Currently we have a hierarchy of pkgdata directories and the code has to put together
a search path and look through each in turn until it finds the data it needs.
This has lead to a number of hardcoded paths and file globing which
is unpredictable and undesirable. Worse, certain tricks that should be
easy like a GL specific package architecture become problematic with the
curretn search paths.
With the modern sstate code, we can do better and construct a single pkgdata
directory for each machine in just the same way as we do for the sysroot. This
is already tried and well tested. With such a single directory, all the code that
iterated through multiple pkgdata directories and simply be removed and give
a significant simplification of the code. Even existing build directories adapt
to the change well since the package contents doesn't change, just the location
they're installed to and the stamp for them.
The only complication is the we need a different shlibs directory for each
multilib. These are only used by package.bbclass and the simple fix is to
add MLPREFIX to the shlib directory name. This means the multilib packages will
repackage and the sstate checksum will change but an existing build directory
will adapt to the changes safely.
It is close to release however I believe the benefits this patch give us
are worth consideration for inclusion and give us more options for dealing
with problems like the GL one. It also sets the ground work well for
shlibs improvements in 1.6.
(From OE-Core rev: 1b8e4abd2d9c0901d38d89d0f944fe1ffd019379)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There are various bits of cruft that have built up around our file accesses. This patch
cleans some of them up, specifically:
* Remove pointless "from __builtin__ import file"
* Use open(), not file()
* Wrap file usage in a with container to ensure files are closed
* Add missing .close() calls in some cases
(From OE-Core rev: a43e0a8ecd0441131e929daf998c3cd454d9c8f3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The current multilib search path code for packagedata is flawed since it
doesn't correctly handle changes in the TARGET_VENDOR/TARGET_OS that
multilib may make. This patch enhances the code to correctly build the
search paths so multilib packagedata is found correctly.
(From OE-Core rev: f50c5d36b2da9b36d56d95a7d89404509a1a3e9b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The bb and os modules are always imported so having these extra import calls
are a waste of space/execution time. They also set a bad example for people
copy and pasting code so clean them up.
(From OE-Core rev: 7d674820958be3a7051ea619effe1a6061d9cbe2)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This happens when tryng to add libgcc-dev to as a multilib package
(e.g. IMAGE_INSTALL_append = " lib32-libgcc-dev")
| Processing task-core-boot...
| Processing fman-ucode...
| Processing dosfstools...
| Processing lib32-libgcc-dev...
| Unable to find package lib32-libgcc-dev (libgcc-dev)!
NOTE: package fsl-image-full-1.0-r1.1.3.6: task do_rootfs: Failed
RPM (or bitbake?) is looking in the tmp/pkgdata, however some of these file
paths are mungned for the multilib scenario:
$ find tmp/pkgdata/ | grep libgcc-dev$
tmp/pkgdata/ppce5500-fsl-linux/runtime/lib32-libgcc-dev
tmp/pkgdata/ppc64e5500-fsl-linux/runtime/libgcc-dev
This patch fixes where we look for these files so they can be found and
properly installed for the multilib root file system
(From OE-Core rev: 4e65ddda5d351bd9f2172e77e656903f61a52f34)
Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If both VAR and VAR_foo are in subd, we need to ensure VAR_foo gets
returned. Currently the code would work randomly. The only current
user is the rpm packaging backend.
(From OE-Core rev: ad098e40e8c57727892819e131bf18308bf244d2)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie
Matthew McClintock
Chris Larson