Skip to contentSkip to conversation

systemd-systemctl-native: correctly set PV

PATCH In review Alexander Kanavin 4 participants 8 messages Mar 11, 2025 – Mar 20, 2025

lore
Share & tools
lore
#0 lore raw

PATCH 1/4 systemd-systemctl-native: correctly set PV

From: Alexander Kanavin <[email protected]>

---
 ...temd-systemctl-native.bb => systemd-systemctl-native_257.3.bb} | 0
 1 file changed, 0 insertions(+), 0 deletions(-)
 rename meta/recipes-core/systemd/{systemd-systemctl-native.bb => systemd-systemctl-native_257.3.bb} (100%)
meta/recipes-core/systemd/systemd-systemctl-native_257.3.bb renamed from meta/recipes-core/systemd/systemd-systemctl-native.bb +0 -0
similarity index 100%
rename from meta/recipes-core/systemd/systemd-systemctl-native.bb
rename to meta/recipes-core/systemd/systemd-systemctl-native_257.3.bb
Signature · 1 line
2.39.5
#1 lore raw

PATCH 2/4 python3-roman-numbers-py: correct upstream version check

From: Alexander Kanavin <[email protected]>

---
 meta/recipes-devtools/python/python3-roman-numerals-py_3.1.0.bb | 1 +
 1 file changed, 1 insertion(+)
meta/recipes-devtools/python/python3-roman-numerals-py_3.1.0.bb +1 -0
index 2f22f15d698..9a76df0de26 100644
@@ -5,6 +5,7 @@ LIC_FILES_CHKSUM = "file://LICENCE.rst;md5=bfcc8b16e42929aafeb9d414360bc2fd"
 
 SRC_URI[sha256sum] = "be4bf804f083a4ce001b5eb7e3c0862479d10f94c936f6c4e5f250aa5ff5bd2d"
 PYPI_PACKAGE = "roman_numerals_py"
+UPSTREAM_CHECK_PYPI_PACKAGE = "roman_numerals_py"
 
 inherit pypi python_flit_core
 
Signature · 1 line
2.39.5
#2 lore raw

PATCH 3/4 gnupg: follow odd-even scheme in upstream version checks

In reply to Alexander Kanavin

From: Alexander Kanavin <[email protected]>

Unfortunately this has been overlooked, and resulted in erroneous updates to testing/development releases.

The check will report an 'unknown' latest version until 2.6.0 is released.

---
 meta/recipes-support/gnupg/gnupg_2.5.4.bb | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
meta/recipes-support/gnupg/gnupg_2.5.4.bb +2 -2
index e6a7e3fb407..0001631d2c5 100644
@@ -11,9 +11,9 @@ LIC_FILES_CHKSUM = "file://COPYING;md5=189af8afca6d6075ba6c9e0aa8077626 \
 
 DEPENDS = "npth libassuan libksba zlib bzip2 readline libgcrypt"
 
-inherit autotools gettext texinfo pkgconfig
+inherit autotools gettext texinfo pkgconfig upstream-version-is-even
 
-UPSTREAM_CHECK_URI = "https://gnupg.org/download/index.html"
+UPSTREAM_CHECK_URI = "https://gnupg.org/ftp/gcrypt/gnupg/"
 SRC_URI = "${GNUPG_MIRROR}/${BPN}/${BPN}-${PV}.tar.bz2 \
            file://0002-use-pkgconfig-instead-of-npth-config.patch \
            file://0004-autogen.sh-fix-find-version-for-beta-checking.patch \
Signature · 1 line
2.39.5
#3 lore raw

Re: [OE-core] [PATCH 3/4] gnupg: follow odd-even scheme in upstream version checks

Quoted 31 lines · Alexander Kanavin via lists.openembedded.org

From: Alexander Kanavin <[email protected]>

Unfortunately this has been overlooked, and resulted in erroneous updates to testing/development releases.

The check will report an 'unknown' latest version until 2.6.0 is released.

---
 meta/recipes-support/gnupg/gnupg_2.5.4.bb | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
meta/recipes-support/gnupg/gnupg_2.5.4.bb +2 -2
index e6a7e3fb407..0001631d2c5 100644
@@ -11,9 +11,9 @@ LIC_FILES_CHKSUM =
 "file://COPYING;md5=189af8afca6d6075ba6c9e0aa8077626 \
 
 DEPENDS = "npth libassuan libksba zlib bzip2 readline libgcrypt"
 
-inherit autotools gettext texinfo pkgconfig
+inherit autotools gettext texinfo pkgconfig upstream-version-is-even
 
-UPSTREAM_CHECK_URI = "https://gnupg.org/download/index.html"
+UPSTREAM_CHECK_URI = "https://gnupg.org/ftp/gcrypt/gnupg/"
 SRC_URI = "${GNUPG_MIRROR}/${BPN}/${BPN}-${PV}.tar.bz2 \
            file://0002-use-pkgconfig-instead-of-npth-config.patch \
Quoted 3 lines

file://0004-autogen.sh-fix-find-version-for-beta-checking.patch \

Could you maybe add a comment next to the version hash saying that the project is using odd/even
convention and that people should only update to stable branches? At least it might avoid the error
next time.

