Re: [OS-BUILD PATCHv3 0/5] Add libperf packages and build kernel tools for Fedora

2024-02-02 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2907#note_1755476971

Sure I approved it. I think we want to resolve this thread as well. I'll leave
that to any of you to resolve it.
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4] redhat: do not compress the full kernel changelog in the src.rpm

2023-12-21 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: do not compress the full kernel changelog in the src.rpm

Bugzilla: INTERNAL
Upstream Status: RHEL only

By default redhat/scripts/expand_srpm.sh adds a .gitignore entry excluding any
*.xz file in dist-git, since those are usually tarballs which can get very 
large.

The problem is that kernel.changelog.xz file recently added gets excluded and
can't be commited to dist-git due that.

To avoid it, since src.rpm and git can compress contents and the updates
are incremental, just use the uncompressed contents when adding it to
the src.rpm, and only compress the file when providing it to the
kernel-doc package.

This fixes commit "redhat: ship all the changelog from source git into 
kernel-doc"
recently merged into kernel-ark.

Reported-by: Justin Forbes 
Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -700,7 +700,7 @@ sources-rh: $(TARBALL) generate-testpatch-tmp setup-source 
dist-configs-check
dracut-virt.conf \
$(SOURCES)/
@cat $$(ls -1 $(SPECPACKAGE_NAME).changelog-* | sort -t '.' -k 3 -n -r) 
\
-   | xz > $(SOURCES)/kernel.changelog.xz
+   > $(SOURCES)/kernel.changelog
@if [ "$(RELEASED_KERNEL)" -ne 0 ]; then \
cp keys/redhatsecureboot{301,501,ca5,ca1}.cer $(SOURCES)/; \
cp keys/secureboot_{ppc,s390}.cer $(SOURCES)/; \
diff --git a/redhat/kernel.spec.template b/redhat/kernel.spec.template
index blahblah..blahblah 100644
--- a/redhat/kernel.spec.template
+++ b/redhat/kernel.spec.template
@@ -799,7 +799,7 @@ BuildRequires: tpm2-tools
 Source0: linux-%{tarfile_release}.tar.xz
 
 Source1: Makefile.rhelver
-Source2: kernel.changelog.xz
+Source2: kernel.changelog
 
 # Name of the packaged file containing signing key
 %ifarch ppc64le
@@ -2918,7 +2918,8 @@ 
docdir=$RPM_BUILD_ROOT%{_datadir}/doc/kernel-doc-%{specversion}-%{pkgrelease}
 # copy the source over
 mkdir -p $docdir
 tar -h -f - --exclude=man --exclude='.*' -c Documentation | tar xf - -C $docdir
-install -m644 %{SOURCE2} $docdir/
+cat %{SOURCE2} | xz > $docdir/kernel.changelog.xz
+chmod 0644 $docdir/kernel.changelog.xz
 
 # with_doc
 %endif

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2841
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3] redhat: do not compress the full kernel changelog in the src.rpm

2023-12-21 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: do not compress the full kernel changelog in the src.rpm

Bugzilla: INTERNAL
Upstream Status: RHEL only

By default redhat/scripts/expand_srpm.sh adds a .gitignore entry excluding any
*.xz file in dist-git, since those are usually tarballs which can get very 
large.

The problem is that kernel.changelog.xz file recently added gets excluded and
can't be commited to dist-git due that.

To avoid it, since src.rpm and git can compress contents and the updates
are incremental, just use the uncompressed contents when adding it to
the src.rpm, and only compress the file when providing it to the
kernel-doc package.

This fixes commit "redhat: ship all the changelog from source git into 
kernel-doc"
recently merged into kernel-ark.

Reported-by: Justin Forbes 
Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -700,7 +700,7 @@ sources-rh: $(TARBALL) generate-testpatch-tmp setup-source 
dist-configs-check
dracut-virt.conf \
$(SOURCES)/
@cat $$(ls -1 $(SPECPACKAGE_NAME).changelog-* | sort -t '.' -k 3 -n -r) 
\
-   | xz > $(SOURCES)/kernel.changelog.xz
+   > $(SOURCES)/kernel.changelog
@if [ "$(RELEASED_KERNEL)" -ne 0 ]; then \
cp keys/redhatsecureboot{301,501,ca5,ca1}.cer $(SOURCES)/; \
cp keys/secureboot_{ppc,s390}.cer $(SOURCES)/; \
diff --git a/redhat/kernel.spec.template b/redhat/kernel.spec.template
index blahblah..blahblah 100644
--- a/redhat/kernel.spec.template
+++ b/redhat/kernel.spec.template
@@ -799,7 +799,7 @@ BuildRequires: tpm2-tools
 Source0: linux-%{tarfile_release}.tar.xz
 
 Source1: Makefile.rhelver
-Source2: kernel.changelog.xz
+Source2: kernel.changelog
 
 # Name of the packaged file containing signing key
 %ifarch ppc64le
@@ -2918,7 +2918,8 @@ 
docdir=$RPM_BUILD_ROOT%{_datadir}/doc/kernel-doc-%{specversion}-%{pkgrelease}
 # copy the source over
 mkdir -p $docdir
 tar -h -f - --exclude=man --exclude='.*' -c Documentation | tar xf - -C $docdir
-install -m644 %{SOURCE2} $docdir/
+cat %{SOURCE2} | xz > $docdir/%{SOURCE2}.xz
+chmod 0644 $docdir/%{SOURCE2}.xz
 
 # with_doc
 %endif

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2841
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2] redhat: fix dist-git/upload for the kernel.changelog.xz file

2023-12-21 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2841#note_1704941100

Yes I can change it that way, let me try.
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2] redhat: fix dist-git/upload for the kernel.changelog.xz file

2023-12-21 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix dist-git/upload for the kernel.changelog.xz file

Bugzilla: INTERNAL
Upstream Status: RHEL only

By default redhat/scripts/expand_srpm.sh adds a .gitignore entry
excluding any *.xz file in dist-git, since those are usually tarballs
which can get very large.

The problem is that kernel.changelog.xz file recently added gets
excluded and can't be commited to dist-git due that. However, since
this changelog file can get large, lets treat it as it was a tarball
as well and upload it instead of directly commiting to dist-git.

This fixes commit "redhat: ship all the changelog from source git into
kernel-doc" recently merged into kernel-ark.

Reported-by: Justin Forbes 
Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -328,6 +328,8 @@ else
   RHPKG_BIN:=rhpkg
 endif
 
+CLOGFILE:=$(SOURCES)/kernel.changelog.xz
+
 TARFILE:=linux-$(SPECTARFILE_RELEASE).tar.xz
 TARBALL:=$(REDHAT)/$(TARFILE)
 
@@ -700,7 +702,7 @@ sources-rh: $(TARBALL) generate-testpatch-tmp setup-source 
dist-configs-check
dracut-virt.conf \
$(SOURCES)/
@cat $$(ls -1 $(SPECPACKAGE_NAME).changelog-* | sort -t '.' -k 3 -n -r) 
\
-   | xz > $(SOURCES)/kernel.changelog.xz
+   | xz > $(CLOGFILE)
@if [ "$(RELEASED_KERNEL)" -ne 0 ]; then \
cp keys/redhatsecureboot{301,501,ca5,ca1}.cer $(SOURCES)/; \
cp keys/secureboot_{ppc,s390}.cer $(SOURCES)/; \
diff --git a/redhat/scripts/rh-dist-git.sh b/redhat/scripts/rh-dist-git.sh
index blahblah..blahblah 100755
--- a/redhat/scripts/rh-dist-git.sh
+++ b/redhat/scripts/rh-dist-git.sh
@@ -43,7 +43,7 @@ echo "Unpacking from SRPM"
 sed -i "/linux-.*.tar.xz/d" "$tmpdir/$SPECPACKAGE_NAME"/{sources,.gitignore};
 sed -i "/kernel-abi-stablelists.*.tar.xz/d" 
"$tmpdir/$SPECPACKAGE_NAME"/{sources,.gitignore};
 sed -i "/kernel-kabi-dw-.*.tar.xz/d" 
"$tmpdir/$SPECPACKAGE_NAME"/{sources,.gitignore};
-upload_list="$TARBALL $KABI_TARBALL $KABIDW_TARBALL"
+upload_list="$CLOGFILE $TARBALL $KABI_TARBALL $KABIDW_TARBALL"
 
 echo "Uploading new tarballs: $upload_list"
 # We depend on word splitting here:
diff --git a/redhat/self-test/data/centos-2585cf9dfaad.el7 
b/redhat/self-test/data/centos-2585cf9dfaad.el7
index blahblah..blahblah 100644
--- a/redhat/self-test/data/centos-2585cf9dfaad.el7
+++ b/redhat/self-test/data/centos-2585cf9dfaad.el7
@@ -12,6 +12,7 @@ BUILD_TARGET=c9s-candidate
 BUILD_TARGET_USERDEF=
 BUMP_RELEASE=yes
 CHANGELOG_EXT=changelog-9.99
+CLOGFILE=../redhat/rpm/SOURCES/kernel.changelog.xz
 CROSS_DISABLED_SUBPACKAGES=--without bpftool --without perf --without tools
 CROSS_PACKAGE_LIST=cross-binutils-common cross-gcc-common diffstat 
glibc-static ncurses-devel numactl-devel rng-tools binutils-aarch64-linux-gnu 
gcc-aarch64-linux-gnu binutils-ppc64le-linux-gnu gcc-ppc64le-linux-gnu 
binutils-s390x-linux-gnu gcc-s390x-linux-gnu
 CROSS_RPMFLAGS=rpmbuild --define "_sourcedir ../redhat/rpm/SOURCES" --define 
"_builddir ../redhat/rpm/BUILD" --define "_srcrpmdir ../redhat/rpm/SRPMS" 
--define "_rpmdir ../redhat/rpm/RPMS" --define "_specdir ../redhat/rpm/SPECS" 
--define "dist .el7"
diff --git a/redhat/self-test/data/centos-2585cf9dfaad.fc25 
b/redhat/self-test/data/centos-2585cf9dfaad.fc25
index blahblah..blahblah 100644
--- a/redhat/self-test/data/centos-2585cf9dfaad.fc25
+++ b/redhat/self-test/data/centos-2585cf9dfaad.fc25
@@ -12,6 +12,7 @@ BUILD_TARGET=c9s-candidate
 BUILD_TARGET_USERDEF=
 BUMP_RELEASE=yes
 CHANGELOG_EXT=changelog-9.99
+CLOGFILE=../redhat/rpm/SOURCES/kernel.changelog.xz
 CROSS_DISABLED_SUBPACKAGES=--without bpftool --without perf --without tools
 CROSS_PACKAGE_LIST=cross-binutils-common cross-gcc-common diffstat 
glibc-static ncurses-devel numactl-devel rng-tools binutils-aarch64-linux-gnu 
gcc-aarch64-linux-gnu binutils-ppc64le-linux-gnu gcc-ppc64le-linux-gnu 
binutils-s390x-linux-gnu gcc-s390x-linux-gnu
 CROSS_RPMFLAGS=rpmbuild --define "_sourcedir ../redhat/rpm/SOURCES" --define 
"_builddir ../redhat/rpm/BUILD" --define "_srcrpmdir ../redhat/rpm/SRPMS" 
--define "_rpmdir ../redhat/rpm/RPMS" --define "_specdir ../redhat/rpm/SPECS" 
--define "dist .fc25"
diff --git a/redhat/self-test/data/centos-78e36f3b0dae.el7 
b/redhat/self-test/data/centos-78e36f3b0dae.el7
index blahblah..blahblah 100644
--- a/redhat/self-test/data/centos-78e36f3b0dae.el7
+++ b/redhat/self-test/data/centos-78e36f3b0dae.el7
@@ -12,6 +12,7 @@ BUILD_TARGET=c9s-candidate
 BUILD_TARGET_USERDEF=
 BUMP_RELEASE=yes
 CHANGELOG_EXT=changelog-9.99
+CLOGFILE=../redhat/rpm/SOURCES/kernel.changelog.xz
 CROSS_DISABLED_SUBPACKAGES=--without bpftool --without perf --without tools
 CROSS_PACKAGE_LIST=cross-binutils-common cross-gcc-common diffstat 
glibc-static ncurses-devel numactl-devel rng-tools binutils-aarch64-linux-gnu 
gcc-aarch64-linux-gnu binutils-ppc64le-linux-gnu gcc-ppc64le-linux-gnu 

[OS-BUILD PATCH] redhat: fix dist-git/upload for the kernel.changelog.xz file

2023-12-21 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix dist-git/upload for the kernel.changelog.xz file

Bugzilla: INTERNAL
Upstream Status: RHEL only

By default redhat/scripts/expand_srpm.sh adds a .gitignore entry
excluding any *.xz file in dist-git, since those are usually tarballs
which can get very large.

The problem is that kernel.changelog.xz file recently added gets
excluded and can't be commited to dist-git due that. However, since
this changelog file can get large, lets treat it as it was a tarball
as well and upload it instead of directly commiting to dist-git.

This fixes commit "redhat: ship all the changelog from source git into
kernel-doc" recently merged into kernel-ark.

Reported-by: Justin Forbes 
Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -328,6 +328,8 @@ else
   RHPKG_BIN:=rhpkg
 endif
 
+CLOGFILE:=$(SOURCES)/kernel.changelog.xz
+
 TARFILE:=linux-$(SPECTARFILE_RELEASE).tar.xz
 TARBALL:=$(REDHAT)/$(TARFILE)
 
@@ -700,7 +702,7 @@ sources-rh: $(TARBALL) generate-testpatch-tmp setup-source 
dist-configs-check
dracut-virt.conf \
$(SOURCES)/
@cat $$(ls -1 $(SPECPACKAGE_NAME).changelog-* | sort -t '.' -k 3 -n -r) 
\
-   | xz > $(SOURCES)/kernel.changelog.xz
+   | xz > $(CLOGFILE)
@if [ "$(RELEASED_KERNEL)" -ne 0 ]; then \
cp keys/redhatsecureboot{301,501,ca5,ca1}.cer $(SOURCES)/; \
cp keys/secureboot_{ppc,s390}.cer $(SOURCES)/; \
diff --git a/redhat/scripts/rh-dist-git.sh b/redhat/scripts/rh-dist-git.sh
index blahblah..blahblah 100755
--- a/redhat/scripts/rh-dist-git.sh
+++ b/redhat/scripts/rh-dist-git.sh
@@ -43,7 +43,7 @@ echo "Unpacking from SRPM"
 sed -i "/linux-.*.tar.xz/d" "$tmpdir/$SPECPACKAGE_NAME"/{sources,.gitignore};
 sed -i "/kernel-abi-stablelists.*.tar.xz/d" 
"$tmpdir/$SPECPACKAGE_NAME"/{sources,.gitignore};
 sed -i "/kernel-kabi-dw-.*.tar.xz/d" 
"$tmpdir/$SPECPACKAGE_NAME"/{sources,.gitignore};
-upload_list="$TARBALL $KABI_TARBALL $KABIDW_TARBALL"
+upload_list="$CLOGFILE $TARBALL $KABI_TARBALL $KABIDW_TARBALL"
 
 echo "Uploading new tarballs: $upload_list"
 # We depend on word splitting here:

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2841
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv3 0/2] redhat: create an empty changelog file when changing its name

2023-12-21 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2838#note_1704887686

Actually I have another fix instead of changing expand_srpm.sh, can you check
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2841 ?
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv3 0/2] redhat: create an empty changelog file when changing its name

2023-12-21 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2838#note_1704863969

Ok I see, I'll fix expand_srpm.sh.
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv3 0/2] redhat: create an empty changelog file when changing its name

2023-12-21 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2838#note_1704808386

If a large kernel.changelog file in dist-git is not a problem, I can compress
it only when adding the file to kernel-doc package, is that ok? Otherwise we
need to change the dist-git .gitignore file.
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3 1/2] redhat: create an empty changelog file when changing its name

2023-12-20 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: create an empty changelog file when changing its name

Bugzilla: INTERNAL
Upstream Status: RHEL only

Up until now, if we rename the spec package name or bump the rhel major
or minor versions, we used to copy the previous changelog file if it
exists into the new one. However, this makes the changelog grow
indefinitely since a stream start, and it starts to get very big. We
do not need to keep the entire history at each new changelog file,
because eg. in a stream life, we always keep the older changelog files
in the git tree.

Thus, instead of copying the old changelog, just create a new empty
file. This also makes sure we trim the rpm changelog at each new minor
release, since it's used as the source of the rpm package changelog
content. And that's another intention with this change: to avoid having
the rpm changelog contents to grow too big, as reported by Brian Masney.

There was also a case where if we are starting a stream or changing the
package name, and there is no old changelog file available, we copied
a kernel.changelog* if it existed. But looks like this case we should
also do the same and trim it to start fresh, the old changelog file can
be kept if wanted in the tree for reference and we don't necessarily
need to copy all entries from the old changelog.

With this there is some simplification as well and we can remove the
CHANGELOG_PREV_EXT variable.

