Chen Qi [Wed, 13 May 2015 06:40:09 +0000 (14:40 +0800)]
populate_sdk_ext: set noexec flag for some tasks
For the SDK, what we are interested in is the sysroot part. So the
only useful task is do_populate_sysroot. Other tasks like do_package
don't need to run. So mark them as 'noexec' for SDK.
This would result in two benefits:
1) The size of extensible SDK is larged reduced. For example, extensible
SDK for core-image-minimal is reduced from 1.8G to 793M.
2) The installation time is reduced as a lot of tasks are not run.
Richard Purdie [Sun, 10 May 2015 11:30:49 +0000 (12:30 +0100)]
insserv: Remove
Remove insserv from OE-Core. It did have uses but we've optimised the sysvinit
scripts as needed and there are other directions init systems are moving now.
It no longer belongs in the core.
BB_ALLOWED_NETWORKS is a list of hosts that the fetcher will be allowed
to use when BB_NO_NETWORK is not set.
If BB_NO_NETWORK is set, then networking is still disabled.
If BB_ALLOWED_NETWORKS is not set, the behaviour remains the same as
today.
If BB_NO_NETWORK is NOT set, and BB_ALLOWED_NETWORKS is configured, then
only the hosts in the list are usable by the fetcher.
eg:
BB_ALLOWED_NETWORKS="yoctoproject.org git.gnu.org"
The fetcher will be able to download from yoctoproject.org, git.gnu.org,
but not ftp.gnu.org or any other hostname that is not in the list.
There is also limited support for wildcards on the beginning of the
hosts, so BB_ALLOWED_NETWORKS="*.gnu.org" with match git.gnu.org and
ftp.gnu.org as well as foo.git.gnu.org
Jason Wessel [Fri, 8 May 2015 18:21:40 +0000 (13:21 -0500)]
bitbake: fetch2: Allow git to use a different CA Bundle with private keys
Setting BB_ENV_EXTRAWHITE to include GIT_SSL_CAINFO is enough to
get "git ls-remote" to work for recipe parsing but it is not enough
to get the fetcher to clone properly.
This is because the fetcher has its own idea about what variables
should be exported in the environment for some operations.
It is desirable to use alternate CA Bundles for internal testing prior
to using public keys for https, so we should allow the GIT_SSL_CAINFO
to pass through.
We remove url entries, views and templates that are not
used in the interface; these are leftovers from the development
process, and are not linked anywhere in the interface nor
mentioned in documentation.
This is a URL refactoring needed to remove the inadvertent
usage of server-side user session in a REST-style API.
We move the parameters that were stored in the user session
to the URL, making navigation more robust.
This refactoring allows a clean (no 500 HTTP return codes) crawl
of the website following inside links.
Patrick Ohly [Fri, 8 May 2015 12:37:30 +0000 (14:37 +0200)]
combo-layer: improve merge commit handling
When the head of a branch is a merge commit, combo-layer did not
record that commit as last_revision because it only considers applied
patches, and the merge commit never gets applied.
This causes problems when the merge commit leads to multiple patches
and the commit id that gets recorded only reaches some of these
patches. The next run then will try to re-apply the other patches.
This special case is now detected and dealt with by bumping
last_revision to the branch commit. The behavior where the head is a
normal commit is intentionally not changed, because some users might
prefer the traditional behavior.
Bruce Ashfield [Fri, 8 May 2015 03:36:15 +0000 (23:36 -0400)]
kernel-yocto: propagate in tree defconfigs to WORKDIR
As reported by Steffen Pankratz <Steffen.Pankratz@elektrobit.com>, the
previous logic of KBUILD_DEFCONFIG processing would not propagate an in
tree defcofig to WORKDIR if one was not already present.
We fix the propagation by copying the in tee config if a defconfig is
not already in WORKDIR.
Additionally we only warn (versus copying) if an in tree configuration
is specified, is different than the WORKDIR version and isn't copied.
Bruce Ashfield [Fri, 8 May 2015 03:36:13 +0000 (23:36 -0400)]
linux-yocto/3.14: fix qemumips build error
Updating the SRCREVs to import the following fix:
mips: define cpu_has_saa in common features include
To avoid build failures such as the following on non-cavium
platforms:
| arch/mips/include/asm/atomic.h: In function 'atomic_add':
| arch/mips/include/asm/atomic.h:52:6: error: 'cpu_has_saa'
| undeclared (first use in this function)
| if (cpu_has_saa) {
| ^
We define a disabled cpu_has_saa unless the machine specific feature
overrides define a value.
