Re: [OE-core] [PATCH v8 0/9] rpm: 4.20.1 -> 6.0.1

[Randy MacLeod <randy.macleod@windriver.com>]

[-- Attachment #1: Type: text/plain, Size: 5337 bytes --]

Hi Everyone,

Sumanth (CCed) has joined Wind River and will pick up where Robert left 
off unless someone has
things underway already. Sumanth doesn't have Robert's deep knowledge of 
rpm but
he is an experienced developer and will make this a top priority 
immediately.


The first step will be to refresh the patchset locally without changing 
much but then
he'll update to either 6.0.2 or 6.1-rc2. What do people advise for the 
version?

../Randy


On 2026-03-14 03:00, Mathieu Dubois-Briand via lists.openembedded.org wrote:
> On Fri Mar 13, 2026 at 7:58 PM CET, Mathieu Dubois-Briand via lists.openembedded.org wrote:
>> On Fri Mar 13, 2026 at 5:54 AM CET, Robert Yang wrote:
>>> Hi Mathieu,
>>>
>>> On 3/12/26 23:52, Mathieu Dubois-Briand via lists.openembedded.org wrote:
>>>> On Thu Mar 12, 2026 at 3:09 PM CET, liezhi.yang wrote:
>>>>> From: Robert Yang<liezhi.yang@windriver.com>
>>>>>
>>>>> * V8:
>>>>>     - Fix build with systemd (Check patch 7 for detailes)
>>>>>     - Make rpmbuild can find the pkgconfig (.pc) files in /usr/share/pkgconfig
>>>>>       (Check patch 7 for detailes)
>>>>>     - Backport a PULL from git/libarchive to fix oe-selftest failures.
>>>>>
>>>>> * V7:
>>>>>     - Add 0001-scripts-pkgconfigdeps.sh-Use-pkg-config-from-PATH.patch:
>>>>>       To fix "bitbake core-image-sato -cpopulate_sdk" error on Ubuntu 24.04 when
>>>>>       pkg-config is installed on host, now the host contamination has been fixed,
>>>>>       and the patch has been submitted to upstream:
>>>>>       https://github.com/rpm-software-management/rpm/pull/4127
>>>>>
>>>>> * V6
>>>>>     - Fix -cpopulate_sdk error when DISTRO = "poky"
>>>>>
>>>>> * V5
>>>>>     - Fix RP's comments: Don't hardcode libc6 into package_rpm.bbclass
>>>>>     - Use SKIP_LDCONFIG_POSTINST_FRAGMENT:${PN} to fix the loop dependencies
>>>>>       between /bin/sh and libc.so.
>>>>>
>>>>> * V4
>>>>>     - Fix loop dependencies issues for libc6 and busybox
>>>>>     - Fix issues for can't find pkg-config files (.pc) on qemux86-64.
>>>>>
>>>>> * V3
>>>>>     - Fix Alexender's comments
>>>>>       a) Suppot rpm 6
>>>>>       b) Move rpm-setup-autosign to rpm-build
>>>>>       c) Remove -DLIBELF_FOUND=False
>>>>>
>>>>>     - Fix build for nativesdk-rpm
>>>>>
>>>>>     - Drop patch for createrepo-c since it has been upgraded (rev:
>>>>>       804a238c8a4e86c2bcfb2192d54cb861d5fd32a6)
>>>>>
>>>>> * V1 & V2
>>>>>     - Initial version
>>>>>
>>>>> // Robert
>>>>>
>>>> Hi Robert,
>>>>
>>>> Thanks for the new version.
>>>>
>>>> I suspect I still need to get some gettext patch with it. Either yours
>>>> or the one from the glibc upgrade series. Is that right?
>>> I think you meat the following oe-selftest error:
>>>
>>> ERROR: glib-2.0-1_2.86.4-r0 do_package_qa: QA Issue: glib-2.0-ptest rdepends on
>>> locale-base-c, but it isn't a build dependency? [build-deps]
>>>
>>>
>>> I can reproduce it with pure master branch: (rev:
>>> 438afb3ded629408172c4bd7d556c5d2abacd631)
>>> PACKAGE_CLASSES = "package_rpm"
>>> OE_FRAGMENTS += 'core/yocto-autobuilder/autobuilder'
>>> MACHINE = 'qemux86-64'
>>>
>>> IMAGE_INSTALL:append = " glibc-utils localedef"
>>> GLIBC_GENERATE_LOCALES = "en_US.UTF-8 fr_FR.UTF-8 en_US.ISO-8859-1 de_DE.UTF-8
>>> fr_FR.ISO-8859-1 zh_HK.BIG5-HKSCS tr_TR.UTF-8"
>>> IMAGE_LINGUAS:append = " en-us fr-fr"
>>>
>>> $ bitbake glibc-2.0
>>>
>>> So I don't think it's caused by rpm 6 or gettext upgrading, and we don't need
>>> other patches.
>>>
>>> I will check how to fix the glib-2.0 issues, but it's not related to rpm 6.
>>>
>>> // Robert
>>>
>> Hi Robert,
>>
>> So yes, I have this issue, and I'm not having it without your series:
>> https://autobuilder.yoctoproject.org/valkyrie/#/builders/35/builds/3399
>> https://autobuilder.yoctoproject.org/valkyrie/#/builders/48/builds/3290
>> https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/3517
>>
>> And some reproducibility error:
>> https://autobuilder.yoctoproject.org/valkyrie/#/builders/37/builds/3579
>>
>> I will try to build just your series on top of master and compare with
>> master. Maybe some strange combination with the glibc upgrade.
>>
> Ok, so here is an a-full build with RPM patches applied directly on top
> of master: we have the same issues as yesterday:
>
> https://autobuilder.yoctoproject.org/valkyrie/#/builders/29/builds/3410
>
> selftests:
> https://autobuilder.yoctoproject.org/valkyrie/#/builders/23/builds/3526
> https://autobuilder.yoctoproject.org/valkyrie/#/builders/35/builds/3420
> https://autobuilder.yoctoproject.org/valkyrie/#/builders/48/builds/3298
>
> reproducible:
> https://autobuilder.yoctoproject.org/valkyrie/#/builders/37/builds/3592
>
> By comparison, here is an a-full build on master:
> https://autobuilder.yoctoproject.org/valkyrie/#/builders/29/builds/3414
>
> I there any test I can do to help you?
>
> Thanks,
> Mathieu
>
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#233072):https://lists.openembedded.org/g/openembedded-core/message/233072
> Mute This Topic:https://lists.openembedded.org/mt/118279443/3616765
> Group Owner:openembedded-core+owner@lists.openembedded.org
> Unsubscribe:https://lists.openembedded.org/g/openembedded-core/unsub [randy.macleod@windriver.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>

-- 
# Randy MacLeod
# Wind River Linux

[-- Attachment #2: Type: text/html, Size: 8242 bytes --]