Reported-by: Brian Masney 
Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -72,7 +72,6 @@ SOURCES:=$(RPM)/SOURCES
 TESTPATCH:=$(REDHAT)/linux-kernel-test.patch
 CHANGELOG_EXT:=changelog-$(RHEL_MAJOR).$(RHEL_MINOR)
 SPECCHANGELOG:=$(SPECPACKAGE_NAME).$(CHANGELOG_EXT)
-CHANGELOG_PREV_EXT:=changelog-$(RHEL_MAJOR).$(shell expr $(RHEL_MINOR) - 1)
 ARCH_LIST=aarch64 ppc64le s390x x86_64
 
 # save some user selectable values to know later if we can override them
@@ -644,13 +643,8 @@ _setup-source: dist-git-version-check
fi
@cp $(REDHAT)/$(SPECFILE).template $(SOURCES)/$(SPECFILE)
@if [ ! -e  $(REDHAT)/$(SPECCHANGELOG) ]; then \
-   if [ -e $(SPECPACKAGE_NAME).$(CHANGELOG_PREV_EXT) ]; then \
-   echo "Creating $(SPECCHANGELOG) as copy of 
$(CHANGELOG_PREV_EXT)"; \
-   cp $(REDHAT)/$(SPECPACKAGE_NAME).$(CHANGELOG_PREV_EXT) 
$(REDHAT)/$(SPECCHANGELOG); \
-   else \
-   echo "Creating $(SPECCHANGELOG) as copy of 
$(REDHAT)/kernel.$(CHANGELOG_EXT)"; \
-   cp $(REDHAT)/kernel.$(CHANGELOG_EXT) $(SPECCHANGELOG); \
-   fi; \
+   echo "Creating an empty $(SPECCHANGELOG) since it does not 
exist yet"; \
+   touch $(REDHAT)/$(SPECCHANGELOG); \
fi
@if [ -z "$(RHSELFTESTDATA)" ]; then \
cp $(REDHAT)/$(SPECCHANGELOG) $(SOURCES)/$(SPECCHANGELOG); \
diff --git a/redhat/self-test/data/centos-2585cf9dfaad.el7 
b/redhat/self-test/data/centos-2585cf9dfaad.el7
index blahblah..blahblah 100644
--- a/redhat/self-test/data/centos-2585cf9dfaad.el7
+++ b/redhat/self-test/data/centos-2585cf9dfaad.el7
@@ -12,7 +12,6 @@ BUILD_TARGET=c9s-candidate
 BUILD_TARGET_USERDEF=
 BUMP_RELEASE=yes
 CHANGELOG_EXT=changelog-9.99
-CHANGELOG_PREV_EXT=changelog-9.98
 CROSS_DISABLED_SUBPACKAGES=--without bpftool --without perf --without tools
 CROSS_PACKAGE_LIST=cross-binutils-common cross-gcc-common diffstat 
glibc-static ncurses-devel numactl-devel rng-tools binutils-aarch64-linux-gnu 
gcc-aarch64-linux-gnu binutils-ppc64le-linux-gnu gcc-ppc64le-linux-gnu 
binutils-s390x-linux-gnu gcc-s390x-linux-gnu
 CROSS_RPMFLAGS=rpmbuild --define "_sourcedir ../redhat/rpm/SOURCES" --define 
"_builddir ../redhat/rpm/BUILD" --define "_srcrpmdir ../redhat/rpm/SRPMS" 
--define "_rpmdir ../redhat/rpm/RPMS" --define "_specdir ../redhat/rpm/SPECS" 
--define "dist .el7"
diff --git a/redhat/self-test/data/centos-2585cf9dfaad.fc25 
b/redhat/self-test/data/centos-2585cf9dfaad.fc25
index blahblah..blahblah 100644
--- a/redhat/self-test/data/centos-2585cf9dfaad.fc25
+++ b/redhat/self-test/data/centos-2585cf9dfaad.fc25
@@ -12,7 +12,6 @@ BUILD_TARGET=c9s-candidate
 BUILD_TARGET_USERDEF=
 BUMP_RELEASE=yes
 CHANGELOG_EXT=changelog-9.99
-CHANGELOG_PREV_EXT=changelog-9.98
 CROSS_DISABLED_SUBPACKAGES=--without bpftool --without perf --without tools
 CROSS_PACKAGE_LIST=cross-binutils-common cross-gcc-common diffstat 
glibc-static ncurses-devel numactl-devel rng-tools binutils-aarch64-linux-gnu 
gcc-aarch64-linux-gnu binutils-ppc64le-linux-gnu gcc-ppc64le-linux-gnu 
binutils-s390x-linux-gnu gcc-s390x-linux-gnu
 CROSS_RPMFLAGS=rpmbuild --define "_sourcedir ../redhat/rpm/SOURCES" --define 
"_builddir ../redhat/rpm/BUILD" --define "_srcrpmdir ../redhat/rpm/SRPMS" 
--define "_rpmdir ../redhat/rpm/RPMS" --define "_specdir ../redhat/rpm/SPECS" 
--define "dist .fc25"
diff --git 

[OS-BUILD PATCHv3 2/2] redhat: ship all the changelog from source git into kernel-doc

2023-12-20 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: ship all the changelog from source git into kernel-doc

Bugzilla: INTERNAL
Upstream Status: RHEL only

In the previous change, the rpm changelog will be trimmed due size
for each rpm binary on each minor release bump. However, someone may
still want to see the full changelog since the start of a given stream,
thus provide it in the src.rpm and through kernel-doc.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -699,6 +699,8 @@ sources-rh: $(TARBALL) generate-testpatch-tmp setup-source 
dist-configs-check
kernel-local \
dracut-virt.conf \
$(SOURCES)/
+   @cat $$(ls -1 $(SPECPACKAGE_NAME).changelog-* | sort -t '.' -k 3 -n -r) 
\
+   | xz > $(SOURCES)/kernel.changelog.xz
@if [ "$(RELEASED_KERNEL)" -ne 0 ]; then \
cp keys/redhatsecureboot{301,501,ca5,ca1}.cer $(SOURCES)/; \
cp keys/secureboot_{ppc,s390}.cer $(SOURCES)/; \
diff --git a/redhat/kernel.spec.template b/redhat/kernel.spec.template
index blahblah..blahblah 100644
--- a/redhat/kernel.spec.template
+++ b/redhat/kernel.spec.template
@@ -799,7 +799,7 @@ BuildRequires: tpm2-tools
 Source0: linux-%{tarfile_release}.tar.xz
 
 Source1: Makefile.rhelver
-
+Source2: kernel.changelog.xz
 
 # Name of the packaged file containing signing key
 %ifarch ppc64le
@@ -2918,6 +2918,7 @@ 
docdir=$RPM_BUILD_ROOT%{_datadir}/doc/kernel-doc-%{specversion}-%{pkgrelease}
 # copy the source over
 mkdir -p $docdir
 tar -h -f - --exclude=man --exclude='.*' -c Documentation | tar xf - -C $docdir
+install -m644 %{SOURCE2} $docdir/
 
 # with_doc
 %endif
@@ -3455,6 +3456,7 @@ fi\
 %{_datadir}/doc/kernel-doc-%{specversion}-%{pkgrelease}/Documentation/*
 %dir %{_datadir}/doc/kernel-doc-%{specversion}-%{pkgrelease}/Documentation
 %dir %{_datadir}/doc/kernel-doc-%{specversion}-%{pkgrelease}
+%{_datadir}/doc/kernel-doc-%{specversion}-%{pkgrelease}/kernel.changelog.xz
 %endif
 
 %if %{with_perf}

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2838
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3 0/2] redhat: create an empty changelog file when changing its name

2023-12-20 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2838

Bugzilla: INTERNAL
Upstream Status: RHEL only

Up until now, if we rename the spec package name or bump the rhel major
or minor versions, we used to copy the previous changelog file if it
exists into the new one. However, this makes the changelog grow
indefinitely since a stream start, and it starts to get very big. We
do not need to keep the entire history at each new changelog file,
because eg. in a stream life, we always keep the older changelog files
in the git tree.

Thus, instead of copying the old changelog, just create a new empty
file. This also makes sure we trim the rpm changelog at each new minor
release, since it's used as the source of the rpm package changelog
content. And that's another intention with this change: to avoid having
the rpm changelog contents to grow too big, as reported by Brian Masney.

There was also a case where if we are starting a stream or changing the
package name, and there is no old changelog file available, we copied
a kernel.changelog* if it existed. But looks like this case we should
also do the same and trim it to start fresh, the old changelog file can
be kept if wanted in the tree for reference and we don't necessarily
need to copy all entries from the old changelog.

With this there is some simplification as well and we can remove the
CHANGELOG_PREV_EXT variable.

Reported-by: Brian Masney 
Signed-off-by: Herton R. Krzesinski 

---
 redhat/self-test/data/centos-2585cf9dfaad.el7  |   1 -
 redhat/self-test/data/centos-2585cf9dfaad.fc25 |   1 -
 redhat/self-test/data/centos-78e36f3b0dae.el7  |   1 -
 redhat/self-test/data/centos-78e36f3b0dae.fc25 |   1 -
 redhat/self-test/data/centos-df0cc57e057f.el7  |   1 -
 redhat/self-test/data/centos-df0cc57e057f.fc25 |   1 -
 redhat/self-test/data/centos-fce15c45d3fb.el7  |   1 -
 redhat/self-test/data/centos-fce15c45d3fb.fc25 |   1 -
 redhat/self-test/data/fedora-2585cf9dfaad.el7  |   1 -
 redhat/self-test/data/fedora-2585cf9dfaad.fc25 |   1 -
 redhat/self-test/data/fedora-78e36f3b0dae.el7  |   1 -
 redhat/self-test/data/fedora-78e36f3b0dae.fc25 |   1 -
 redhat/self-test/data/fedora-df0cc57e057f.el7  |   1 -
 redhat/self-test/data/fedora-df0cc57e057f.fc25 |   1 -
 redhat/self-test/data/fedora-fce15c45d3fb.el7  |   1 -
 redhat/self-test/data/fedora-fce15c45d3fb.fc25 |   1 -
 redhat/self-test/data/rhel-2585cf9dfaad.el7|   1 -
 redhat/self-test/data/rhel-2585cf9dfaad.fc25   |   1 -
 redhat/self-test/data/rhel-78e36f3b0dae.el7|   1 -
 redhat/self-test/data/rhel-78e36f3b0dae.fc25   |   1 -
 redhat/self-test/data/rhel-df0cc57e057f.el7|   1 -
 redhat/self-test/data/rhel-df0cc57e057f.fc25   |   1 -
 redhat/self-test/data/rhel-fce15c45d3fb.el7|   1 -
 redhat/self-test/data/rhel-fce15c45d3fb.fc25   |   1 -
 redhat/Makefile|  12 
 redhat/kernel.spec.template|   4 +++-
 26 files changed, 7 insertions(+), 33 deletions(-)
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2 1/2] redhat: create an empty changelog file when changing its name

2023-12-19 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: create an empty changelog file when changing its name

Bugzilla: INTERNAL
Upstream Status: RHEL only

Up until now, if we rename the spec package name or bump the rhel major
or minor versions, we used to copy the previous changelog file if it
exists into the new one. However, this makes the changelog grow
indefinitely since a stream start, and it starts to get very big. We
do not need to keep the entire history at each new changelog file,
because eg. in a stream life, we always keep the older changelog files
in the git tree.

Thus, instead of copying the old changelog, just create a new empty
file. This also makes sure we trim the rpm changelog at each new minor
release, since it's used as the source of the rpm package changelog
content. And that's another intention with this change: to avoid having
the rpm changelog contents to grow too big, as reported by Brian Masney.

There was also a case where if we are starting a stream or changing the
package name, and there is no old changelog file available, we copied
a kernel.changelog* if it existed. But looks like this case we should
also do the same and trim it to start fresh, the old changelog file can
be kept if wanted in the tree for reference and we don't necessarily
need to copy all entries from the old changelog.

With this there is some simplification as well and we can remove the
CHANGELOG_PREV_EXT variable.

Reported-by: Brian Masney 
Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -72,7 +72,6 @@ SOURCES:=$(RPM)/SOURCES
 TESTPATCH:=$(REDHAT)/linux-kernel-test.patch
 CHANGELOG_EXT:=changelog-$(RHEL_MAJOR).$(RHEL_MINOR)
 SPECCHANGELOG:=$(SPECPACKAGE_NAME).$(CHANGELOG_EXT)
-CHANGELOG_PREV_EXT:=changelog-$(RHEL_MAJOR).$(shell expr $(RHEL_MINOR) - 1)
 ARCH_LIST=aarch64 ppc64le s390x x86_64
 
 # save some user selectable values to know later if we can override them
@@ -644,13 +643,8 @@ _setup-source: dist-git-version-check
fi
@cp $(REDHAT)/$(SPECFILE).template $(SOURCES)/$(SPECFILE)
@if [ ! -e  $(REDHAT)/$(SPECCHANGELOG) ]; then \
-   if [ -e $(SPECPACKAGE_NAME).$(CHANGELOG_PREV_EXT) ]; then \
-   echo "Creating $(SPECCHANGELOG) as copy of 
$(CHANGELOG_PREV_EXT)"; \
-   cp $(REDHAT)/$(SPECPACKAGE_NAME).$(CHANGELOG_PREV_EXT) 
$(REDHAT)/$(SPECCHANGELOG); \
-   else \
-   echo "Creating $(SPECCHANGELOG) as copy of 
$(REDHAT)/kernel.$(CHANGELOG_EXT)"; \
-   cp $(REDHAT)/kernel.$(CHANGELOG_EXT) $(SPECCHANGELOG); \
-   fi; \
+   echo "Creating an empty $(SPECCHANGELOG) since it does not 
exist yet"; \
+   touch $(REDHAT)/$(SPECCHANGELOG); \
fi
@if [ -z "$(RHSELFTESTDATA)" ]; then \
cp $(REDHAT)/$(SPECCHANGELOG) $(SOURCES)/$(SPECCHANGELOG); \
diff --git a/redhat/self-test/data/centos-2585cf9dfaad.el7 
b/redhat/self-test/data/centos-2585cf9dfaad.el7
index blahblah..blahblah 100644
--- a/redhat/self-test/data/centos-2585cf9dfaad.el7
+++ b/redhat/self-test/data/centos-2585cf9dfaad.el7
@@ -12,7 +12,6 @@ BUILD_TARGET=c9s-candidate
 BUILD_TARGET_USERDEF=
 BUMP_RELEASE=yes
 CHANGELOG_EXT=changelog-9.99
-CHANGELOG_PREV_EXT=changelog-9.98
 CROSS_DISABLED_SUBPACKAGES=--without bpftool --without perf --without tools
 CROSS_PACKAGE_LIST=cross-binutils-common cross-gcc-common diffstat 
glibc-static ncurses-devel numactl-devel rng-tools binutils-aarch64-linux-gnu 
gcc-aarch64-linux-gnu binutils-ppc64le-linux-gnu gcc-ppc64le-linux-gnu 
binutils-s390x-linux-gnu gcc-s390x-linux-gnu
 CROSS_RPMFLAGS=rpmbuild --define "_sourcedir ../redhat/rpm/SOURCES" --define 
"_builddir ../redhat/rpm/BUILD" --define "_srcrpmdir ../redhat/rpm/SRPMS" 
--define "_rpmdir ../redhat/rpm/RPMS" --define "_specdir ../redhat/rpm/SPECS" 
--define "dist .el7"
diff --git a/redhat/self-test/data/centos-2585cf9dfaad.fc25 
b/redhat/self-test/data/centos-2585cf9dfaad.fc25
index blahblah..blahblah 100644
--- a/redhat/self-test/data/centos-2585cf9dfaad.fc25
+++ b/redhat/self-test/data/centos-2585cf9dfaad.fc25
@@ -12,7 +12,6 @@ BUILD_TARGET=c9s-candidate
 BUILD_TARGET_USERDEF=
 BUMP_RELEASE=yes
 CHANGELOG_EXT=changelog-9.99
-CHANGELOG_PREV_EXT=changelog-9.98
 CROSS_DISABLED_SUBPACKAGES=--without bpftool --without perf --without tools
 CROSS_PACKAGE_LIST=cross-binutils-common cross-gcc-common diffstat 
glibc-static ncurses-devel numactl-devel rng-tools binutils-aarch64-linux-gnu 
gcc-aarch64-linux-gnu binutils-ppc64le-linux-gnu gcc-ppc64le-linux-gnu 
binutils-s390x-linux-gnu gcc-s390x-linux-gnu
 CROSS_RPMFLAGS=rpmbuild --define "_sourcedir ../redhat/rpm/SOURCES" --define 
"_builddir ../redhat/rpm/BUILD" --define "_srcrpmdir ../redhat/rpm/SRPMS" 
--define "_rpmdir ../redhat/rpm/RPMS" --define "_specdir ../redhat/rpm/SPECS" 
--define "dist .fc25"
diff --git 

[OS-BUILD PATCHv2 0/2] redhat: create an empty changelog file when changing its name

2023-12-19 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2838

Bugzilla: INTERNAL
Upstream Status: RHEL only

Up until now, if we rename the spec package name or bump the rhel major
or minor versions, we used to copy the previous changelog file if it
exists into the new one. However, this makes the changelog grow
indefinitely since a stream start, and it starts to get very big. We
do not need to keep the entire history at each new changelog file,
because eg. in a stream life, we always keep the older changelog files
in the git tree.

Thus, instead of copying the old changelog, just create a new empty
file. This also makes sure we trim the rpm changelog at each new minor
release, since it's used as the source of the rpm package changelog
content. And that's another intention with this change: to avoid having
the rpm changelog contents to grow too big, as reported by Brian Masney.

There was also a case where if we are starting a stream or changing the
package name, and there is no old changelog file available, we copied
a kernel.changelog* if it existed. But looks like this case we should
also do the same and trim it to start fresh, the old changelog file can
be kept if wanted in the tree for reference and we don't necessarily
need to copy all entries from the old changelog.

With this there is some simplification as well and we can remove the
CHANGELOG_PREV_EXT variable.

