Remove 0001-Set-use_tcl-to-be-empty-string-if-tcl-is-disabled.patch.
This patch is obsolete and not needed because the current graphviz
configure.ac has correct logic of checking use_tcl. This use_tcl
variable needs to be set when '--disable-tcl' is set, otherwise,
things will behave as if no option is supplied and the configure
process will check tcl automatically.
This patch is problematic because its logic against the current version
is wrong. The recipe has already explicitly set '--disable-tcl', so
the configure process should not do automatic checking for tcl at do_configure.
This patch fixes do_configure error when host has tcl8.6-dev installed. The
error is like below:
QA Issue: This autoconf log indicates errors, it looked at host include and/or
library paths while determining system capabilities.
Rerun configure task after fixing this. [configure-unsafe]
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
Graphviz 2.36 before 10.0.0 has an out-of-bounds read via a
crafted config6a file. NOTE: exploitability may be uncommon
because this file is typically owned by root.
CVE-2023-46045-0003.patch is the CVE fix and CVE-2023-46045-0001.patch,
CVE-2023-46045-0002.patch are dependent commits to fix the CVE.
Reference:
https://nvd.nist.gov/vuln/detail/CVE-2023-46045
Upstream patches:
361f274ca93f31704cafa95f977f5d
Signed-off-by: Yogita Urade <yogita.urade@windriver.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
Remove the CXX flag "-O2" for GCC 13 and 14. There's a bug with ARM GCC
that breaks the iteration of "types" in the createMeshShaderMiscTestsEXT
function. This issue is not present for clang or x86_64 GCC 14.
It seems that the array is not initialized before the first iteration.
In testing this can result in a random value being used. This can
manifest in LINES type being processed twice, resulting in the following
error:
FATAL ERROR: Failed to initialize dEQP: Test case with non-unique name
'no_lines' added to group 'misc'.
Signed-off-by: Randolph Sapp <rs@ti.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
The recipe used in the meta-openembedded is a different st package compared to the one which has the CVE issue.
Package used in meta-embedded: https://st.suckless.org/
Package with CVE issue: https://www.npmjs.com/package/st
No action required.
Signed-off-by: Ninette Adhikari <ninette@thehoodiefirm.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Fixes
ERROR: ydotool-1.0.4+git-r0 do_package: QA Issue: ydotool: Files/directories were installed but not shipped in any package:
/usr/lib
/usr/lib/systemd
/usr/lib/systemd/user
/usr/lib/systemd/user/ydotoold.service
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Changelog:
===========
-Fix stems reader processing code broken in 3.3.114.
-Update xml2stems to handle newer "loc" vs "fl" xml tags.
-Change preg_regex_c_1 decl to use regex_t* as datatype.
-Move gtkwave.appdata.xml to
-io.github.gtkwave.GTKWave.metainfo.xml.
-Fixed popen security advisories:
-Fixed FST security advisories:
-Fixed evcd2vcd security advisories:
-Fixed VCD security advisories:
-Fixed VZT security advisories:
-Fixed LXT2 security advisories:
-Remove FST_DO_MISALIGNED_OPS.
-Update lz4 to current version from github.
-Change LZ4_compress to LZ4_compress_default.
-Update libghw.c/.h to latest upstream version.
-Fix for -Wsign-compare in fstapi.c.
-Security fixes for GHW.
-Fix left shift of a negative number warning in fstapi.c.
-Fix ctrl-A behavior for SST filter entry.
-Fix for bad shmat return value in main.c.
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The following paths have been replaced with PYTHON_SITEPACKAGES_DIR:
- "${libdir}/${PYTHON_DIR}/site-packages"
- "${libdir}/python${PYTHON_BASEVERSION}/site-packages"
- "${libdir}/python*/site-packages"
- "${libdir}/python3.*/site-packages"
Signed-off-by: alperak <alperyasinak1@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
It fixes an issue where the fbdev driver fails to set up a buffer due to
buffer size mismatch. It happens when the actual screen resolution is
smaller than the default one 800x480 because the actual resolution was
being set after lv_display_set_buffers.
