- Compass-ci - mailweb.openeuler.org

[Compass-ci]valgrind包申请添加工程配置中%gcc_secure_exclude 选项
by 王凯 16 Sep '25

16 Sep '25

一.背景 25.09转测发现valgrind 存在安全编译选项 SP 不满足，参考其他社区添加-fstack-protector-strong编译选项修复,会报错undefined reference to `__stack_chk_fail'，排查是安装gcc_secure 导致冲突，验证不使用gcc_secure可以正常构建通过 issue : https://gitee.com/src-openeuler/valgrind/issues/ICXUJZ 报错日志：https://eulermaker.compass-ci.openeuler.openatom.cn/api/result/rpmbuild/2025-09-16/dc-64g/openeuler-25.09-amd64-x86_64/x86_64-common-gitee.com-src-openeuler-val/cbs.10473177/dmesg 2025-09-16 10:31:33 gcc -B /home/lkp/rpmbuild/BUILD/valgrind-3.25.1/shared/libgcc/ -m64 -O2 -g -Wall -Wmissing-prototypes -Wshadow -Wpointer-arith -Wstrict-prototypes -Wmissing-declarations -Wno-unused-result -Wcast-align -Wcast-qual -Wwrite-strings -Wempty-body -Wformat -Wformat-signedness -Wformat-security -Wignored-qualifiers -Wmissing-parameter-type -Wlogical-op -Wenum-conversion -Wimplicit-fallthrough=2 -Wold-style-declaration -finline-functions -fno-strict-aliasing -fno-builtin -O -g -fno-omit-frame-pointer -fno-strict-aliasing -fpic -fno-builtin -fno-ipa-icf -O2 -g -grecord-gcc-switches -pipe -Wall -Werror=format-security -Wp,-U_FORTIFY_SOURCE,-D_FORTIFY_SOURCE=2 -Wp,-D_GLIBCXX_ASSERTIONS -specs=/usr/lib/rpm/generic-hardened-cc1 -m64 -mtune=generic -fasynchronous-unwind-tables -fstack-clash-protection -nodefaultlibs -shared -Wl,-z,interpose,-z,initfirst -m64 -Wl,--whole-archive ../coregrind/libreplacemalloc_toolpreload-amd64-linux.a -Wl,--no-whole-archive -Wl,-z,relro -Wl,--as-needed -specs=/usr/lib/rpm/generic-hardened-ld -specs=/usr/lib/rpm/openEuler/openEuler-package-notes -o vgpreload_memcheck-amd64-linux.so vgpreload_memcheck_amd64_linux_so-mc_replace_strmem.o 2025-09-16 10:31:33 make[3]: Leaving directory '/home/lkp/rpmbuild/BUILD/valgrind-3.25.1/memcheck' 2025-09-16 10:31:36 /usr/bin/ld: /tmp/ccHMRr6s.ltrans6.ltrans.o: in function `expandInput': 2025-09-16 10:31:36 /home/lkp/rpmbuild/BUILD/valgrind-3.25.1/memcheck/../coregrind/m_errormgr.c:1993:(.text+0x59b): undefined reference to `__stack_chk_fail' 2025-09-16 10:31:36 /usr/bin/ld: /tmp/ccHMRr6s.ltrans6.ltrans.o: in function `vgPlain_parse_enum_set': 2025-09-16 10:31:36 /home/lkp/rpmbuild/BUILD/valgrind-3.25.1/memcheck/../coregrind/m_libcbase.c:637:(.text+0x1474): undefined reference to `__stack_chk_fail' 2025-09-16 10:31:36 /usr/bin/ld: /tmp/ccHMRr6s.ltrans6.ltrans.o: in function `bm_qsort.lto_priv.0': 2025-09-16 10:31:36 /home/lkp/rpmbuild/BUILD/valgrind-3.25.1/memcheck/../coregrind/m_libcbase.c:933:(.text+0x1ac9): undefined reference to `__stack_chk_fail' 2025-09-16 10:31:36 /usr/bin/ld: /tmp/ccHMRr6s.ltrans6.ltrans.o: in function `vgPlain_stat': 2025-09-16 10:31:36 /home/lkp/rpmbuild/BUILD/valgrind-3.25.1/memcheck/../coregrind/m_libcfile.c:605:(.text+0x2626): undefined reference to `__stack_chk_fail' 2025-09-16 10:31:36 /usr/bin/ld: /tmp/ccHMRr6s.ltrans6.ltrans.o: in function `vgPlain_fstat': 2025-09-16 10:31:36 /home/lkp/rpmbuild/BUILD/valgrind-3.25.1/memcheck/../coregrind/m_libcfile.c:689:(.text+0x2888): undefined reference to `__stack_chk_fail' 2025-09-16 10:31:36 /usr/bin/ld: /tmp/ccHMRr6s.ltrans6.ltrans.o:/home/lkp/rpmbuild/BUILD/valgrind-3.25.1/memcheck/../coregrind/m_libcfile.c:1133: more undefined references to `__stack_chk_fail' follow 2025-09-16 10:31:36 collect2: error: ld returned 1 exit status 2025-09-16 10:31:36 make[3]: *** [Makefile:1195: memcheck-amd64-linux] Error 1 2025-09-16 10:31:36 make[3]: Entering directory '/home/lkp/rpmbuild/BUILD/valgrind-3.25.1/memcheck' 2025-09-16 10:31:36 ../coregrind/link_tool_exe_linux 0x58000000 gcc -B /home/lkp/rpmbuild/BUILD/valgrind-3.25.1/shared/libgcc/ -g -grecord-gcc-switches -pipe -Wall -Werror=format-security -Wp,-U_FORTIFY_SOURCE,-D_FORTIFY_SOURCE=2 -Wp,-D_GLIBCXX_ASSERTIONS -specs=/usr/lib/rpm/generic-hardened-cc1 -m64 -mtune=generic -fasynchronous-unwind-tables -fstack-clash-protection -Wl,-z,relro -Wl,--as-needed -specs=/usr/lib/rpm/generic-hardened-ld -specs=/usr/lib/rpm/openEuler/openEuler-package-notes -o memcheck-amd64-linux -flto -fuse-linker-plugin -m64 -O2 -g -Wall -Wmissing-prototypes -Wshadow -Wpointer-arith -Wstrict-prototypes -Wmissing-declarations -Wno-unused-result -Wcast-align -Wcast-qual -Wwrite-strings -Wempty-body -Wformat -Wformat-signedness -Wformat-security -Wignored-qualifiers -Wmissing-parameter-type -Wlogical-op -Wenum-conversion -Wimplicit-fallthrough=2 -Wold-style-declaration -finline-functions -fno-strict-aliasing -fno-builtin -fomit-frame-pointer -O2 -static -nodefaultlibs -nostartfiles -u _start -m64 memcheck_amd64_linux-mc_leakcheck.o memcheck_amd64_linux-mc_malloc_wrappers.o memcheck_amd64_linux-mc_main.o memcheck_amd64_linux-mc_main_asm.o memcheck_amd64_linux-mc_translate.o memcheck_amd64_linux-mc_machine.o memcheck_amd64_linux-mc_errors.o ../coregrind/libcoregrind-amd64-linux.a ../VEX/libvex-amd64-linux.a -lgcc ../coregrind/libgcc-sup-amd64-linux.a 2025-09-16 10:31:36 make[3]: Leaving directory '/home/lkp/rpmbuild/BUILD/valgrind-3.25.1/memcheck' 2025-09-16 10:31:36 make[2]: *** [Makefile:1506: all-recursive] Error 1 2025-09-16 10:31:36 make[1]: *** [Makefile:942: all-recursive] Error 1 2025-09-16 10:31:36 make: *** [Makefile:805: all] Error 2 2025-09-16 10:31:36 error: Bad exit status from /var/tmp/rpm-tmp.BhN0wy (%build) 2025-09-16 10:31:36 Bad exit status from /var/tmp/rpm-tmp.BhN0wy (%build) 二.期望添加的配置在openEuler-25.09:everything、openEuler-master:everything、openEuler-24.03-LTS-Next:everything工程的%gcc_secure_exclude添加valgrind.spec配置三.涉及工程范围 Eulermaker中master、25.09、24.03-LTS系列分支涉及工程及后续工程配置

