Matt Reimer [Thu, 17 Aug 2006 18:46:59 +0000 (18:46 +0000)]
LAB-kernel: Generate a LAB kernel suitable for flashing using the HTC
rescue bootloader. The kernel is installed to tmp/deploy/images and
its filename ends with .htc.
Holger Freyther [Thu, 17 Aug 2006 14:40:54 +0000 (14:40 +0000)]
classes/tinderclient.bbclass: Fixes. truncate does not work this way...
Fixup truncate to forget the the status of previous runs. We will
simply write("") to the file and empty it this way. int(f.read()) will
fail then
Holger Freyther [Thu, 17 Aug 2006 13:44:31 +0000 (13:44 +0000)]
classes/tinderclient.bbclass: Add workaround for bitbake -k and save the status
Using the continue option of BitBake can lead to some issues with the
Tinderbox. Even if one build failed the color of the result will be green
as the BuildCompleted event will be send by BitBake.
As a workaround we will store the failure in a tinder-status file in case
of failure (e.g. no Provider or PkgFailed) and will reread on the start.
When the build starts we will truncate the file and remove 'the' error
Jamie Lenehan [Thu, 17 Aug 2006 04:19:48 +0000 (04:19 +0000)]
site/i{3,4,5}-linux: Add entries for ipsec-tools so it will build.
i686-linux was already ok, but i{3,4,5}86-linux were missing the appropriate
entries. Note that ipsec-tools does not currently build with gcc 4.x.
Jamie Lenehan [Wed, 16 Aug 2006 15:12:51 +0000 (15:12 +0000)]
openssl: Update the endianess test to handle the situation where
only {big,litte}endian is set to "no" and neither of them is set to "yes".
Previously in the case where only one of the entries was present and set to
"no", openssl would just assume little endian, even if it was saying no to
little endian. Similar fix as for net-snmp as per #1320.
Jamie Lenehan [Wed, 16 Aug 2006 14:54:39 +0000 (14:54 +0000)]
hping2 2.0.0rc3: Update the endianess test to handle the situation where only
{big,litte}endian is set to "no" and neither of them is set to "yes". Saying
it's "not big endian" seems a bit backwards but that's what some site files
seem to get by with. Same fix as for net-snmp as per #1320.
Jamie Lenehan [Wed, 16 Aug 2006 14:52:12 +0000 (14:52 +0000)]
net-snmp 5.1.2: Update the endianess test to handle the situation where only
{big,litte}endian is set to "no" and neither of them is set to "yes". Saying
it's "not big endian" seems a bit backwards but that's what some site files
seem to get by with. Closes #1320 "net-snmp-5.1.2 fails on uclibc x86 and
ppc".
Graeme Gregory [Mon, 14 Aug 2006 09:20:29 +0000 (09:20 +0000)]
xmms_1.2.10.bb : add an icon and .desktop file for xmms
Call via a wrapper scipt so we can insert XLIB_SKIP_ARGB_VISUALS=1 before
xmms so that double size mode works correctly (a tip gleaned from gentoo FAQ)
Jamie Lenehan [Mon, 14 Aug 2006 09:03:21 +0000 (09:03 +0000)]
ndisc6: Add ndisc6 which contains tools for IPV6 neighbor discover, IPV6
router discovery, an IPv4/IPv6 performance measurement tool, and an IPv6
UDP+TCP traceroute implementation.
Jamie Lenehan [Mon, 14 Aug 2006 08:51:54 +0000 (08:51 +0000)]
iputils 20020927: Permissions are not carried over from install to the
packages, so for things that need SUID we need to set that in postinst. Add
this for ping, ping6 and traceroute6. Also modify that package list so we
append instead of replace it.
Graeme Gregory [Sat, 12 Aug 2006 22:38:06 +0000 (22:38 +0000)]
xmms_1.2.10.bb : improve autotools situation so we dont need to override
do_configure and do_compile.
Enable vorbis and mikmod as per TODO
stage the dev files so we can build plugins.
madwifi: removed -mapcs-32 and -malignment-traps from patches.
these are not supported for gcc4 and after discussion in #oe it
was considered safe to remove them as even on gcc3 they apparently do
not have any effect above arm7
Graeme Gregory [Sat, 12 Aug 2006 10:50:03 +0000 (10:50 +0000)]
task-angstrom.bb : add util-linux-mount/umount to RDEPENDS of base so that
filesystems get mounted on boot. mount -at nfs has problems with busybox
that stops this working.