Bruce Ashfield [Fri, 8 May 2015 03:36:14 +0000 (23:36 -0400)]
linux-yocto: fix race between checkout and meta data generation
There are two tasks that must run before a linux-yocto kernel is built.
- Kernel checkout and relocation to work-shared (kernel_checkout)
- Meta data gathering and configuration prep (kernel_metadata)
The current task definitions for both are simply "before do_patch",
which is correct, but kernel_checkout must run before and not race with
kernel_metadata.
So we set the definition of kernel_checkout to be more specific and
enforce the proper ordering.
Bruce Ashfield [Fri, 8 May 2015 03:36:12 +0000 (23:36 -0400)]
linux-yocto/3.19: Braswell support and bug fixes
Updating to include the following commits:
a4d0c407cced dmaengine: dw: don't handle interrupt when dmaengine is not used e92b2ce791b2 dmaengine: dw: define DW_DMA_MAX_NR_MASTERS 931304a6567e dmaengine: dw: provide DMA capabilities 468bad4f7a6f dmaengine: dw: Split device_control b4afd7710db7 i2c: i801: Use managed pcim_* PCI device initialization and reservation d81a8a11ecba i2c: i801: Remove pci_enable_device() call from i801_resume() 40e18604e70c i2c: i801: Use managed devm_* memory and irq allocation b54f65dbe57b i2c: i801: Remove i801_driver forward declaration e95740d4d079 i2c: i801: Don't break user-visible strings 423e98721e04 ACPI: Introduce has_acpi_companion() 291f620dc052 i2c: designware: Suppress error message if platform_get_irq() < 0 cf5ff51a8e3f i2c: designware-pci: no need to provide clk_khz 4f583ce420d3 i2c: designware-pci: remove Moorestown support e000c549c9d8 i2c: designware: Add Intel Baytrail PMIC I2C bus support 7ffbd9ca19a7 i2c: designware: fixup return handling of wait_for_completion_timeout 5758d5a1df32 i2c: designware: Do not calculate SCL timing parameters needlessly 2f58fcae92db i2c: designware: Add i2c bus locking support 84a73e51e900 i2c: designware: use {readl|writel}_relaxed instead of readl/writel f672bb8424e6 serial: 8250_dw: Fix get_mctrl behaviour 91bd64585489 serial: 8250: add support for ACPI-probed serial port for X-Gene platform 1190cba71f09 serial:8250:8250_pci: delete unneeded quirk entries 6405a4b71451 serial:8250:8250_pci: fix redundant entry report for WCH_CH352_2S 0a1a31bbbc19 serial: 8250_pci: remove one useless explicit type conversion 4edc52a55f82 intel_idle: Add support for the Airmont Core in the Cherrytrail and Braswell SOCs 934f85e8bfdb x86/irq, ACPI: Implement ACPI driver to support IOAPIC hotplug c6a3440252a8 ACPI / LPSS: check the result of ioremap() 6aacc0c931b7 pinctrl: update direction_output function of cherryview driver cb4a43a2177d pinctrl: cherryview: Configure HiZ pins to be input when requested as GPIOs 0df22c007ce1 pinctrl: intel: drop owner assignment from platform_drivers 07b16f04700b pinctrl: cherryview: Save and restore pin configs over system sleep e8e5cfffa231 pinctrl: baytrail: Save pin context over system sleep 04cb3cc0ff21 pinctrl: baytrail: Rework interrupt handling 4cac25d2574d pinctrl: baytrail: Clear interrupt triggering from pins that are in GPIO mode eacab9ab234a pinctrl: baytrail: Relax GPIO request rules
Ross Burton [Thu, 7 May 2015 14:30:54 +0000 (15:30 +0100)]
bind: disable the HTTP statistics service by default
Rename the "libxml2" PACKAGECONFIG to "httpstats" so that is is meaningful, and
disable it by default as a web frontend to the server statistics shouldn't be
enabled by default.
Roy Li [Mon, 4 May 2015 08:43:39 +0000 (16:43 +0800)]
bind: upgrade to 9.10.2
1. Remove two unneeded CVE patches, 9.10.2 fixed the CVE-2015-1349 also
2. Remove bind-subdirs-run-serially.patch and cross-build-fix.patch,
similar fixes are merged into 9.10.2
3. update the dont-test-on-host.patch
4. update the Copyright file checksum, since the date in it has been changed.
