Bruce Ashfield [Wed, 6 Oct 2010 03:44:59 +0000 (23:44 -0400)]
kern_tools: update SRCREV for defconfig fixes
Fixes [BUGID #399]
During BSP work it was noticed that some defconfigs weren't being
pulled into BSPs. This was due to a bad test that ended up
allowing a defconfig to exclude itself. The test has been tightened
to only kernel feature files, and defconfigs are now always
pulled in.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Darren Hart [Wed, 29 Sep 2010 08:42:42 +0000 (01:42 -0700)]
Export ALL_PROXY for bitbake commands
We already export http,ftp,https proxy environment variables. Some environments,
GNOME for instance, place the socks proxy in ALL_PROXY and all_proxy. Export it
as well.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Mark Hatle [Wed, 6 Oct 2010 23:48:16 +0000 (18:48 -0500)]
pseudo: Update to the latest version of pseudo
[BUGID #403]
The latest version of pseudo has a workaround for an extended attribute
problem that may show up with certain configurations of modern coreutils.
Specifically cp -p does not use chmod to change the permissions, but instead
uses fsetxattr to change the permissions. fsetxattr (and extended attributes)
are not supported within pseudo.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Richard Purdie [Wed, 6 Oct 2010 14:17:14 +0000 (15:17 +0100)]
qt4-tools-native: Fix compile failures
fontconfig/freetype are not listed in the dependencies, neither are any of the
X components yet this recipe appears to be building the whole UI framework, not
just the tools.
It turns out this is a dependency for uic3 and qdbus. Since at this time we don't
require these, remove them from the build.
Its likely cross versions of those specific tools would need to be added due to
their dependency on the windowing system compilation.
[BUGID #390]
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
Joshua Lock [Tue, 5 Oct 2010 11:44:08 +0000 (12:44 +0100)]
bitbake/fetch: if mirror fetching fails, ensure exception is raised
We catch any exception raised by the fetchers go() method and attempt to work
around it by trying any (post) mirrors which are configured. However, should
the mirrors fail the exception is lost and the fetch is assumed to have
completed successfully.
Instead, save the exception and if the local file does not exist after trying
the mirrors re-raise the exception.
Bruce Ashfield [Mon, 4 Oct 2010 20:51:15 +0000 (16:51 -0400)]
kernel: update SRCREVs for emenlow and blacksand/Kconfig
Fixes [BUGID #395, #396]
Add the emenlow-standard branch to the SRCREV list, and update
all the branch revisions to include the fix for defconfig processing,
that was exposed by work on the blacksand BSP.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Bruce Ashfield [Mon, 4 Oct 2010 04:46:27 +0000 (00:46 -0400)]
createme/patchme: use kgit-checkpoint to restore meta data
Fixes [BUGID: #390]
The custom de-checkpoint routines used in createme/patchme
destroyed data each run. This meant that incremental and
expermental work couldn't be done directly in the kernel source
tree itself. Using the more robust kgit-checkpoint allows that
data to be persistent (until a clean is done).
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Bruce Ashfield [Mon, 4 Oct 2010 03:27:15 +0000 (23:27 -0400)]
kern_tools: allow patching on autocreated/empty branches
Fixes [BUGID #388]
Even if we aren't creating the branch from scratch, it is possible
that it has never been patched before. So let's make sure the
infrastructure for patching is always in place.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
x11-common: correct the syntax for the -mouse arg to Xfbdev
Fixes [BUGID #99]
The mouse option as passed to Xfbdev was correct for most
targets, except mips (and a PS/2 mouse). Modifying the
syntax of the option fixes the mips case, and has no impact
on other users.
Signed-off-by: Richard Griffiths <rgriffit@windriver.com>
The mouse, usb and cursor devices don't work for the
qemumips machine. There's no easy way to 'undo' the
defaults, or dynamically remove them, so we simply
don't use the default UI options for this machine type.
Mouse and pointer are provided via ps/2 and not via
the usb and wacom tablet driver.
