Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merge latest upstream #20

Closed

Conversation

erickshepherdNI
Copy link

No merge conflicts. Merged by checking out upstream/scarthgap and the merging with origin/nilrt/master/next using the Ours strategy in order to avoid merge conflicts and upstream commits that were not being properly merge. Previous discussion can be found on the old PR here: #19

AB#2755018

Testing:

  • Ran "bitbake packagefeed-ni-core"
  • Ran "bitbake packagegroup-ni-desirable"
  • Ran "bitbake package-index && bitbake nilrt-base-system-image"
  • Reimaged a cRIO with the new base image and successfully booted it

fmntf and others added 22 commits April 5, 2024 08:22
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
…vesdk

Since packagegroup.bbclass is using inherit_defer in oe-core, some classes are
now appearing after native resulting in QA errors

QA Issue: nativesdk-packagegroup-qt5-toolchain-host: native/nativesdk class is not inherited last, this can result in unexpected behaviour. Classes inherited after native/nativesdk: allarch.bbclass [native-last]

Signed-off-by: Chaitanya Vadrevu <chaitanya.vadrevu@ni.com>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
'inherit_defer' was added in scarthgap and since meta-qt5 requires using it,
mark the layer as compatible with only scarthgap.

Signed-off-by: Chaitanya Vadrevu <chaitanya.vadrevu@ni.com>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
* the old one was copy pasted from:
  qtquicktimeline_git.bb:SRCREV = 29394e35db43acb20e7b4d5e978a733f5b4232a6

Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
…ild with gcc-14

Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
* Accepted was replaced with Backport in gatesgarth:
  https://docs.yoctoproject.org/migration-guides/migration-3.2.html#miscellaneous-changes

ERROR: qt5-creator-4.15.2+git-r0 do_patch: QA Issue: Malformed Upstream-Status in patch
/OE/build/oe-core/meta-qt5/recipes-qt/qt5/qt5-creator/0002-clangformat-Fix-build-with-LLVM-13.patch
Please correct according to https://docs.yoctoproject.org/contributor-guide/recipe-style-guide.html#patch-upstream-status :
Upstream-Status: Accepted [https://code.qt.io/cgit/qt-creator/qt-creator.git/commit/?id=14eb5b382cabf90352634393c19bf87a4c15766a] [patch-status]
ERROR: qt5-creator-4.15.2+git-r0 do_patch: QA Issue: Malformed Upstream-Status in patch
/OE/build/oe-core/meta-qt5/recipes-qt/qt5/qt5-creator/0003-clangformat-Fix-build-with-LLVM-13.patch
Please correct according to https://docs.yoctoproject.org/contributor-guide/recipe-style-guide.html#patch-upstream-status :
Upstream-Status: Accepted [https://code.qt.io/cgit/qt-creator/qt-creator.git/commit/?id=55b91a76172a3235b4879daf0b675519d5b02db7] [patch-status]

Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Commit b61ea2c stripped host paths for the
CFLAGS and CXXFLAGS entries in qmake.conf, but QMAKE_LFLAGS entry in the
target's qmake.conf still contains a host path.

Fix that by using the same mechanism as the previous commit.

Signed-off-by: James Minor <james.minor@ni.com>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
The buildpaths QA check is producing the following warnings:
WARNING: qtbase-5.15.13+git-r0 do_package_qa: QA Issue: File /usr/bin/.debug/qmake in package qtbase-dbg contains reference to TMPDIR [buildpaths]
WARNING: qtbase-5.15.13+git-r0 do_package_qa: QA Issue: File /usr/bin/qmake in package qtbase-tools contains reference to TMPDIR [buildpaths]
WARNING: qtbase-5.15.13+git-r0 do_package_qa: QA Issue: File
/usr/src/debug/qtbase/5.15.13+git/src/corelib/global/qconfig.cpp in package
qtbase-src contains reference to TMPDIR [buildpaths]

To remove the host path from qmake, set various path strings in configure.pri to
something else.

Signed-off-by: James Minor <james.minor@ni.com>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Some of the auto tests use SRCDIR which will fail on a target when compiled and
run as a ptest.  Since those tests don't pass completely anyway and so that host
paths won't leak into these binaries, remove them.

Signed-off-by: James Minor <james.minor@ni.com>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Justin Bronder <jsbronder@cold-front.org>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Justin Bronder <jsbronder@cold-front.org>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
PyQt v5.15.10 has been released. This is a bug-fix release and also adds
support for Python v3.12.

Signed-off-by: Justin Bronder <jsbronder@cold-front.org>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Reinstate support for older OE releases by using inherit_defer only
for OE Scarthgap release, while retaining the old behavior for all
the recently supported older OE releases.

Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
This reverts commit 6e24497.

meta-qt5/meta-qt5#561 added a work around
to keep compatibility with older releases.
OE dunfell builds have to inherit setuptools3, while anything newer
has to inherit python_setuptools_build_meta. While OE dunfell is now
EOL, this layer still claims to support it, so make it buildable and
avoid breakage until the layer stops to support OE dunfell for good.
Once that happens, this commit should be reverted.