1 0

kata-containers 和 kata-containers-go 申请root权限构建
by xiadanni 05 Sep '25

05 Sep '25

一、背景 openEuler-24.03-LTS-SP1和openEuler-25.09分支的kata-containers在EBS工程构建时因权限不足导致日志存在报错信息： cp拷贝目录和文件时因为权限问题失败 [cid:image001.png@01DC1BF9.27512BF0] 另外，openEuler-24.03-LTS-Next、master分支的kata-containers-go在EBS工程构建时也存在上述权限不足的报错信息；经个人工程验证，加上root权限能解决上述问题，且当前master、openEuler-20.03-LTS-SP4、openEuler-22.03-LTS-SP3、openEuler-22.03-LTS-SP4、openEuler-24.03-LTS、openEuler-24.03-LTS-Next、openEuler-24.03-LTS-SP2分支的kata-containers构建也存在类似问题。二、期望添加的配置 1.在master、openEuler-25.09、openEuler-20.03-LTS-SP4、openEuler-22.03-LTS-SP3、openEuler-22.03-LTS-SP4、openEuler-24.03-LTS、openEuler-24.03-LTS-Next、openEuler-24.03-LTS-SP1、openEuler-24.03-LTS-SP2分支的everything工程的root权限配置中添加kata-containers 2.在openEuler-24.03-LTS-Next分支的everything工程和master分支的factory工程root权限配置中添加kata-containers-go 三、涉及工程范围 1.Eulermaker中master、openEuler-25.09、openEuler-20.03-LTS-SP4、openEuler-22.03-LTS-SP3、openEuler-22.03-LTS-SP4、openEuler-24.03-LTS、openEuler-24.03-LTS-Next、openEuler-24.03-LTS-SP1、openEuler-24.03-LTS-SP2分支kata-containers包编译工程 2. Eulermaker中master、openEuler-24.03-LTS-Next分支kata-containers-go包编译工程

1 0

[PATCH] lib/stats.rb: forcely show more stats if --no-hide-noises is set in sbin/compare
by Zheng Zengkai 06 Aug '25