5. replace the hardcode lib dir with $base_libdir in bind-add-crosscripts*.patch
Michael Wood [Tue, 21 Apr 2015 10:59:37 +0000 (11:59 +0100)]
bitbake: toaster: Refactor and expand layer add remove mechanism
We have multiple pages which have buttons to add and remove layers this
patch adds functionality to libtoaster to abstract this and implements
it in the pages affected. We handle loading and showing the dependencies
dialog here too and generating the notification messages.
Also implemented is using the selectmachine api from the projectapp to
avoid having to handle this in each page that allows selecting machines.
A small number of jshint issues, help text and the machine page name
have also been fixed.
Use the project page to select the machine rather than setting it and
then redirecting to the project page. This will also avoid having to have a
special handler in the machines page it's self.
Michael Wood [Fri, 10 Apr 2015 17:15:03 +0000 (18:15 +0100)]
bitbake: toaster: Move project context variables to common scope
We have a bunch of context data which are used in multiple pages so it
makes more sense to have this in a single place libtoaster.ctx that's
accessible from each page rather than request it from every page.
bitbake: toasterui: close build on BuildCompleted event
Moving the code to close the build and instantiate a fresh
buildinfohelper object to BuildCompleted event, as the
CommandCompleted/Failed/Exit events come in too early.
Paul Eggleton [Thu, 7 May 2015 13:52:25 +0000 (14:52 +0100)]
oe-selftest: devtool: add a proper test to see if tap devices exist
Check up front in test_devtool_deploy_target whether the tap devices
exist and skip if not. If we don't do this we get a significantly less
comprehensible error via pexpect.
Paul Eggleton [Thu, 7 May 2015 13:52:24 +0000 (14:52 +0100)]
devtool: fix for rename of gcc-source
After OE-Core commit 67db7182faf6742b0d971d61d8c5ba34f69d2e12, PV is
appended to the end of the gcc-source PN, thus we need to handle that in
devtool and the corresponding test.
Khem Raj [Wed, 6 May 2015 07:04:59 +0000 (00:04 -0700)]
libart-lgpl: Fix cross compiling
We use standard ISO C data types and get rid of configure guess work
which is iffy in cross compiled environment. Cleans up the re-build as well
since we were deleting art_config.h without considering if the file is
there in first place or not.
Ed Bartosh [Wed, 6 May 2015 20:28:39 +0000 (23:28 +0300)]
kernel.bbclass: Fix race condition
Race condition between do_compile_kernelmodules and do_shared_workdir
tasks occurs when do_compile_kernelmodules changes files in
include/generated/* while do_shared_workdir tries to copy them to
shared working directory.
Fixed race by moving do_shared_workdir after do_compile but before
do_compile_kernelmodules.
Paul Eggleton [Wed, 6 May 2015 16:19:27 +0000 (17:19 +0100)]
classes/buildstats: misc cleanups
* Rename BNFILE and DEVFILE so that they are clearly specific to this
class, since they end up in the global scope
* Use "with open" when opening files consistently
* Use getVar('PF', True) instead of expand('${PF}')
* Drop some unnecessary assignments
Paul Eggleton [Wed, 6 May 2015 16:19:26 +0000 (17:19 +0100)]
classes/buildstats: don't catch all errors from mkdirhier
mkdirhier() doesn't raise an exception if the directory exists, so if it
does raise one we should just be failing, otherwise we're just going
to hit errors later anyway.
Cristian Iorga [Wed, 6 May 2015 15:21:47 +0000 (18:21 +0300)]
runqemu-internal: correctly set format for root FS as raw
qemu guesses via probing the format of root FS, but
gives a warning and restricts write operations on block 0.
So fix it by setting correctly the format as raw.
python3: Disable pip to fix issues on the Autobuilders
As Python-3.4 pip is required and installed by default, pips installation
is bootstrapped using python wheels, this makes it complicated to patch,
by default the installation uses an absolute path to /tmp, on most cases
this would not be a problem, but since on the Autobuilders several Python
installations (for different archs) may be happening at the same time,
this shared resource utilization results in errors at do_install(), this
patch disables pip installation by default for now, until we have a
python3-pip package which we can easily patch in these situations.
python3-native: Disable pip on native builds to fix Autobuilder issues
As Python-3.4 pip is required and installed by default, pips installation
is bootstrapped using python wheels, this makes it complicated to patch
and causes errors on the Autobuilders, since we don't need pip for
python3-native, this disables it fixing the issues.
python3-native: Fix pip install issue due to unclean build directory
When installing python3-native sometimes pips default build
directory (which is on the host and is user dependant) is left unclean,
due to this, when python3-core is being installed it tries to use
the same directory producing an error, this explicitly removes
what the previous installation might have left behind, fixing the issue.