Reported-by: Brian Masney 
Signed-off-by: Herton R. Krzesinski 

---
 redhat/self-test/data/centos-2585cf9dfaad.el7  |   1 -
 redhat/self-test/data/centos-2585cf9dfaad.fc25 |   1 -
 redhat/self-test/data/centos-78e36f3b0dae.el7  |   1 -
 redhat/self-test/data/centos-78e36f3b0dae.fc25 |   1 -
 redhat/self-test/data/centos-df0cc57e057f.el7  |   1 -
 redhat/self-test/data/centos-df0cc57e057f.fc25 |   1 -
 redhat/self-test/data/centos-fce15c45d3fb.el7  |   1 -
 redhat/self-test/data/centos-fce15c45d3fb.fc25 |   1 -
 redhat/self-test/data/fedora-2585cf9dfaad.el7  |   1 -
 redhat/self-test/data/fedora-2585cf9dfaad.fc25 |   1 -
 redhat/self-test/data/fedora-78e36f3b0dae.el7  |   1 -
 redhat/self-test/data/fedora-78e36f3b0dae.fc25 |   1 -
 redhat/self-test/data/fedora-df0cc57e057f.el7  |   1 -
 redhat/self-test/data/fedora-df0cc57e057f.fc25 |   1 -
 redhat/self-test/data/fedora-fce15c45d3fb.el7  |   1 -
 redhat/self-test/data/fedora-fce15c45d3fb.fc25 |   1 -
 redhat/self-test/data/rhel-2585cf9dfaad.el7|   1 -
 redhat/self-test/data/rhel-2585cf9dfaad.fc25   |   1 -
 redhat/self-test/data/rhel-78e36f3b0dae.el7|   1 -
 redhat/self-test/data/rhel-78e36f3b0dae.fc25   |   1 -
 redhat/self-test/data/rhel-df0cc57e057f.el7|   1 -
 redhat/self-test/data/rhel-df0cc57e057f.fc25   |   1 -
 redhat/self-test/data/rhel-fce15c45d3fb.el7|   1 -
 redhat/self-test/data/rhel-fce15c45d3fb.fc25   |   1 -
 redhat/Makefile|  12 
 redhat/kernel.spec.template|   3 ++-
 26 files changed, 6 insertions(+), 33 deletions(-)
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2 2/2] redhat: ship all the changelog from source git into kernel-doc

2023-12-19 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: ship all the changelog from source git into kernel-doc

Bugzilla: INTERNAL
Upstream Status: RHEL only

In the previous change, the rpm changelog will be trimmed due size
for each rpm binary on each minor release bump. However, someone may
still want to see the full changelog since the start of a given stream,
thus provide it in the src.rpm and through kernel-doc.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -699,6 +699,8 @@ sources-rh: $(TARBALL) generate-testpatch-tmp setup-source 
dist-configs-check
kernel-local \
dracut-virt.conf \
$(SOURCES)/
+   @cat $$(ls -1 $(SPECPACKAGE_NAME).changelog-* | sort -t '.' -k 3 -n -r) 
\
+   | xz > $(SOURCES)/kernel.changelog.xz
@if [ "$(RELEASED_KERNEL)" -ne 0 ]; then \
cp keys/redhatsecureboot{301,501,ca5,ca1}.cer $(SOURCES)/; \
cp keys/secureboot_{ppc,s390}.cer $(SOURCES)/; \
diff --git a/redhat/kernel.spec.template b/redhat/kernel.spec.template
index blahblah..blahblah 100644
--- a/redhat/kernel.spec.template
+++ b/redhat/kernel.spec.template
@@ -799,7 +799,7 @@ BuildRequires: tpm2-tools
 Source0: linux-%{tarfile_release}.tar.xz
 
 Source1: Makefile.rhelver
-
+Source2: kernel.changelog.xz
 
 # Name of the packaged file containing signing key
 %ifarch ppc64le
@@ -2918,6 +2918,7 @@ 
docdir=$RPM_BUILD_ROOT%{_datadir}/doc/kernel-doc-%{specversion}-%{pkgrelease}
 # copy the source over
 mkdir -p $docdir
 tar -h -f - --exclude=man --exclude='.*' -c Documentation | tar xf - -C $docdir
+cp %{SOURCE2} $docdir/
 
 # with_doc
 %endif

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2838
--
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCH] redhat: create an empty changelog file when changing its name

2023-12-19 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: create an empty changelog file when changing its name

Bugzilla: INTERNAL
Upstream Status: RHEL only

Up until now, if we rename the spec package name or bump the rhel major
or minor versions, we used to copy the previous changelog file if it
exists into the new one. However, this makes the changelog grow
indefinitely since a stream start, and it starts to get very big. We
do not need to keep the entire history at each new changelog file,
because eg. in a stream life, we always keep the older changelog files
in the git tree.

Thus, instead of copying the old changelog, just create a new empty
file. This also makes sure we trim the rpm changelog at each new minor
release, since it's used as the source of the rpm package changelog
content. And that's another intention with this change: to avoid having
the rpm changelog contents to grow too big, as reported by Brian Masney.

There was also a case where if we are starting a stream or changing the
package name, and there is no old changelog file available, we copied
a kernel.changelog* if it existed. But looks like this case we should
also do the same and trim it to start fresh, the old changelog file can
be kept if wanted in the tree for reference and we don't necessarily
need to copy all entries from the old changelog.

With this there is some simplification as well and we can remove the
CHANGELOG_PREV_EXT variable.

Reported-by: Brian Masney 
Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -72,7 +72,6 @@ SOURCES:=$(RPM)/SOURCES
 TESTPATCH:=$(REDHAT)/linux-kernel-test.patch
 CHANGELOG_EXT:=changelog-$(RHEL_MAJOR).$(RHEL_MINOR)
 SPECCHANGELOG:=$(SPECPACKAGE_NAME).$(CHANGELOG_EXT)
-CHANGELOG_PREV_EXT:=changelog-$(RHEL_MAJOR).$(shell expr $(RHEL_MINOR) - 1)
 ARCH_LIST=aarch64 ppc64le s390x x86_64
 
 # save some user selectable values to know later if we can override them
@@ -644,13 +643,8 @@ _setup-source: dist-git-version-check
fi
@cp $(REDHAT)/$(SPECFILE).template $(SOURCES)/$(SPECFILE)
@if [ ! -e  $(REDHAT)/$(SPECCHANGELOG) ]; then \
-   if [ -e $(SPECPACKAGE_NAME).$(CHANGELOG_PREV_EXT) ]; then \
-   echo "Creating $(SPECCHANGELOG) as copy of 
$(CHANGELOG_PREV_EXT)"; \
-   cp $(REDHAT)/$(SPECPACKAGE_NAME).$(CHANGELOG_PREV_EXT) 
$(REDHAT)/$(SPECCHANGELOG); \
-   else \
-   echo "Creating $(SPECCHANGELOG) as copy of 
$(REDHAT)/kernel.$(CHANGELOG_EXT)"; \
-   cp $(REDHAT)/kernel.$(CHANGELOG_EXT) $(SPECCHANGELOG); \
-   fi; \
+   echo "Creating an empty $(SPECCHANGELOG) since it does not 
exist yet"; \
+   touch $(REDHAT)/$(SPECCHANGELOG); \
fi
@if [ -z "$(RHSELFTESTDATA)" ]; then \
cp $(REDHAT)/$(SPECCHANGELOG) $(SOURCES)/$(SPECCHANGELOG); \
diff --git a/redhat/self-test/data/centos-2585cf9dfaad.el7 
b/redhat/self-test/data/centos-2585cf9dfaad.el7
index blahblah..blahblah 100644
--- a/redhat/self-test/data/centos-2585cf9dfaad.el7
+++ b/redhat/self-test/data/centos-2585cf9dfaad.el7
@@ -12,7 +12,6 @@ BUILD_TARGET=c9s-candidate
 BUILD_TARGET_USERDEF=
 BUMP_RELEASE=yes
 CHANGELOG_EXT=changelog-9.99
-CHANGELOG_PREV_EXT=changelog-9.98
 CROSS_DISABLED_SUBPACKAGES=--without bpftool --without perf --without tools
 CROSS_PACKAGE_LIST=cross-binutils-common cross-gcc-common diffstat 
glibc-static ncurses-devel numactl-devel rng-tools binutils-aarch64-linux-gnu 
gcc-aarch64-linux-gnu binutils-ppc64le-linux-gnu gcc-ppc64le-linux-gnu 
binutils-s390x-linux-gnu gcc-s390x-linux-gnu
 CROSS_RPMFLAGS=rpmbuild --define "_sourcedir ../redhat/rpm/SOURCES" --define 
"_builddir ../redhat/rpm/BUILD" --define "_srcrpmdir ../redhat/rpm/SRPMS" 
--define "_rpmdir ../redhat/rpm/RPMS" --define "_specdir ../redhat/rpm/SPECS" 
--define "dist .el7"
diff --git a/redhat/self-test/data/centos-2585cf9dfaad.fc25 
b/redhat/self-test/data/centos-2585cf9dfaad.fc25
index blahblah..blahblah 100644
--- a/redhat/self-test/data/centos-2585cf9dfaad.fc25
+++ b/redhat/self-test/data/centos-2585cf9dfaad.fc25
@@ -12,7 +12,6 @@ BUILD_TARGET=c9s-candidate
 BUILD_TARGET_USERDEF=
 BUMP_RELEASE=yes
 CHANGELOG_EXT=changelog-9.99
-CHANGELOG_PREV_EXT=changelog-9.98
 CROSS_DISABLED_SUBPACKAGES=--without bpftool --without perf --without tools
 CROSS_PACKAGE_LIST=cross-binutils-common cross-gcc-common diffstat 
glibc-static ncurses-devel numactl-devel rng-tools binutils-aarch64-linux-gnu 
gcc-aarch64-linux-gnu binutils-ppc64le-linux-gnu gcc-ppc64le-linux-gnu 
binutils-s390x-linux-gnu gcc-s390x-linux-gnu
 CROSS_RPMFLAGS=rpmbuild --define "_sourcedir ../redhat/rpm/SOURCES" --define 
"_builddir ../redhat/rpm/BUILD" --define "_srcrpmdir ../redhat/rpm/SRPMS" 
--define "_rpmdir ../redhat/rpm/RPMS" --define "_specdir ../redhat/rpm/SPECS" 
--define "dist .fc25"
diff --git 

Re: [OS-BUILD PATCHv6 0/3] redhat: Switch KABI file compression to xz

2023-10-30 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2760#note_1626469984

yes, I was reading tar docs, and found it tries to auto-detect the format
automatically, first by signature, if it fails then by name suffix (.xz) of
the file. So ok, should be fine.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv6 0/3] redhat: Switch KABI file compression to xz

2023-10-30 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2760#note_1626455222

Hmm that's strange, unless I missed something, shouldn't this be '-xJvf'?
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] Fix binutils breakage

2023-10-18 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2749#note_1609004744

