PR#48: April CPU update - rpms/java-latest-openjdk

rpms / java-latest-openjdk

#48 April CPU update

Merged 2 years ago by jerboaa. Opened 2 years ago by pmikova.

Unknown source rawhide into rawhide

Petra Mikova • 2 years ago

ed8c45a

.gitignore

file modified

		`@@ -14,3 +14,4 @@`
		`/tapsets-icedtea-3.15.0.tar.xz`
		`/jdk-updates-jdk15u-jdk-15.0.2+7.tar.xz`
		`/openjdk-jdk16-jdk-16+36.tar.xz`
		`+ /openjdk-jdk16u-jdk-16.0.1+9.tar.xz`

java-latest-openjdk.spec

file modified

+5 -8

		`@@ -274,7 +274,7 @@`
		`# New Version-String scheme-style defines`
		`%global featurever 16`
		`%global interimver 0`
		`- %global updatever 0`
		`+ %global updatever 1`
		`%global patchver 0`
		`# If you bump featurever, you must bump also vendor_version_string`
		`# Used via new version scheme. JDK 16 was`
		`@@ -297,8 +297,8 @@`
		`%global origin_nice OpenJDK`
		`%global top_level_dir_name %{origin}`
		`%global top_level_dir_name_backup %{top_level_dir_name}-backup`
		`- %global buildver 36`
		`- %global rpmrelease 2`
		`+ %global buildver 9`
		`+ %global rpmrelease 1`
		`# Priority must be 8 digits in total; up to openjdk 1.8, we were using 18..... so when we moved to 11, we had to add another digit`
		`%if %is_system_jdk`
		`# Using 10 digits may overflow the int used for priority, so we combine the patch and build versions`
		`@@ -1115,8 +1115,8 @@`

		`# to regenerate source0 (jdk) run update_package.sh`
		`# update_package.sh contains hard-coded repos, revisions, tags, and projects to regenerate the source archives`
		`- #Source0: openjdk-jdk%{featurever}u-jdk-%{filever}+%{buildver}%{?tagsuffix:-%{tagsuffix}}.tar.xz`
		`- Source0: openjdk-jdk%{featurever}-jdk-%{filever}+%{buildver}.tar.xz`
		`+ Source0: openjdk-jdk%{featurever}u-jdk-%{filever}+%{buildver}%{?tagsuffix:-%{tagsuffix}}.tar.xz`
		`+ #Source0: openjdk-jdk%{featurever}-jdk-%{filever}+%{buildver}.tar.xz`

		`# Use 'icedtea_sync.sh' to update the following`
		`# They are based on code contained in the IcedTea project (3.x).`
		`@@ -1165,8 +1165,6 @@`
		`Patch4: pr3183-rh1340845-support_fedora_rhel_system_crypto_policy.patch`
		`# Depend on pcs-lite-libs instead of pcs-lite-devel as this is only in optional repo`
		`Patch6: rh1684077-openjdk_should_depend_on_pcsc-lite-libs_instead_of_pcsc-lite-devel.patch`
		`- # JDK-8259949: Use i686 instead of i586 on x86 when -fcf-protection is passed to the compiler, as CMOV is needed`
		`- Patch8: jdk8259949-allow_cf-protection_on_x86.patch`

		`#############################################`
		`#`
		`@@ -1523,7 +1521,6 @@`
		`%patch3 -p1`
		`%patch4 -p1`
		`%patch6 -p1`
		`- %patch8 -p1`
		`popd # openjdk`

		`%patch1000`

jdk8259949-allow_cf-protection_on_x86.patch

file removed

-27

		`@@ -1,27 +0,0 @@`
		`- diff --git openjdk.orig/make/autoconf/flags-cflags.m4 openjdk/make/autoconf/flags-cflags.m4`
		`- --- openjdk.orig/make/autoconf/flags-cflags.m4`
		`- +++ openjdk/make/autoconf/flags-cflags.m4`
		`- @@ -715,9 +715,21 @@`
		`- # CFLAGS PER CPU`
		`- if test "x$TOOLCHAIN_TYPE" = xgcc \|\| test "x$TOOLCHAIN_TYPE" = xclang; then`
		`- # COMMON to gcc and clang`
		`- + AC_MSG_CHECKING([if $1 is x86])`
		`- if test "x$FLAGS_CPU" = xx86; then`
		`- - # Force compatibility with i586 on 32 bit intel platforms.`
		`- - $1_CFLAGS_CPU="-march=i586"`
		`- + AC_MSG_RESULT([yes])`
		`- + AC_MSG_CHECKING([if control flow protection is enabled by additional compiler flags])`
		`- + if echo "${EXTRA_CFLAGS}${EXTRA_CXXFLAGS}${EXTRA_ASFLAGS}" \| ${GREP} -q 'fcf-protection' ; then`
		`- + # cf-protection requires CMOV and thus i686`
		`- + $1_CFLAGS_CPU="-march=i686"`
		`- + AC_MSG_RESULT([yes, forcing ${$1_CFLAGS_CPU}])`
		`- + else`
		`- + # Force compatibility with i586 on 32 bit intel platforms.`
		`- + $1_CFLAGS_CPU="-march=i586"`
		`- + AC_MSG_RESULT([no, forcing ${$1_CFLAGS_CPU}])`
		`- + fi`
		`- + else`
		`- + AC_MSG_RESULT([no])`
		`- fi`
		`- fi`
		`-`