Signed-off-by: Richard Griffiths <rgriffit@windriver.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Daniel Díaz [Sun, 3 Oct 2010 20:45:14 +0000 (15:45 -0500)]
ncurses: Fix export of LD_LIBRARY_PATH.
This fixes the following error on do_install:
> line 175: export: `=': not a valid identifier
> line 175: export: `[...]/usr/lib': not a valid identifier
Signed-off-by: Daniel Díaz <yosoy@danieldiaz.org> Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
Mark Hatle [Thu, 30 Sep 2010 20:50:42 +0000 (15:50 -0500)]
insane: Update the insane class to check for host contamination
[BUGID #376]
Update to the insane class to detect the new gcc and binutils poison
messages located within the config.log. (Continue to scan for the old
style message.)
Add a new compile and install log check to the package_qa that scans the
logs for the same types of messages "unsafe for cross-compilation".
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Mark Hatle [Thu, 30 Sep 2010 20:04:18 +0000 (15:04 -0500)]
build.py: Add a symlink for the logging
[BUGID #375]
When configuring for the log file output, we generate a symlink to the
target filename. This link uses the same file naming, but without the
active pid, making it easier to see which log file is the last one
generated.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Mark Hatle [Thu, 30 Sep 2010 20:04:31 +0000 (15:04 -0500)]
siteconfig: Fix mkdir
[BUGID #379]
In a strange configuration, siteconfig could have already generated it's
directory, but decided it needs to regenerate the site config. The lack
of -p in the mkdir causes a failure.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Mark Hatle [Thu, 30 Sep 2010 20:03:59 +0000 (15:03 -0500)]
binutils: Update the poison patch
[BUGID #374]
Previously the poison directories patch was present, but not enabled due
to the lack of the configure option being set, and also the fact configure
itself was not being patched.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Mark Hatle [Wed, 29 Sep 2010 20:18:01 +0000 (15:18 -0500)]
gcc: Update poisoned include path checking
[BUGID #374]
The poison directory patch that was included with gcc-4.5.0 was not previously
enabled due to the lack of the configure file changes.
The patch has been updated to include the configure fragment. It was also noted
that this patch preformed nearly the same functions as the
zecke-no-host-includes patch, but with slightly different directories.
The directories scanned were added from the zecke-no-host-includes patch to
the new gcc-poison-dir-extend.patch.
The other difference with the zecke patch is that poisoned headers is no longer
an immediate fatal error. There may be instances where someone wants to do
this.
Adding -Werror=poison-system-directories to the CFLAGS would restore the
behavior.
Also fix a small problem where --help=warnings on gcc wouldn't return the
poison-system-directories as a valid option, even though it was.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
qemu: introduce vmware vga FIFO rewind patch to fix qemu hang
In some circumstance guest driver got interrupted before inserting all args
for one command, so that qemu get an invalid args and hang. GL patch
doesn't consume these missing args, which further cause FIFO disorder.
This commit reverts wrong behavior of GL patch, and introduces a qemu
upstream patch to rewind FIFO unpon detecting incomplete command.
linux-libc-header: Sync CONNECTOR_MAX_MSG_SIZE with wrs kernel
Currently default header is linux-libc-header, which leads the dismatch
with wrs kernel.
The different headers for same file cause trouble between kerenl module and
user level binary. E.g. segfault in v86d due to different
CONNECTOR_MAX_MSG_SIZE in connector.h from the one used by uvesafb.
This fix is work around to sync with wrs kernel headers, and should be removed
after switching to linux-libc-headers-wrs.
Kevin Tian [Wed, 29 Sep 2010 00:54:33 +0000 (08:54 +0800)]
gcc: fix check for target libc ssp support
gcc uses hardcoded path "${with-build-sysroot}/usr/include" to check target
libc ssp support. Based on GLIBC version strings in features.h in that search
path, gcc knows whether target (e)glibc implements stack protector itself.
However this breaks meta-toolchain, which actually has target libc headers
installed under {with-build-sysroot}/opt/... This way features.h is not found
and thus gcc-crosssdk-intermediate thinks that target (e)glibc doesn't support ssp.