06 Aug '25

Reuse the "--no-hide-noises" option and allow to bypass some checks for stats and show more stats comparisions if "--no-hide-noises" option is given when using lkp compare. It is useful when tester wants to see more stats comparisions between different runs, for example when the change ratio is less than 10%. Before the patch: [root@localhost /]# lkp compare -twa -d commit 5.10.0-4a4dbce046f7+ 5.10.0-47abaacd89b8+ info: "export LKP_GIT_WORK_TREE=/path/to/project/git/repo/" to enable parsing non-sha1 commit names. tests: 1 [root@localhost /]# lkp compare -twa --no-hide-noises -d commit 5.10.0-4a4dbce046f7+ 5.10.0-47abaacd89b8+ info: "export LKP_GIT_WORK_TREE=/path/to/project/git/repo/" to enable parsing non-sha1 commit names. tests: 1 [root@localhost /]# After the patch: [root@localhost /]# lkp compare -twa --no-hide-noises -d commit 5.10.0-4a4dbce046f7+ 5.10.0-47abaacd89b8+ info: "export LKP_GIT_WORK_TREE=/path/to/project/git/repo/" to enable parsing non-sha1 commit names. tests: 1 suite/path_params/tbox_group/run: unixbench/syscall/localhost 5.10.0-4a4dbce04 5.10.0-47abaacd89b8+ ---------------- -------------------------- %stddev change %stddev \ | \ 62340 -8% 57520 unixbench.System_Call_Overhead fail:runs %reproduction fail:runs | | | 2:2 0% 3:3 dmesg.tstage.last 2:2 0% 3:3 kmsg.eid.ACPI:IORT:[Firmware_Bug]:[map(____ptrval____)]conflicting_mapping_for_input_ID 2:2 0% 3:3 kmsg.eid.ACPI:IORT:[Firmware_Bug]:applying_workaround 2:2 0% 3:3 kmsg.eid.integrity:Unable_to_open_file:/etc/keys/x509_ima.der(-#) 2:2 0% 3:3 kmsg.eid.integrity:Unable_to_open_file:/etc/keys/x509_evm.der(-#) 2:2 0% 3:3 kmsg.eid.hisi_sas_v3_hw#:#:#:#hw_queues 2:2 0% 3:3 kmsg.eid.hisi_sas_v3_hw###:#hw_queues 158:2 0% 237:3 kmsg.msg.ACPI:IORT:[Firmware_Bug]:[map(____ptrval____)]conflicting_mapping_for_input_ID.message 100:2 0% 150:3 kmsg.msg.ACPI:IORT:[Firmware_Bug]:applying_workaround.message 114:2 0% 171:3 kmsg.msg.integrity:Unable_to_open_file:/etc/keys/x509_ima.der(-#).message 114:2 0% 171:3 kmsg.msg.integrity:Unable_to_open_file:/etc/keys/x509_evm.der(-#).message 84:2 0% 126:3 kmsg.msg.hisi_sas_v3_hw#:#:#:#hw_queues.message 84:2 0% 126:3 kmsg.msg.hisi_sas_v3_hw###:#hw_queues.message 0.72:2 0% 1.08:3 kmsg.ts.ACPI:IORT:[Firmware_Bug]:[map(____ptrval____)]conflicting_mapping_for_input_ID 0.72:2 0% 1.08:3 kmsg.ts.ACPI:IORT:[Firmware_Bug]:applying_workaround 83.19:2 -4% 124.69:3 kmsg.ts.last 28.85:2 -2% 43.22:3 kmsg.ts.integrity:Unable_to_open_file:/etc/keys/x509_evm.der(-#) 28.85:2 -2% 43.22:3 kmsg.ts.integrity:Unable_to_open_file:/etc/keys/x509_ima.der(-#) 33.27:2 -24% 49.18:3 kmsg.ts.hisi_sas_v3_hw#:#:#:#hw_queues 48.59:2 -171% 67.75:3 kmsg.ts.hisi_sas_v3_hw###:#hw_queues [root@localhost /]# Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- lib/stats.rb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/stats.rb b/lib/stats.rb index 3ef5c937a..be2331936 100755 --- a/lib/stats.rb +++ b/lib/stats.rb @@ -630,7 +630,7 @@ def __get_changed_stats(a, b, is_incomplete_run, options) b.each_key { |k| a[k] = [0] * cols_a unless a.include?(k) } # rubocop:disable Style/CombinableLoops a.each do |k, v| - is_force_stat = options["force_#{k}"] + is_force_stat = options["force_#{k}"] || $opt_no_hide_noises next if v[-1].is_a?(String) next if options['perf'] && !perf_metric?(k) -- 2.20.1

1 0

[PATCH] sbin/split-job: get_kernel_info and init job result_root elements
by Zheng Zengkai 04 Aug '25