Signed-off-by: Jaeyoon Jung <jaeyoon.jung@lge.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Move the default PACKAGECONFIG setting from both recipes into
lv-conf.inc to avoid duplication. No functional change.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The PV has been appended because the source has been pulled from
git as of specific revision past 9.0.1 release of LVGL. The current
source is pulled as of tagged release 9.1.0, drop the extra git
suffix.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
This ALLOW_EMPTY was a workaround for LVGL built as static library.
The lvgl is now built as shared library, drop this remnant.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Update LVGL to final 9.1.0 . Drop the XRGB8888 fix which
is part of the release.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The LVGL demo is configured slightly differently than the LVGL itself.
Logging and printing is enabled, memory size is higher and compressed
fonts are enabled. Reinstate the settings, otherwise the demo keeps
crashing at least with SDL2 backend.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Changelog:
===========
-configure: Use AC_SYS_LARGEFILE to enable large file support
-Add -version option
-updates -help option to not require successfully opening a display
-Print which arguments were unknown before giving usage message
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Changelog:
==========
-gitlab CI: stop requiring Signed-off-by in commits
-configure: Use AC_SYS_LARGEFILE to enable large file support
-Use reallocarray if available, provide local version if not
-Silence -Wanalyzer-out-of-bounds warnings from gcc 13
-Variable scope reductions, as suggested by cppcheck
-unifdef __UNIXOS2__
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* use [[:space:]] instead of [ \t]
* remove global replace, as there is only one instance to replace for
each expression
* use extended regular expressions and remove escape characters
Signed-off-by: Christophe Chapuis <chris.chapuis@gmail.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The libdrm include is only needed when "drm" is selected in
PACKAGECONFIG.
In addition, move the TARGET_CFLAG addition to lv-conf.inc, where the
DRM configuration is actually selected for LVGL.
Signed-off-by: Christophe Chapuis <chris.chapuis@gmail.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Use [ \t] to detect optionnal spaces at the beginning of each
line, and after the define.
Signed-off-by: Christophe Chapuis <chris.chapuis@gmail.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
When DRM is activated for LVGL, it adds a dependency on drm.h.
As for lvgl-demo-fb, add an include path to fix this usecase.
Signed-off-by: Christophe Chapuis <chris.chapuis@gmail.com>
Acked-by: Martin Jansa <martin.jansa@gmail.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
As it can be usefull to customize these other variables, let's
add them in lv-conf.inc.
Signed-off-by: Christophe Chapuis <chris.chapuis@gmail.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Use back reference instead of repeating the search pattern. Later on
this could lead to a more generic way of replacing values in the .h file.
Signed-off-by: Christophe Chapuis <chris.chapuis@gmail.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
As ${includedir}${PN} is already listed in FILES, there is no
need to add one of its subdirectories.
Signed-off-by: Christophe Chapuis <chris.chapuis@gmail.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Add an install append to copy the generated lv_conf.h as part of the
lvgl package, so that it will be found and used by the recipes that
want to use lvgl.
Signed-off-by: Christophe Chapuis <chris.chapuis@gmail.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Fix a typo in the first sed instruction, where the #if 0 is
never changed to #if 1 at the beginning of lv_conf.h.
Signed-off-by: Christophe Chapuis <chris.chapuis@gmail.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* as oe-core did in:
https://git.openembedded.org/openembedded-core/commit/?id=d4c346e8ab
* when people are have to maintain own PRs for recipes in oe-core, they
might add them for meta-oe recipes at the same time when upgrading
to next LTS
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Add missing version suffix to shared libraries. Currently the filename of
generated shared libraries is only liblvgl.so, which prevents coexistence
of different versions of LVGL on the same system. Set VERSION and SOVERSION
to make cmake add the version suffix to generated shared libraries. That
changes the filename to liblvgl.so.9.0.0 and includes symlink with major
ABI version, i.e. liblvgl.so.9 .
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Replace the current awful sed hacking with real patches bound for upstream.
The sed hacking was simply not maintainable anymore.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Build LVGL as shared library instead of static library. In case
there are multiple applications linking against LVGL, this is more
efficient, and it also makes it possible to update only the LVGL
library separately from the applications.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The ARGB8888 framebuffer format for base canvas makes little sense
as the base canvas is unlikely to be transparent and require alpha.
Use XRGB8888 framebuffer format which is more widely supported by
DRM drivers as base plane pixel format.
This is identical to upstream fix for this issue which already landed:
c71e5f84bb
The upstream fix just never made it into the LVGL 9.0.0 for some reason,
an repeated upstream fix is being worked in now.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Configure all the LVGL assertions to enabled or disabled based on OE
DEBUG_BUILD variable. This way, debug builds come with assertions
which are expensive, while regular builds do not.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Expose SDL2 fullscreen mode as configuration option. The default
is disabled, which matches the SDL2 backend configuration in LVGL.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The lv-conf.inc is much more fitting name, since there is now only one
configuration file and that is called lv_conf.h . No functional change.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade LVGL and matching recipes to LVGL 9 series. Use latest git HEAD
instead of 9.0.0 release to pull in fixes which landed after the 9.0.0
release and which are of interest to Linux deployments.
There are significant breaking changes, which are listed below.
The lv-drivers and lv-lib-png PNG library seem to be part of the main LVGL
repository, drop the now unnecessary recipes.
Configuration is now done in lv_conf.h only, there is no more lv_drv_conf.h,
rework lv-drivers.inc accordingly. USE_SDL_GPU support has been renamed to
LV_USE_DRAW_SDL and newly depends on libsdl2-image, SDL_DOUBLE_BUFFERED to
LV_SDL_BUF_COUNT=2 .
All configuration options from lvgl_%.bb, LV_COLOR_DEPTH, LV_MEM_CUSTOM as
well as LV_TICK_CUSTOM, LV_TICK_CUSTOM_INCLUDE, LV_TICK_CUSTOM_SYS_TIME_EX
are all moved into lv-drivers.inc , so the configuration is done in one
place, using the same set of configuration options.
Wayland support is gone, drop it from lv-drivers.inc. Evdev input device
selection via EVDEV_NAME is also gone, the selection is now done using
API call lv_evdev_create(). DRM card selection via DRM_CARD is also gone,
the selection is now done using API call lv_linux_drm_set_file(). Move
LVGL_CONFIG_DRM_CARD into lvgl-demo-fb, which calls lv_linux_drm_set_file().
The lvgl-demo-fb recipe is adjusted to be compatible with LVGL 9 again,
this makes some of the sed adjustments really awful, so this part will
be replaced by upstream patches in a separate commit to avoid growing
this commit even more.
The lvgl-demo-fb recipe is also no longer using git submodules, but
instead fetches both its source and LVGL using git fetcher separately.
This is needed to build the SDL backend successfully, which requires
newer LVGL than what the repository points to via its submodule.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
This recipe is barely compatible with LVGL 8.3, the repository
seems to be unreachable now, drop the recipe.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
remove STAGING_DIR_HOST and S from the waylandpp-targets.cmake file
Signed-off-by: Markus Volk <f_l_k@t-online.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
remove STAGING_DIR_HOST from the LibVNCServerTargets.cmake file
Signed-off-by: Markus Volk <f_l_k@t-online.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Changelog:
===========
-Apple's Diffie-Hellman authentication (security type 30) has been fixed.
-A new client connection no longer causes a DNS lookup.
-Clients are now allowed to request more than 32 encodings
-Zlib streams are now preserved when a client switches between encodings
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