Kind regards,
Marta

#4 lore raw

Re: [OE-core] [PATCH 3/4] gnupg: follow odd-even scheme in upstream version checks

Quoted 3 lines · Marta Rybczynska

Could you maybe add a comment next to the version hash saying that the project is using odd/even convention and that people should only update to stable branches? At least it might avoid the error next time.

This change is not only for version checks, it also affects version upgrade attempts with devtool or AUH, that from now on would only consider even versions.

We have many other recipes with this restriction, and generally people don't attempt to manually update to something else.

Alex

#5 lore raw

PATCH 4/4 gcr: exclude x.y.9z pre-releases from upstream version check

In reply to Alexander Kanavin

From: Alexander Kanavin <[email protected]>

---
 meta/recipes-gnome/gcr/gcr_4.3.0.bb | 1 +
 1 file changed, 1 insertion(+)
meta/recipes-gnome/gcr/gcr_4.3.0.bb +1 -0
index dcd8658d9cb..3fc8ec0a643 100644
@@ -15,6 +15,7 @@ CFLAGS += "-D_GNU_SOURCE"
 
 GTKDOC_MESON_OPTION = "gtk_doc"
 inherit gnomebase gtk-icon-cache gi-docgen features_check vala gobject-introspection gettext mime mime-xdg
+UPSTREAM_CHECK_REGEX = "gcr-(?P<pver>\d+\.\d+\.(?!9\d+)\d+)"
 
 REQUIRED_DISTRO_FEATURES = "${@bb.utils.contains('DISTRO_FEATURES', 'x11', 'opengl', '', d)}"
 
Signature · 1 line
2.39.5
#6 lore raw
In reply to Alexander Kanavin

I don't know who else has run into this, but having

    DISTRO_FEATURES_NATIVE="systemd"

produces circular dependency between
systemd-systemctl-native and util-linux-native. FWIW, I do not know why I had it, but I did since 2020. With that setting removed, the build can start.

The build warns about other obsolete settings,
can this be added to the set of warnings somehow?

2025. 03. 11. 11:04 keltezéssel, Alexander Kanavin via lists.openembedded.org írta:

Quoted 21 lines

From: Alexander Kanavin <[email protected]>

---
  ...temd-systemctl-native.bb => systemd-systemctl-native_257.3.bb} | 0
  1 file changed, 0 insertions(+), 0 deletions(-)

rename meta/recipes-core/systemd/{systemd-systemctl-native.bb => systemd-systemctl-native_257.3.bb} (100%)

meta/recipes-core/systemd/systemd-systemctl-native_257.3.bb renamed from meta/recipes-core/systemd/systemd-systemctl-native.bb +0 -1
similarity index 100%
rename from meta/recipes-core/systemd/systemd-systemctl-native.bb
rename to meta/recipes-core/systemd/systemd-systemctl-native_257.3.bb
 
-=-=-=-=-=-=-=-=-=-=-=-

Links: You receive all messages sent to this group.
View/Reply Online (#212550): https://lists.openembedded.org/g/openembedded-core/message/212550
Mute This Topic: https://lists.openembedded.org/mt/111637143/3617728
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

#7 lore raw

Any reason you want to build native systemd?

Slava

Quoted 23 lines · Zoltan Boszormenyi via lists.openembedded.org

I don't know who else has run into this, but having

    DISTRO_FEATURES_NATIVE="systemd"

produces circular dependency between
systemd-systemctl-native and util-linux-native. FWIW, I do not know why I had it, but I did since 2020. With that setting removed, the build can start.

The build warns about other obsolete settings,
can this be added to the set of warnings somehow?

2025. 03. 11. 11:04 keltezéssel, Alexander Kanavin via lists.openembedded.org írta:

Quoted 12 lines

From: Alexander Kanavin <[email protected]>

---
  ...temd-systemctl-native.bb => systemd-systemctl-native_257.3.bb} | 0

  1 file changed, 0 insertions(+), 0 deletions(-)

  rename meta/recipes-core/systemd/{systemd-systemctl-native.bb => systemd-systemctl-native_257.3.bb} (100%)

meta/recipes-core/systemd/systemd-systemctl-native_257.3.bb renamed from meta/recipes-core/systemd/systemd-systemctl-native.bb +0 -0
similarity index 100%
rename from meta/recipes-core/systemd/systemd-systemctl-native.bb
rename to meta/recipes-core/systemd/systemd-systemctl-native_257.3.bb