From: "Böszörményi Zoltán" <zboszor@gmail.com>
To: Alexander Kanavin <alex.kanavin@gmail.com>
Cc: Alexandre Belloni <alexandre.belloni@bootlin.com>,
openembedded-core@lists.openembedded.org,
Ross Burton <Ross.Burton@arm.com>,
Richard Purdie <richard.purdie@linuxfoundation.org>
Subject: Re: [OE-core] [PATCH v6] rpm: Allow setting platform macro settings externally
Date: Thu, 3 Aug 2023 12:59:13 +0200 [thread overview]
Message-ID: <2b5bc564-f05c-3c14-688e-50b3ba62eb2e@gmail.com> (raw)
In-Reply-To: <CANNYZj-hE=Cct9bLOhy_kWA9NZqnpUAJ3ZTXW2RnoXzz65B1fw@mail.gmail.com>
2023. 08. 03. 9:55 keltezéssel, Alexander Kanavin írta:
> I applied the patch and checked what the difference in install tree
> is. I see that the only difference is addition of
> usr/lib/rpm/platform/qemux86_64-linux/macros
>
> I'd like to step back for a bit, and clarify:
>
> What is the mechanism at runtime for finding that file?
This will show:
strace rpmbuild -ba something.spec
When /etc/rpm/platform exists, its content is read and
/usr/lib/rpm/platform/`cat /etc/rpm/platform`/macros will be used
for setting up essential macro values, even if this file does not exist.
rpmbuild will look no further for other fallback "macros" file, like
/usr/lib/rpm/platform/x86_64-linux/macros in my case.
For any missing macro, its reference is substituted literally:
[root@localhost ~]# rpmbuild --eval="%optflags"
%optflags
[root@localhost ~]# mv /etc/rpm/platform /etc/rpm/platform.dontuse
[root@localhost ~]# rpmbuild --eval="%optflags"
-O2 -g
When running rpmbuild, %configure macro in the specfile will fail
because of this reason. RPMOPTFLAGS will contain %optflags literally.
> How does rpm
> tooling arrive at the path it needs to read? How is 'qemux86_64-linux'
> computed? Are there alternative locations that macros are also taken
> from?
>
> And once that is clarified, the next question would be:
> Can we instruct rpm to read macros from a default location for a
> self-hosted target?
Yes. /etc/rpm/platform simply must not exist.
> You might also just teach that developer who is unwilling to learn new
> stuff to write recipes. It's about the same amount of effort as
> writing rpm .spec files :)
Nah, it's futile. It's a developer at a 3rd party company and I have
no way to talk to them directly. Their stuff is closed source and
they don't really deviate from "mainstream" distros for supporting it.
>
> Alex
>
>
>
> On Thu, 3 Aug 2023 at 08:09, Böszörményi Zoltán <zboszor@gmail.com> wrote:
>> 2023. 08. 02. 21:54 keltezéssel, Alexander Kanavin írta:
>>> I think the culprit is the usage of MACHINE_ARCH. You can't have that
>>> in regular recipes, as that makes them machine-specific, rather than
>>> tune (i.e. architecture) specific.
>> So, how to fix it?
>>
>> Would putting this into the rpm recipe be enough and acceptable?
>>
>> PACKAGE_ARCH = "${MACHINE_ARCHJ}"
>>
>> Reasoning: the patch that makes rpm carry the
>> $MACHINE_ARCH}-linux platform specific macros file
>> does make rpm machine-specific.
>>
>>> Alex
>>>
>>> On Wed, 2 Aug 2023 at 21:28, Böszörményi Zoltán <zboszor@gmail.com> wrote:
>>>> Some clues please. What does this error mean? How to solve it?