The way to do it in this case should be:
```TPROGS_CFLAGS="%{?build_hostcflags}" %{make} %{?_smp_mflags} ARCH=$Arch V=1
M=samples/bpf/ VMLINUX_H="${RPM_VMLINUX_H}" || true```

Pass the variable before the make invocation. So make does not override it and
will append. However, that may still Makefile modification if it doesn't use
override or doesn't append with ```+=```.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2] redhat/kernel.spec.template: simplify __modsign_install_post

2023-10-11 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2734#note_1599021686

I think mod-sign.sh will not like it. It does use ```xargs -r``` but may exit
with an error if no signing is done. You may want check and adjust if needed
mod-sign.sh and then I think we can remove the special case for zfcpdump, plus
it can then allow other variants to not have any modules.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH 1/3] redhat: Use kspdx-tool output for the License: field

2023-08-24 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2648#note_1527728838

Well there are still firmware being included in the linux sources, one
example: ```drivers/net/appletalk/cops_ffdrv.h```

But probably cases like this should be handled upstream (moving or removing
them..., or adding a proper SPDX identifier). From the License tag I think we
need to remove the Redistributable... info since it's not a valid identifier.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv5 0/8] redhat: Add realtime build and packaging logic

2023-06-05 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2403#note_1418892403

thanks this looks solved now, thus resolving this thread.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCH 0/3] redhat: add support for Jira issues in changelog

2023-06-01 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2496

This MR syncs genlog.py changes done in centos stream
aiming at Jira support.

Signed-off-by: Herton R. Krzesinski 

---
 redhat/scripts/genspec/genlog.py |  93 +++
 1 files changed, 63 insertions(+), 30 deletions(-)
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCH 3/3] redhat/genlog.py: add support to list/process zstream Jira tickets

2023-06-01 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat/genlog.py: add support to list/process zstream Jira tickets

Bugzilla: INTERNAL
Upstream Status: RHEL only

Even if we don't have the zstream process finalized yet for Jira, our
maintainer tooling already has some support for it and it can look for
Z-JIRA tags already. Also, we already have at least one case where
we have a Z-Stream bug with a Y-Stream Jira issue, and we need to make
sure the resolves line don't list the Y-Stream Jira issue.

This change adds support to list Z-Stream Jira issues in the changelog
while at the same time it also makes sure to avoid adding Y-Stream tickets
when we are doing Z-Stream work to the Resolves line.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/scripts/genspec/genlog.py b/redhat/scripts/genspec/genlog.py
index blahblah..blahblah 100755
--- a/redhat/scripts/genspec/genlog.py
+++ b/redhat/scripts/genspec/genlog.py
@@ -63,11 +63,13 @@ def parse_commit(commit):
 bug_set = set()
 zbug_set = set()
 jira_set = set()
+zjira_set = set()
 for line in lines[1:]:
 # Metadata in git notes has priority over commit log
-# If we found any BZ/ZBZ/CVE in git notes, we ignore commit log
+# If we found any BZ/ZBZ/JIRA/ZJIRA/CVE in git notes,
+# we ignore the commit log
 if line == "^^^NOTES-END^^^":
-if bug_set or zbug_set or cve_set:
+if bug_set or zbug_set or jira_set or zjira_set or cve_set:
 break
 
 # Process Bugzilla and ZStream Bugzilla entries
@@ -79,21 +81,23 @@ def parse_commit(commit):
 
 # Process Jira issues
 jira_set.update(find_ji_in_line(line, 'JIRA'))
+zjira_set.update(find_ji_in_line(line, 'Z-JIRA'))
 
-return (log_entry, cve_set, bug_set, zbug_set, jira_set)
+return (log_entry, cve_set, bug_set, zbug_set, jira_set, zjira_set)
 
 
 if __name__ == "__main__":
 all_bzs = set()
 all_zbzs = set()
 all_jiras = set()
+all_zjiras = set()
 commits = sys.stdin.read().split('\0')
 for c in commits:
 if not c:
 continue
-log_item, cves, bugs, zbugs, jiras = parse_commit(c)
+log_item, cves, bugs, zbugs, jiras, zjiras = parse_commit(c)
 entry = f"{log_item}"
-if bugs or zbugs or jiras:
+if bugs or zbugs or jiras or zjiras:
 entry += " ["
 if zbugs:
 entry += " ".join(sorted(zbugs))
@@ -106,18 +110,32 @@ if __name__ == "__main__":
 entry += " " if bugs or zbugs else ""
 entry += " ".join(sorted(jiras))
 all_jiras.update(jiras)
+if zjiras:
+entry += " " if bugs or zbugs or jiras else ""
+entry += " ".join(sorted(zjiras))
+all_zjiras.update(zjiras)
 entry += "]"
 if cves:
 entry += " {" + " ".join(sorted(cves)) + "}"
 print(entry)
 
-resolved_bzs = all_zbzs if all_zbzs else all_bzs
+# If we are doing Z-Stream work, we are addressing Z-Stream tickets
+# and not Y-Stream tickets, so we must make sure to list on Resolves
+# line only the Z-Stream tickets
+resolved_bzs = set()
+resolved_jiras = set()
+if all_zbzs or all_zjiras:
+resolved_bzs = all_zbzs
+resolved_jiras = all_zjiras
+else:
+resolved_bzs = all_bzs
+resolved_jiras = all_jiras
 print("Resolves: ", end="")
 for i, bzid in enumerate(sorted(resolved_bzs)):
 if i:
 print(", ", end="")
 print(f"rhbz#{bzid}", end="")
-for j, jid in enumerate(sorted(all_jiras)):
+for j, jid in enumerate(sorted(resolved_jiras)):
 if j or resolved_bzs:
print(", ", end="")
 print(f"{jid}", end="")

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2496
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCH 1/3] redhat: add support for Jira issues in changelog

2023-06-01 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: add support for Jira issues in changelog

Bugzilla: INTERNAL
Upstream Status: RHEL only

RHEL will switch from Bugzilla to Jira for tickets. This adds support
for parsing and listing Jira issues in the changelog.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/scripts/genspec/genlog.py b/redhat/scripts/genspec/genlog.py
index blahblah..blahblah 100755
--- a/redhat/scripts/genspec/genlog.py
+++ b/redhat/scripts/genspec/genlog.py
@@ -15,20 +15,26 @@
 import re
 import sys
 
-def find_bz_in_line(line, prefix):
-"""Return bug number from properly formated Bugzilla: line."""
-# BZs must begin with '{prefix}: ' and contain a complete BZ URL or id
-_bugs = set()
+def find_ticket_in_line(line, prefix, tkt_re, tkt_groups):
+"""Return ticket referenced in the given line"""
+_tkts = set()
 if not line.startswith(f"{prefix}: "):
-return _bugs
+return _tkts
+for match in tkt_re.finditer(line[len(f"{prefix}:"):]):
+for group in tkt_groups:
+if match.group(group):
+tid = match.group(group).strip()
+_tkts.add(tid)
+return _tkts
+
+def find_bz_in_line(line, prefix):
 bznum_re = re.compile(r'(?P \d{4,8})|'
 r'( 
http(s)?://bugzilla\.redhat\.com/(show_bug\.cgi\?id=)?(?P\d{4,8}))')
-for match in bznum_re.finditer(line[len(f"{prefix}:"):]):
-for group in [ 'bug_ids', 'url_bugs' ]:
-if match.group(group):
-bid = match.group(group).strip()
-_bugs.add(bid)
-return _bugs
+return find_ticket_in_line(line, prefix, bznum_re, [ 'bug_ids', 'url_bugs' 
])
+
+def find_ji_in_line(line, prefix):
+ji_re = re.compile(r' 
https://issues\.redhat\.com/(?:browse|projects/RHEL/issues)/(?PRHEL-\d{1,8})\s*$')
+return find_ticket_in_line(line, prefix, ji_re, [ 'jira_id' ])
 
 def find_cve_in_line(line):
 """Return cve number from properly formated CVE: line."""
@@ -56,6 +62,7 @@ def parse_commit(commit):
 cve_set = set()
 bug_set = set()
 zbug_set = set()
+jira_set = set()
 for line in lines[1:]:
 # Metadata in git notes has priority over commit log
 # If we found any BZ/ZBZ/CVE in git notes, we ignore commit log
@@ -70,28 +77,35 @@ def parse_commit(commit):
 # Grab CVE tags if they are present
 cve_set.update(find_cve_in_line(line))
 
-return (log_entry, sorted(cve_set), sorted(bug_set), sorted(zbug_set))
+# Process Jira issues
+jira_set.update(find_ji_in_line(line, 'JIRA'))
+
+return (log_entry, sorted(cve_set), sorted(bug_set), sorted(zbug_set), 
sorted(jira_set))
 
 
 if __name__ == "__main__":
 all_bzs = []
 all_zbzs = []
+all_jiras = []
 commits = sys.stdin.read().split('\0')
 for c in commits:
 if not c:
 continue
-log_item, cves, bugs, zbugs = parse_commit(c)
+log_item, cves, bugs, zbugs, jiras = parse_commit(c)
 entry = f"{log_item}"
-if bugs or zbugs:
+if bugs or zbugs or jiras:
 entry += " ["
 if zbugs:
 entry += " ".join(zbugs)
 all_zbzs.extend(zbugs)
-if bugs and zbugs:
-entry += " "
 if bugs:
+entry += " " if zbugs else ""
 entry += " ".join(bugs)
 all_bzs.extend(bugs)
+if jiras:
+entry += " " if bugs or zbugs else ""
+entry += " ".join(jiras)
+all_jiras.extend(jiras)
 entry += "]"
 if cves:
 entry += " {" + " ".join(cves) + "}"
@@ -106,5 +120,9 @@ if __name__ == "__main__":
 if i:
 print(", ", end="")
 print(f"rhbz#{bzid}", end="")
+for j, jid in enumerate(all_jiras):
+if j or len(resolved_bzs) > 0:
+   print(", ", end="")
+print(f"{jid}", end="")
 print("\n")
 

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2496
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCH 2/3] redhat: fix duplicate jira issues in the resolves line

2023-06-01 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix duplicate jira issues in the resolves line

Bugzilla: INTERNAL
Upstream Status: RHEL only

The current genlog.py code does not check for duplicated jira issues
added to the resolves line. For bugzilla numbers, that check was present
when creating the resolved_bzs list.

To resolve it, convert everything to sets (which can't have duplicated
items), as such we don't need to do additional checks when printing the
resolves line. Also, make sure we sort not only the per changelog item
numbers, but also the items in the resolves line.

Fixes: 552fe563b55b ("redhat: add support for Jira issues in changelog")
Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/scripts/genspec/genlog.py b/redhat/scripts/genspec/genlog.py
index blahblah..blahblah 100755
--- a/redhat/scripts/genspec/genlog.py
+++ b/redhat/scripts/genspec/genlog.py
@@ -80,13 +80,13 @@ def parse_commit(commit):
 # Process Jira issues
 jira_set.update(find_ji_in_line(line, 'JIRA'))
 
-return (log_entry, sorted(cve_set), sorted(bug_set), sorted(zbug_set), 
sorted(jira_set))
+return (log_entry, cve_set, bug_set, zbug_set, jira_set)
 
 
 if __name__ == "__main__":
-all_bzs = []
-all_zbzs = []
-all_jiras = []
+all_bzs = set()
+all_zbzs = set()
+all_jiras = set()
 commits = sys.stdin.read().split('\0')
 for c in commits:
 if not c:
@@ -96,32 +96,29 @@ if __name__ == "__main__":
 if bugs or zbugs or jiras:
 entry += " ["
 if zbugs:
-entry += " ".join(zbugs)
-all_zbzs.extend(zbugs)
+entry += " ".join(sorted(zbugs))
+all_zbzs.update(zbugs)
 if bugs:
 entry += " " if zbugs else ""
-entry += " ".join(bugs)
-all_bzs.extend(bugs)
+entry += " ".join(sorted(bugs))
+all_bzs.update(bugs)
 if jiras:
 entry += " " if bugs or zbugs else ""
-entry += " ".join(jiras)
-all_jiras.extend(jiras)
+entry += " ".join(sorted(jiras))
+all_jiras.update(jiras)
 entry += "]"
 if cves:
-entry += " {" + " ".join(cves) + "}"
+entry += " {" + " ".join(sorted(cves)) + "}"
 print(entry)
 
-resolved_bzs = []
-for bzid in (all_zbzs if all_zbzs else all_bzs):
-if not bzid in resolved_bzs:
-resolved_bzs.append(bzid)
+resolved_bzs = all_zbzs if all_zbzs else all_bzs
 print("Resolves: ", end="")
-for i, bzid in enumerate(resolved_bzs):
+for i, bzid in enumerate(sorted(resolved_bzs)):
 if i:
 print(", ", end="")
 print(f"rhbz#{bzid}", end="")
-for j, jid in enumerate(all_jiras):
-if j or len(resolved_bzs) > 0:
+for j, jid in enumerate(sorted(all_jiras)):
+if j or resolved_bzs:
print(", ", end="")
 print(f"{jid}", end="")
 print("\n")

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2496
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv4 0/6] redhat: Add realtime build and packaging logic

2023-05-30 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2403#note_1411398674

Hi Clark,

this is different from what we have on centos9/9.3: the name of the package is
rt-debug.

Using kernel-rt_debug instead of kernel-rt-debug breaks the upgrade path from
the previously separate kernel-rt and from 9.3

Can you sync your patch with the centos9 version? You also may need to bring
the 64k arm support to ark, not sure. On centos/9.3, it was a dependency
because of the order it was merged.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv9 0/3] redhat/kernel.spec.template: Add global compression variables

2023-04-26 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2173#note_1369001617

I would go further on that, I see little value in using variables. It's very
rare to switch compression algorithms over time. The only way I see it's
useful is if we want to support users switching compressors, but then you're
right here we may want to cover it fully like wrapping the compression level
argument into another variable. I'm acking the MR anyway as is though, as I'm
fine with either approach.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2] redhat/Makefile: fix default values for dist-brew's DISTRO and DIST

2023-03-02 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2339#note_1298900961

ah, I think you need to run ```make dist-self-test-data```, and add/push the
changes to redhat/self-test/data/*, can you try that?
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2] redhat/Makefile: fix default values for dist-brew's DISTRO and DIST

2023-03-02 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2339#note_1298757389

This last version looks good to me, thank you.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] redhat/Makefile: fix default values for dist-brew's DISTRO and DIST

2023-03-01 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2339#note_1297506964

About changing the DISTRO, may be it's be useful, although we default for RHEL
as usually a RHEL developer wants to test the build on brew (rhel) instead of
centos even if we are building/integrating it first on centos. So that's why I
think no one complained so far, the maintainer (eg. me) is usually the one
which uses the DISTRO value (only at the dist-git sync step), but that has
been working on the centos side so far passing through the command line, so I
didn't noticed it was broken too since I don't use Makefile.variables.

To determine the where the value comes, we could use origin like ```ifeq
("$(origin BUILD_TARGET)", "command line")```, but I didn't think what the
logic should be.

I think we should keep DIST as is on Makefile.variables and logic is done in
Makefile? I think could be possible? I'm not sure if there is a problem. Right
now we change the DIST in Makefile.variables when kernel-ark is forked to
RHEL, for example, centos9 kernel src.git sets it to DIST ?= .el9 in
Makefile.variables.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] redhat/kernel.spec.template: Fix RHEL systemd-boot-unsigned dependency for

2023-02-28 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2341#note_1295896026

As Aquini pointed out: https://bugzilla.redhat.com/show_bug.cgi?id=2157663 -
they changed it on rhel9 now, surprisingly now breaking any kernel build on
rhel9
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] redhat/Makefile: fix default values for dist-brew's DISTRO and DIST

2023-02-28 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2339#note_1295688724

This doesn't work, because it'll override anything you set inside
redhat/Makefile.variables. For example if I do this:

```
diff --git a/redhat/Makefile.variables b/redhat/Makefile.variables
index c20fef88a468..1381ffbb7c1d 100644
--- a/redhat/Makefile.variables
+++ b/redhat/Makefile.variables
@@ -30,7 +30,7 @@ BUILD_PROFILE ?=
 # default values can be found in redhat/Makefile and are dependent on the
 # target OS.  The target OS can be changed by setting the DIST or DISTRO
 # variables.
-BUILD_TARGET ?=
+BUILD_TARGET ?= rawhide

 # BUMP_RELEASE determines whether the pkgrelease is bumped when you call
 # 'make dist-release'.  It should be set to "yes" if you expect to do
```

Your patch will override it, even if I want to change the value in
Makefile.variables.

Probably what we want to do is to change the ifdefs. For example:

```
  ifndef BUILD_TARGET
```

to

```
ifeq ("$(BUILD_TARGET)", "")
```

It looks like we should be using ifeq instead. I remember something about
ifdef
not being ideal with make and exported variables.

So switch to ifeq, and see if the remaining changes are really needed or not
here.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv3 0/4] redhat: update merge.py to handle merge.pl corner cases

2023-01-31 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2262#note_1259584001

Yes I agree we could do improvements/changes to the script or cleanup the
comments in another MR, since it's out of scope for this one, so approving
this.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv3 0/4] redhat: update merge.py to handle merge.pl corner cases

2023-01-31 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2262#note_1259579704

Yes indeed. I think we can keep the logic the same as .pl right now, and may
be we can improve the .py script later. Since original concern is solved,
resolving this thread.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv3] kernel.spec: add llvm-devel build requirement

2023-01-31 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2264#note_1259440012

You should specify here only llvm-devel: when you install llvm-devel it
already requires llvm so it's redundant to list both.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2 0/3] redhat: update merge.py to handle merge.pl corner cases

2023-01-30 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2262#note_1258074197

Looks like this part doesn't translate exactly what the perl part implemented:

```
if (defined($configname) && !exists($configvalues{$configname})) {
$configvalues{$configname} = $str;
$configoptions[$configcounter] = $configname;
$configcounter ++;
}
```

If the overrides file has the same config option listed twice, the perl seemed
to ignore it. Eg.

CONFIG_FOO = y
CONFIG_FOO = n

configvalues gets CONFIG_FOO = y
next iteration configvalues is not assigned because CONFIG_FOO is already
assigned previously

The python script is not checking this, always assign what it reads
(```overrides[c] = line```)

Not sure if this is a real problem or not, but just pointing out. We can
resolve this thread if it's not an issue.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCH] redhat: ignore rpminspect runpath report on urandom_read selftest binaries

2023-01-26 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: ignore rpminspect runpath report on urandom_read selftest binaries

Upstream Status: RHEL only

Since a recent bpf update on the kernel, rpminspect started to report the
following issue in the urandom_read selftest binaries:
/usr/libexec/kselftests/bpf/(...)urandom_read has an invalid-looking DT_RPATH 
on : .

This is expected since upstream commit:
00a0fa2d7d49 ("selftests/bpf: Add urandom_read shared lib and USDTs")

As the selftest binary uses shared library linked from the current directory.

Thus add the reported binaries to the ignore list in rpminspect configuration
so this doesn't get reported every rpminspect CI run.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/scripts/gating/rpminspect.yaml 
b/redhat/scripts/gating/rpminspect.yaml
index blahblah..blahblah 100644
--- a/redhat/scripts/gating/rpminspect.yaml
+++ b/redhat/scripts/gating/rpminspect.yaml
@@ -24,3 +24,8 @@ patches:
 ignore_list:
 - linux-kernel-test.patch
 - patch-%%SPECKVERSION%%.%%SPECKPATCHLEVEL%%-redhat.patch
+
+runpath:
+ignore:
+- /usr/libexec/kselftests/bpf/urandom_read
+- /usr/libexec/kselftests/bpf/no_alu32/urandom_read

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2270
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2] Revert "redhat: fix elf got hardening for vm tools"

2023-01-25 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2236#note_1252361900

@dzickusrh gitlab considers your block rule as invalid here since you're
blocking your own MR, anyway, I think you still need to unblock this, can you
"/unblock" this MR?
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv6] redhat/kernel.spec.template: Parallelize compression

2023-01-23 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2254#note_1249574743

@prarit didn't see changes compared to latest version, didn't you want it to
be ```-n 16```?
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2] Revert "redhat: fix elf got hardening for vm tools"

2023-01-23 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2236#note_1249510524

We agreed in the last ark meeting we need this patch and the CFLAGS underlying
issue will be fixed once upstream patch
https://www.spinics.net/lists/kernel/msg4653901.html lands.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv5] redhat/kernel.spec.template: Parallelize compression

2023-01-23 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2254#note_1249483495

I don't think we need to overengineer this. I agree fork costs can be
something to avoid, but we don't need a separate script. If there is a
heuristic for it we should calculate it and and give to ```-n``` option. But
I'm ok the way it is right now. If you want to determine an heuristic and
measure the difference feel free to.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] redhat/kernel.spec.template: Parallelize compression

2023-01-20 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2254#note_1248025244

Do you really need an external script here? I think you can use -n argument of
xargs, eg:

```find $RPM_BUILD_ROOT/lib/modules/ -type f -name '*.ko' | xargs -n 1
-P${RPM_BUILD_NCPUS} -r xz; \```
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv7 2/4] redhat/Makefile.variables: Remove mention of Makefile.rhpkg

2023-01-16 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2240#note_1241321402

I just noticed this, you mention removing Makefile.rhpkg from the comment, but
you kept it and instead removed the ../Makefile.rhelver file from the comment.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv6 0/4] makefile: Add -Werror support for RHEL

2023-01-16 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2240#note_1241308951

Well, but your code in change in redhat/configs/build_configs.sh sets
CONFIG_WERROR when you set ENABLE_WERROR, are you aiming a use case where
build_configs.sh doesn't run?
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv6 0/4] makefile: Add -Werror support for RHEL

2023-01-16 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2240#note_1241302470

@prarit looks even better without need the -Wno-error flag, however do you
really need to add Werror again here? Since if you already have ENABLE_WERROR
set, it will flip CONFIG_WERROR=y in which case the kernel Makefile already
takes care of adding the flag.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2] Revert "redhat: fix elf got hardening for vm tools"

2022-12-22 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2236#note_1219437543

And yes, building kernel tools with proper cflags/ldflags have been a
nightmare, since each Makefile implements its own way of using flags, it's not
standard, not everyone uses EXTRA_CFLAGS etc.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2] Revert "redhat: fix elf got hardening for vm tools"

2022-12-22 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2236#note_1219436857

I'm not familiar with the original problem, but familiar with rpminspect in
general. The problem is that the build of the tool does not respect the
CFLAGS/LDFLAGS given by rpm with hardening options etc., so later rpminspect
fails since the binaries were not built with proper cflags/ldflags.

In this case looking quickly here, there are two problems: the makefile
(tools/vm/Makefile) does not accept user specified CFLAGS, and for the
tools/lib/api that the slabinfo/page_owner_sort uses, that uses
tools/lib/api/Makefile which has EXTRA_CFLAGS usage...

So I think you need two options here to really fix this, something like this
patch:

```
$ git diff
diff --git a/tools/vm/Makefile b/tools/vm/Makefile
index 9860622cbb15..6c1da51f4177 100644
--- a/tools/vm/Makefile
+++ b/tools/vm/Makefile
@@ -8,8 +8,8 @@ TARGETS=page-types slabinfo page_owner_sort
 LIB_DIR = ../lib/api
 LIBS = $(LIB_DIR)/libapi.a

-CFLAGS = -Wall -Wextra -I../lib/
-LDFLAGS = $(LIBS)
+CFLAGS += -Wall -Wextra -I../lib/
+LDFLAGS += $(LIBS)

 all: $(TARGETS)

```

Then you can build the tools with:
EXTRA_CFLAGS="${RPM_OPT_FLAGS}" %{tools_make} slabinfo page_owner_sort

You can verify that the proper flags are being used passing V=1, for example:
EXTRA_CFLAGS="${RPM_OPT_FLAGS}" %{tools_make} V=1 slabinfo page_owner_sort
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH 0/3] redhat/kernel.spec.template: Add global compression variables

2022-12-04 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2173#note_1195644358

I'm not sure you can really change the symvers file compression away from gzip
right now. At least looking quickly here, it seems grub2 on Fedora 36 has a
script that references it:

```
$ grep -r symvers /usr/lib/kernel/install.d/*
/usr/lib/kernel/install.d/20-grub.install:# symvers is
symvers-.gz symlink, needs a special treatment
/usr/lib/kernel/install.d/20-grub.install:
i="$KERNEL_DIR/symvers.gz"
/usr/lib/kernel/install.d/20-grub.install:rm -f
"/boot/symvers-${KERNEL_VERSION}.gz"
/usr/lib/kernel/install.d/20-grub.install:ln -s "$i"
"/boot/symvers-${KERNEL_VERSION}.gz"
/usr/lib/kernel/install.d/20-grub.install:restorecon
"/boot/symvers-${KERNEL_VERSION}.gz"
/usr/lib/kernel/install.d/20-grub.install:# symvers is
symvers-.gz symlink, needs a special treatment
/usr/lib/kernel/install.d/20-grub.install:rm -f
"/boot/symvers-${KERNEL_VERSION}.gz"
$ rpm -qf /usr/lib/kernel/install.d/20-grub.install
grub2-common-2.06-54.fc36.noarch
```
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] redhat: genlog.sh should expect genlog.py in the current directory

2022-11-30 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2167#note_1191028584

I think that's the way to do it. It's like an script, you can't remove the
genlog.sh to a separate place due the CLOGF temp file.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] redhat/Makefile: Fix RHJOBS grep warning

2022-11-16 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2148#note_1174863504

So there is a problem in this expression I think, but it was already an
existing problem before this change.

If you have eg.: ```MAKEFLAGS="-j45"```, be at the beginning or -j being the
only option in the MAKEFLAGS, the expression will not match.