04 Aug '25

In order to avoid potential problems while assigning values to dimension keys(like commit,compiler,kconfig) from option parameters, assign default values to these dimension variables. Implement a helper function get_kernel_info() to retrieve the kernel version and the version of the compiler that is used to compile the kernel. Invoke the get_kernel_info() during collecting distribution information of local system. Initialize the dimension commit key with value kernelversion got from the distroinfo if no commit option is setting. Initialize the dimension compiler key with value kernelcompiler if it is not assigned. Initialize the dimension kconfig key with value "defconfig" if no kconfig option is setting, as we can't guarantee that we can get the right kconfig file name in various environment. Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- lib/detect-system.sh | 24 ++++++++++++++++++++++++ lib/distro_info.rb | 9 ++++++--- sbin/split-job | 5 +++-- 3 files changed, 33 insertions(+), 5 deletions(-) diff --git a/lib/detect-system.sh b/lib/detect-system.sh index 384480b29..da3cdea63 100755 --- a/lib/detect-system.sh +++ b/lib/detect-system.sh @@ -108,6 +108,30 @@ detect_libc_version() return 1 } +get_kernel_compiler() +{ + local system_dir=${1:-/} + local proc_version=$(<${system_dir}proc/version) + local pattern1='gcc[[:space:]].GCC.[[:space:]]*([0-9]+\.[0-9]+\.[0-9]+)' + local pattern2='gcc[[:space:]]+version[[:space:]]+([0-9]+\.[0-9]+\.[0-9]+)' + export _kernel_compiler='unknown' + + if [[ "$proc_version" =~ $pattern1 ]]; then + _kernel_compiler="${BASH_REMATCH[1]}" + elif [[ "$proc_version" =~ $pattern2 ]]; then + _kernel_compiler="${BASH_REMATCH[1]}" + fi +} + +get_kernel_info() +{ + local system_dir=${1:-/} + export _kernel_version='unknown' + + _kernel_version=$(command uname -r) + get_kernel_compiler $system_dir +} + detect_system() { local system_dir=${1:-/} diff --git a/lib/distro_info.rb b/lib/distro_info.rb index b17c77fa7..066be440c 100755 --- a/lib/distro_info.rb +++ b/lib/distro_info.rb @@ -6,7 +6,7 @@ module LKP # # DistroInfo is singleton, and provide information to distribution information of local system - # Include: system name, system version, system arch + # Include: system name, system version, system arch, kernel version, kernel compiler # In the backend, it's invoking detect-system.sh to get environment information. # Example of properties on debian # p systemName => Debian @@ -18,12 +18,12 @@ module LKP # class DistroInfo include Singleton - attr_reader :systemname, :systemnamel, :systemversion, :systemarch + attr_reader :systemname, :systemnamel, :systemversion, :systemarch, :kernelversion, :kernelcompiler def initialize(rootfs = '/') path_to_script = "#{LKP_SRC}/lib/detect-system.sh" - @systemname, @systemnamel, @systemversion, @systemarch = %x[ + @systemname, @systemnamel, @systemversion, @systemarch, @kernelversion, @kernelcompiler = %x[ export LKP_SRC=#{LKP_SRC} . #{path_to_script} detect_system #{rootfs} @@ -31,6 +31,9 @@ module LKP echo $_system_name_lowercase echo $_system_version get_system_arch #{rootfs} + get_kernel_info #{rootfs} + echo $_kernel_version + echo $_kernel_compiler ].split end end diff --git a/sbin/split-job b/sbin/split-job index bef7e970f..63b68bc84 100755 --- a/sbin/split-job +++ b/sbin/split-job @@ -97,8 +97,9 @@ ARGV.each do |jobfile| jobfile = Job.find_jobfile(jobfile) jobfile = refine_job_file(jobfile) if @opt_compatible jobs.load(jobfile, expand_template: true) || next - jobs['kconfig'] = @opt_kconfig if @opt_kconfig - jobs['commit'] = @opt_commit if @opt_commit + jobs['kconfig'] = @opt_kconfig || "defconfig" + jobs['commit'] = @opt_commit || distroinfo.kernelversion + jobs['compiler'] ||= distroinfo.kernelcompiler jobs['testbox'] = @opt_testbox jobs['arch'] ||= distroinfo.systemarch jobs['tbox_group'] = tbox_group(@opt_testbox) -- 2.20.1

1 0

[PATCH] lib/result.rb: Restore PATH_SCHEME of LOCAL_RUN in ResultPath
by Zheng Zengkai 31 Jul '25

31 Jul '25

Commit cd1d0c6cff23 ("run-local: simplify result path") changes LOCAL_RUN path scheme to "%w[path_params run]" like below: /root/.cache/lkp/result/unixbench/1-100-syscall/0 /root/.cache/lkp/result/unixbench/1-100-syscall/1 /root/.cache/lkp/result/unixbench/96-100-syscall/0 According to the lkp compare usage, user can use the -d option with "commit, kconfig, fs, etc" dimensions to compare results generated by different kernel versions with different commits, kconfigs, or rootfs, it's useful for evaluating the impact of different dimensions(like commits) on performance by running benchmarks on local server. " [root@localhost ~]# lkp compare --help Usage: compare [options] RESULT_ROOT... options: -d, --dimension DIMENSION DIMENSION to compare: commit, kconfig, fs, etc. " But with the simplified result path scheme for local run, this usage is not supported, as the remove of commit, kconfig and rootfs elements, the result_root path will not contain these sub-directories and when setup_result_root_hash greps "dim value" in the "~/.cache/lkp/paths", nothing will be matched, then $_result_root_tuple is empty and no stats can be compared and shown. This patch changes the PATH_SCHEME of LOCAL_RUN to support this usage. Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- lib/result.rb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/result.rb b/lib/result.rb index 492e43b82..8260c8671 100755 --- a/lib/result.rb +++ b/lib/result.rb @@ -28,7 +28,7 @@ class ResultPath < Hash AXIS_KEYS = (MAXIS_KEYS + ['run']).freeze PATH_SCHEME = { - 'LOCAL_RUN' => %w[path_params run], + 'LOCAL_RUN' => %w[path_params tbox_group rootfs kconfig compiler commit run], 'default' => %w[path_params tbox_group rootfs kconfig compiler commit run], 'kvm:default' => %w[path_params tbox_group rootfs kconfig compiler commit run], 'health-stats' => %w[path_params run], -- 2.43.0

