fri, 23 Nov 2007 19:54:17 +0000 (19:54 +0000)]
klibc 1.1.1: Fix multimachine build.
* klibc fishes in kernel sources, and this is per-machine stuff, so
klibc must be per-machine too. This is of course not ideal, as klibc is
*not* machine-dependent, but at least it's on par with uclibc, which has the
same issue.
Koen Kooi [Fri, 23 Nov 2007 18:04:45 +0000 (18:04 +0000)]
xserver-kdrive-common: use Xglamo for gta02, remove some of the bogus Xserver checks
* if a Xserver is in /usr/bin doesn't mean the it's the right server for the job
Henning Heinold [Thu, 22 Nov 2007 20:53:35 +0000 (20:53 +0000)]
midpath-cldc: rproviding virtual/cldc-api-1.1
* set in midpath-cldc: RPROVIDES = "virtual/cldc-api-1.1"
so we are in sync with cacao-cldc
* bump PR to 1
Koen Kooi [Thu, 22 Nov 2007 18:27:22 +0000 (18:27 +0000)]
busybox: Enable support for -t & -n options for read command in angstrom configs.
* These are timeout and num chars to read options, both are useful
when writing UI interaction code, with busybox-static being targetted
at writing bootloader tools and stuff.
Paul Sokolovsky [Thu, 22 Nov 2007 18:04:08 +0000 (18:04 +0000)]
bitbake.conf: Reformat and alphabetize ASSUME_PROVIDED list.
* According to http://lists.linuxtogo.org/pipermail/openembedded-devel/2007-November/003493.html ,
ASSUME_PROVIDED in the bitbake.conf is the authoritative location of list
of package which expected to be installed on the host for OE to work properly.
Make the list more prominent and easy to maintain.
Paul Sokolovsky [Thu, 22 Nov 2007 17:34:24 +0000 (17:34 +0000)]
initramfs-module-bootmenu: Interactive boot location selection for initramfs-uniboot.
* For now, supports booting from ext2/ext3 partitions and from ext2/ext3 loopback
images placed on vfat partition.
Paul Sokolovsky [Thu, 22 Nov 2007 17:26:10 +0000 (17:26 +0000)]
initrdscripts: module-initfs: Be silent on loading non-existent module.
* Use -q for modprobe, and as that apparently doesn't help bitbake's modprobe,
redirect output to /dev/null.