sources

file modified

+1 -1

		`@@ -1,2 +1,2 @@`
		`SHA512 (tapsets-icedtea-3.15.0.tar.xz) = c752a197cb3d812d50c35e11e4722772be40096c81d2a57933e0d9b8a3c708b9c157b8108a4e33a06ca7bb81648170994408c75d6f69d5ff12785d0c31009671`
		`- SHA512 (openjdk-jdk16-jdk-16+36.tar.xz) = 8fc618b960eb6d982d408dd5607dea61f230895b2e2b233793308048ae004868595f501e56a1ab79c3d6aa5b638e06f58383c34d0a09fb321f48678a3a89f5f4`
		`+ SHA512 (openjdk-jdk16u-jdk-16.0.1+9.tar.xz) = ff06d5c97377cac4d5e8921766c61d4d96dc0c968913263b5371f0d3d0e98bb6a5ccf5b1cbb3ddaff3380bf4499ff6501e73f96e0e922b2294d1f7c1ec3eee23`

pmikova commented 2 years ago

Generated new sources and removed obsoleted patch.

Please verify i bumped the correct version number for the security update, as I wasn't sure which should i use.

Scratch build in progress: https://koji.fedoraproject.org/koji/taskinfo?taskID=66408602

Thanks!

zuul commented 2 years ago

Build failed. More information on how to proceed and troubleshoot errors available at https://fedoraproject.org/wiki/Zuul-based-ci

check-for-arches : SUCCESS in 1m 41s
rpm-scratch-build : SUCCESS in 1h 16m 31s
rpm-scratch-build-s390x : SUCCESS in 1h 21m 47s (non-voting)
rpm-scratch-build-ppc64le : SUCCESS in 3h 23m 09s (non-voting)
rpm-scratch-build-i686 : SUCCESS in 1h 00m 09s (non-voting)
rpm-scratch-build-armv7hl : SUCCESS in 1h 20m 02s (non-voting)
rpm-scratch-build-aarch64 : SUCCESS in 2h 45m 50s (non-voting)
rpm-linter : FAILURE in 9m 27s
rpm-rpminspect : POST_FAILURE in 19m 11s (non-voting)
check-for-tests : SUCCESS in 41s
rpm-install-test : SUCCESS in 4m 22s
rpm-test : SKIPPED

jvanek commented 2 years ago

The 16.0.1+9 is afaik proper tag to use.
The removed patch was clearly already in upstream sources?

pmikova commented 2 years ago

My issue was, if out of:

%global interimver 0
%global updatever 0
%global patchver 0

was updatever the one to use, so the final build: java-latest-openjdk-16.0.1.0.9-1.rolling.fc35 is versioned properly. Because it was recently renamed and is slightly confusing which 0 is the security update version.

The removed patch was https://bugs.openjdk.java.net/browse/JDK-8259949 which should be present in 16.0.1, according to the https://bugs.openjdk.java.net/browse/JDK-8263047. The link to the patch commit is in the bug, so afaik it is the source.

Thanks!

jvanek commented 2 years ago

Scratch failed onl on aarch64. but AFTER hte build finsihed. It seesm liek infrastructure glitch:
https://koji.fedoraproject.org/koji/getfile?taskID=66418033&volume=DEFAULT&name=build.log&offset=-4000

jvanek commented 2 years ago

The manual scratch is ok

Edited 2 years ago by jvanek

ahughes commented 2 years ago

Looks good to me (though I usually start rpmrelease at 0, personal choice)
Yes, I pushed the patch to the 16u tree ages ago. Would have preferred it made GA but apparently wasn't high enough priority.
updatever is the right one to bump. The naming is the same as used by JEP 322.

Edited 2 years ago by ahughes