1 0

[PATCH] sbin/compare: Fix undefined method for false in sort_stats
by Zheng Zengkai 30 Jul '25

30 Jul '25

when doing lkp compare, it complains: /home/lkp-tests/sbin/compare:826:in `block in sort_stats': undefined method `=~' for false:FalseClass (NoMethodError) from /home/lkp-tests/sbin/compare:814:in `each' from /home/lkp-tests/sbin/compare:814:in `sort_by' from /home/lkp-tests/sbin/compare:814:in `sort_by!' from /home/lkp-tests/sbin/compare:814:in `sort_stats' from /home/lkp-tests/sbin/compare:860:in `<main>' After checking the code and operator priority in ruby, since the 'Logical NOT operator !' has a higher priority than the 'Regex match operator =~', there is a mistake introduced by the commit 8fa05e8aae9f ("cleanup rubpcop issue Style/AndOr"). Add Parentheses to fix it. Fixes: 8fa05e8aae9f ("cleanup rubpcop issue Style/AndOr") Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- sbin/compare | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sbin/compare b/sbin/compare index 11c4fbb05..f5db5ef33 100755 --- a/sbin/compare +++ b/sbin/compare @@ -823,7 +823,7 @@ def sort_stats(stats) if x =~ ABS_CHANGE_STATS_RE [stat_priority(fields[0]), fields[1], abs_change] - elsif fields.size >= 3 && !fields[1] =~ /^node\d+$/ + elsif fields.size >= 3 && !(fields[1] =~ /^node\d+$/) # when a monitor can be broken into sub-groups of stats # under which values can be logically compared [stat_priority(fields[0]), fields[1], rel_change] -- 2.20.1

1 0

kuasar master、openEuler-25.03、openEuler-24.03-sp2分支申请root权限构建
by xiadanni 20 May '25

20 May '25

一、背景 OpenEuler-2403sp2分支kuasar-cc-1.0.0-23.oe2403sp2.aarch64.rpm包失败。报错日志： [cid:image002.png@01DBC96C.D1A67680] cp拷贝二进制文件因为权限问题失败 [cid:image004.png@01DBC96C.D1A67680] 制作kuasar rootfs时存在mount报错问题经私人工程验证，加上root权限能解决上述问题，且当前master、openEuler-25.03分支的kuasar构建也存在类似问题。二、期望添加的配置在master、openEuler-25.03、openEuler-24.03-sp2分支的everything工程的root权限配置中添加kuasar 三、涉及工程范围 1.Eulermaker中master、openEuler-25.03、openEuler-24.03-sp2分支kuasar包编译工程

1 0

openEuler 2503版本构建部分使能mold相关说明
by yinchuang 27 Feb '25

27 Feb '25

各位伙伴好：背景：为了提升openEuler软件包的编译效率，从而进一步提升门禁和开发者本地开发效率，我们在2503版本使用mold链接器来缩短C/C++库的链接时间。 Mold（A Modern Linker）是一种对比其他链接器（GNU ld、GNU gold、LLVM lld）链接效率更高的开源链接器, 其通过更高效的Identical Comdat Folding (ICF)算法、更好的并行编程，更高效的多线程库等技术来提升链接的效率，具体可参考：https://github.com/rui314/mold/blo… 具体方案：由于mold链接器本身存在一定的功能欠缺（对kernel的不支持，对linker script支持不完善）我们决定对一些编译时间较长且不受mold影响的包打开mold 链接： 1. 在2503版本上白名单内的软件包默认配置使用mold进行链接，白名单可参见：https://gitee.com/src-openeuler/openE… 2. 为了提供更加灵活的链接器配置方式，各软件包可在spec文件中覆写_ld_use变量来切换链接器，如： 1) %define _ld_use %{nil} 取消因为软件包在白名单中所添加的mold选项 2) %define _ld_use -fuse-ld=xxx 切换到不同的链接器，注意当定义多个-fuse-ld 选项时编译器会取最后一个配置项 3. 版本上会预装mold链接器来提供各软件包对mold的调用收益：经统计通过使能mold的软件包编译效率总体可提升5+% 我们会持续在社区讨论mold使能可行性与风险，感谢大家。 openEuler Compiler SIG 印闯

1 0

[Compass-ci]mysql包申请添加openEuler-20.03-LTS-SP4:everything工程配置中use_root选项
by 王凯 27 Aug '24