>>>>
>>>> 2023. 08. 02. 15:41 keltezéssel, Alexandre Belloni írta:
>>>>> Hello,
>>>>>
>>>>> This causes
>>>>> sstatetests.SStateHashSameSigs3.test_sstate_sametune_samesigs to fail:
>>>>>
>>>>> https://autobuilder.yoctoproject.org/typhoon/#/builders/87/builds/5572/steps/14/logs/stdio
>>>>> https://autobuilder.yoctoproject.org/typhoon/#/builders/86/builds/5538/steps/14/logs/stdio
>>>>> https://autobuilder.yoctoproject.org/typhoon/#/builders/79/builds/5547/steps/15/logs/stdio
>>>>> https://autobuilder.yoctoproject.org/typhoon/#/builders/80/builds/5495/steps/14/logs/stdio
>>>>>
>>>>> On 01/08/2023 15:29:57+0200, Zoltan Boszormenyi wrote:
>>>>>> Feed platform settings to installplatform externally. Based on the patch
>>>>>> submitted under https://github.com/rpm-software-management/rpm/pull/2585
>>>>>>
>>>>>> Patch against INSTALL was backported for rpm 4.18.1
>>>>>>
>>>>>> Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com>
>>>>>> ---
>>>>>> ...atform-macro-settings-externally-258.patch | 119 ++++++++++++++++++
>>>>>> meta/recipes-devtools/rpm/rpm_4.18.1.bb | 20 +++
>>>>>> 2 files changed, 139 insertions(+)
>>>>>> create mode 100644 meta/recipes-devtools/rpm/files/0001-Allow-setting-platform-macro-settings-externally-258.patch
>>>>>>
>>>>>> diff --git a/meta/recipes-devtools/rpm/files/0001-Allow-setting-platform-macro-settings-externally-258.patch b/meta/recipes-devtools/rpm/files/0001-Allow-setting-platform-macro-settings-externally-258.patch
>>>>>> new file mode 100644
>>>>>> index 0000000000..a32ce38533
>>>>>> --- /dev/null
>>>>>> +++ b/meta/recipes-devtools/rpm/files/0001-Allow-setting-platform-macro-settings-externally-258.patch
>>>>>> @@ -0,0 +1,119 @@
>>>>>> +From e58c607dcdb303ff3ba1b22fd23eb20e374b6ed3 Mon Sep 17 00:00:00 2001
>>>>>> +From: =?UTF-8?q?Zolt=C3=A1n=20B=C3=B6sz=C3=B6rm=C3=A9nyi?=
>>>>>> + <zboszor@gmail.com>
>>>>>> +Date: Fri, 28 Jul 2023 12:16:56 +0200
>>>>>> +Subject: [PATCH] Allow setting platform macro settings externally (#2585)
>>>>>> +MIME-Version: 1.0
>>>>>> +Content-Type: text/plain; charset=UTF-8
>>>>>> +Content-Transfer-Encoding: 8bit
>>>>>> +
>>>>>> +* Allow setting platform macro settings externally
>>>>>> +
>>>>>> +By default, rpm installs a series of default platforms based on
>>>>>> +the CPU architecture names in subdirectories called
>>>>>> +
>>>>>> + /usr/lib/platform/<arch>-<os>
>>>>>> +
>>>>>> +This is enough for regular Linux distributions. However, some
>>>>>> +distributions may use more specific platform names that refer to
>>>>>> +particular computer systems, like SBCs or specific CPU tuning when
>>>>>> +compiling.
>>>>>> +
>>>>>> +If the platform subdirectory does not exist in /usr/lib/platform
>>>>>> +then rpmbuild does not work.
>>>>>> +
>>>>>> +Allow creating such custom platform subdirectory with feeding
>>>>>> +the necessary data using external variables: RPM_CUSTOM_ARCH,
>>>>>> +RPM_CUSTOM_ISANAME, RPM_CUSTOM_ISABITS, RPM_CUSTOM_CANONARCH
>>>>>> +and RPM_CUSTOM_CANONCOLOR
>>>>>> +
>>>>>> +Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com>
>>>>>> +Upstream-Status: Backport [https://github.com/rpm-software-management/rpm/commit/fde03ae33d55b160a31a6c54946880dcdf8fd0d6]
>>>>>> +
>>>>>> +---------
>>>>>> +
>>>>>> +Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com>
>>>>>> +Co-authored-by: Florian Festi <ffesti@redhat.com>
>>>>>> +---
>>>>>> + INSTALL | 37 +++++++++++++++++++++++++++++++++++++
>>>>>> + installplatform | 9 ++++++++-
>>>>>> + 2 files changed, 45 insertions(+), 1 deletion(-)
>>>>>> +
>>>>>> +diff --git a/INSTALL b/INSTALL
>>>>>> +index 13d0d8a83..5e035cf6a 100644
>>>>>> +--- a/INSTALL
>>>>>> ++++ b/INSTALL
>>>>>> +@@ -148,6 +148,43 @@ and then install with:
>>>>>> +
>>>>>> + make install
>>>>>> +
>>>>>> ++
>>>>>> ++By default, rpm installs a series of default platforms based on the CPU
>>>>>> ++architecture names in subdirectories called
>>>>>> ++
>>>>>> ++ /usr/lib/platform/<arch>-<os>
>>>>>> ++
>>>>>> ++This is enough for many distributions. However, some distributions
>>>>>> ++may use more specific platform names that refer to particular computer
>>>>>> ++systems, like SBCs or specific CPU tuning when compiling. Examples of such
>>>>>> ++platform names are: "genericx86_64", "intel_skylake_64", "raspberrypi_armv7",
>>>>>> ++"raspberrypi_armv8", etc.