Later when building eglibc-nativesdk, undefined reference to "__stack_chk_guard"
occurs which was caused by:
o eglibc do_configure found that gcc-crosssdk-intermediate supports ssp,
and thus enable -fstack-protector for nscd
o eglibc itself supports stack smash proctection for some architectures such
as i386, x86-64, etc. It's expected to use its own method to provide stack
protection, instead of relying on gcc. So eglibc rtld.os doesn't export
__stack_chk_guard to other modules
o then when installing nscd objects, gcc-crosssdk-intermediate sees the
flag "-fstack-protector", while it thought this eglibc doesn't implement
ssp itself, so gcc turns to the alternative to find a valid
__stack_chk_guard exported. eglibc doesn'g export it, while
gcc-crosssdk-intermediate itself disables libssp.
Then the undefined reference happens. If enabling libssp for gcc-crosssdk-
intermediate, it may also work-around this issue. But the ideal fix is still
to replace hard coded path with the actual one where target libc gets installed.
glibc-nativesdk doesn't encounter this issue because it thinks gcc doesn't
support ssp, and thus doesn't enable "-fstack-protector" for nscd. Don't know
the reason yet
This fix [BUGID #366]
Signed-off-by: Dexuan Cui <dexuan.cui@intel.com> Signed-off-by: Kevin Tian <kevin.tian@intel.com>
Yu Ke [Thu, 30 Sep 2010 12:54:49 +0000 (20:54 +0800)]
initrdscripts: remove the incorrect moblin message (Bug 368)
When install the live image into netbook/emenlow, the install tool prompt:
# Found drive at /dev/sda. Do you want to install moblin there ? [y/n]
The "moblin" here should be replaced by "poky".
Yu Ke [Thu, 30 Sep 2010 08:58:10 +0000 (16:58 +0800)]
qmmp: add qmmp into image (Bug 301)
qmmp is not included any image before, because it depends on one moblin recipe taglib. Since now the taglib is included into meta/recipes-support, it is safely to add qmmp into the task-poky-qt.
This commit also add an build time dependency for qmmp, because qmmp use some head files from taglib. Without this, qmmp build failure is occationally observed.
Bruce Ashfield [Sat, 25 Sep 2010 05:39:02 +0000 (01:39 -0400)]
linux-wrs: validate commits when forcing branches
Fixes [BUGID #372]
If for some unknown reason, the kernel git repository hasn't been properly
updated and is *older* than the revisions listed in default-revisions the
branch sanity checking kicks in and tries to force branches to the right
revs. The problem is, the revisions it is looking for don't exist in the
tree. As a result, error messages are thrown about invalid commit IDs.
These aren't helpful, and are simply confusing for the user.
Instead we can test for the commit ID, and if it isn't valid, indicate
that the commit ID isn't valid and that the tree is potentially out
of sync.
This situation is not common, but it is an easy test and the extra
sanity checking is useful.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Mark Hatle [Wed, 29 Sep 2010 17:17:36 +0000 (12:17 -0500)]
rootfs_rpm.bbclass: Add Dirnames configuration
[BUGID #327]
RPM attempts to validate all of the directories on the system are
owned by a package. While the root "/" directory was not owned, so
in some cases an error could be thrown. Resolve this by informing
RPM that yes, "/" is in fact a directory on the system that was
properly constructed and is "owned".
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
sstate: list directories in child first order in the manifest file
The directory list in the manifest file needs to be sorted such that child
directories are deleted first. Fortunately as the list is generated by walking
the directory tree achieving this is as simple as reversing the directory list.
Then for each command "/path/cmd" listed in ALTERNATIVE_LINKS, below is done
automatically:
${D}/path/cmd is renamed to ${D}/path/cmd.{PN}
a new alternative named 'cmd' is created which:
links /path/cmd to /path/cmd.{PN} with priority specified in ALTERNATIVE_PRIORITY
This way the recipe with multiple alternatives could be simplified a lot. There
are still some cases where above assumptions may break, but I expect more recipes
should benefit from this simple enhancement