27 Aug '24

一.背景 20.03-LTS-SP4分支mysql源码包因尺寸超出上传门禁限制，使用wget下载覆盖空文件源码包，这个过程中需要被赋予root权限。报错日志：https://eulermaker.compass-ci.openeuler.openatom.cn/package/build-record?osProject=wk333:openEuler-20.03-LTS-SP4:everything-testmysql0823&packageName=mysql&jobId=cbs.6075654 2024-08-23 17:30:56 + wget -qO /home/lkp/rpmbuild/SOURCES/mysql-boost-8.0.38.tar.gz https://user-repo.openeuler.openatom.cn/lfs-tar/mysql/mysql-boost-8.0.38.ta… 2024-08-23 17:30:56 /home/lkp/rpmbuild/SOURCES/mysql-boost-8.0.38.tar.gz: Permission denied 2024-08-23 17:30:56 error: Bad exit status from /var/tmp/rpm-tmp.ulSf9W (%prep) 二.期望添加的配置在openEuler-20.03-LTS-SP4:everything工程的use_root配置中添加mysql 三.涉及工程范围 1.Eulermaker中20.03-SP4分支mysql包编译工程

1 0

Compass-CI使用相关问题
by 黄顺聪 12 Aug '24

12 Aug '24

1.申请账号未通过，同时没有任何未通过原因邮件反馈。原始邮件如下，希望优化申请流程，对申请失败情况有个简要的回复。 2.https://gitee.com/wu_fengguang/lkp-tests/tree/master/doc 文档与实际不符（存在部… 3.希望提供一个本地部署Compass-CI的操作文档 4.lkp-test希望出一期运行调用逻辑说明以及新增用例demo演示，降低入门成本发自我的企业微信

1 0

[Compass-ci]takari-archiver包申请添加openEuler-master:everything工程配置中use_root选项
by dear wang 16 Jul '24

16 Jul '24

一.背景 master分支takari-archiver包因takari-pom包升级后jdkversion被配置为11，重新归档的takari系列包提供的jar包版本全部升级为55版本，且系列包高度耦合，需要从takari-archiver包打破耦合，编译工程中需要删除55版本的jar包重新编译。这个过程中需要被赋予root权限。报错日志：https://eulermaker.compass-ci.openeuler.openatom.cn/api/result/rpmbuild/2024-07-16/dc-64g/openeuler-mainline-aarch64/aarch64-common-gitee.com-src-openeuler-ta/cbs.5701400/dmesg 二.期望添加的配置在openEuler-master:everything工程的use_root配置中添加takari-archiver 三.涉及工程范围 1.Eulermaker中master分支takari-archiver包编译工程

1 0

[Compas-ci]libabigail包申请去除openEuler-master:everything工程配置中use git lfs选项
by 707078654 09 Jul '24

09 Jul '24

一.背景 master分支的 libabigail 包已将tar包放至 user-repo （https://user-repo.openeuler.openatom.cn/lfs-tar/libabigail/libabigail-2.5.tar.xz）路径，不再采用git lfs方式获取tar包，且该选项会导致clone代码失败，申请移除use_git_lfs配置报错日志：https://eulermaker.compass-ci.openeuler.openatom.cn/package/build-record?osProject=xuping:openEuler-master:everything-0709-bak&packageName=libabigail&jobId=cbs.5628463 二.期望添加的配置移除openEuler-master:everything工程use_git_lfs配置中的libabigail 三.涉及工程范围 1.Eulermaker中master 分支libabigail 包编译工程 2.后续从 master分支拉取出来的新分支均需继承该工程许萍 18902872027 Best wishes

1 0

【openEuler社区OBS下线计划】
by xiasenlin 12 Jun '24

12 Jun '24

各位openeuler社区的maintainer、 committer和contributor们好：社区CICD sig时刻关注openEuler官方镜像仓中各维护版本最新EOL变更情况，近期openEuler-20.03-LTS-SP1和openEuler-22.03-LTS正式停维，至此OBS（https://build.openeuler.openatom.cn/）不再承担社区在维版本/开发版本工程日构建（请前往EulerMaker：https://eulermaker.compass-ci.openeuler.openatom.cn/），我们制定如下社区OBS构建系统下线计划：（1）个人数据备份（2024/6/12~2024/6/15）：在此期间，请还在使用社区OBS构建系统的个人用户，尽快备份数据至本地；（2） Worker释放（2024/6/17）：除预留2台worker（arm:1; x86:1）以备紧急情况使用外，其余worker全部释放；（3）静默期（2024/6/17~2024/7/17）：在此期间，社区OBS前端可访问，部分信息可查看，但CICD sig不再主动维护；（4）下线（2024/7/18）：所有backend及worker全部释放，社区OBS构建系统正式下线； PS： 1．社区OBS构建系统从2019年部署上线以来，承担了openEuler5个LTS&SPx版本（openEuler-20.03-LTS/SP1/SP2/SP3; openEuler-22.03-LTS）和3个创新版本（openEuler-20.09/21.03/21.09）的日构建和发布，感谢参与OBS环境搭建和维护的基础设施 sig，也感谢社区开发者在使用过程中的包容和反馈！ 2． CICD sig推出社区自研统一构建系统（EulerMaker），OBS影响深远，在此向OBS上游社区开发者&贡献者致敬！