Modifies:
python3-native_3.3.3.bb -> python3-native_3.4.2.bb: Updates checksums, version, patches
and SRC_URIs to use xz format, LICENSE did not change, dates were updated, adds fix for
ctypes/libffi needed for pip <- required since Python 3.4 (see PEP #453).
python3_3.3.3.bb -> python3_3.4.2.bb: Updates checksums, version, patches and SRC_URIs
to use xz format, adds fix for ctypes/libffi, fixes oe_multilib_header() call,
adds python3-core as RDEPENDS to python3-misc to avoid QA issue.
generate-manifest-3.4.py - > generate-manifest-3.4.py: fixes reprlib,
adds _sitebuiltins and _collections_abc <- formerly part of collections, separated since 3.4
python-3.3-manifest.inc -> python-3.4-manifest.inc
Peter Seebach [Mon, 4 May 2015 20:55:49 +0000 (15:55 -0500)]
pseudo 1.6.5: less pointlessly chatty
There was a stupid logic error controlling the diagnostic for
a "possible" mismatch involving trailing slashes and whether or
not a node was believed to be a directory. Specifically, a diagnostic
got printed any time a lookup for a directory *didn't* have a
trailing slash, as well as in the (actually intended) case where
a non-directory lookup *did*.
No other changes, but that one is probably significant.
Scott Rifenbark [Tue, 28 Apr 2015 21:25:01 +0000 (14:25 -0700)]
dev-manual: Updates to add Luna support for Eclipse
Adding Luna and dropping Juno. Pretty much went through the
"Working With Eclipse" section and dropped Juno and added
Luna. A major deletion was the BitBake Commander section using
Hob. Todor told me to drop this section completely. Note that
I also added a couple new YoctoProjectTools that appear on the
menu: SystemTap and yocto-bsp.
Scott Rifenbark [Tue, 28 Apr 2015 21:38:51 +0000 (14:38 -0700)]
ref-manual: Updated the "Image Features" section
Removed the requirement that debug-tweaks is needed in order to
individually use the allow-empty-passowrk, empty-root-password,
and post-install-logging features. These three features can be
individually turned on an off regardless of debug-tweaks.
Scott Rifenbark [Mon, 27 Apr 2015 22:20:28 +0000 (15:20 -0700)]
ref-manual; mega-manual: Edits to further define packaging process
In response to community input where developers were attempting to
locate information on how to write packages out to a directory
other than ${DEPLOY_DIR}/<package_type>, I updated the "Package
Feeds" section, added several new DEPLOY_DIR_* variables, updated
four classes, and updated four tasks. Here are some details:
* Made changes to the "Package Feeds" section to provide more
accurate information in the package feeds directory structure
used by the build system in build/tmp. These changes included
updating the figure itself and some explanatory text.
* Updated the DEPLOY_DIR variable description.
* Added new variable descriptions for DEPLOY_DIR_DEB, DEPLOY_DIR_IPK,
DEPLOY_DIR_RPM, and DEPLOY_DIR_TAR.
* Updated the related classes: package_deb, package_ipk, package_rpm,
and package_tar.
* Updated the related tasks: do_package_write_deb, do_package_write_ipk,
do_package_write_rpm, and do_package_write_tar.
Richard Purdie [Sun, 3 May 2015 10:27:24 +0000 (11:27 +0100)]
directfb: Improve CFLAGS handling for x86
The problem with filter_out is that it expands the variable meaning overrides
to sub components can get lost. This has happening to the security flags
meaning directfb failed in lsb builds.
Use _remove instead of filter_out since it gives much more predictable results.
BlueZ 4.x and associated recipes
are now obsolete.
(bluez4,bluez-hcidump, gst-plugin-bluetooth).
Will be moved into recipes-connectivity
collection of meta-openembedded.
BlueZ 4.x is still usable in poky/YP.
Richard Purdie [Sat, 2 May 2015 09:04:37 +0000 (10:04 +0100)]
multilib: Tweak value of PN used for OVERRIDES
Currently, PN is used in overrides which is expanded to have a MLPREFIX. This
means and pn- overrides without the prefix would be ignored which is not what
is usually expected.
We noticed huge problems using poky-lsb with multilib since the per recipe
overrides were not applied. This adds in handling for PN with and without
the prefix. This should unbreak world-lsb builds on the autobuilder.
projects / thirdparty / openembedded / openembedded-core-contrib.git / log