I think the expression need to be changed to: ```grep -Eo "(^|[ ])-j[0-9]*"```
So it matches either at the beginning, or as an option in the middle of other
options with the space.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2] redhat: Force remove tmp file

2022-11-07 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2135#note_1163073892

I see. But I mean you could have folded it in just one line:
```rm -f vmlinux.o .tmp_vmlinux.btf```
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] redhat: Test tmp file before deleting

2022-11-07 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2135#note_1162936472

Wouldn't it be better to just use ```rm -f``` instead of adding the if?
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCH] redhat: fix the branch we pull from the documentation tree

2022-11-04 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix the branch we pull from the documentation tree

Bugzilla: INTERNAL
Upstream Status: RHEL only

It has been some time since the documentation project has a main
branch instead of master, so fix that.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/scripts/merge-subtrees.sh b/redhat/scripts/merge-subtrees.sh
index blahblah..blahblah 100755
--- a/redhat/scripts/merge-subtrees.sh
+++ b/redhat/scripts/merge-subtrees.sh
@@ -8,5 +8,5 @@ entries="
 
 echo "$entries" | while read -r name url; do
[ -z "$name" ] && continue
-   git subtree pull --prefix="$name" "$url"  master
+   git subtree pull --prefix="$name" "$url" main
 done

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2130
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2 0/6] redhat: python replacement for merge.pl

2022-10-27 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2117#note_1151621759

Yes, python2 doesn't exist anymore, and python3 is the default. At least it
hasn't been a problem.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv2 0/6] redhat: python replacement for merge.pl

2022-10-27 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2117#note_1151619466

Wouldn't it be better to just do what we already do eg. for redhat/genlog.py
(where it's called directly from redhat/genlog.sh)?

```
Add #!/usr/bin/python3 to the first line of merge.py
make it executable
call it directly from buildconfigs.sh
```
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] redhat/configs: Disable CONFIG_FW_LOADER_COMPRESS

2022-10-24 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2112#note_1146809414

Fedora enabled firmware compression. So when we inherited things for RHEL 9,
the kernel had to followup and enable it, otherwise it breaks firmware
loading. So yes, we can't disable it, otherwise it'll break firmware loading,
as @jmflinuxtx mentions above.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH] redhat: set LC_ALL=C before sorting config content

2022-09-19 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2011#note_1105589193

sorry I don't know. I could trace it back to a commit in rhel-7 that
@jwilsonrh did (commit 41ce43f3f46207e6e2a3223fddf57ab7e0b99155 - "[redhat]
configs: adapt tooling for new config option layout"). Not sure if Jarod will
remember about this... But from the logic I believe the sort is done because
the script concatenate some config files and may be for cosmetic reason to
have a sane final config file.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv5 0/9] ark: rework changelog generation process during dist-release

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980#note_1067972199

Check out the updated MR, example of the entry now:

```
+* Wed Aug 17 2022 Herton R. Krzesinski 
[6.0.0-0.rc1.3cc40a443a04.15]
+- Linux v6.0.0-0.rc1.3cc40a443a04
+Resolves:
+
```
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv5 9/9] redhat: change the changelog item for upstream merges

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: change the changelog item for upstream merges

We do merges from upstream in ark, not real rebases. Reword the
changelog item as suggested by Don Zickus and Justin Forbes.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -50,7 +50,7 @@ fi
 if [ "$MARKER" != "$LAST_MARKER" ]; then
# genlog.py always adds a Resolves: line, thus we
# can insert the rebase changelog item before it
-   sed -i "s/\(^Resolves:.*\)/- $MARKER rebase\n\1/" "$clogf"
+   sed -i "s/\(^Resolves:.*\)/- Linux 
v${SPECVERSION}${UPSTREAMBUILD:+-}${UPSTREAMBUILD%.}\n\1/" "$clogf"
 fi
 
 # during rh-dist-git genspec runs again and generates empty changelog

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv5 8/9] redhat: fix dist-release build number test

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix dist-release build number test

Since previous commit, when fixing the build number bump, this test will
fail, now the build number and package release number should match as
this test was intended to check.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/self-test/2001-dist-release.bats 
b/redhat/self-test/2001-dist-release.bats
index blahblah..blahblah 100644
--- a/redhat/self-test/2001-dist-release.bats
+++ b/redhat/self-test/2001-dist-release.bats
@@ -43,7 +43,6 @@ _dist-release_test_2() {
# Strip off ...kernel-VV.PP.SS-:
pkgrelease=${title##*kernel-+([5-9]).+([0-9]).+([0-9])-}
build=$(BUILD= DIST=.fc33 make dist-dump-variables | grep -E "^BUILD=" 
| cut -d"=" -f2 | xargs)
-   ((build--))
echo "pkgrelease=$pkgrelease"
echo "build=$build"
run _dist-release_test_2

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv5 7/9] redhat: fix release number bump when dist-release-changed runs

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix release number bump when dist-release-changed runs

Right now inside the dist-release-changed target, we run a sub make
invocation calling the dist-release-finish target, after bumping
the release number with new_release.sh. The problem is that the
BUILD number still pointed to the old pre-bump number inside the
dist-release-finish invocation, and the problem is that we don't
reload the RHEL_RELEASE number in redhat/Makefile if BUILD is already
defined. Since it looks like that check was added to avoid overriding
the BUILD number set when RHSELFTESTDATA is defined, lets change
the if check there so BUILD has the updated value when dist-release-finish
runs through dist-release-changed.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -58,7 +58,7 @@ SPECKPATCHLEVEL:=$(shell $(GIT) show $(HEAD):Makefile | sed 
-ne '/^PATCHLEVEL\ =
 SPECKSUBLEVEL:=$(shell $(GIT) show $(HEAD):Makefile | sed -ne '/^SUBLEVEL\ =\ 
/{s///;p;q}')
 SPECKEXTRAVERSION:=$(shell $(GIT) show $(HEAD):Makefile | sed -ne 
'/^EXTRAVERSION\ =\ /{s///;p;q}')
 GITID:= $(shell $(GIT) log --max-count=1 --pretty=format:%H $(HEAD))
-ifndef BUILD
+ifndef RHSELFTESTDATA
   BUILD:=$(RHEL_RELEASE)
 endif
 FLAVOR=

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv5 6/9] redhat: use new genlog.sh script to detect changes for dist-release

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: use new genlog.sh script to detect changes for dist-release

The current dist-release process has a flaw, where both its subtargets
(dist-release-changed and dist-release-finish) depends on setup-source
target to run (so genspec.sh is run twice). To avoid having to do the
work twice and do and revert changes before dist-release-finish is
called, we now detect changes by creating the changelog using genlog.sh
first to detect if there are really changes, and do the marker check
using an if condition instead of relying on git to see if there are
changes.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -572,18 +572,18 @@ dist-rpm-baseonly: dist-sources do-rpmbuild
 # unless you know what you're doing, you don't want to use the next four ones
 dist-release-finish: setup-source
@cp $(SOURCES)/$(SPECCHANGELOG) $(REDHAT)/$(SPECCHANGELOG)
+   @echo $(MARKER) > $(REDHAT)/marker
@$(GIT) add $(REDHAT)/$(SPECCHANGELOG)
@$(GIT) add $(REDHAT)/marker
@$(GIT) commit -s ../Makefile.rhelver $(REDHAT)/marker 
$(REDHAT)/$(SPECCHANGELOG) $(SPECFILE).template -m "[redhat] $(RELEASETAG)"
-dist-release-changed: setup-source
-   @cp $(SOURCES)/$(SPECCHANGELOG) $(REDHAT)/$(SPECCHANGELOG)
-   @echo $(MARKER) > $(REDHAT)/marker
-   @# if neither changelog nor marker was updated, skip bumping a release
-   @$(GIT) update-index -q --really-refresh
-   @if $(GIT) diff-index --quiet HEAD; then \
+dist-release-changed:
+   @CLOGF="$$(mktemp)"; \
+   trap 'rm -f "$$CLOGF"' SIGHUP SIGINT SIGTERM EXIT; \
+   $(REDHAT)/genlog.sh "$$CLOGF"; \
+   if [ "$(MARKER)" == "$$(cat $(REDHAT)/marker)" ] && \
+  [ "$$(wc -l $$CLOGF | cut -d ' ' -f 1)" -le 3 ]; then \
echo "Nothing changed, skipping updates"; \
else \
-   $(GIT) checkout -- $(REDHAT)/$(SPECCHANGELOG); \
$(REDHAT)/scripts/new_release.sh; \
$(MAKE) dist-release-finish; \
fi

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv5 5/9] redhat: move changelog addition to the spec file back into genspec.sh

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: move changelog addition to the spec file back into genspec.sh

In the next change, we want to detect at the Makefile level if we have new
changelog entries before we touch the spec file, so it makes sense to have
changelog addition being done in genspec.sh.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -36,13 +36,13 @@ git log --topo-order --no-merges -z "$GIT_NOTES" 
"$GIT_FORMAT" \
 
 if [ "$HIDE_REDHAT" = "1" ]; then
grep -v -e "^- \[redhat\]" "$clogf" |
-   sed -e 's!\[Fedora\]!!g' > "$clogf.stripped"
-   cp "$clogf.stripped" "$clogf"
+   sed -e 's!\[Fedora\]!!g' > "$clogf.tmp"
+   mv -f "$clogf.tmp" "$clogf"
 fi
 
 if [ "$HIDE_UNSUPPORTED_ARCH" = "1" ]; then
-   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.stripped"
-   cp "$clogf.stripped" "$clogf"
+   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.tmp"
+   mv -f "$clogf.tmp" "$clogf"
 fi
 
 # If the markers aren't the same then this a rebase.
@@ -60,17 +60,4 @@ if [ "$LENGTH" = 0 ]; then
echo -n > "$clogf"
 fi
 
-cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
-mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
-
-# genlog.py generates Resolves lines as well, strip these from RPM changelog
-grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
-
-test -f "$SOURCES/$SPECFILE" &&
-   sed -i -e "
-   /%%SPECCHANGELOG%%/r $clogf.stripped
-   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
-
 echo "MARKER is $MARKER"
-
-rm -f "$clogf".stripped;
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -88,9 +88,20 @@ if [ -n "$RHSELFTESTDATA" ]; then
 fi
 
 clogf=$(mktemp)
-trap 'rm -f "$clogf"' SIGHUP SIGINT SIGTERM EXIT
+trap 'rm -f "$clogf" "$clogf".stripped' SIGHUP SIGINT SIGTERM EXIT
 "${0%/*}"/genlog.sh "$clogf"
 
+cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
+mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
+
+# genlog.py generates Resolves lines as well, strip these from RPM changelog
+grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
+
+test -f "$SOURCES/$SPECFILE" &&
+   sed -i -e "
+   /%%SPECCHANGELOG%%/r $clogf.stripped
+   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
+
 if [ "$DISTRO" == "fedora" ]; then
# The tarball in the SRPM contains only the upstream sources.
 

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv5 4/9] redhat: always add a rebase entry when ark merges from upstream

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: always add a rebase entry when ark merges from upstream

We have some complex logic to parse changelog entries to choose whether
we add a rebase entry or not to the changelog. Since genlog.py is adding
a Resolves entry always now, simply replace that logic with a
replacement using sed to add a rebase entry when the markers differ.
The marker check doesn't work at the moment due Makefile logic but
that will be fixed in the following changes.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -48,14 +48,9 @@ fi
 # If the markers aren't the same then this a rebase.
 # This means we need to zap entries that are already present in the changelog.
 if [ "$MARKER" != "$LAST_MARKER" ]; then
-   # awk trick to get all unique lines
-   awk '!seen[$0]++' "$SOURCES/$SPECCHANGELOG" "$clogf" > "$clogf.unique"
-   # sed trick to get the end of the changelog minus the line
-   sed -e '1,/# END OF CHANGELOG/ d' "$clogf.unique" > "$clogf.tmp"
-   # Add an explicit entry to indicate a rebase.
-   echo "" > "$clogf"
-   echo -e "- $MARKER rebase" | cat "$clogf.tmp" - >> "$clogf"
-   rm "$clogf.tmp" "$clogf.unique"
+   # genlog.py always adds a Resolves: line, thus we
+   # can insert the rebase changelog item before it
+   sed -i "s/\(^Resolves:.*\)/- $MARKER rebase\n\1/" "$clogf"
 fi
 
 # during rh-dist-git genspec runs again and generates empty changelog

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv5 2/9] redhat: don't hardcode temporary changelog file

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: don't hardcode temporary changelog file

We always create a changelog file inside the sources directory, which
is temporary and always deleted at the end. Instead hardcoding the
temporary file name, create a real temporary file to be used with
genlog.sh script, given as its first argument.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -1,7 +1,7 @@
 #!/bin/bash
 
 LAST_MARKER=$(cat "${REDHAT}"/marker)
-clogf="$SOURCES/changelog"
+clogf="$1"
 # hide [redhat] entries from changelog
 HIDE_REDHAT=1;
 # hide entries for unsupported arches
@@ -70,8 +70,7 @@ fi
 # create empty file to avoid adding extra header to changelog
 LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')
 if [ "$LENGTH" = 0 ]; then
-   rm -f "$clogf"
-   touch "$clogf"
+   echo -n > "$clogf"
 fi
 
 cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
@@ -87,4 +86,4 @@ test -f "$SOURCES/$SPECFILE" &&
 
 echo "MARKER is $MARKER"
 
-rm -f "$clogf"{,.stripped};
+rm -f "$clogf".stripped;
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -87,7 +87,9 @@ if [ -n "$RHSELFTESTDATA" ]; then
exit 0
 fi
 
-"${0%/*}"/genlog.sh
+clogf=$(mktemp)
+trap 'rm -f "$clogf"' SIGHUP SIGINT SIGTERM EXIT
+"${0%/*}"/genlog.sh "$clogf"
 
 if [ "$DISTRO" == "fedora" ]; then
# The tarball in the SRPM contains only the upstream sources.

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv5 3/9] redhat: drop merge ark patches hack

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: drop merge ark patches hack

Changelog generation carries a temporary hack to add a merge ark-patches
entry in the changelog. I think we can drop this today.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -58,14 +58,6 @@ if [ "$MARKER" != "$LAST_MARKER" ]; then
rm "$clogf.tmp" "$clogf.unique"
 fi
 
-# HACK temporary hack until single tree workflow
-# Don't reprint all the ark-patches again.
-if [ -n "$(git log --oneline --first-parent --grep="Merge ark patches" 
"$lasttag"..)" ]; then
-   # Throw away the clogf and just print the summary merge
-   echo "" > "$clogf"
-   echo "- Merge ark-patches" >> "$clogf"
-fi
-
 # during rh-dist-git genspec runs again and generates empty changelog
 # create empty file to avoid adding extra header to changelog
 LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv5 1/9] redhat: split changelog generation from genspec.sh

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: split changelog generation from genspec.sh

This creates a new genlog.sh script to split out the changelog
generation code from genspec.sh. The splitted code is kept
as is.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
new file mode 100755
index blahblah..blahblah 100755
--- /dev/null
+++ b/redhat/genlog.sh
@@ -0,0 +1,90 @@
+#!/bin/bash
+
+LAST_MARKER=$(cat "${REDHAT}"/marker)
+clogf="$SOURCES/changelog"
+# hide [redhat] entries from changelog
+HIDE_REDHAT=1;
+# hide entries for unsupported arches
+HIDE_UNSUPPORTED_ARCH=1;
+# override LC_TIME to avoid date conflicts when building the srpm
+LC_TIME=
+
+GIT_FORMAT="--format=- %s (%an)%n%N%n^^^NOTES-END^^^%n%b"
+GIT_NOTES="--notes=refs/notes/${RHEL_MAJOR}.${RHEL_MINOR}*"
+
+lasttag=$(git rev-list --first-parent --grep="^\[redhat\] 
kernel-${SPECKVERSION}.${SPECKPATCHLEVEL}" --max-count=1 HEAD)
+# if we didn't find the proper tag, assume this is the first release
+if [[ -z $lasttag ]]; then
+if [[ -z ${MARKER//[0-9a-f]/} ]]; then
+# if we're doing an untagged release, just use the marker
+echo "Using $MARKER"
+lasttag=$MARKER
+else
+   lasttag=$(git describe --match="$MARKER" --abbrev=0)
+fi
+fi
+echo "Gathering new log entries since $lasttag"
+# master is expected to track mainline.
+
+cname="$(git var GIT_COMMITTER_IDENT |sed 's/>.*/>/')"
+cdate="$(LC_ALL=C date +"%a %b %d %Y")"
+cversion="[$BASEVERSION]";
+echo "* $cdate $cname $cversion" > "$clogf"
+
+git log --topo-order --no-merges -z "$GIT_NOTES" "$GIT_FORMAT" \
+   ^"${UPSTREAM}" "$lasttag".. -- ':!/redhat/rhdocs' | "${0%/*}"/genlog.py 
>> "$clogf"
+
+if [ "$HIDE_REDHAT" = "1" ]; then
+   grep -v -e "^- \[redhat\]" "$clogf" |
+   sed -e 's!\[Fedora\]!!g' > "$clogf.stripped"
+   cp "$clogf.stripped" "$clogf"
+fi
+
+if [ "$HIDE_UNSUPPORTED_ARCH" = "1" ]; then
+   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.stripped"
+   cp "$clogf.stripped" "$clogf"
+fi
+
+# If the markers aren't the same then this a rebase.
+# This means we need to zap entries that are already present in the changelog.
+if [ "$MARKER" != "$LAST_MARKER" ]; then
+   # awk trick to get all unique lines
+   awk '!seen[$0]++' "$SOURCES/$SPECCHANGELOG" "$clogf" > "$clogf.unique"
+   # sed trick to get the end of the changelog minus the line
+   sed -e '1,/# END OF CHANGELOG/ d' "$clogf.unique" > "$clogf.tmp"
+   # Add an explicit entry to indicate a rebase.
+   echo "" > "$clogf"
+   echo -e "- $MARKER rebase" | cat "$clogf.tmp" - >> "$clogf"
+   rm "$clogf.tmp" "$clogf.unique"
+fi
+
+# HACK temporary hack until single tree workflow
+# Don't reprint all the ark-patches again.
+if [ -n "$(git log --oneline --first-parent --grep="Merge ark patches" 
"$lasttag"..)" ]; then
+   # Throw away the clogf and just print the summary merge
+   echo "" > "$clogf"
+   echo "- Merge ark-patches" >> "$clogf"
+fi
+
+# during rh-dist-git genspec runs again and generates empty changelog
+# create empty file to avoid adding extra header to changelog
+LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')
+if [ "$LENGTH" = 0 ]; then
+   rm -f "$clogf"
+   touch "$clogf"
+fi
+
+cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
+mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
+
+# genlog.py generates Resolves lines as well, strip these from RPM changelog
+grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
+
+test -f "$SOURCES/$SPECFILE" &&
+   sed -i -e "
+   /%%SPECCHANGELOG%%/r $clogf.stripped
+   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
+
+echo "MARKER is $MARKER"
+
+rm -f "$clogf"{,.stripped};
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -1,16 +1,6 @@
 #!/bin/bash
 # shellcheck disable=SC2153
 
-LAST_MARKER=$(cat "${REDHAT}"/marker)
-clogf="$SOURCES/changelog"
-# hide [redhat] entries from changelog
-HIDE_REDHAT=1;
-# hide entries for unsupported arches
-HIDE_UNSUPPORTED_ARCH=1;
-# override LC_TIME to avoid date conflicts when building the srpm
-LC_TIME=
-# STAMP=$(echo $MARKER | cut -f 1 -d '-' | sed -e "s/v//"); # unused
-
 UPSTREAM=$(git rev-parse -q --verify origin/"${UPSTREAM_BRANCH}" || \
   git rev-parse -q --verify "${UPSTREAM_BRANCH}")
 
@@ -97,83 +87,7 @@ if [ -n "$RHSELFTESTDATA" ]; then
exit 0
 fi
 
-GIT_FORMAT="--format=- %s (%an)%n%N%n^^^NOTES-END^^^%n%b"
-GIT_NOTES="--notes=refs/notes/${RHEL_MAJOR}.${RHEL_MINOR}*"
-
-lasttag=$(git rev-list --first-parent --grep="^\[redhat\] 
kernel-${SPECKVERSION}.${SPECKPATCHLEVEL}" --max-count=1 HEAD)
-# if we didn't find the proper tag, assume this is the first release
-if [[ -z $lasttag ]]; then
-if [[ -z 

[OS-BUILD PATCHv5 0/9] ark: rework changelog generation process during dist-release

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980

This series try to improve the changelog generation process done
during a dist-release target run. To avoid doing hacks and having
to run setup-source/genspec.sh twice, this series splits the changelog
creation from genspec.sh and changes the Makefile to use the splitted
code.

Signed-off-by: Herton R. Krzesinski 

---
 redhat/self-test/2001-dist-release.bats |   1 -
 redhat/Makefile |  16 +++---
 redhat/genlog.sh|  63 +
 redhat/genspec.sh   |  81 +---
 4 files changed, 74 insertions(+), 87 deletions(-)
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv4 0/8] ark: rework changelog generation process during dist-release

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980#note_1067946428

Ok I'll rework the changelog entry. About the Resolves line, it's kept only in
the kernel.changelog file, the rpm/dist-git should have it removed. We need to
keep the Resolves line in the redhat/kernel.changelog* since it's later used.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv4 0/8] ark: rework changelog generation process during dist-release

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980#note_1067924279

I did a simulation here, using following steps, inside a kernel-ark checkout:

```
git checkout -b tmp b4f81e09279e3c37243c35a40947bab779eabd55
git merge split-loggen # this is my branch with changes of this MR
make LOCALVERSION="" dist-release
git merge origin/master
make LOCALVERSION="" dist-release
```

And this is the result of the last release commit with just the master merge:

```
$ git show -p f4a834d051e7cf13707b42475355cd79b79d2533
commit f4a834d051e7cf13707b42475355cd79b79d2533 (HEAD -> tmp)
Author: Herton R. Krzesinski 
Date:   Wed Aug 17 12:50:22 2022 -0300

[redhat] kernel-6.0.0-0.rc1.3cc40a443a04.15

Signed-off-by: Herton R. Krzesinski 

diff --git a/Makefile.rhelver b/Makefile.rhelver
index 5e65c2fc923e..39d4dc1fe57d 100644
--- a/Makefile.rhelver
+++ b/Makefile.rhelver
@@ -12,7 +12,7 @@ RHEL_MINOR = 99
 #
 # Use this spot to avoid future merge conflicts.
 # Do not trim this comment.
-RHEL_RELEASE = 14
+RHEL_RELEASE = 15

 #
 # ZSTREAM
diff --git a/redhat/kernel.changelog-9.99 b/redhat/kernel.changelog-9.99
index 8d861210ed9e..337912c32c5f 100644
--- a/redhat/kernel.changelog-9.99
+++ b/redhat/kernel.changelog-9.99
@@ -1,3 +1,7 @@
+* Wed Aug 17 2022 Herton R. Krzesinski 
[6.0.0-0.rc1.3cc40a443a04.15]
+- 3cc40a443a04 rebase
+Resolves:
+
 * Wed Aug 17 2022 Herton R. Krzesinski  [6.0.0-0.rc1.14]
 - redhat: fix dist-release build number test (Herton R. Krzesinski)
 - redhat: fix release number bump when dist-release-changed runs (Herton R.
Krzesinski)
diff --git a/redhat/marker b/redhat/marker
index 60cf3a175526..6fa36a12cba0 100644
--- a/redhat/marker
+++ b/redhat/marker
@@ -1 +1 @@
-v6.0-rc1
+3cc40a443a04
```
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv4 0/8] ark: rework changelog generation process during dist-release

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980#note_1067764364

Yes, the solution is to verify a temporary changelog instead of
changing/reverting/changing again using the Makefile target hacks: so
genlog.py always creates a header (the rpm changelog header for the changes
with the author and timestamp) plus a Resolves line, plus a blank line. Thus
that's 3 lines and why the check uses wc, so if there is only that, there are
no changes, and we use that to check for changes. Also due ark rebases there
is also the marker check.

I see genlog.sh vs .py might be confusing, but didn't have a better name.
Moving all code into python also may not be desirable, the python part does a
good job of parsing the git changelog (it replaced the previous awk script we
had), but the other stuff seems better to be handled through shell. And yes,
genlog.py is only called by genlog.sh.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv4 0/8] ark: rework changelog generation process during dist-release

2022-08-17 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980#note_1067615447

Yes, we call genlog.sh twice, part of the check to see if there are changes
and if we need a release bump/commit. What is not called anymore twice is the
full genspec.sh/setup-source target.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCH 0/2] RHMAINTAINERS: Add new maintainers for Automotive

2022-08-16 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1989#note_1066767563

I expect once MR above you mention is merged to documentation tree, Prarit
will later sync with ark and you then don't need this MR for kernel-ark,
@prarit is this corect?
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


Re: [OS-BUILD PATCHv4 0/8] ark: rework changelog generation process during dist-release

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980#note_1065030651

Noticed a small error where marker would not updated, so did a new pushed with
it fixed.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4 6/8] redhat: use new genlog.sh script to detect changes for dist-release

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: use new genlog.sh script to detect changes for dist-release

The current dist-release process has a flaw, where both its subtargets
(dist-release-changed and dist-release-finish) depends on setup-source
target to run (so genspec.sh is run twice). To avoid having to do the
work twice and do and revert changes before dist-release-finish is
called, we now detect changes by creating the changelog using genlog.sh
first to detect if there are really changes, and do the marker check
using an if condition instead of relying on git to see if there are
changes.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -572,18 +572,18 @@ dist-rpm-baseonly: dist-sources do-rpmbuild
 # unless you know what you're doing, you don't want to use the next four ones
 dist-release-finish: setup-source
@cp $(SOURCES)/$(SPECCHANGELOG) $(REDHAT)/$(SPECCHANGELOG)
+   @echo $(MARKER) > $(REDHAT)/marker
@$(GIT) add $(REDHAT)/$(SPECCHANGELOG)
@$(GIT) add $(REDHAT)/marker
@$(GIT) commit -s ../Makefile.rhelver $(REDHAT)/marker 
$(REDHAT)/$(SPECCHANGELOG) $(SPECFILE).template -m "[redhat] $(RELEASETAG)"
-dist-release-changed: setup-source
-   @cp $(SOURCES)/$(SPECCHANGELOG) $(REDHAT)/$(SPECCHANGELOG)
-   @echo $(MARKER) > $(REDHAT)/marker
-   @# if neither changelog nor marker was updated, skip bumping a release
-   @$(GIT) update-index -q --really-refresh
-   @if $(GIT) diff-index --quiet HEAD; then \
+dist-release-changed:
+   @CLOGF="$$(mktemp)"; \
+   trap 'rm -f "$$CLOGF"' SIGHUP SIGINT SIGTERM EXIT; \
+   $(REDHAT)/genlog.sh "$$CLOGF"; \
+   if [ "$(MARKER)" == "$$(cat $(REDHAT)/marker)" ] && \
+  [ "$$(wc -l $$CLOGF | cut -d ' ' -f 1)" -le 3 ]; then \
echo "Nothing changed, skipping updates"; \
else \
-   $(GIT) checkout -- $(REDHAT)/$(SPECCHANGELOG); \
$(REDHAT)/scripts/new_release.sh; \
$(MAKE) dist-release-finish; \
fi

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4 7/8] redhat: fix release number bump when dist-release-changed runs

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix release number bump when dist-release-changed runs

Right now inside the dist-release-changed target, we run a sub make
invocation calling the dist-release-finish target, after bumping
the release number with new_release.sh. The problem is that the
BUILD number still pointed to the old pre-bump number inside the
dist-release-finish invocation, and the problem is that we don't
reload the RHEL_RELEASE number in redhat/Makefile if BUILD is already
defined. Since it looks like that check was added to avoid overriding
the BUILD number set when RHSELFTESTDATA is defined, lets change
the if check there so BUILD has the updated value when dist-release-finish
runs through dist-release-changed.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -58,7 +58,7 @@ SPECKPATCHLEVEL:=$(shell $(GIT) show $(HEAD):Makefile | sed 
-ne '/^PATCHLEVEL\ =
 SPECKSUBLEVEL:=$(shell $(GIT) show $(HEAD):Makefile | sed -ne '/^SUBLEVEL\ =\ 
/{s///;p;q}')
 SPECKEXTRAVERSION:=$(shell $(GIT) show $(HEAD):Makefile | sed -ne 
'/^EXTRAVERSION\ =\ /{s///;p;q}')
 GITID:= $(shell $(GIT) log --max-count=1 --pretty=format:%H $(HEAD))
-ifndef BUILD
+ifndef RHSELFTESTDATA
   BUILD:=$(RHEL_RELEASE)
 endif
 FLAVOR=

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4 8/8] redhat: fix dist-release build number test

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix dist-release build number test

Since previous commit, when fixing the build number bump, this test will
fail, now the build number and package release number should match as
this test was intended to check.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/self-test/2001-dist-release.bats 
b/redhat/self-test/2001-dist-release.bats
index blahblah..blahblah 100644
--- a/redhat/self-test/2001-dist-release.bats
+++ b/redhat/self-test/2001-dist-release.bats
@@ -43,7 +43,6 @@ _dist-release_test_2() {
# Strip off ...kernel-VV.PP.SS-:
pkgrelease=${title##*kernel-+([5-9]).+([0-9]).+([0-9])-}
build=$(BUILD= DIST=.fc33 make dist-dump-variables | grep -E "^BUILD=" 
| cut -d"=" -f2 | xargs)
-   ((build--))
echo "pkgrelease=$pkgrelease"
echo "build=$build"
run _dist-release_test_2

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4 5/8] redhat: move changelog addition to the spec file back into genspec.sh

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: move changelog addition to the spec file back into genspec.sh

In the next change, we want to detect at the Makefile level if we have new
changelog entries before we touch the spec file, so it makes sense to have
changelog addition being done in genspec.sh.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -36,13 +36,13 @@ git log --topo-order --no-merges -z "$GIT_NOTES" 
"$GIT_FORMAT" \
 
 if [ "$HIDE_REDHAT" = "1" ]; then
grep -v -e "^- \[redhat\]" "$clogf" |
-   sed -e 's!\[Fedora\]!!g' > "$clogf.stripped"
-   cp "$clogf.stripped" "$clogf"
+   sed -e 's!\[Fedora\]!!g' > "$clogf.tmp"
+   mv -f "$clogf.tmp" "$clogf"
 fi
 
 if [ "$HIDE_UNSUPPORTED_ARCH" = "1" ]; then
-   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.stripped"
-   cp "$clogf.stripped" "$clogf"
+   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.tmp"
+   mv -f "$clogf.tmp" "$clogf"
 fi
 
 # If the markers aren't the same then this a rebase.
@@ -60,17 +60,4 @@ if [ "$LENGTH" = 0 ]; then
echo -n > "$clogf"
 fi
 
-cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
-mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
-
-# genlog.py generates Resolves lines as well, strip these from RPM changelog
-grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
-
-test -f "$SOURCES/$SPECFILE" &&
-   sed -i -e "
-   /%%SPECCHANGELOG%%/r $clogf.stripped
-   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
-
 echo "MARKER is $MARKER"
-
-rm -f "$clogf".stripped;
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -88,9 +88,20 @@ if [ -n "$RHSELFTESTDATA" ]; then
 fi
 
 clogf=$(mktemp)
-trap 'rm -f "$clogf"' SIGHUP SIGINT SIGTERM EXIT
+trap 'rm -f "$clogf" "$clogf".stripped' SIGHUP SIGINT SIGTERM EXIT
 "${0%/*}"/genlog.sh "$clogf"
 
+cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
+mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
+
+# genlog.py generates Resolves lines as well, strip these from RPM changelog
+grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
+
+test -f "$SOURCES/$SPECFILE" &&
+   sed -i -e "
+   /%%SPECCHANGELOG%%/r $clogf.stripped
+   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
+
 if [ "$DISTRO" == "fedora" ]; then
# The tarball in the SRPM contains only the upstream sources.
 

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4 4/8] redhat: always add a rebase entry when ark merges from upstream

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: always add a rebase entry when ark merges from upstream

We have some complex logic to parse changelog entries to choose whether
we add a rebase entry or not to the changelog. Since genlog.py is adding
a Resolves entry always now, simply replace that logic with a
replacement using sed to add a rebase entry when the markers differ.
The marker check doesn't work at the moment due Makefile logic but
that will be fixed in the following changes.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -48,14 +48,9 @@ fi
 # If the markers aren't the same then this a rebase.
 # This means we need to zap entries that are already present in the changelog.
 if [ "$MARKER" != "$LAST_MARKER" ]; then
-   # awk trick to get all unique lines
-   awk '!seen[$0]++' "$SOURCES/$SPECCHANGELOG" "$clogf" > "$clogf.unique"
-   # sed trick to get the end of the changelog minus the line
-   sed -e '1,/# END OF CHANGELOG/ d' "$clogf.unique" > "$clogf.tmp"
-   # Add an explicit entry to indicate a rebase.
-   echo "" > "$clogf"
-   echo -e "- $MARKER rebase" | cat "$clogf.tmp" - >> "$clogf"
-   rm "$clogf.tmp" "$clogf.unique"
+   # genlog.py always adds a Resolves: line, thus we
+   # can insert the rebase changelog item before it
+   sed -i "s/\(^Resolves:.*\)/- $MARKER rebase\n\1/" "$clogf"
 fi
 
 # during rh-dist-git genspec runs again and generates empty changelog

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4 2/8] redhat: don't hardcode temporary changelog file

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: don't hardcode temporary changelog file

We always create a changelog file inside the sources directory, which
is temporary and always deleted at the end. Instead hardcoding the
temporary file name, create a real temporary file to be used with
genlog.sh script, given as its first argument.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -1,7 +1,7 @@
 #!/bin/bash
 
 LAST_MARKER=$(cat "${REDHAT}"/marker)
-clogf="$SOURCES/changelog"
+clogf="$1"
 # hide [redhat] entries from changelog
 HIDE_REDHAT=1;
 # hide entries for unsupported arches
@@ -70,8 +70,7 @@ fi
 # create empty file to avoid adding extra header to changelog
 LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')
 if [ "$LENGTH" = 0 ]; then
-   rm -f "$clogf"
-   touch "$clogf"
+   echo -n > "$clogf"
 fi
 
 cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
@@ -87,4 +86,4 @@ test -f "$SOURCES/$SPECFILE" &&
 
 echo "MARKER is $MARKER"
 
-rm -f "$clogf"{,.stripped};
+rm -f "$clogf".stripped;
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -87,7 +87,9 @@ if [ -n "$RHSELFTESTDATA" ]; then
exit 0
 fi
 
-"${0%/*}"/genlog.sh
+clogf=$(mktemp)
+trap 'rm -f "$clogf"' SIGHUP SIGINT SIGTERM EXIT
+"${0%/*}"/genlog.sh "$clogf"
 
 if [ "$DISTRO" == "fedora" ]; then
# The tarball in the SRPM contains only the upstream sources.

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4 3/8] redhat: drop merge ark patches hack

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: drop merge ark patches hack

Changelog generation carries a temporary hack to add a merge ark-patches
entry in the changelog. I think we can drop this today.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -58,14 +58,6 @@ if [ "$MARKER" != "$LAST_MARKER" ]; then
rm "$clogf.tmp" "$clogf.unique"
 fi
 
-# HACK temporary hack until single tree workflow
-# Don't reprint all the ark-patches again.
-if [ -n "$(git log --oneline --first-parent --grep="Merge ark patches" 
"$lasttag"..)" ]; then
-   # Throw away the clogf and just print the summary merge
-   echo "" > "$clogf"
-   echo "- Merge ark-patches" >> "$clogf"
-fi
-
 # during rh-dist-git genspec runs again and generates empty changelog
 # create empty file to avoid adding extra header to changelog
 LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4 0/8] ark: rework changelog generation process during dist-release

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980

This series try to improve the changelog generation process done
during a dist-release target run. To avoid doing hacks and having
to run setup-source/genspec.sh twice, this series splits the changelog
creation from genspec.sh and changes the Makefile to use the splitted
code.

Signed-off-by: Herton R. Krzesinski 

---
 redhat/self-test/2001-dist-release.bats |   1 -
 redhat/Makefile |  16 +++---
 redhat/genlog.sh|  63 +
 redhat/genspec.sh   |  81 +---
 4 files changed, 74 insertions(+), 87 deletions(-)
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv4 1/8] redhat: split changelog generation from genspec.sh

2022-08-15 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: split changelog generation from genspec.sh

This creates a new genlog.sh script to split out the changelog
generation code from genspec.sh. The splitted code is kept
as is.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
new file mode 100755
index blahblah..blahblah 100755
--- /dev/null
+++ b/redhat/genlog.sh
@@ -0,0 +1,90 @@
+#!/bin/bash
+