1 0

[compass-ci]iproute 包申请使用root权限构建
by 穆丰演 31 May '24

31 May '24

一、背景 openEuler-22.03-LTS-SP4分支的iproute包使用普通用户权限在执行%check阶段会失败 PR链接：https://gitee.com/src-openeuler/iproute/pulls/189 报错日志：https://eulermaker.compass-ci.openeuler.openatom.cn/package/build-record?osProject=openEuler-22.03-LTS-SP4:aarch64:iproute:189&packageName=iproute&jobId=cbs.5289261 需要root权限构建二、期望添加的配置申请iproute包使用root权限构建三、涉及工程范围 1、Eulermaker 中openEuler-22.03-LTS-SP4分支的iproute工程 2、Eulermaker 中openEuler-22.03-LTS-SP4分支、openEuler-22.03-LTS-SP3、openEuler-24.03-LTS及后续的分支需要root权限进行%check构建 | | 穆丰演 | | 邮箱：mfymfy55555(a)163.com |

1 0

[compass-ci]iproute 包申请使用root权限构建
by 穆丰演 30 May '24

30 May '24

一、背景 openEuler-22.03-LTS-SP4分支的iproute包使用普通用户权限在执行%check阶段会失败 PR链接：https://gitee.com/src-openeuler/iproute/pulls/189 报错日志：https://eulermaker.compass-ci.openeuler.openatom.cn/package/build-record?osProject=openEuler-22.03-LTS-SP4:aarch64:iproute:189&packageName=iproute&jobId=cbs.5289261 需要root权限构建二、期望添加的配置申请iproute包使用root权限构建三、涉及工程范围 1、Eulermaker 中openEuler-22.03-LTS-SP4分支的iproute工程 2、Eulermaker 中openEuler-22.03-LTS-SP4分支、openEuler-22.03-LTS-SP3、openEuler-24.03-LTS及后续的分支需要root权限进行%check构建 | | 穆丰演 | | 邮箱：mfymfy55555(a)163.com |

1 0

[Compass-ci]nodejs-commander包申请使用root权限构建
by 王凯 23 May '24

23 May '24

一.背景 22.03-LTS-SP4分支的 nodejs-commander包使用普通用户执行%check阶段会失败 issue:https://gitee.com/src-openeuler/nodejs-commander/issues/I9REYW?from=p… 报错日志：https://eulermaker.compass-ci.openeuler.openatom.cn/package/build-record?osProject=openEuler-22.03-LTS-SP4:everything&packageName=nodejs-commander&jobId=cbs.5169077 问题同24.03-LTS，执行测试时使用了 /tmp/stderr 文件，该文件在构建时已被root用户创建，所以普通用户无权限写入二.期望添加的配置申请nodejs-commander使用root权限构建三.涉及工程范围 1.Eulermaker中22.03-LTS-SP4分支nodejs-commander编译工程 2.Eulermaker中22.03-LTS-SP3、22.03-LTS-SP2、22.03-LTS-SP1、22.03-LTS-Next、20.03-LTS-SP4分支工程，如果会打开%check编译，均需添加root权限

1 0

[Compass-ci]conntrack-tools包申请使用root权限构建选项
by sunhai (C) 11 May '24

11 May '24

一.背景 1．2403-LTS分支的conntrack-tools包在工程执行check阶段时的测试用例失败，需要使用root权限构建 2．因为sh run-test.sh由于普通用户权限受限执行不了该操作，所以需要root权限 3．在22.03及其以上都涉及修改，所以会涉及到22.03及24.03相关的版本二.期望添加的配置申请conntrack-tools包使用root权限构建三.涉及工程范围 1.Eulermaker/OBS中master/2403-LTS/2403-LTS-Next/2203-LTS/2203-LTS-SP1/2203-LTS-SP2/2203-LTS-SP3/2203-LTS-Next/分支conntrack-tools包编译工程 2.后续从master分支拉取出来的新分支均需继承该工程

1 0

[Compass-ci]nodejs-commander包申请使用root权限构建
by 王凯 11 May '24

11 May '24

一.背景 24.03-LTS分支的 nodejs-commander包使用普通用户执行%check阶段会失败报错日志：https://eulermaker.compass-ci.openeuler.openatom.cn/package/build-record?osProject=openEuler-24.03-LTS:everything&packageName=nodejs-commander&jobId=cbs.4372115 定位到执行测试时使用了 /tmp/stderr 文件，该文件在构建时已被root用户创建，所以普通用户无权限写入二.期望添加的配置申请nodejs-commander使用root权限构建三.涉及工程范围 1.Eulermaker中24.03-LTS分支nodejs-commander编译工程 2.后续从master分支拉取出来的新分支均需继承该工程

1 0

【compass-ci】rust包申请使用root权限构建
by 杨超豪 26 Apr '24

26 Apr '24

【背景】 1、rust在24.03-LTS分支在Eulermaker构建失败，构建失败issue: 【EulerMaker】rust 在openEuler-24.03-LTS:everything中构建失败 · Issue #I9IYXP · src-openEuler/rust - Gitee.com。 2、目前构建环境无人力定位，rust构建可通过添加root权限来规避构建问题，故申请root权限。【期望添加的配置】申请rust使用root权限构建【涉及工程范围】 1、 Eulermaker中24.03-LTS分支rust编译工程 2、后续从master分支拉取处理的新分支均需继承该工程