Signed-off-by: Marek Vasut <marex@denx.de>
CVE fixed:
-CVE-2024-25580: qtbase: potential buffer overflow when reading KTX images
Upstream-Status: Backport from https://download.qt.io/official_releases/qt/5.15/CVE-2024-25580-qtbase-5.15.diff

Signed-off-by: Rohini Sangam <rsangam@mvista.com>
Signed-off-by: Siddharth Doshi <sdoshi@mvista.com>
    
- do not use ${} in nested shell functions to mix bitbake's shell variables
- git rid of 'local' which is not a POSIX variant
    
Signed-off-by: Valek Andrej <andrej.v@skyrain.eu>
* it's already failing to build with gcc-14:
  qtwebkit/5.15.13+git/recipe-sysroot/usr/include/c++/14.1.0/type_traits:1237:52: error: non-constant condition for static assertion
  qtwebkit/5.15.13+git/recipe-sysroot/usr/include/c++/14.1.0/type_traits:1237:52: error: 'constexpr std::true_type std::__is_complete_or_unbounded(__type_identity<_Tp>) [with _Tp = WebCore::ContentSecurityPolicyHashAlgorithm; long unsigned int <anonymous> = 4; true_type = true_type]' used before its definition
  qtwebkit/5.15.13+git/recipe-sysroot/usr/include/c++/14.1.0/type_traits:1237:52: error: non-constant condition for static assertion
  qtwebkit/5.15.13+git/recipe-sysroot/usr/include/c++/14.1.0/type_traits:1237:52: error: 'constexpr std::true_type std::__is_complete_or_unbounded(__type_identity<_Tp>) [with _Tp = WTF::Vector<unsigned char>; long unsigned int <anonymous> = 16; true_type = true_type]' used before its definition

* but this bump includes some of the .patch files we won't need here:
  $ git log --oneline 10cd6a106e1c461c774ca166a67b8c835c755ef7..ac8ebc6c3a56064f88f5506e5e3783ab7bee2456
  ac8ebc6c3 (origin/5.212) Enable QML tests
  5a99e5d4c Import QtWebKit commit 5d89eef9bc689abb59ef799b445c35b1e322b3f1
  685ed792f Import QtWebKit commit 143b469188bde226d5e70cff5f01699b8a29d865
  8967b4f87 Import QtWebKit commit 6faf5d547055321f0d518fe83766499cac5d8664
  08eaf858c Import QtWebKit commit cf37bc44f798d3c5a26d17086ada92688b9bfe6d
  412e257d0 Workaround for Windows Store python3.exe
  e16357a52 Import QtWebKit commit f966d667c14ddcfe983f4a31fc80a1edbb6ada10

  0005-Let-Bison-generate-the-header-directly-to-fix-build-.patch
  0006-Disable-code-related-to-HTTP-2-when-Qt-is-configured.patch
  0007-Fix-compilation-with-Python-3.9-avoid-passing-encodi.patch
  were applied upstream and most of
  0006-Riscv-Add-support-for-riscv.patch/0009-Riscv-Add-support-for-riscv.patch
  as well (not sure if we need the last chunk)

Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Merged by checking out upstream/scarthgap and merging with origin/nilrt/master/next using the Ours strategy in order to preserve commits from upstream that were not being merged correctly.

Signed-off-by: Erick Shepherd <erick.shepherd@ni.com>
Copy link

@chaitu236 chaitu236 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Merged by checking out upstream/scarthgap and merging with origin/nilrt/master/next using the Ours strategy in order to preserve commits from upstream that were not being merged correctly.

Changes look fine but can you split this merge commit message into multiple lines at 72 chars boundary? - ref

@chaitu236
Copy link

Reworded commit message and merged as ef05ff4.

@chaitu236 chaitu236 closed this Sep 6, 2024
@chaitu236
Copy link

Some background for posterity's sake

We're trying to merge upstream/scarthgap into nilrt/master/next.
Both these branches have common ancestor b8e1ae8 (layer.conf: Update for the scarthgap release series).

We send some changes upstream which were merged into upstream/scarthgap - we then cherry-picked these commits into nilrt/master/next. There are no changes in nilrt/master/next (after the branches diverged) other than these cherry-picked commits.

upstream/scarthgap reverted one of the commits - ref d8eeef0. Now when we merge upstream/scarthgap into nilrt/master/next, this revert is not applied - git for some reason keeps the original commit changes.

image

Interestingly, merging bb65104 (the commit before the revert) into nilrt/master/next and then merging upstream/scarthgap gets the repo into the state that we want - i.e., contents of revert are applied and both trees don't have any diff.

Since we don't have any NI specific changes that are not already in upstream/scarthgap, we ended up doing git checkout upstream/scarthgap && git merge -s ours nilrt/master/next - which ties together the two branches with a merge commit but throws away anything in nilrt/master/next branch.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

8 participants