All, We're respinning the M3 release candidate due to a few issues around persist db, build history and how the autobuilder handles it. The rc1 build is being built from current master and should be

You need wildcards: PREFERRED_VERSION_gcc = "4.5%" So it will match on 4.5.4+svnr189152, which is the actual version of the 4.5 toolchain you have available. Bruce

HI, If I set the values as follows: PREFERRED_VERSION_gcc = "4.5" PREFERRED_VERSION_gcc-cross = "4.5" PREFERRED_VERSION_gcc-cross-initial = "4.5" PREFERRED_VERSION_gcc-cross-intermediate =

I have a clone and started a build. When I have some results .. I'll send more email. Cheers, Bruce

Hi Kishore, Could you break this up into 3 patches: - one moving the mga recipe to common/recipes-graphics/xorg-driver (you put it in xorg-server) - one adding the new XSERVER_IA32_MATROX_MGA

Hi, I switched to using gcc-4.6 from the meta-openembedded layer, and still get the same issue, i.e. no login or bash prompt with init=/bin/sh zImage starting: loaded at 0x00800000 (sp:

<snip> Damn! Found the problem, just a typo :-) | -KMACHINE_portux9g20 = "portuxg20" | +KMACHINE_portuxg20 = "portuxg20" Yes it's the same BSP and the kconf_check warnings are persistent! |

back here: http://www.mail-archive.com/yocto@.../msg07820.html i pointed out that the dev manual doesn't accurately reflect the situation with SSH servers in qemu images. i don't see

In my experience .. it is always just the version, not the package that you want to specify, so from all of those, you'd drop the gcc_ .. etc, from the string. Bruce

Hi Bruce, Ahh, ok, that worked. Along similar lines, if I had to explicitly set the gcc version, is it correct to specify the following variables? PREFERRED_VERSION_gcc =

Have you tried dropping the "linux-libc-headers" ? i.e. PREFERRED_VERSION_linux-libc-headers = "3.3" Bruce

Hi, In my local.conf I have set the following variable PREFERRED_VERSION_linux-libc-headers = "linux-libc-headers_3.3" However, when I run bitbake linux-libc-headers -c fetchall, it complains that the

Is this the same BSP producing the kconf check warnings on denzil ? I ran tests this morning and denzil itself is clean, so there's definitely something wrong in the layer. If this is the same BSP, I

Hi @all, I'm having some trouble building a BSP supporting two different machines. The machines in question are the Stamp9g20 and PortuxG20 from taskit. These machines are just slightly different

You could.

Hi Khem, Ok, just reconfigured all my layers to use the latest poky and meta-openembedded updates, and set the preferred versions for gcc to 4.6, and will give it a try. I just have one question

Hi everyone, I've been using Hob lately and I'm completely blown away by how useful it is! Kudos to everyone: from those insisting there exist such a tool to those working on it :-) While using it,

It could be one of the updates merged back to the -stable release. I'll fire up a denzil build in the next while to see if I get the same results. Cheers, Bruce -- "Thou shalt not follow the NULL

Mainly connectivity and multimedia packages. Signed-off-by: Cristian Iorga <cristian.iorga@...> --- .../conf/distro/include/upstream_tracking.inc | 30 ++++++++++---------- 1 file

Hmm ok but since I use denzil, and denzil provides the bbclass *and* the kernel revision for the meta branch ... why don't they match? Cheers, Markus