>>>>>> ++
>>>>>> ++If the platform name is put into /etc/rpm/platform, then rpmbuild uses it
>>>>>> ++and the only macros file rpmbuild looks for is
>>>>>> ++
>>>>>> ++ /usr/lib/platform/`cat /etc/rpm/platform`-<os>/macros
>>>>>> ++
>>>>>> ++If this file does not exist, many rpm macros will not have their expected
>>>>>> ++values set and e.g. %configure will fail when trying to run rpmbuild.
>>>>>> ++
>>>>>> ++To allow creating the macros file for such a custom platform, the shell
>>>>>> ++variables listed below must be set. If RPM_CUSTOM_ARCH is not set, the rest
>>>>>> ++is ignored.
>>>>>> ++
>>>>>> ++ export RPM_CUSTOM_ARCH=genericx86_64
>>>>>> ++ export RPM_CUSTOM_ISANAME=x86
>>>>>> ++ export RPM_CUSTOM_ISABITS=64
>>>>>> ++ export RPM_CUSTOM_CANONARCH=x86_64
>>>>>> ++ export RPM_CUSTOM_CANONCOLOR=0 # to use /usr/lib for %_libdir
>>>>>> ++ export RPM_CUSTOM_CANONCOLOR=3 # to use /usr/lib64 for %_libdir
>>>>>> ++
>>>>>> ++ make install
>>>>>> ++
>>>>>> ++This also creates and installs the new platform file e.g.
>>>>>> ++/usr/lib/platform/genericx86_64-linux/macros
>>>>>> ++
>>>>>> ++
>>>>>> + Rpm comes with an automated self-test suite. The test-suite relies heavily
>>>>>> + on fakechroot (https://github.com/dex4er/fakechroot/) and cannot be executed
>>>>>> + without it. Provided that fakechroot was found during configure,
>>>>>> +diff --git a/installplatform b/installplatform
>>>>>> +index a5ad7c5b8..59f57697b 100755
>>>>>> +--- a/installplatform
>>>>>> ++++ b/installplatform
>>>>>> +@@ -11,7 +11,7 @@ VENDOR="${4}"
>>>>>> + OS="${5}"
>>>>>> + RPMRC_GNU="${6}"
>>>>>> +
>>>>>> +-for ARCH in noarch `grep ^arch_canon $RPMRC | cut -d: -f2`; do
>>>>>> ++for ARCH in noarch `grep ^arch_canon $RPMRC | cut -d: -f2` ${RPM_CUSTOM_ARCH:+custom}; do
>>>>>> + RPMRC_OPTFLAGS="`sed -n 's/^optflags: '$ARCH' //p' $RPMRC`"
>>>>>> + RPMRC_OPTFLAGS="`echo $RPMRC_OPTFLAGS | sed -e 's, ,\ ,g'`"
>>>>>> + case $RPMRC_OPTFLAGS in
>>>>>> +@@ -30,6 +30,13 @@ for ARCH in noarch `grep ^arch_canon $RPMRC | cut -d: -f2`; do
>>>>>> + CANONCOLOR=
>>>>>> + FILTER=cat
>>>>>> + case "${ARCH}" in
>>>>>> ++ custom)
>>>>>> ++ ARCH=$RPM_CUSTOM_ARCH
>>>>>> ++ ISANAME=$RPM_CUSTOM_ISANAME
>>>>>> ++ ISABITS=$RPM_CUSTOM_ISABITS
>>>>>> ++ CANONARCH=$RPM_CUSTOM_CANONARCH
>>>>>> ++ CANONCOLOR=$RPM_CUSTOM_CANONCOLOR
>>>>>> ++ ;;
>>>>>> + sparc64*)
>>>>>> + ISANAME=sparc
>>>>>> + ISABITS=64
>>>>>> +--
>>>>>> +2.41.0
>>>>>> +
>>>>>> diff --git a/meta/recipes-devtools/rpm/rpm_4.18.1.bb b/meta/recipes-devtools/rpm/rpm_4.18.1.bb
>>>>>> index 95a9e92f96..0b8ebebb51 100644
>>>>>> --- a/meta/recipes-devtools/rpm/rpm_4.18.1.bb
>>>>>> +++ b/meta/recipes-devtools/rpm/rpm_4.18.1.bb
>>>>>> @@ -40,6 +40,7 @@ SRC_URI = "git://github.com/rpm-software-management/rpm;branch=rpm-4.18.x;protoc
>>>>>> file://0001-python-Use-Py_hash_t-instead-of-long-in-hdr_hash.patch \
>>>>>> file://fix-declaration.patch \
>>>>>> file://ea3187cfcf9cac87e5bc5e7db79b0338da9e355e.patch \