+LAST_MARKER=$(cat "${REDHAT}"/marker)
+clogf="$SOURCES/changelog"
+# hide [redhat] entries from changelog
+HIDE_REDHAT=1;
+# hide entries for unsupported arches
+HIDE_UNSUPPORTED_ARCH=1;
+# override LC_TIME to avoid date conflicts when building the srpm
+LC_TIME=
+
+GIT_FORMAT="--format=- %s (%an)%n%N%n^^^NOTES-END^^^%n%b"
+GIT_NOTES="--notes=refs/notes/${RHEL_MAJOR}.${RHEL_MINOR}*"
+
+lasttag=$(git rev-list --first-parent --grep="^\[redhat\] 
kernel-${SPECKVERSION}.${SPECKPATCHLEVEL}" --max-count=1 HEAD)
+# if we didn't find the proper tag, assume this is the first release
+if [[ -z $lasttag ]]; then
+if [[ -z ${MARKER//[0-9a-f]/} ]]; then
+# if we're doing an untagged release, just use the marker
+echo "Using $MARKER"
+lasttag=$MARKER
+else
+   lasttag=$(git describe --match="$MARKER" --abbrev=0)
+fi
+fi
+echo "Gathering new log entries since $lasttag"
+# master is expected to track mainline.
+
+cname="$(git var GIT_COMMITTER_IDENT |sed 's/>.*/>/')"
+cdate="$(LC_ALL=C date +"%a %b %d %Y")"
+cversion="[$BASEVERSION]";
+echo "* $cdate $cname $cversion" > "$clogf"
+
+git log --topo-order --no-merges -z "$GIT_NOTES" "$GIT_FORMAT" \
+   ^"${UPSTREAM}" "$lasttag".. -- ':!/redhat/rhdocs' | "${0%/*}"/genlog.py 
>> "$clogf"
+
+if [ "$HIDE_REDHAT" = "1" ]; then
+   grep -v -e "^- \[redhat\]" "$clogf" |
+   sed -e 's!\[Fedora\]!!g' > "$clogf.stripped"
+   cp "$clogf.stripped" "$clogf"
+fi
+
+if [ "$HIDE_UNSUPPORTED_ARCH" = "1" ]; then
+   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.stripped"
+   cp "$clogf.stripped" "$clogf"
+fi
+
+# If the markers aren't the same then this a rebase.
+# This means we need to zap entries that are already present in the changelog.
+if [ "$MARKER" != "$LAST_MARKER" ]; then
+   # awk trick to get all unique lines
+   awk '!seen[$0]++' "$SOURCES/$SPECCHANGELOG" "$clogf" > "$clogf.unique"
+   # sed trick to get the end of the changelog minus the line
+   sed -e '1,/# END OF CHANGELOG/ d' "$clogf.unique" > "$clogf.tmp"
+   # Add an explicit entry to indicate a rebase.
+   echo "" > "$clogf"
+   echo -e "- $MARKER rebase" | cat "$clogf.tmp" - >> "$clogf"
+   rm "$clogf.tmp" "$clogf.unique"
+fi
+
+# HACK temporary hack until single tree workflow
+# Don't reprint all the ark-patches again.
+if [ -n "$(git log --oneline --first-parent --grep="Merge ark patches" 
"$lasttag"..)" ]; then
+   # Throw away the clogf and just print the summary merge
+   echo "" > "$clogf"
+   echo "- Merge ark-patches" >> "$clogf"
+fi
+
+# during rh-dist-git genspec runs again and generates empty changelog
+# create empty file to avoid adding extra header to changelog
+LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')
+if [ "$LENGTH" = 0 ]; then
+   rm -f "$clogf"
+   touch "$clogf"
+fi
+
+cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
+mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
+
+# genlog.py generates Resolves lines as well, strip these from RPM changelog
+grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
+
+test -f "$SOURCES/$SPECFILE" &&
+   sed -i -e "
+   /%%SPECCHANGELOG%%/r $clogf.stripped
+   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
+
+echo "MARKER is $MARKER"
+
+rm -f "$clogf"{,.stripped};
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -1,16 +1,6 @@
 #!/bin/bash
 # shellcheck disable=SC2153
 
-LAST_MARKER=$(cat "${REDHAT}"/marker)
-clogf="$SOURCES/changelog"
-# hide [redhat] entries from changelog
-HIDE_REDHAT=1;
-# hide entries for unsupported arches
-HIDE_UNSUPPORTED_ARCH=1;
-# override LC_TIME to avoid date conflicts when building the srpm
-LC_TIME=
-# STAMP=$(echo $MARKER | cut -f 1 -d '-' | sed -e "s/v//"); # unused
-
 UPSTREAM=$(git rev-parse -q --verify origin/"${UPSTREAM_BRANCH}" || \
   git rev-parse -q --verify "${UPSTREAM_BRANCH}")
 
@@ -97,83 +87,7 @@ if [ -n "$RHSELFTESTDATA" ]; then
exit 0
 fi
 
-GIT_FORMAT="--format=- %s (%an)%n%N%n^^^NOTES-END^^^%n%b"
-GIT_NOTES="--notes=refs/notes/${RHEL_MAJOR}.${RHEL_MINOR}*"
-
-lasttag=$(git rev-list --first-parent --grep="^\[redhat\] 
kernel-${SPECKVERSION}.${SPECKPATCHLEVEL}" --max-count=1 HEAD)
-# if we didn't find the proper tag, assume this is the first release
-if [[ -z $lasttag ]]; then
-if [[ -z 

[OS-BUILD PATCHv3 8/8] redhat: fix dist-release build number test

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix dist-release build number test

Since previous commit, when fixing the build number bump, this test will
fail, now the build number and package release number should match as
this test was intended to check.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/self-test/2001-dist-release.bats 
b/redhat/self-test/2001-dist-release.bats
index blahblah..blahblah 100644
--- a/redhat/self-test/2001-dist-release.bats
+++ b/redhat/self-test/2001-dist-release.bats
@@ -43,7 +43,6 @@ _dist-release_test_2() {
# Strip off ...kernel-VV.PP.SS-:
pkgrelease=${title##*kernel-+([5-9]).+([0-9]).+([0-9])-}
build=$(BUILD= DIST=.fc33 make dist-dump-variables | grep -E "^BUILD=" 
| cut -d"=" -f2 | xargs)
-   ((build--))
echo "pkgrelease=$pkgrelease"
echo "build=$build"
run _dist-release_test_2

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3 7/8] redhat: fix release number bump when dist-release-changed runs

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix release number bump when dist-release-changed runs

Right now inside the dist-release-changed target, we run a sub make
invocation calling the dist-release-finish target, after bumping
the release number with new_release.sh. The problem is that the
BUILD number still pointed to the old pre-bump number inside the
dist-release-finish invocation, and the problem is that we don't
reload the RHEL_RELEASE number in redhat/Makefile if BUILD is already
defined. Since it looks like that check was added to avoid overriding
the BUILD number set when RHSELFTESTDATA is defined, lets change
the if check there so BUILD has the updated value when dist-release-finish
runs through dist-release-changed.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -58,7 +58,7 @@ SPECKPATCHLEVEL:=$(shell $(GIT) show $(HEAD):Makefile | sed 
-ne '/^PATCHLEVEL\ =
 SPECKSUBLEVEL:=$(shell $(GIT) show $(HEAD):Makefile | sed -ne '/^SUBLEVEL\ =\ 
/{s///;p;q}')
 SPECKEXTRAVERSION:=$(shell $(GIT) show $(HEAD):Makefile | sed -ne 
'/^EXTRAVERSION\ =\ /{s///;p;q}')
 GITID:= $(shell $(GIT) log --max-count=1 --pretty=format:%H $(HEAD))
-ifndef BUILD
+ifndef RHSELFTESTDATA
   BUILD:=$(RHEL_RELEASE)
 endif
 FLAVOR=

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3 6/8] redhat: use new genlog.sh script to detect changes for dist-release

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: use new genlog.sh script to detect changes for dist-release

The current dist-release process has a flaw, where both its subtargets
(dist-release-changed and dist-release-finish) depends on setup-source
target to run (so genspec.sh is run twice). To avoid having to do the
work twice and do and revert changes before dist-release-finish is
called, we now detect changes by creating the changelog using genlog.sh
first to detect if there are really changes, and do the marker check
using an if condition instead of relying on git to see if there are
changes.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -575,15 +575,14 @@ dist-release-finish: setup-source
@$(GIT) add $(REDHAT)/$(SPECCHANGELOG)
@$(GIT) add $(REDHAT)/marker
@$(GIT) commit -s ../Makefile.rhelver $(REDHAT)/marker 
$(REDHAT)/$(SPECCHANGELOG) $(SPECFILE).template -m "[redhat] $(RELEASETAG)"
-dist-release-changed: setup-source
-   @cp $(SOURCES)/$(SPECCHANGELOG) $(REDHAT)/$(SPECCHANGELOG)
-   @echo $(MARKER) > $(REDHAT)/marker
-   @# if neither changelog nor marker was updated, skip bumping a release
-   @$(GIT) update-index -q --really-refresh
-   @if $(GIT) diff-index --quiet HEAD; then \
+dist-release-changed:
+   @CLOGF="$$(mktemp)"; \
+   trap 'rm -f "$$CLOGF"' SIGHUP SIGINT SIGTERM EXIT; \
+   $(REDHAT)/genlog.sh "$$CLOGF"; \
+   if [ "$(MARKER)" == "$$(cat $(REDHAT)/marker)" ] && \
+  [ "$$(wc -l $$CLOGF | cut -d ' ' -f 1)" -le 3 ]; then \
echo "Nothing changed, skipping updates"; \
else \
-   $(GIT) checkout -- $(REDHAT)/$(SPECCHANGELOG); \
$(REDHAT)/scripts/new_release.sh; \
$(MAKE) dist-release-finish; \
fi

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3 5/8] redhat: move changelog addition to the spec file back into genspec.sh

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: move changelog addition to the spec file back into genspec.sh

In the next change, we want to detect at the Makefile level if we have new
changelog entries before we touch the spec file, so it makes sense to have
changelog addition being done in genspec.sh.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -36,13 +36,13 @@ git log --topo-order --no-merges -z "$GIT_NOTES" 
"$GIT_FORMAT" \
 
 if [ "$HIDE_REDHAT" = "1" ]; then
grep -v -e "^- \[redhat\]" "$clogf" |
-   sed -e 's!\[Fedora\]!!g' > "$clogf.stripped"
-   cp "$clogf.stripped" "$clogf"
+   sed -e 's!\[Fedora\]!!g' > "$clogf.tmp"
+   mv -f "$clogf.tmp" "$clogf"
 fi
 
 if [ "$HIDE_UNSUPPORTED_ARCH" = "1" ]; then
-   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.stripped"
-   cp "$clogf.stripped" "$clogf"
+   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.tmp"
+   mv -f "$clogf.tmp" "$clogf"
 fi
 
 # If the markers aren't the same then this a rebase.
@@ -60,17 +60,4 @@ if [ "$LENGTH" = 0 ]; then
echo -n > "$clogf"
 fi
 
-cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
-mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
-
-# genlog.py generates Resolves lines as well, strip these from RPM changelog
-grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
-
-test -f "$SOURCES/$SPECFILE" &&
-   sed -i -e "
-   /%%SPECCHANGELOG%%/r $clogf.stripped
-   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
-
 echo "MARKER is $MARKER"
-
-rm -f "$clogf".stripped;
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -88,9 +88,20 @@ if [ -n "$RHSELFTESTDATA" ]; then
 fi
 
 clogf=$(mktemp)
-trap 'rm -f "$clogf"' SIGHUP SIGINT SIGTERM EXIT
+trap 'rm -f "$clogf" "$clogf".stripped' SIGHUP SIGINT SIGTERM EXIT
 "${0%/*}"/genlog.sh "$clogf"
 
+cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
+mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
+
+# genlog.py generates Resolves lines as well, strip these from RPM changelog
+grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
+
+test -f "$SOURCES/$SPECFILE" &&
+   sed -i -e "
+   /%%SPECCHANGELOG%%/r $clogf.stripped
+   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
+
 if [ "$DISTRO" == "fedora" ]; then
# The tarball in the SRPM contains only the upstream sources.
 

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3 1/8] redhat: split changelog generation from genspec.sh

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: split changelog generation from genspec.sh

This creates a new genlog.sh script to split out the changelog
generation code from genspec.sh. The splitted code is kept
as is.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
new file mode 100755
index blahblah..blahblah 100755
--- /dev/null
+++ b/redhat/genlog.sh
@@ -0,0 +1,90 @@
+#!/bin/bash
+
+LAST_MARKER=$(cat "${REDHAT}"/marker)
+clogf="$SOURCES/changelog"
+# hide [redhat] entries from changelog
+HIDE_REDHAT=1;
+# hide entries for unsupported arches
+HIDE_UNSUPPORTED_ARCH=1;
+# override LC_TIME to avoid date conflicts when building the srpm
+LC_TIME=
+
+GIT_FORMAT="--format=- %s (%an)%n%N%n^^^NOTES-END^^^%n%b"
+GIT_NOTES="--notes=refs/notes/${RHEL_MAJOR}.${RHEL_MINOR}*"
+
+lasttag=$(git rev-list --first-parent --grep="^\[redhat\] 
kernel-${SPECKVERSION}.${SPECKPATCHLEVEL}" --max-count=1 HEAD)
+# if we didn't find the proper tag, assume this is the first release
+if [[ -z $lasttag ]]; then
+if [[ -z ${MARKER//[0-9a-f]/} ]]; then
+# if we're doing an untagged release, just use the marker
+echo "Using $MARKER"
+lasttag=$MARKER
+else
+   lasttag=$(git describe --match="$MARKER" --abbrev=0)
+fi
+fi
+echo "Gathering new log entries since $lasttag"
+# master is expected to track mainline.
+
+cname="$(git var GIT_COMMITTER_IDENT |sed 's/>.*/>/')"
+cdate="$(LC_ALL=C date +"%a %b %d %Y")"
+cversion="[$BASEVERSION]";
+echo "* $cdate $cname $cversion" > "$clogf"
+
+git log --topo-order --no-merges -z "$GIT_NOTES" "$GIT_FORMAT" \
+   ^"${UPSTREAM}" "$lasttag".. -- ':!/redhat/rhdocs' | "${0%/*}"/genlog.py 
>> "$clogf"
+
+if [ "$HIDE_REDHAT" = "1" ]; then
+   grep -v -e "^- \[redhat\]" "$clogf" |
+   sed -e 's!\[Fedora\]!!g' > "$clogf.stripped"
+   cp "$clogf.stripped" "$clogf"
+fi
+
+if [ "$HIDE_UNSUPPORTED_ARCH" = "1" ]; then
+   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.stripped"
+   cp "$clogf.stripped" "$clogf"
+fi
+
+# If the markers aren't the same then this a rebase.
+# This means we need to zap entries that are already present in the changelog.
+if [ "$MARKER" != "$LAST_MARKER" ]; then
+   # awk trick to get all unique lines
+   awk '!seen[$0]++' "$SOURCES/$SPECCHANGELOG" "$clogf" > "$clogf.unique"
+   # sed trick to get the end of the changelog minus the line
+   sed -e '1,/# END OF CHANGELOG/ d' "$clogf.unique" > "$clogf.tmp"
+   # Add an explicit entry to indicate a rebase.
+   echo "" > "$clogf"
+   echo -e "- $MARKER rebase" | cat "$clogf.tmp" - >> "$clogf"
+   rm "$clogf.tmp" "$clogf.unique"
+fi
+
+# HACK temporary hack until single tree workflow
+# Don't reprint all the ark-patches again.
+if [ -n "$(git log --oneline --first-parent --grep="Merge ark patches" 
"$lasttag"..)" ]; then
+   # Throw away the clogf and just print the summary merge
+   echo "" > "$clogf"
+   echo "- Merge ark-patches" >> "$clogf"
+fi
+
+# during rh-dist-git genspec runs again and generates empty changelog
+# create empty file to avoid adding extra header to changelog
+LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')
+if [ "$LENGTH" = 0 ]; then
+   rm -f "$clogf"
+   touch "$clogf"
+fi
+
+cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
+mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
+
+# genlog.py generates Resolves lines as well, strip these from RPM changelog
+grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
+
+test -f "$SOURCES/$SPECFILE" &&
+   sed -i -e "
+   /%%SPECCHANGELOG%%/r $clogf.stripped
+   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
+
+echo "MARKER is $MARKER"
+
+rm -f "$clogf"{,.stripped};
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -1,16 +1,6 @@
 #!/bin/bash
 # shellcheck disable=SC2153
 
-LAST_MARKER=$(cat "${REDHAT}"/marker)
-clogf="$SOURCES/changelog"
-# hide [redhat] entries from changelog
-HIDE_REDHAT=1;
-# hide entries for unsupported arches
-HIDE_UNSUPPORTED_ARCH=1;
-# override LC_TIME to avoid date conflicts when building the srpm
-LC_TIME=
-# STAMP=$(echo $MARKER | cut -f 1 -d '-' | sed -e "s/v//"); # unused
-
 UPSTREAM=$(git rev-parse -q --verify origin/"${UPSTREAM_BRANCH}" || \
   git rev-parse -q --verify "${UPSTREAM_BRANCH}")
 
@@ -97,83 +87,7 @@ if [ -n "$RHSELFTESTDATA" ]; then
exit 0
 fi
 
-GIT_FORMAT="--format=- %s (%an)%n%N%n^^^NOTES-END^^^%n%b"
-GIT_NOTES="--notes=refs/notes/${RHEL_MAJOR}.${RHEL_MINOR}*"
-
-lasttag=$(git rev-list --first-parent --grep="^\[redhat\] 
kernel-${SPECKVERSION}.${SPECKPATCHLEVEL}" --max-count=1 HEAD)
-# if we didn't find the proper tag, assume this is the first release
-if [[ -z $lasttag ]]; then
-if [[ -z 

[OS-BUILD PATCHv3 4/8] redhat: always add a rebase entry when ark merges from upstream

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: always add a rebase entry when ark merges from upstream

We have some complex logic to parse changelog entries to choose whether
we add a rebase entry or not to the changelog. Since genlog.py is adding
a Resolves entry always now, simply replace that logic with a
replacement using sed to add a rebase entry when the markers differ.
The marker check doesn't work at the moment due Makefile logic but
that will be fixed in the following changes.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -48,14 +48,9 @@ fi
 # If the markers aren't the same then this a rebase.
 # This means we need to zap entries that are already present in the changelog.
 if [ "$MARKER" != "$LAST_MARKER" ]; then
-   # awk trick to get all unique lines
-   awk '!seen[$0]++' "$SOURCES/$SPECCHANGELOG" "$clogf" > "$clogf.unique"
-   # sed trick to get the end of the changelog minus the line
-   sed -e '1,/# END OF CHANGELOG/ d' "$clogf.unique" > "$clogf.tmp"
-   # Add an explicit entry to indicate a rebase.
-   echo "" > "$clogf"
-   echo -e "- $MARKER rebase" | cat "$clogf.tmp" - >> "$clogf"
-   rm "$clogf.tmp" "$clogf.unique"
+   # genlog.py always adds a Resolves: line, thus we
+   # can insert the rebase changelog item before it
+   sed -i "s/\(^Resolves:.*\)/- $MARKER rebase\n\1/" "$clogf"
 fi
 
 # during rh-dist-git genspec runs again and generates empty changelog

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3 3/8] redhat: drop merge ark patches hack

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: drop merge ark patches hack

Changelog generation carries a temporary hack to add a merge ark-patches
entry in the changelog. I think we can drop this today.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -58,14 +58,6 @@ if [ "$MARKER" != "$LAST_MARKER" ]; then
rm "$clogf.tmp" "$clogf.unique"
 fi
 
-# HACK temporary hack until single tree workflow
-# Don't reprint all the ark-patches again.
-if [ -n "$(git log --oneline --first-parent --grep="Merge ark patches" 
"$lasttag"..)" ]; then
-   # Throw away the clogf and just print the summary merge
-   echo "" > "$clogf"
-   echo "- Merge ark-patches" >> "$clogf"
-fi
-
 # during rh-dist-git genspec runs again and generates empty changelog
 # create empty file to avoid adding extra header to changelog
 LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3 2/8] redhat: don't hardcode temporary changelog file

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: don't hardcode temporary changelog file

We always create a changelog file inside the sources directory, which
is temporary and always deleted at the end. Instead hardcoding the
temporary file name, create a real temporary file to be used with
genlog.sh script, given as its first argument.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -1,7 +1,7 @@
 #!/bin/bash
 
 LAST_MARKER=$(cat "${REDHAT}"/marker)
-clogf="$SOURCES/changelog"
+clogf="$1"
 # hide [redhat] entries from changelog
 HIDE_REDHAT=1;
 # hide entries for unsupported arches
@@ -70,8 +70,7 @@ fi
 # create empty file to avoid adding extra header to changelog
 LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')
 if [ "$LENGTH" = 0 ]; then
-   rm -f "$clogf"
-   touch "$clogf"
+   echo -n > "$clogf"
 fi
 
 cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
@@ -87,4 +86,4 @@ test -f "$SOURCES/$SPECFILE" &&
 
 echo "MARKER is $MARKER"
 
-rm -f "$clogf"{,.stripped};
+rm -f "$clogf".stripped;
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -87,7 +87,9 @@ if [ -n "$RHSELFTESTDATA" ]; then
exit 0
 fi
 
-"${0%/*}"/genlog.sh
+clogf=$(mktemp)
+trap 'rm -f "$clogf"' SIGHUP SIGINT SIGTERM EXIT
+"${0%/*}"/genlog.sh "$clogf"
 
 if [ "$DISTRO" == "fedora" ]; then
# The tarball in the SRPM contains only the upstream sources.

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv3 0/8] ark: rework changelog generation process during dist-release

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski on gitlab.com
Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980

This series try to improve the changelog generation process done
during a dist-release target run. To avoid doing hacks and having
to run setup-source/genspec.sh twice, this series splits the changelog
creation from genspec.sh and changes the Makefile to use the splitted
code.

Signed-off-by: Herton R. Krzesinski 

---
 redhat/self-test/2001-dist-release.bats |   1 -
 redhat/Makefile |  15 ++---
 redhat/genlog.sh|  63 +
 redhat/genspec.sh   |  81 +---
 4 files changed, 73 insertions(+), 87 deletions(-)
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2 4/8] redhat: always add a rebase entry when ark merges from upstream

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: always add a rebase entry when ark merges from upstream

We have some complex logic to parse changelog entries to choose whether
we add a rebase entry or not to the changelog. Since genlog.py is adding
a Resolves entry always now, simply replace that logic with a
replacement using sed to add a rebase entry when the markers differ.
The marker check doesn't work at the moment due Makefile logic but
that will be fixed in the following changes.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -48,14 +48,9 @@ fi
 # If the markers aren't the same then this a rebase.
 # This means we need to zap entries that are already present in the changelog.
 if [ "$MARKER" != "$LAST_MARKER" ]; then
-   # awk trick to get all unique lines
-   awk '!seen[$0]++' "$SOURCES/$SPECCHANGELOG" "$clogf" > "$clogf.unique"
-   # sed trick to get the end of the changelog minus the line
-   sed -e '1,/# END OF CHANGELOG/ d' "$clogf.unique" > "$clogf.tmp"
-   # Add an explicit entry to indicate a rebase.
-   echo "" > "$clogf"
-   echo -e "- $MARKER rebase" | cat "$clogf.tmp" - >> "$clogf"
-   rm "$clogf.tmp" "$clogf.unique"
+   # genlog.py always adds a Resolves: line, thus we
+   # can insert the rebase changelog item before it
+   sed -i "s/\(^Resolves:.*\)/- $MARKER rebase\n\1/" "$clogf"
 fi
 
 # during rh-dist-git genspec runs again and generates empty changelog

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2 8/8] redhat: fix dist-release build number test

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix dist-release build number test

Since previous commit, when fixing the build number bump, this test will
fail, now the build number and package release number should match as
this test was intended to check.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/self-test/2001-dist-release.bats 
b/redhat/self-test/2001-dist-release.bats
index blahblah..blahblah 100644
--- a/redhat/self-test/2001-dist-release.bats
+++ b/redhat/self-test/2001-dist-release.bats
@@ -43,7 +43,6 @@ _dist-release_test_2() {
# Strip off ...kernel-VV.PP.SS-:
pkgrelease=${title##*kernel-+([5-9]).+([0-9]).+([0-9])-}
build=$(BUILD= DIST=.fc33 make dist-dump-variables | grep -E "^BUILD=" 
| cut -d"=" -f2 | xargs)
-   ((build--))
echo "pkgrelease=$pkgrelease"
echo "build=$build"
run _dist-release_test_2

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2 7/8] redhat: fix release number bump when dist-release-changed runs

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: fix release number bump when dist-release-changed runs

Right now inside the dist-release-changed target, we run a sub make
invocation calling the dist-release-finish target, after bumping
the release number with new_release.sh. The problem is that the
BUILD number still pointed to the old pre-bump number inside the
dist-release-finish invocation, and the problem is that we don't
reload the RHEL_RELEASE number in redhat/Makefile if BUILD is already
defined. Since it looks like that check was added to avoid overriding
the BUILD number set when RHSELFTESTDATA is defined, lets change
the if check there so BUILD has the updated value when dist-release-finish
runs through dist-release-changed.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -58,7 +58,7 @@ SPECKPATCHLEVEL:=$(shell $(GIT) show $(HEAD):Makefile | sed 
-ne '/^PATCHLEVEL\ =
 SPECKSUBLEVEL:=$(shell $(GIT) show $(HEAD):Makefile | sed -ne '/^SUBLEVEL\ =\ 
/{s///;p;q}')
 SPECKEXTRAVERSION:=$(shell $(GIT) show $(HEAD):Makefile | sed -ne 
'/^EXTRAVERSION\ =\ /{s///;p;q}')
 GITID:= $(shell $(GIT) log --max-count=1 --pretty=format:%H $(HEAD))
-ifndef BUILD
+ifndef RHSELFTESTDATA
   BUILD:=$(RHEL_RELEASE)
 endif
 FLAVOR=

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2 6/8] redhat: use new genlog.sh script to detect changes for dist-release

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: use new genlog.sh script to detect changes for dist-release

The current dist-release process has a flaw, where both its subtargets
(dist-release-changed and dist-release-finish) depends on setup-source
target to run (so genspec.sh is run twice). To avoid having to do the
work twice and do and revert changes before dist-release-finish is
called, we now detect changes by creating the changelog using genlog.sh
first to detect if there are really changes, and do the marker check
using an if condition instead of relying on git to see if there are
changes.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/Makefile b/redhat/Makefile
index blahblah..blahblah 100644
--- a/redhat/Makefile
+++ b/redhat/Makefile
@@ -575,15 +575,15 @@ dist-release-finish: setup-source
@$(GIT) add $(REDHAT)/$(SPECCHANGELOG)
@$(GIT) add $(REDHAT)/marker
@$(GIT) commit -s ../Makefile.rhelver $(REDHAT)/marker 
$(REDHAT)/$(SPECCHANGELOG) $(SPECFILE).template -m "[redhat] $(RELEASETAG)"
-dist-release-changed: setup-source
-   @cp $(SOURCES)/$(SPECCHANGELOG) $(REDHAT)/$(SPECCHANGELOG)
-   @echo $(MARKER) > $(REDHAT)/marker
-   @# if neither changelog nor marker was updated, skip bumping a release
-   @$(GIT) update-index -q --really-refresh
-   @if $(GIT) diff-index --quiet HEAD; then \
+dist-release-changed:
+   @CLOGF="$$(mktemp)"; \
+   echo $$CLOGF; \
+   trap 'rm -f "$$CLOGF"' SIGHUP SIGINT SIGTERM EXIT; \
+   $(REDHAT)/genlog.sh "$$CLOGF"; \
+   if [ "$(MARKER)" == "$$(cat $(REDHAT)/marker)" ] && \
+  [ "$$(wc -l $$CLOGF | cut -d ' ' -f 1)" -le 3 ]; then \
echo "Nothing changed, skipping updates"; \
else \
-   $(GIT) checkout -- $(REDHAT)/$(SPECCHANGELOG); \
$(REDHAT)/scripts/new_release.sh; \
$(MAKE) dist-release-finish; \
fi

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2 3/8] redhat: drop merge ark patches hack

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: drop merge ark patches hack

Changelog generation carries a temporary hack to add a merge ark-patches
entry in the changelog. I think we can drop this today.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -58,14 +58,6 @@ if [ "$MARKER" != "$LAST_MARKER" ]; then
rm "$clogf.tmp" "$clogf.unique"
 fi
 
-# HACK temporary hack until single tree workflow
-# Don't reprint all the ark-patches again.
-if [ -n "$(git log --oneline --first-parent --grep="Merge ark patches" 
"$lasttag"..)" ]; then
-   # Throw away the clogf and just print the summary merge
-   echo "" > "$clogf"
-   echo "- Merge ark-patches" >> "$clogf"
-fi
-
 # during rh-dist-git genspec runs again and generates empty changelog
 # create empty file to avoid adding extra header to changelog
 LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2 5/8] redhat: move changelog addition to the spec file back into genspec.sh

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: move changelog addition to the spec file back into genspec.sh

In the next change, we want to detect at the Makefile level if we have new
changelog entries before we touch the spec file, so it makes sense to have
changelog addition being done in genspec.sh.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -36,13 +36,13 @@ git log --topo-order --no-merges -z "$GIT_NOTES" 
"$GIT_FORMAT" \
 
 if [ "$HIDE_REDHAT" = "1" ]; then
grep -v -e "^- \[redhat\]" "$clogf" |
-   sed -e 's!\[Fedora\]!!g' > "$clogf.stripped"
-   cp "$clogf.stripped" "$clogf"
+   sed -e 's!\[Fedora\]!!g' > "$clogf.tmp"
+   mv -f "$clogf.tmp" "$clogf"
 fi
 
 if [ "$HIDE_UNSUPPORTED_ARCH" = "1" ]; then
-   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.stripped"
-   cp "$clogf.stripped" "$clogf"
+   grep -E -v "^- 
\[(alpha|arc|arm|avr32|blackfin|c6x|cris|frv|h8300|hexagon|ia64|m32r|m68k|metag|microblaze|mips|mn10300|openrisc|parisc|score|sh|sparc|tile|um|unicore32|xtensa)\]"
 "$clogf" > "$clogf.tmp"
+   mv -f "$clogf.tmp" "$clogf"
 fi
 
 # If the markers aren't the same then this a rebase.
@@ -60,17 +60,4 @@ if [ "$LENGTH" = 0 ]; then
echo -n > "$clogf"
 fi
 
-cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
-mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
-
-# genlog.py generates Resolves lines as well, strip these from RPM changelog
-grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
-
-test -f "$SOURCES/$SPECFILE" &&
-   sed -i -e "
-   /%%SPECCHANGELOG%%/r $clogf.stripped
-   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
-
 echo "MARKER is $MARKER"
-
-rm -f "$clogf".stripped;
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -88,9 +88,20 @@ if [ -n "$RHSELFTESTDATA" ]; then
 fi
 
 clogf=$(mktemp)
-trap 'rm -f "$clogf"' SIGHUP SIGINT SIGTERM EXIT
+trap 'rm -f "$clogf" "$clogf".stripped' SIGHUP SIGINT SIGTERM EXIT
 "${0%/*}"/genlog.sh "$clogf"
 
+cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
+mv -f "$clogf.full" "$SOURCES/$SPECCHANGELOG"
+
+# genlog.py generates Resolves lines as well, strip these from RPM changelog
+grep -v -e "^Resolves: " "$SOURCES/$SPECCHANGELOG" > "$clogf".stripped
+
+test -f "$SOURCES/$SPECFILE" &&
+   sed -i -e "
+   /%%SPECCHANGELOG%%/r $clogf.stripped
+   /%%SPECCHANGELOG%%/d" "$SOURCES/$SPECFILE"
+
 if [ "$DISTRO" == "fedora" ]; then
# The tarball in the SRPM contains only the upstream sources.
 

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


[OS-BUILD PATCHv2 2/8] redhat: don't hardcode temporary changelog file

2022-08-10 Thread Herton R. Krzesinski (via Email Bridge)
From: Herton R. Krzesinski 

redhat: don't hardcode temporary changelog file

We always create a changelog file inside the sources directory, which
is temporary and always deleted at the end. Instead hardcoding the
temporary file name, create a real temporary file to be used with
genlog.sh script, given as its first argument.

Signed-off-by: Herton R. Krzesinski 

diff --git a/redhat/genlog.sh b/redhat/genlog.sh
index blahblah..blahblah 100755
--- a/redhat/genlog.sh
+++ b/redhat/genlog.sh
@@ -1,7 +1,7 @@
 #!/bin/bash
 
 LAST_MARKER=$(cat "${REDHAT}"/marker)
-clogf="$SOURCES/changelog"
+clogf="$1"
 # hide [redhat] entries from changelog
 HIDE_REDHAT=1;
 # hide entries for unsupported arches
@@ -70,8 +70,7 @@ fi
 # create empty file to avoid adding extra header to changelog
 LENGTH=$(grep -c "^-" "$clogf" | awk '{print $1}')
 if [ "$LENGTH" = 0 ]; then
-   rm -f "$clogf"
-   touch "$clogf"
+   echo -n > "$clogf"
 fi
 
 cat "$clogf" "$SOURCES/$SPECCHANGELOG" > "$clogf.full"
@@ -87,4 +86,4 @@ test -f "$SOURCES/$SPECFILE" &&
 
 echo "MARKER is $MARKER"
 
-rm -f "$clogf"{,.stripped};
+rm -f "$clogf".stripped;
diff --git a/redhat/genspec.sh b/redhat/genspec.sh
index blahblah..blahblah 100755
--- a/redhat/genspec.sh
+++ b/redhat/genspec.sh
@@ -87,7 +87,9 @@ if [ -n "$RHSELFTESTDATA" ]; then
exit 0
 fi
 
-"${0%/*}"/genlog.sh
+clogf=$(mktemp)
+trap 'rm -f "$clogf"' SIGHUP SIGINT SIGTERM EXIT
+"${0%/*}"/genlog.sh "$clogf"
 
 if [ "$DISTRO" == "fedora" ]; then
# The tarball in the SRPM contains only the upstream sources.

--
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue


  1   2   3   4   >