1 0

[Compass-ci]KubeOS包申请使用root权限构建
by liyuanrong 23 Apr '24

23 Apr '24

【背景】 1、KubeOS在24.03-LTS分支在Eulermaker构建失败，构建失败issue: https://gitee.com/src-openeuler/KubeOS/issues/I9H8W6?from=project-issue。 2、与工程负责人定位发现是由于KubeOS构建需root权限，否则执行不了install操作，Eulermaker验证如图，构建成功为手动添加root权限，构建失败为取消root权限，所以KubeOS构建需要root权限 [cid:image003.png@01DA94F7.B4868A40] 【期望添加的配置】申请KubeOS使用root权限构建【涉及工程范围】 1、 Eulermaker/OBS中master/24.03-LTS/24.03-LTS-Next/22.03-LTS/22.03-LTS-SP1/22.03-LTS-SP2/22.03-LTS-SP3/22.03-LTS-SP4/22.03-LTS-Next/分支KubeOS编译工程 2、后续从master分支拉取处理的新分支均需继承该工程

2 1

[Compass-ci]at-spi2-core包申请使用root权限构建选项
by sunhai (C) 18 Apr '24

18 Apr '24

一.背景 1．2203-LTS分支的at-spi2-core包在obs工程执行check阶段时的测试用例失败，需要使用root权限构建 2．因为涉及的memory用例需要执行/usr/libexec/at-spi-bus-launcher，由于该文件at-spi2-core本包提供，所以在执行用例之前将当前编译出来的文件install至/usr/libexec/目录下来解决用例失败的问题，由于普通用户权限受限执行不了该操作，所以需要root权限 3．在22.03及其以上都会后期修改，所以会涉及到22.03及24.03相关的版本二.期望添加的配置申请at-spi2-core包使用root权限构建三.涉及工程范围 1.Eulermaker/OBS中master/2403-LTS/2403-LTS-Next/2203-LTS/2203-LTS-SP1/2203-LTS-SP2/2203-LTS-SP3/2203-LTS-Next/分支at-spi2-core包编译工程 2.后续从master分支拉取出来的新分支均需继承该工程

2 1

mcelog编译时开启root权限申请
by gaoruoshu 26 Mar '24

26 Mar '24

Hi maintainer, 申请为mcelog软件所有分支添加x86架构的root编译权限，该软件的测试用例必须通过root用例才能执行，如图： [cid:image001.png@01DA7F65.0983D4B0] https://eulermaker.compass-ci.openeuler.openatom.cn/package/build-record?os… 当前22.03相关分支已经为root权限构建，不会报错，而24.03-LTS分支在进行x86架构软件编译会被阻塞，现申请对该软件所有分支（含Next分支及后续预期拉取的分支）添加root编译权限。 BR, gaoruoshu

2 1

thunderbird 包申请添加 openEuler-master:everything 工程配置中 use_root 选项
by misaka00251 05 Mar '24

05 Mar '24

一. 背景 master 分支 thunderbird 源码包因尺寸超出上传门禁限制，目前被分割成两个子包，在构建过程中，通过 cat 连接出完成的源码包。这个过程中需要被赋予 root 权限。报错日志: https://eulermaker.compass-ci.openeuler.openatom.cn/package/build-record?os… 二. 期望添加的配置在 openEuler-master:everything 工程的 use root 配置中添加 thunderbird 三. 涉及工程范围 1. Eulermaker 中 master 分支 thunderbird 包编译工程 2. 后续从 master 分支拉取出来的新分支均需继承该工程

1 0

sig-CICD双周例会
by openEuler conference 21 Feb '24

21 Feb '24

您好！ sig-CICD 邀请您参加 2024-02-22 16:00 召开的Zoom会议(自动录制) 会议主题：sig-CICD双周例会会议链接：https://us06web.zoom.us/j/83244204317?pwd=4kh08KeNKVuDLcXQpZAwbNBjrPuSas.1 会议纪要：https://etherpad.openeuler.org/p/sig-CICD-meetings 更多资讯尽在：https://www.openeuler.org/zh/ Hello! sig-CICD invites you to attend the Zoom conference(auto recording) will be held at 2024-02-22 16:00, The subject of the conference is sig-CICD双周例会, You can join the meeting at https://us06web.zoom.us/j/83244204317?pwd=4kh08KeNKVuDLcXQpZAwbNBjrPuSas.1. Add topics at https://etherpad.openeuler.org/p/sig-CICD-meetings. More information: https://www.openeuler.org/en/

1 0

sig-CICD双周例会
by openEuler conference 21 Feb '24

21 Feb '24

您好！ sig-CICD 邀请您参加 2024-02-22 16:00 召开的Zoom会议(自动录制) 会议主题：sig-CICD双周例会会议链接：https://us06web.zoom.us/j/87042950235?pwd=4GdObEi4ivj4twRBQHw8uXAxS3dWYG.1 会议纪要：https://etherpad.openeuler.org/p/sig-CICD-meetings 更多资讯尽在：https://www.openeuler.org/zh/ Hello! sig-CICD invites you to attend the Zoom conference(auto recording) will be held at 2024-02-22 16:00, The subject of the conference is sig-CICD双周例会, You can join the meeting at https://us06web.zoom.us/j/87042950235?pwd=4GdObEi4ivj4twRBQHw8uXAxS3dWYG.1. Add topics at https://etherpad.openeuler.org/p/sig-CICD-meetings. More information: https://www.openeuler.org/en/

1 0