>>>>>> + file://0001-Allow-setting-platform-macro-settings-externally-258.patch \
>>>>>> "
>>>>>>
>>>>>> PE = "1"
>>>>>> @@ -128,10 +129,29 @@ do_install:append:class-nativesdk() {
>>>>>> EOF
>>>>>> }
>>>>>>
>>>>>> +def rpm_isaname(d):
>>>>>> + import re
>>>>>> + arch = d.getVar('TARGET_ARCH')
>>>>>> + if re.match("^i.86$", arch) or re.match("^x86.*64$", arch):
>>>>>> + return "x86"
>>>>>> + # Add more platform tweaks for ISANAME as needed
>>>>>> + return arch
>>>>>> +
>>>>>> +# Add the custom BSP platform "macros" file
>>>>>> +do_install:prepend:class-target() {
>>>>>> + export RPM_CUSTOM_ARCH="${MACHINE_ARCH}"
>>>>>> + export RPM_CUSTOM_ISANAME="${@rpm_isaname(d)}"
>>>>>> + export RPM_CUSTOM_ISABITS="${SITEINFO_BITS}"
>>>>>> + export RPM_CUSTOM_CANONARCH="${TARGET_ARCH}"
>>>>>> + # CANONCOLOR determines whether /usr/lib or /usr/lib64 is used for a 64-bit platform
>>>>>> + export RPM_CUSTOM_CANONCOLOR="${@bb.utils.contains('DISTRO_FEATURES', 'multilib', '3', '0', d)}"
>>>>>> +}
>>>>>> +
>>>>>> # Rpm's make install creates var/tmp which clashes with base-files packaging
>>>>>> do_install:append:class-target() {
>>>>>> rm -rf ${D}/var
>>>>>> }
>>>>>> +
>>>>>> do_install:append:class-nativesdk() {
>>>>>> rm -rf ${D}${SDKPATHNATIVE}/var
>>>>>> # Ensure find-debuginfo is located correctly inside SDK
>>>>>> --
>>>>>> 2.41.0
>>>>>>
>>>>>> -=-=-=-=-=-=-=-=-=-=-=-
>>>>>> Links: You receive all messages sent to this group.
>>>>>> View/Reply Online (#185245): https://lists.openembedded.org/g/openembedded-core/message/185245
>>>>>> Mute This Topic: https://lists.openembedded.org/mt/100483574/3617179
>>>>>> Group Owner: openembedded-core+owner@lists.openembedded.org
>>>>>> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [alexandre.belloni@bootlin.com]
>>>>>> -=-=-=-=-=-=-=-=-=-=-=-
>>>>>>
next prev parent reply other threads:[~2023-08-03 10:59 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-08-01 13:29 [PATCH v6] rpm: Allow setting platform macro settings externally Zoltán Böszörményi
2023-08-02 13:41 ` [OE-core] " Alexandre Belloni
2023-08-02 19:28 ` Böszörményi Zoltán
2023-08-02 19:54 ` Alexander Kanavin
2023-08-03 6:09 ` Böszörményi Zoltán
2023-08-03 7:55 ` Alexander Kanavin
2023-08-03 10:59 ` Böszörményi Zoltán [this message]
2023-08-03 11:15 ` Alexander Kanavin
2023-08-03 12:17 ` Böszörményi Zoltán
[not found] ` <1777D06DC603E04E.27883@lists.openembedded.org>
2023-08-03 8:55 ` Alexander Kanavin
[not found] ` <1777D3B13579ECA3.27883@lists.openembedded.org>
2023-08-03 10:19 ` Alexander Kanavin
2023-08-03 11:13 ` Böszörményi Zoltán
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=2b5bc564-f05c-3c14-688e-50b3ba62eb2e@gmail.com \
--to=zboszor@gmail.com \
--cc=Ross.Burton@arm.com \
--cc=alex.kanavin@gmail.com \
--cc=alexandre.belloni@bootlin.com \
--cc=openembedded-core@lists.openembedded.org \
--cc=richard.purdie@linuxfoundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox