Re: [Openembedded-architecture] [OE-core] Where are we at?

On 9 May 2024, at 19:58, Ross Burton via lists.openembedded.org wrote: > > > >> On 9 May 2024, at 10:37, Richard Purdie via lists.openembedded.org >> wrote: >> Intermittent do_compile failure in libportal: >> https://autobuilder.yoctoproject.org/typhoon/#/

[OE-core] [PATCH] libportal: fix rare build race

There is a build race where the libportal.vapi is not always generated before it is needed to build libportal-gtk*.vapi. Backport the fix from upstream. [ YOCTO #15479 ] Signed-off-by: Ross Burton --- ...build-race-when-building-GTK-vapi-fi.patch | 49 +++ .../libportal

Re: [OE-core] [PATCH v2] glib/gobject-introspection: update 2.78.4 -> 2.80.0, 1.78.1 -> 1.80.0

On 10 May 2024, at 10:52, Alexander Kanavin wrote: > > On Tue, 7 May 2024 at 17:47, Ross Burton wrote: >> packages/cortexa57-poky-linux/glib-2.0/glib-2.0-utils: FILELIST: added >> "/usr/bin/gi-compile-repository /usr/bin/gi-inspect-typelib >> /usr/bin/gi-d

Re: [OE-core] Where are we at?

> On 9 May 2024, at 10:37, Richard Purdie via lists.openembedded.org > wrote: > Intermittent do_compile failure in libportal: > https://autobuilder.yoctoproject.org/typhoon/#/builders/117/builds/4760/steps/12/logs/stdio Root caused this, fix incoming. > debuginfod keeps breaking: >

Re: [OE-core] [PATCH 1/6] enchant2: upgrade 2.6.9 -> 2.7.2

Can you respin this to 2.7.3, so we get the vala/gcc14 fix? Thanks, Ross > On 7 May 2024, at 03:23, Anuj Mittal via lists.openembedded.org > wrote: > > Signed-off-by: Anuj Mittal > --- > .../enchant/{enchant2_2.6.9.bb => enchant2_2.7.2.bb}| 2 +- > 1 file changed, 1 insertion(+),

Re: [OE-core] [PATCH] enchant2: Disable incompatible-pointer-types warning as error

On 9 May 2024, at 05:14, Khem Raj via lists.openembedded.org wrote: > +#api.vala:43:18: error: assignment to 'gchar **' {aka 'char **'} from > incompatible pointer type 'const gchar * const*' {aka 'const char * const*'} > [-Wincompatible-pointer-types] > + > +CFLAGS +=

Re: [OE-core] [PATCHv1] Fix building error with '-O2' in sysroot path

Feels like we should both backport that _and_ pass —disable-O-flag-munging, as we control the compiler flags. Ross > On 8 May 2024, at 11:48, simit.ghane via lists.openembedded.org > wrote: > > Characters like '-O2' or '-Ofast' will be replaced by '-O1' and '-O0' > respectively when

Re: [OE-core] [PATCH v2] glib/gobject-introspection: update 2.78.4 -> 2.80.0, 1.78.1 -> 1.80.0

> On 2 May 2024, at 11:32, Alexander Kanavin via lists.openembedded.org > wrote: > -GTKDOC_MESON_OPTION = "gtk_doc" > +inherit meson gettext gi-docgen pkgconfig ptest-gnome > upstream-version-is-even bash-completion gio-module-cache manpages > gobject-introspection-data You’ll want to set

Re: [OE-core] [PATCH v2] glib/gobject-introspection: update 2.78.4 -> 2.80.0, 1.78.1 -> 1.80.0

> On 2 May 2024, at 11:32, Alexander Kanavin via lists.openembedded.org > wrote: > +++ > b/meta/recipes-core/glib-2.0/files/0001-girepository-introspection-correctly-install-.gir-fi.patch > @@ -0,0 +1,88 @@ > +From 4e6be5d15e5d2f8e494d22a3d49b623064f9c1a5 Mon Sep 17 00:00:00 2001 > +From:

Re: [OE-core] [PATCH v2] glib/gobject-introspection: update 2.78.4 -> 2.80.0, 1.78.1 -> 1.80.0

Hi Alex, Thanks for digging into this! packages/cortexa57-poky-linux/glib-2.0/glib-2.0-utils: FILELIST: added "/usr/bin/gi-compile-repository /usr/bin/gi-inspect-typelib /usr/bin/gi-decompile-typelib” packages/cortexa57-poky-linux/glib-2.0/glib-2.0-utils: PKGSIZE changed from 770935 to

[OE-core] [PATCH 2/2] gstreamer1.0: skip another known flaky test

The baseparse:parser_pull_short_read test is known to be unreliable, according to the list of known bad tests in gst-devtools. Also clean up an incorrect comment. Signed-off-by: Ross Burton --- meta/recipes-multimedia/gstreamer/gstreamer1.0/run-ptest | 9 +++-- 1 file changed, 7 insertions

[OE-core] [PATCH 1/2] curl: locale-base-en-us isn't glibc-specific

The musl-locales package provides this too, so we can depend without a libc override. Signed-off-by: Ross Burton --- meta/recipes-support/curl/curl_8.7.1.bb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/meta/recipes-support/curl/curl_8.7.1.bb b/meta/recipes-support/curl

[OE-core] [PATCH] openssl: fix pkgconfig path problems

f725a81c, 2009 Signed-off-by: Ross Burton --- meta/recipes-connectivity/openssl/openssl_3.3.0.bb | 6 +- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/meta/recipes-connectivity/openssl/openssl_3.3.0.bb b/meta/recipes-connectivity/openssl/openssl_3.3.0.bb index 2cdaf4c75d5

Re: [OE-core] [PATCH] openssl: fix pkg-config file libdir from /usr back to /usr/lib

On 3 May 2024, at 11:43, Martin Jansa via lists.openembedded.org wrote: > Isn't it easier to fix tpm2-openssl then? > > I had similar issue in chip which was using "pkg-config > --variable=prefix openssl" and didn't expect it to be empty in "if > prefix[-4] == '/usr'" and rather patched chip

Re: [OE-core] [PATCH] u-boot: Enable devupstream

On 2 May 2024, at 12:02, Richard Purdie via lists.openembedded.org wrote: >> +BBCLASSEXTEND += "devupstream:target" >> +SRCREV:class-devupstream = "${AUTOREV}" >> +PV:class-devupstream = "dev+git" > > Doesn't that mean parsing would hit the network by default every time? I _believe_ it only

[OE-core] [PATCH] insane: handle dangling symlinks in the libdir QA check

. [ YOCTO #13949 ] Signed-off-by: Ross Burton --- meta/classes-global/insane.bbclass | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/meta/classes-global/insane.bbclass b/meta/classes-global/insane.bbclass index e963001d09a..c32dfffd836 100644 --- a/meta/classes-global/insa

Re: [OE-core] [PATCH][master scarthgap] procps: fix build with new glibc but old kernel headers

Hold this, v2 incoming. It breaks on clang which exposes another problem with the current code… Ross > On 26 Apr 2024, at 11:40, Ross Burton via lists.openembedded.org > wrote: > > If you're building procps with a newer glibc (with pidfd_open()) but > older kernel headers (

[OE-core] [PATCH][master scarthgap] procps: fix build with new glibc but old kernel headers

If you're building procps with a newer glibc (with pidfd_open()) but older kernel headers (say 4.x, before __NR_pidfd_open) then procps will fail to build because of a typo in configure.ac. Signed-off-by: Ross Burton --- .../procps/procps/pidfd.patch | 25

Re: [OE-core] btrfs-tools-native build python error

On 25 Apr 2024, at 13:29, Sheng Lean Tan via lists.openembedded.org wrote: > > Hi, > Good day to openembedded community! > > This is my first post - hope this is the right place to ask for help > regarding a build issue, as refer from here: >

Re: [OE-Core] [PATCH 27/34] python3-setuptools: upgrade 69.1.1 -> 69.2.0

On 23 Apr 2024, at 13:37, Ross Burton via lists.openembedded.org wrote: > > Somehow, this distributes a copy of the build tree: > > Changes to packages/aarch64-linux/python3-setuptools-native (sysroot): > /usr/lib/python3.12/site-packages/build was added > /usr/lib/python

Re: [OE-Core] [PATCH 27/34] python3-setuptools: upgrade 69.1.1 -> 69.2.0

Somehow, this distributes a copy of the build tree: Changes to packages/aarch64-linux/python3-setuptools-native (sysroot): /usr/lib/python3.12/site-packages/build was added /usr/lib/python3.12/site-packages/build/lib was added /usr/lib/python3.12/site-packages/build/lib/_distutils_hack was

Re: [OE-core] [PATCH 19/19] connman: submit 0002-resolve-musl-does-not-implement-res_ninit.patch upstream

On 19 Apr 2024, at 13:19, Alexander Kanavin via lists.openembedded.org wrote: > -Upstream-Status: Pending > +Upstream-Status: Submitted [to conn...@lists.linux.dev,mar...@holtmann.org] Links to the actual posts make it much easier to find any discussion in the future:

Re: [OE-core] [PATCH 14/19] libtraceevent: submit meson.patch upstream

On 19 Apr 2024, at 13:19, Alexander Kanavin via lists.openembedded.org wrote: > -Upstream-Status: Pending > +Upstream-Status: Submitted [via email to > linux-trace-de...@vger.kernel.org,rost...@goodmis.org] A link to the thread makes it easier to find any discussion in a years time:

Re: [OE-core] [PATCH v2] shadow: install manpages

On 11 Apr 2024, at 23:36, Dan McGregor via lists.openembedded.org wrote: > +do_install:append:class-nativesdk() { > + oe_runmake -C ${B}/man DESTDIR="${D}" sbindir="${base_sbindir}" > usbindir="${sbindir}" install-man > +} > + > +do_install:append:class-target() { > + oe_runmake -C ${B}/man

[OE-core] [PATCH] npth: remove redundant FILES

This is covered by the default rules. Signed-off-by: Ross Burton --- meta/recipes-support/npth/npth_1.7.bb | 2 -- 1 file changed, 2 deletions(-) diff --git a/meta/recipes-support/npth/npth_1.7.bb b/meta/recipes-support/npth/npth_1.7.bb index 861f0b92c80..0cf35565a36 100644 --- a/meta/recipes

[OE-core] [PATCH 4/5] npth: remove obsolete pkgconfig patch

npth now provides a pkg-config itself[1] so we don't need to patch one in anymore. [1] https://git.gnupg.org/cgi-bin/gitweb.cgi?p=npth.git;a=commit;h=f3cbbc86cb49a1729d6df573bd76a5559e7f636b Signed-off-by: Ross Burton --- .../recipes-support/npth/npth/pkgconfig.patch | 59

[OE-core] [PATCH 5/5] npth: remove obsolete binconfig-disabled inherit

npth no longer installs the npth-config script[1] that we disabled, so there's no need to inherit the class. [1] https://git.gnupg.org/cgi-bin/gitweb.cgi?p=npth.git;a=commit;h=2b6a8e5369ed37d64d1bafd3a546ec52e25acd26 Signed-off-by: Ross Burton --- meta/recipes-support/npth/npth_1.7.bb | 2

[OE-core] [PATCH 3/5] npth: update homepage

Set the homepage to the actual homepage, not that of the older GNU Pth. Also fix some whitespace in the license fields. Signed-off-by: Ross Burton --- meta/recipes-support/npth/npth_1.7.bb | 7 +++ 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/meta/recipes-support/npth

[OE-core] [PATCH 2/5] ruby: remove obsolete build dependences

/ruby/ruby/commit/172077232e7f40b4dfedf6a5434c252b3d326fab Signed-off-by: Ross Burton --- meta/recipes-devtools/ruby/ruby_3.3.0.bb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/meta/recipes-devtools/ruby/ruby_3.3.0.bb b/meta/recipes-devtools/ruby/ruby_3.3.0.bb index 17eb6d73c2b

[OE-core] [PATCH 1/5] waffle: remove dependency on udev

Waffle 1.8.0 dropped the use of libudev and instead uses libdrm[1] in the gbm code. Update the depends. [1] https://gitlab.freedesktop.org/mesa/waffle/-/commit/6bf58d4bb9044a7c65818d5de5d4ca747766e77f Signed-off-by: Ross Burton --- meta/recipes-graphics/waffle/waffle_1.8.0.bb | 2 +- 1 file

Re: [OE-core] [PATCH 02/19] package_rpm: remove support for DIRFILES

> On 19 Apr 2024, at 17:38, Ross Burton wrote: > > On 19 Apr 2024, at 13:19, Alexander Kanavin via lists.openembedded.org > wrote: >> >> This was added here: >> https://git.yoctoproject.org/poky/commit/?id=2f42ef8d8fb1febf28252b98884cebabc931f720 >>

Re: [OE-core] [PATCH 02/19] package_rpm: remove support for DIRFILES

On 19 Apr 2024, at 13:19, Alexander Kanavin via lists.openembedded.org wrote: > > This was added here: > https://git.yoctoproject.org/poky/commit/?id=2f42ef8d8fb1febf28252b98884cebabc931f720 > > It's undocumented, untested, rpm-specific, and currently > broken, and to begin with looks like a

Re: [OE-core] [PATCH] at-spi2-core: set CVE_PRODUCT

On 17 Apr 2024, at 15:21, Emil Kronborg via lists.openembedded.org wrote: > > Signed-off-by: Emil Kronborg > --- > meta/recipes-support/atk/at-spi2-core_2.52.0.bb | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/meta/recipes-support/atk/at-spi2-core_2.52.0.bb >

Re: [OE-core] [PATCH 1/3] oe-build-perf-report: Add apache echarts to make report interactive

Hi Ninette, First, this is awesome, thanks very much! One thing I noticed is that the ‘rootfs size’ and ’tmpdir size’ tooltips are trying to format the data as a timestamp instead of a size. Cheers, Ross -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply

Re: [OE-core] [PATCH][kirkstone] openssl: fix crash on aarch64 if BTI is enabled but no Crypto instructions

On 29 Mar 2024, at 13:31, Steve Sakoman wrote: > > Sorry for the delayed response, but I found the same issue with the > kirkstone version that Vivek did. Sorry for the very delayed response, I somehow missed this reply in my inbox. The patched code doesn’t exist in kirkstone so this isn’t

Re: [oe-core][PATCH 2/2] packagegroup-core-boot: recommend ifupdown

On 11 Apr 2024, at 15:58, Joe Slater via lists.openembedded.org wrote: > This was accidental since I removed the qemuall/interfaces in order to use > the main one with qemu. I'm sending new patches. Maybe, though, we don't > really need the qemuall stuff. The last time I looked we did need

Re: [oe-core][PATCH 2/2] packagegroup-core-boot: recommend ifupdown

On 11 Apr 2024, at 10:33, Richard Purdie via lists.openembedded.org wrote: >> --- a/meta/recipes-core/init-ifupdown/init-ifupdown-1.0/qemuall/interfaces >> +++ /dev/null >> @@ -1,5 +0,0 @@ >> -# /etc/network/interfaces -- configuration file for ifup(8), ifdown(8) >> - >> -# The loopback

[OE-core] [PATCH] eudev: update Upstream-Status on netifnames.patch

Signed-off-by: Ross Burton --- meta/recipes-core/udev/eudev/netifnames.patch | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/meta/recipes-core/udev/eudev/netifnames.patch b/meta/recipes-core/udev/eudev/netifnames.patch index 8f6e9a1538b..4f8e54d12d4 100644 --- a/meta/recipes

Re: [OE-core][PATCH] systemd: make predictable name mac policy opt-out

On 8 Apr 2024, at 08:04, Peter Marko via lists.openembedded.org wrote: > > From: Peter Marko > > Even the patch says it's inappropriate for upstream, > and it's also inappropriate for some downstream projects, too. > So make it possible to opt-out on it. I’m looking at these patches because

Re: [OE-core][PATCH] systemd: make predictable name mac policy opt-out

On 8 Apr 2024, at 08:04, Peter Marko via lists.openembedded.org wrote: > + ${@bb.utils.contains('PACKAGECONFIG', 'predictable-if-mac', > 'file://0001-NamePolicy.patch', '', d)} \ There’s a few other places which use the pni-names DISTRO_FEATURE (which probably needs to be

Re: [oe-core][PATCH 1/1] init-ifupdown: add predictable interface names

On 10 Jan 2024, at 21:10, Joe Slater via lists.openembedded.org wrote: > -# Wired or wireless interfaces > -auto eth0 > +# Wired or wireless interfaces including predictable names > +auto /eth0 Obviously we’re days away from release but I think we need to revert or rewrite this. The /name

[OE-core] [PATCH] curl: fix quoting when disabling flaky tests

From: Ross Burton The list of test labels to disable shouldn't be quoted, and this meant that tests were running when they should not. [ YOCTO #15268 ] Signed-off-by: Ross Burton --- meta/recipes-support/curl/curl/run-ptest | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git

[OE-core] [PATCH] glib-2.0: skip a timing sensitive ptest

From: Ross Burton The /timeout/rounding test is sensitive to system load, as it expects timeouts to trigger in windows that on an idle system are realistic but not when running inside a qemu-system on a loaded system. [ YOCTO #14464 ] Signed-off-by: Ross Burton --- .../glib-2.0/glib-2.0/skip

[yocto] New mailing list for layer patches

Hi, At the moment if a layer doesn’t have enough volume to justify a dedicated mailing list for patches (see, for example, meta-...@lists.yoctoproject.org ) then the convention is that the patches can be sent to yo...@lists.yoctoproject.org

Re: [OE-core] [PATCH v2] curl: improve reproducibility

On 27 Mar 2024, at 17:14, Oleh Matiusha via lists.openembedded.org wrote: > +do_install:append:class-target() { > + fix_absolute_paths > +} > + > +do_install:append:class-nativesdk() { > + fix_absolute_paths > +} Would it be best to just always do this sed? A native build won’t have a

[OE-core] [PATCH][nanbield] openssl: fix crash on aarch64 if BTI is enabled but no Crypto instructions

From: Ross Burton On aarch64, if the processor doesn't have the Crypto instructions then OpenSSL will fall back onto the "bit-sliced" assembler routines. When branch protection (BTI) was enabled in OpenSSL these routines were missed, so if BTI is available libssl will immediately

[OE-core] [PATCH][kirkstone] openssl: fix crash on aarch64 if BTI is enabled but no Crypto instructions

From: Ross Burton On aarch64, if the processor doesn't have the Crypto instructions then OpenSSL will fall back onto the "bit-sliced" assembler routines. When branch protection (BTI) was enabled in OpenSSL these routines were missed, so if BTI is available libssl will immediately

[OE-core] [PATCH] openssl: fix crash on aarch64 if BTI is enabled but no Crypto instructions

From: Ross Burton On aarch64, if the processor doesn't have the Crypto instructions then OpenSSL will fall back onto the "bit-sliced" assembler routines. When branch protection (BTI) was enabled in OpenSSL these routines were missed, so if BTI is available libssl will immediately

[OE-core] [PATCH] classes/qemuboot: add depends on qemu-system-native and qemu-helper-native

From: Ross Burton Any image that inherits qemuboot must also add image dependencies on qemu-system-native and qemu-helper-native, otherwise the image won't be able to be booted. Currently this is done by conf/machine/include/qemu.inc, but not every machine that uses qemuboot includes that file

