- Aug 15, 2016
-
-
Trevor Woerner authored
Allow chromium to build with gcc-6. Signed-off-by:
Trevor Woerner <twoerner@gmail.com>
-
Trevor Woerner authored
The unistd-2 patch was no longer being referenced by the recipe, so remove it. Signed-off-by:
Trevor Woerner <twoerner@gmail.com>
-
- Jun 15, 2016
-
-
Christopher N. Hesse authored
The correct dependencies on egl/wayland-egl are set in line 91: PACKAGECONFIG[use-egl] = ",,virtual/egl virtual/libgles2" We must not assign "" to the config in order not to break the build. Currently the 'bitbake chromium' task fails with the following message: Log data follows: | DEBUG: Executing python function sysroot_cleansstate | DEBUG: Python function sysroot_cleansstate finished | DEBUG: Executing shell function do_configure | Updating projects from gyp files... | Package egl was not found in the pkg-config search path. | Perhaps you should add the directory containing `egl.pc' | to the PKG_CONFIG_PATH environment variable | No package 'egl' found | Package wayland-egl was not found in the pkg-config search path. | Perhaps you should add the directory containing `wayland-egl.pc' | to the PKG_CONFIG_PATH environment variable | No package 'wayland-egl' found
-
Martin Jansa authored
Signed-off-by:
Martin Jansa <Martin.Jansa@gmail.com>
-
Michael Davis authored
Previously a build of i686 would fail looking for include.gypi. With this new directory it can be found by both i686 and i586 targetted builds. Signed-off-by:
Michael Davis <michael.davis@essvote.com>
-
- May 20, 2016
-
-
Carlos Rafael Giani authored
Signed-off-by:
Carlos Rafael Giani <dv@pseudoterminal.org> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- May 17, 2016
-
-
Carlos Rafael Giani authored
Not upgrading to M49 or M50 in this commit since this is currently the latest version that is supported by ozone-wayland Signed-off-by:
Carlos Rafael Giani <dv@pseudoterminal.org> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Carlos Rafael Giani authored
This eases the chromium browser upgrade process while ensuring that the chromium embedded framework is still supported. With this new structure, chromium.inc contains common code for cef3 and the browser, while chromium-browser.inc contains common code for browser releases (and itself includes chromium.inc). The actual .bb files then mainly set several values (and may contain release specific functions). Signed-off-by:
Carlos Rafael Giani <dv@pseudoterminal.org> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Carlos Rafael Giani authored
Signed-off-by:
Carlos Rafael Giani <dv@pseudoterminal.org> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Mar 16, 2016
-
-
Martin Jansa authored
* the old fix wasn't enough and it was still failing with warnings about unused variables in gtests Signed-off-by:
Martin Jansa <Martin.Jansa@gmail.com>
-
Martin Jansa authored
Signed-off-by:
Martin Jansa <Martin.Jansa@gmail.com>
-
- Mar 07, 2016
-
-
Martin Jansa authored
* for whatever reason setting -Wstrict-overflow just for problematic line doesn't work and it needs to be adjusted on function definition * the errors are also report line number where the function starts not the specific line which causes that warning ../../third_party/WebKit/Source/wtf/dtoa/bignum.cc: In member function 'void WTF::double_conversion::Bignum::Align(const WTF::double_conversion::Bignum&)': ../../third_party/WebKit/Source/wtf/dtoa/bignum.cc:715:31: error: assuming signed overflow does not occur when simplifying conditional to constant [-Werror=strict-overflow] for (int i = 0; i < zero_digits; ++i) { ^ ../../third_party/WebKit/Source/wtf/dtoa/bignum.cc:715:31: error: assuming signed overflow does not occur when simplifying conditional to constant [-Werror=strict-overflow] ../../third_party/WebKit/Source/wtf/dtoa/bignum.cc: In member function 'void WTF::double_conversion::Bignum::AssignDecimalString(WTF::double_conversion::Vector<const char>)': ../../third_party/WebKit/Source/wtf/dtoa/bignum.cc:105:10: error: assuming signed overflow does not occur when assuming that (X + c) < X is always false [-Werror=strict-overflow] void Bignum::AssignDecimalString(Vector<const char> value) { ^ ../../ui/gfx/image/image_util.cc: In function 'bool gfx::VisibleMargins(const gfx::ImageSkia&, int*, int*)': ../../ui/gfx/image/image_util.cc:50:6: error: assuming signed overflow does not occur when assuming that (X - c) <= X is always true [-Werror=strict-overflow] bool VisibleMargins(const ImageSkia& image, int* leading, int* trailing) { ^ Signed-off-by:
Martin Jansa <Martin.Jansa@gmail.com>
-
Martin Jansa authored
* d.getVar now takes 2 parameters, not 1 Signed-off-by:
Martin Jansa <Martin.Jansa@gmail.com>
-
- Jan 07, 2016
-
-
Trevor Woerner authored
If you are creating a kiosk, then enable this PACKAGECONFIG and the browser will start up fullscreen and without menu bars etc. Signed-off-by:
Trevor Woerner <twoerner@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Trevor Woerner authored
If the chromium run-helper script (google-chrome) is run as part of a (sysvinit) init script, shell variables $USER and $HOME will not be set (leading to unwanted behaviour). Replace the use of $USER with `whoami` and replace the use of $HOME with the OE bitbake variable ROOT_HOME. In this way the chromium run-helper script will work as expected in both instances (as part of an init script and from the cmdline). Signed-off-by:
Trevor Woerner <twoerner@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Trevor Woerner authored
Add more information to the "disable-api-keys-info-bar" PACKAGECONFIG description regarding another way to disable the "Google API keys are missing" warning. Signed-off-by:
Trevor Woerner <twoerner@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Jan 04, 2016
-
-
Trevor Woerner authored
Improve the visual appearance of the PACKAGECONFIG options by adding a title to the PACKAGECONFIG option section, wrapping lines to 80 columns, aligning the left justification for subsequent lines, moving the default settings for each option to a more prominent location, and moving all the items affected by PACKAGECONFIG options to be closer to where these options are listed and explained. Signed-off-by:
Trevor Woerner <twoerner@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Trevor Woerner authored
The gl PACKAGECONFIG is "use-egl", not "use-gl". Adjust comment to include a space after initial comment delimiter (#). Close parenthesis at end of comment. Signed-off-by:
Trevor Woerner <twoerner@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Dec 11, 2015
-
-
Andre McCurdy authored
Signed-off-by:
Andre McCurdy <armccurdy@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Andre McCurdy authored
Signed-off-by:
Andre McCurdy <armccurdy@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Nov 10, 2015
-
-
Max Krummenacher authored
The link errors addressed with commit '7887e66d fix build with gcc 4.9.4' occur only when linking with gold. Thus apply the fix only if distro features contain ld-is-gold. Signed-off-by:
Max Krummenacher <max.oss.09@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Nov 09, 2015
-
-
Max Krummenacher authored
During the final link we get errors like this: .../4.9.4/ld: warning: hidden symbol 'sqlite3_temp_directory' in obj/.../sqlite.sqlite3.o) is referenced by DSO .../usr/lib/libsoftokn3.so .../4.9.4/ld: error: treating warnings as errors collect2: error: ld returned 1 exit status Use chromium.bb solution to not treat linker warnings as errors. Signed-off-by:
Max Krummenacher <max.oss.09@gmail.com> Upstream-Status: Inappropriate configuration Signed-off-by:
Max Krummenacher <max.oss.09@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Max Krummenacher authored
Use gypi file for i586 to create the configuration for the x86-64 architecture. Signed-off-by:
Max Krummenacher <max.oss.09@gmail.com> Upstream-Status: Inappropriate configuration Signed-off-by:
Max Krummenacher <max.oss.09@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Max Krummenacher authored
A cef3 build with gcc5 fails with warnings treated as errors. bignum.cc, image_util.cc: strict-overflow warning gtest-typed-test.h: unused-variable warning SaturatedArithmetic.h: parentheses warning Compiles for qemux86 with gcc 5.2.0 and still compiles on armv7a with gcc linaro-4.9.4. Note that the resulting binary has not been tested on a target. Signed-off-by:
Max Krummenacher <max.oss.09@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Max Krummenacher authored
Signed-off-by:
Max Krummenacher <max.oss.09@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Max Krummenacher authored
A chromium build with gcc5 fails with warnings treated as errors. bignum.cc, image_util.cc: strict-overflow warning Compiles and runs on qemux86-64 with gcc 5.2.0 and still compiles on armv7a with gcc linaro-4.9.4. Signed-off-by:
Max Krummenacher <max.oss.09@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Oct 28, 2015
-
-
Max Krummenacher authored
Explicitely include a needed header to fix compile time error. sys/stat.h used to be indirectly included from another header but no longer is. Signed-off-by:
Max Krummenacher <max.oss.09@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Sep 23, 2015
-
-
Carlos Alberto Lopez Perez authored
* Remove the ENABLE_X11 variable and rename ENABLE_WAYLAND to CHROMIUM_ENABLE_WAYLAND. Make the old ENABLE_X11 be just "not CHROMIUM_ENABLE_WAYLAND". * Allow configuring this in the local.conf file. * Add a couple of sanity checks and two notes that will let the user be sure if the ozone-wayland patches will be applied or not. This will be checked before running the do_fetch step of Chromium. That way we avoid printing any warning or doing any check when parsing the recipe. So we only warn or abort when Chromium is actually going to be built. Signed-off-by:
Carlos Alberto Lopez Perez <clopez@igalia.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Carlos Alberto Lopez Perez authored
Signed-off-by:
Carlos Alberto Lopez Perez <clopez@igalia.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Carlos Alberto Lopez Perez authored
* Add also a patch that workarounds a build error when building in Debug mode and using GCC. This patch has no effect when building in Release mode because is guarded within ifdef(debug) blocks. Signed-off-by:
Carlos Alberto Lopez Perez <clopez@igalia.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Aug 18, 2015
-
-
Julien Brianceau \(jbriance\) authored
Do not let ninja choose the number of jobs to run in parallel. Signed-off-by:
Julien Brianceau <jbriance@cisco.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Andre McCurdy authored
Support i686 without needing to duplicate the i586 over-ride. Signed-off-by:
Andre McCurdy <armccurdy@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
Andre McCurdy authored
Support i686 without needing to duplicate the i586 over-ride. Signed-off-by:
Andre McCurdy <armccurdy@gmail.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Jul 23, 2015
-
-
Carlos Alberto Lopez Perez authored
Otherwise chromium.inc will overwrite the list of DEPENDS. Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Jul 07, 2015
-
-
Gary Thomas authored
OE-core now warns if PACKAGECONFIG is used to set an option that does not have a corresponding PACKAGECONFIG[option]="xxx" line. This recipe makes use of many such options & this patch suppresses those warnings by listing the options. Signed-off-by:
Gary Thomas <gary@mlbassoc.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- Jun 17, 2015
-
-
Gary Thomas authored
This recipe should be using the python interpreter as built by bitbake, rather than relying on the host python. In some cases, the host python may be incomplete or just plain old, yielding errors during the configure step such as: | from sgmllib import SGMLParser, SGMLParseError | ImportError: No module named sgmllib | ninja: build stopped: subcommand failed. Adding 'inherit pythonnative' solves this problem on all build hosts. Signed-off-by:
Gary Thomas <gary@mlbassoc.com> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-
- May 28, 2015
-
-
Zoltan Kuscsik authored
Wrong location of cef caused QA errors. Change-Id: I56743864f9e7d47e7c87fb26c478cd812d5d6961 Signed-off-by:
Zoltan Kuscsik <zoltan.kuscsik@linaro.org> Signed-off-by:
Khem Raj <raj.khem@gmail.com>
-
- May 14, 2015
-
-
Khem Raj authored
Use specific domains for installation, so they can be built and installed in a single build Signed-off-by:
Khem Raj <raj.khem@gmail.com>
-
- May 13, 2015
-
-
Nikolay Dimitrov authored
Change CHROMIUM_BUILD_TYPE to Bitbake variable, otherwise Chromium files are not installed in the holding area, and Chromium fails to start on the target. Signed-off-by:
Nikolay Dimitrov <picmaster@mail.bg> Signed-off-by:
Khem Raj <raj.khem@gmail.com>
-
- May 08, 2015
-
-
Khem Raj authored
CEF rework caused few problems where we missed packaging needed resource files and misplaced some icu data files Use loops to install similar files Tested on x86 emulator making sure google-chrome starts up properly Signed-off-by:
Khem Raj <raj.khem@gmail.com> Acked-by:
Otavio Salvador <otavio@ossystems.com.br> Signed-off-by:
Otavio Salvador <otavio@ossystems.com.br>
-