[OE-core] [PATCH 1/2] tcl: improve run-ptest

From: Ross Burton By setting ERROR_ON_FAILURES we don't need to grep the output to know if the tests failed. By default the log runner will print the failed test case, so we don't need to store the log at all. Set the skipped tests across multiple lines so that it's easier to see what skips

[OE-core] [PATCH 2/2] tcl: skip I/O channel 46.1

From: Ross Burton This test, which is in both the io and chanio test suites, has short timeouts which can trigger on loaded systems. [ YOCTO #15407 #15421 ] Signed-off-by: Ross Burton --- meta/recipes-devtools/tcltk/tcl/run-ptest | 2 ++ 1 file changed, 2 insertions(+) diff --git a/meta

Re: [OE-core] [PATCH v2] pypi.bbclass: remove vendor from CVE_PRODUCT

On 20 Mar 2024, at 16:08, Emil Kronborg via lists.openembedded.org wrote: > > By specifying the CVE vendor as python, some CVEs are not found. For > instance, the CVE_PRODUCT for python3-pyopenssl becomes > python:pyopenssl, which yields no matches in the NIST NVD database > because the correct

Re: [OE-core] [PATCH v2] file: add CVE_PRODUCT

On 20 Mar 2024, at 16:08, Emil Kronborg via lists.openembedded.org wrote: > > Having only file as the CVE product is too generic. What we actually > want is file from file_project to match the correct CVE(s). There’s also file:file, for example https://nvd.nist.gov/vuln/detail/CVE-2007-2799.

Re: [OE-core] [PATCH v2] python3-pytest: add CVE_PRODUCT

On 20 Mar 2024, at 16:09, Emil Kronborg via lists.openembedded.org wrote: > > For some reason, the CVE product is just called py and not pytest in the > NIST NVD database. Since the database only accept keywords with at least > 3 characters, the CVE vendor must also be specified. I can only

Re: [oe-core][RFC][PATCHv2] glib-2.0: update 2.78.4 -> 2.80.0

On 20 Mar 2024, at 11:25, f_...@t-online.de wrote: > > On Wed, Mar 20 2024 at 11:15:35 AM +00:00:00, Ross Burton > wrote: >> Are you testing with gobject-introspection? > > > yes, my test was done with gobject-introspection and I ended up with > > | Run-time de

Re: [oe-core][RFC][PATCHv2] glib-2.0: update 2.78.4 -> 2.80.0

Are you testing with gobject-introspection? The API docs don't use gtk-doc either anymore. I’ve a WIP branch at poky-contrib:ross/glib280, this upgrade is less than trivial. (the branch desperately needs rebasing and squashing somewhat, and G-I is still utterly broken) Ross

Re: [OE-core] [master] [PATCH] cve-check: Add provision to exclude classes

On 15 Mar 2024, at 19:52, Dhairya Nagodra -X (dnagodra - E-INFO CHIPS INC at Cisco) wrote: >> What’s the use-case for this? Note that you can control whether cve-check >> runs per-layer already, if that’s useful. > > Currently, the CVE report is generated for all packages associated with the

Re: [OE-core] Call for contributions to the 5.0 release notes

On 15 Mar 2024, at 08:12, Michael Opdenacker via lists.openembedded.org wrote: > > Greetings, > > If you contributed new features to the upcoming release, or if you are > interested in figuring out what's new in 5.0, that would be great if you > could help us by submitting bits of release

Re: [OE-core] [PATCH] openssh: enable sshd.service by default

On 7 Mar 2024, at 20:08, Emil Kronborg via lists.openembedded.org wrote: > > Socket activation is prone to DoS (denial of service) because too many > connections will permanently deactivate sshd.socket [1]. Also, since > socket units do not allow setting Restart, accepting new connections can >

Re: [OE-core] [master] [PATCH] cve-check: Add provision to exclude classes

On 3 Mar 2024, at 17:53, Dhairya Nagodra via lists.openembedded.org wrote: > > From: Dhairya Nagodra > > - There are times when exluding a package that inherits a particular > class/classes may be desired. > - This provides the framework for that via the variable: >

[OE-core] [PATCH 3/3] classes/pypi: don't expose PYPI_ARCHIVE_NAME

From: Ross Burton This variable is only used when constructing a SRC_URI and some recipes think that it's the correct value to assign if the PyPi package name isn't the same as the recipe name, when PYPI_PACKAGE is actually all that needs to be set. Also document the variables we expect

[OE-core] [PATCH 2/3] python-*: don't set PYPI_ARCHIVE_NAME and S when PYPI_PACKAGE is sufficient

From: Ross Burton There's no need to explicitly set PYPI_ARCHIVE_NAME and S when PYPI_PACKAGE is set correctly. Signed-off-by: Ross Burton --- .../python/python3-jsonschema-specifications_2023.12.1.bb | 6 ++ meta/recipes-devtools/python/python3-poetry-core_1.9.0.bb | 4 ++-- meta

[OE-core] [PATCH 1/3] python3_pip517: just count wheels in the directory, not subdirectories

From: Ross Burton The install task uses a recursive find to check that it can only find one wheel, but then does a non-recursive glob to install. This can lead to false-failures if PEP517_WHEEL_PATH points at a directory with subdirectories. Solve this mismatch by making the find non-recusive

[OE-core] [PATCH 3/3] gstreamer1.0: improve test reliability

From: Ross Burton First, libcheck has the ability to increase all test timeouts by an arbitrary multiplier. Because we run our tests on loaded build machines, increase all timeouts by 10x to reduce the chance of load causing failures. Second, use GST_CHECKS_IGNORE to list test cases that should

[OE-core] [PATCH 1/3] curl: improve run-ptest

From: Ross Burton There's no need to run the output of runtests.pl through a sed to get automake-style output, as you can pass -am to get this formatting. Don't run timing dependent tests, as the ptests can run on loaded systems. Add a dependency on the en_US locale because some of the tests

[OE-core] [PATCH 2/3] curl: increase test timeouts

From: Ross Burton We often see multiple curl tests fail during ptest runs, the actual test varies but the output is like this: FAIL: 337: protoc! There was no content at all in the file log/3/server.input. Server glitch? Total curl failure? Returned: 28 Error code 28

Re: [OE-core] [PATCH] insane: add patch-status to default ERROR_QA for all layers

On 14 Mar 2024, at 07:09, Martin Jansa via lists.openembedded.org wrote: > > * it's enabled for patches in oe-core for very long time and I was using > it for many other layers as well, so most layers should be in good > shape > > * it's also possible to disable it for individual layer as

Re: [OE-core] [PATCH 4/4] linux-firmware: remove pointless -license packages

I retract this final patch of the series: there’s a few licenses which specify that the text needs to be distributed alongside the firmware and not just in “documentation or similar”. The rest of the patches are good, however. Ross > On 7 Mar 2024, at 12:05, Ross Burton

[OE-core] [PATCH 4/4] linux-firmware: remove pointless -license packages

From: Ross Burton The linux-firmware recipe goes to a lot of effort to attempt to package the license texts into separate packages, and add dependencies so that installing a piece of firmware will also ship the relevant license text. However, licence.bbclass can already do this, because

[OE-core] [PATCH 3/4] linux-firmware: remove pointless linux-firmware-gplv2-license package

From: Ross Burton The GPLv2 text is the standard text, so the -carl9170 package can just set LICENSE=GPL-2.0-or-later and the custom license handling removed. Confirmed in the source that the intended license is v2-or-later, not v2-only as WHENCE says. Signed-off-by: Ross Burton

[OE-core] [PATCH 2/4] linux-firmware: set LICENSE field for -liquidui and -mellanox

From: Ross Burton Signed-off-by: Ross Burton --- meta/recipes-kernel/linux-firmware/linux-firmware_20231211.bb | 2 ++ 1 file changed, 2 insertions(+) diff --git a/meta/recipes-kernel/linux-firmware/linux-firmware_20231211.bb b/meta/recipes-kernel/linux-firmware/linux-firmware_20231211.bb

[OE-core] [PATCH 1/4] linux-firmware: add support for deduplicating the firmware

From: Ross Burton This can a non-trivial amount of disk space, but requires rdfind from meta-oe. As duplicate pieces of firmware become links, this reduces the size of individual packages and adds dependencies: linux-firmware-adsp-sst: PKGSIZE changed from 6362678 to 560 (-12%) linux

[OE-core] [PATCH] rxvt: add rxvt to desktop entry name

From: Ross Burton If an image ends up with both matchbox-terminal and rxvt installed then there are two Terminal applications that only differ in the description. Rename the rxvt desktop entry to specifically identify itself as rxvt. Signed-off-by: Ross Burton --- meta/recipes-sato/rxvt

Re: [OE-core] [PATCH 1/2] kernel-module-dirs.bbclass: Add class

On 4 Mar 2024, at 15:57, Peter Kjellerstedt via lists.openembedded.org wrote: > +modulesloaddir ??= "${@bb.utils.contains('DISTRO_FEATURES', 'systemd', > '${nonarch_libdir}', '${sysconfdir}', d)}/modules-load.d" > +modprobedir ??= "${@bb.utils.contains('DISTRO_FEATURES', 'systemd', >

Re: [OE-core] [yocto-security] OE-core CVE metrics for master on Sun 03 Mar 2024 01:00:01 AM HST

On 3 Mar 2024, at 13:18, Peter Marko via lists.yoctoproject.org wrote: > > I already mentioned this last week. > https://lists.openembedded.org/g/openembedded-core/message/196199 > > I think that partial NVD DB update is not working properly as things which > were corrected by NVD are still

[OE-core] [PATCH] linux-firmware: split out more firmware pieces

From: Ross Burton Split _most_ of the rtl_nic drivers to a subpackage. Notably this does not include rtl8168 as this is already split into a separate package. These drivers are fairly common so this lets people install the essential firmware easier. Also split out some large firmware

[OE-core] [PATCH] linux-yocto: update yocto-kernel-cache to fix CONFIG_DEBUG_CREDENTIALS warnings

From: Ross Burton Incorporate the fix for upstream removing CONFIG_DEBUG_CREDENTIALS in the stable branch. Signed-off-by: Ross Burton --- meta/recipes-kernel/linux/linux-yocto_6.6.bb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/meta/recipes-kernel/linux/linux-yocto_6.6

[OE-core] [PATCH] gstreamer1.0: skip a test that is known to be flaky

From: Ross Burton The aggregator testcase test_infinite_seek_50_src_live is known upstream to be flaky[1] and when this fails in their CI they just ignore it. It's failing often on our autobuilder, so disable the test case for now until upstream have resolved this issue. [ YOCTO #15054 ] [1

[OE-core] [PATCH 1/2] libexif: remove unused version_underscore

From: Ross Burton This function hasn't been used since the upgrade to 0.6.23 in 2021[1] [1] oe-core 65a54f72e15f59bdf9d8c4618f3ef8510541d134 Signed-off-by: Ross Burton --- meta/recipes-support/libexif/libexif_0.6.24.bb | 3 --- 1 file changed, 3 deletions(-) diff --git a/meta/recipes

[OE-core] [PATCH 2/2] libexif: improve ptest packaging

From: Ross Burton The NLS tests don't actually test anything useful and were not being ran anyway, so don't install them. Install the test binaries using libtool so that we install the actual binaries instead of libtool wrapper scripts. This means that if we also tell the Makefile to use /bin

[OE-core] [PATCH] core-image-initramfs-boot: install all kernel modules

From: Ross Burton The point of a "just boot" initramfs is to contain the kernel modules so that the kernel can find the real root filesystem, so explicitly add them. Signed-off-by: Ross Burton --- meta/recipes-core/images/core-image-initramfs-boot.bb | 2 +- 1 file changed, 1 inser

[OE-core] [RFC PATCH] Add genericarm64 MACHINE using upstream defconfig

From: Ross Burton This is a new 64-bit "generic" Arm machine, that expects the hardware to be SystemReady IR compatible. This is slightly forward-leaning as there's not a _lot_ of SystemReady hardware in the wild, but most modern boards are and the number will only grow. Also, this i

Re: [OE-core] [PATCH] sanity.bbclass: raise_sanity_error if /tmp is noexec

On 21 Feb 2024, at 07:18, ChenQi wrote: > I just noticed the change. I can't find the V2 in my mailbox, so I'm going to > reply here. > I'm a little concerned about forcing such requirement here. It does not seem > *necessary*. > As far as I know, the whole oe-core does not need /tmp to be

Re: [OE-core] [PATCHv2 5/5] lib/oe/patch: Use git notes to store the filenames for the patches

On 19 Feb 2024, at 15:03, Peter Kjellerstedt wrote: >> This is all that is needed for the new code to work correctly with an old >> worktree, right? > > Well, it relies on the fallback (i.e., the exception clause) in > GitApplyTree.getNotes() > that reads from the commit message in case a Git

Re: [OE-core] [PATCHv2 5/5] lib/oe/patch: Use git notes to store the filenames for the patches

On 19 Feb 2024, at 01:28, Peter Kjellerstedt via lists.openembedded.org wrote: > +@staticmethod > +def addNote(repo, ref, key, value=None): > +note = key + (": %s" % value if value else "") > +notes_ref = GitApplyTree.notes_ref > +runcmd(["git", "config",

Re: [OE-core] [PATCHv2 3/5] devtool: Make use of oe.patch.GitApplyTree.commitIgnored()

On 19 Feb 2024, at 01:28, Peter Kjellerstedt via lists.openembedded.org wrote: > > This makes use of the oe.patch.GitApplyTree.commitIgnored() function to > create commits that shall be ignored by `devtool finish`. If we’re using notes to store the filename info, couldn’t the ignores state

[OE-core] [PATCH] meson: set the sysroot in the cross files

From: Ross Burton When cross compiling we have a sysroot, so by telling Meson where it is explicitly it can do the right thing. Currently this means it just sets PKG_CONFIG_SYSROOT_DIR to the value that we already set it to, but in the future this can be used to stop Meson looking directly

[OE-core] [PATCH] libffi: upgrade to 3.4.5

From: Ross Burton Drop the clang/VFP patch, merged in 00116b6. Drop the FILES for ${libdir}/libffi-${PV}, this file isn't installed. License-Update: Copyright year updated to 2024. Signed-off-by: Ross Burton --- ...m-sysv-reverted-clang-VFP-mitigation.patch | 102

Re: [OE-core] [PATCH v4 3/3] cups: use LOCALE_PATHS to split localized HTML templates

On 6 Dec 2023, at 16:06, Jonathan GUILLOT via lists.openembedded.org wrote: > +LOCALE_PATHS += "${datadir}/cups/templates” This doesn’t quite do what you expect. Listing the packages generated by cups I see: cups-locale-da: /usr/share/cups/templates/da/add-class.tmpl

Re: [OE-core] [PATCH] qemuarm64: maximise the emulated CPU capabilities

On 15 Feb 2024, at 06:40, Mikko Rapeli wrote: > optee xtests are triggering a reproducible RCU stall in our testing with > --cpu max,pauth-impdef=on and qemu-system 8.2.1 from Debian unstable. > This does not happen with --cpu cortex-a57 on this qemu-machine version. > [sigh] Can you file a

Re: [oe-core][PATCH] mesa: update 23.3.5 -> 24.0.0

Mesa 24.0.0 is - as clearly indicated on the web site - a development snapshot. 24.0.1 is now out, can you upgrade to that? Ross > On 14 Feb 2024, at 18:14, Markus Volk via lists.openembedded.org > wrote: > > - refresh 0001-meson.build-check-for-all-linux-host_os-combinations.patch > -

[yocto] Future package versions

Hi, A somewhat recurring theme about this time in the release cycle is people asking what version of a recipe will be in the next release, especially more now because the next release is a LTS. So to avoid the relevant people repeating themselves, I’ve created a wiki page:

Re: [OE-core] [PATCH v3 2/3] lib/oe/package: add LOCALE_EXTRA_PATHS to add extra locations for locales

I like the idea of this patch (hardcoding behaviour is bad), but I think it could be generalised some more. > On 6 Dec 2023, at 16:03, Jonathan GUILLOT via lists.openembedded.org > wrote: > -localedir = os.path.join(dvar + datadir, 'locale') > +localedirs = [os.path.join(dvar +

[OE-core] [PATCH] sanity: require bitbake 2.7.2 for the inherit_defer statement

From: Ross Burton Signed-off-by: Ross Burton --- meta/conf/sanity.conf | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/meta/conf/sanity.conf b/meta/conf/sanity.conf index 435455d9fb4..c3e6c5179c6 100644 --- a/meta/conf/sanity.conf +++ b/meta/conf/sanity.conf @@ -3,7 +3,7

Re: [OE-core] [PATCH 4/4] grub2: upgrade 2.06 -> 2.12

On 23 Jan 2024, at 08:47, Anuj Mittal via lists.openembedded.org wrote: > > Drop patches that have been upstreamed. Refresh others. I was nervous but I’m pleased to say: Tested-by: Ross Burton mailto:ross.bur...@arm.com>> For the upgrade on the fvp-base aarch64 machine. Thanks

Re: [OE-core] [PATCH] pulseaudio: Update to 17.0

On 23 Jan 2024, at 14:08, Fabio Estevam via lists.openembedded.org wrote: > +FILES:${PN}-server = "${bindir}/pulseaudio ${bindir}/start-* ${datadir} > ${sysconfdir} ${bindir}/pactl */udev/rules.d/*.rules */*/udev/rules.d/*.rules > ${systemd_user_unitdir}/*” It’s best not to add entire trees

[OE-core] [PATCH] libssh2: backport fix for CVE-2023-48795

From: Ross Burton Backport the upstream fix for CVE-2023-48795. Signed-off-by: Ross Burton --- .../libssh2/libssh2/CVE-2023-48795.patch | 466 ++ .../recipes-support/libssh2/libssh2_1.11.0.bb | 1 + 2 files changed, 467 insertions(+) create mode 100644 meta/recipes

Re: [OE-core] [yocto-security] OE-core CVE metrics for master on Sun 21 Jan 2024 01:00:01 AM HST

And the rest of the CVEs: On 21 Jan 2024, at 11:18, Steve Sakoman via lists.yoctoproject.org wrote: > CVE-2023-25584 (CVSS3: 7.1 HIGH): > binutils:binutils-cross-testsuite:binutils-cross-x86_64:binutils-native > https://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2023-25584 * Part of 2.40,

Re: [OE-core] [PATCH 3/5] zlib: ignore CVE-2023-6992

On 22 Jan 2024, at 14:16, Marko, Peter wrote: > > Hi Ross, > > I think this one is better - > https://lists.openembedded.org/g/openembedded-core/message/193603 > I'm not sure why it was not picked up yet after 9 days, but It's CPE which is > not matching, not our configuration options… Ah I

[OE-core] [PATCH 1/5] cve_check: handle CVE_STATUS being set to the empty string

From: Ross Burton Handle CVE_STATUS[...] being set to an empty string just as if it was not set at all. This is needed for evaluated CVE_STATUS values to work, i.e. when setting not-applicable-config if a PACKAGECONFIG is disabled. Signed-off-by: Ross Burton --- meta/lib/oe/cve_check.py | 4

  1   2   3   4   5   6   7   8   9   10   >