On Wed, 2023-11-08 at 13:17 +, Jose Quaresma wrote:
Joakim Tjernlund
mailto:joakim.tjernl...@infinera.com>> escreveu
no dia quarta, 8/11/2023 à(s) 11:50:
On Wed, 2023-11-08 at 11:22 +, Jose Quaresma wrote:
Hi Joakim,
Joakim Tjernlund via lists.openembedded.or
On Tue, 2024-01-23 at 20:14 +0100, Alexander Kanavin wrote:
> On Tue, 23 Jan 2024 at 20:11, Joakim Tjernlund
> wrote:
> >
> > On Tue, 2024-01-23 at 20:02 +0100, Alexander Kanavin wrote:
> > > Host tools can influence reproducibility of target items too, so I
> > > would not want to use the same
I setup above test for reproducible builds and build one pkg(u-boot) and try to
force a
non-reproducible build to test the framework:
--- a/meta/recipes-bsp/u-boot/u-boot.inc
+++ b/meta/recipes-bsp/u-boot/u-boot.inc
@@ -9,7 +9,7 @@ inherit uboot-config uboot-extlinux-config uboot-sign deploy
avin wrote:
> The short execution time is suspicious. The test is written so that
> build B doesn't reuse the cache from build A, and if it does because
> of your local tweaks, then you're probably just comparing the A
> package with itself.
>
> Alex
>
> On Mon, 18 Dec 2
ion time is suspicious. The test is written so that
> > > build B doesn't reuse the cache from build A, and if it does because
> > > of your local tweaks, then you're probably just comparing the A
> > > package with itself.
> > >
> > > Alex
> > &g
Yes, there is an xxx-st where the reproducible build is performed and there is
no A/B dies in there.
I added some logging:
--- a/meta/lib/oeqa/selftest/cases/reproducible.py
+++ b/meta/lib/oeqa/selftest/cases/reproducible.py
@@ -205,8 +205,8 @@ class ReproducibleTests(OESelftestTestCase):
On Wed, 2023-11-29 at 21:37 +, Richard Purdie wrote:
> On Wed, 2023-11-29 at 07:17 -1000, Steve Sakoman wrote:
> > On Wed, Nov 29, 2023 at 1:11 AM Joakim Tjernlund via
> > lists.openembedded.org
> > wrote:
> > >
> > > Hi Peter :)
> > >
> &g
If /etc/passwd is a symlink, sed -i on same file will replace the
symlink with a new file. Prevent that by adding --follow-symlinks
option to sed
Signed-off-by: Joakim Tjernlund
---
meta/classes/rootfs-postcommands.bbclass | 4 ++--
meta/classes/useradd_base.bbclass| 2 +-
2 files
These patches adds the possibility to store passwd/shadow files
in a sub dir, like /etc/pwdb
In a RO Root FS one can bind mount a writeable dir on /etc/pwdb
to support password changes etc.
Joakim Tjernlund (4):
[meta classes] sed -i destroys symlinks
base-passwd: Add PW_SUBDIR
pseudo: Add
Add support for creating passwd files in a /etc subdir
Set PW_SUBIR to pwdb to get passwd files in /etc/pwdb
Signed-off-by: Joakim Tjernlund
---
.../base-passwd/base-passwd_3.5.29.bb | 24 ---
1 file changed, 16 insertions(+), 8 deletions(-)
diff --git
Add support for creating passwd files in a /etc subdir
Set PW_SUBIR to pwdb to get passwd files in /etc/pwdb
Signed-off-by: Joakim Tjernlund
---
meta/recipes-devtools/pseudo/pseudo.inc | 11 ---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git
Add support for creating passwd files in a /etc subdir
Set PW_SUBIR to pwdb to get passwd files in /etc/pwdb
Signed-off-by: Joakim Tjernlund
---
.../0001-Define-SUBUID_FILE-SUBGID_FILE.patch | 92 +++
meta/recipes-extended/shadow/shadow.inc | 30 +-
2 files changed,
cke
> >
> > On Sun, 2023-11-26 at 21:21 +, Peter Kjellerstedt wrote:
> > > > -Original Message-
> > > > From: openembedded-core@lists.openembedded.org > c...@lists.openembedded.org> On Behalf Of Joakim Tjernlund via
> > lists.openembedded.org
t; > Hi Peter :)
> > > >
> > > > All good comments, will fix accordingly. Not sure how PW_DIR ?=
> > > > "${sysconfdir}" will work though.
> > > >
> > > > How do you envision one should set PW_DIR in distro .conf or
> > l
On Wed, 2023-11-29 at 11:56 +, Ross Burton wrote:
> On 14 Nov 2023, at 20:34, Joakim Tjernlund via lists.openembedded.org
> wrote:
> >
> > These patches adds the possibility to store passwd/shadow files
> > in a sub dir, like /etc/pwdb
> > In a RO Root FS one
23-11-26 at 21:21 +, Peter Kjellerstedt wrote:
> > -Original Message-
> > From: openembedded-core@lists.openembedded.org
> > On Behalf Of Joakim Tjernlund
> > via lists.openembedded.org
> > Sent: den 24 november 2023 15:11
> > To: openembedded-core
In poky/meta/lib/oeqa/selftest/context.py there is:
if "buildhistory.bbclass" in self.tc.td["BBINCLUDED"]:
self.tc.logger.error("You have buildhistory enabled already and
this isn't recommended for selftest, please disable it first.")
raise OEQAPreRun
if
On Fri, 2024-01-26 at 14:33 +0100, Alexander Kanavin via lists.openembedded.org
wrote:
> On Fri, 26 Jan 2024 at 14:21, Joakim Tjernlund
> wrote:
> > I will try workaround these somehow but will not go so far as creating a
> > new MACHINE just to
> > run these tests.
>
> Why would you need a
On Tue, 2024-01-23 at 20:02 +0100, Alexander Kanavin wrote:
> Host tools can influence reproducibility of target items too, so I
> would not want to use the same set in both builds. Checking for
> reproducibility is simply something that needs a powerful machine.
We would like to start gently,
On Fri, 2024-01-26 at 14:33 +0100, Alexander Kanavin via lists.openembedded.org
wrote:
> On Fri, 26 Jan 2024 at 14:21, Joakim Tjernlund
> wrote:
> > I will try workaround these somehow but will not go so far as creating a
> > new MACHINE just to
> > run these tests.
>
> Why would you need a
On Fri, 2024-01-26 at 15:55 +0100, Alexander Kanavin wrote:
> On Fri, 26 Jan 2024 at 15:52, Joakim Tjernlund
> wrote:
> > A thought just occurred to me, these vars(SANITY_TESTED_DISTROS,
> > buildhistory etc.),
> > do they need to be off in the original machine when building up the SSTATE ?
> >
On Mon, 2024-01-22 at 21:39 +0100, Alexander Kanavin wrote:
> On Mon, 22 Jan 2024 at 17:49, Joakim Tjernlund
> wrote:
> >
> > On Mon, 2024-01-22 at 17:19 +0100, Alexander Kanavin wrote:
> > > On Mon, 22 Jan 2024 at 16:59, Joakim Tjernlund
> > > wrote:
> > > > Maybe but what about the missing
Whatever I do, ReproducibleTests always succeeds and I think I got a clue.
Looking at the logs below and the files after build one can see thate there
are no ...reproducibleA/tmp or ...reproducibleB/tmp dirs in the FS.
Furthermore, deploy_A and deploy_B are the same. This all ends up with the
what about the missing reproducibleA/tmp and reproducibleB/tmp ?
Would these be missing if sstate is used for both?
In the log there is this:
INFO - Building reproducibleB (sstate NOT allowed)...
Jocke
>
> Alex
>
> On Mon, 22 Jan 2024 at 16:01, Joakim Tjernlund via
> lists.openembedded.or
On Mon, 2024-01-22 at 17:19 +0100, Alexander Kanavin wrote:
> On Mon, 22 Jan 2024 at 16:59, Joakim Tjernlund
> wrote:
> > Maybe but what about the missing reproducibleA/tmp and reproducibleB/tmp ?
> > Would these be missing if sstate is used for both?
> > In the log there is this:
> > INFO -
On Fri, 2024-01-26 at 16:18 +0200, Mikko Rapeli wrote:
> Hi,
>
> On Fri, Jan 26, 2024 at 02:04:18PM +0000, Joakim Tjernlund via
> lists.openembedded.org wrote:
> > On Fri, 2024-01-26 at 14:33 +0100, Alexander Kanavin via
> > lists.openembedded.org wrote:
> >
On Fri, 2024-01-26 at 13:18 +0100, Alexander Kanavin wrote:
> On Fri, 26 Jan 2024 at 12:09, Joakim Tjernlund via
> lists.openembedded.org
> wrote:
> > These are hardcoded settings too of which some could need need tewaking,
> > DISTRO_FEATURES for sure.
> >
>
On Wed, 2023-11-15 at 14:45 +0100, Alexander Kanavin wrote:
> On Wed, 15 Nov 2023 at 14:42, Joakim Tjernlund
> wrote:
> > > Pending... what exactly?
> >
> > Pending merge
>
> If a patch is pending merge, the correct status is 'Submitted' (with a
> link). If it's already been merged, the correct
On Wed, 2023-11-15 at 13:35 +0100, Alexander Kanavin wrote:
> On Tue, 14 Nov 2023 at 21:36, Joakim Tjernlund via
> lists.openembedded.org
> wrote:
> > Add support for creating passwd files in a /etc subdir
> > +Subject: [PATCH] Define SUBUID_FILE/SUBGID_FILE
> > +
&g
If /etc/passwd is a symlink, sed -i on same file will replace the
symlink with a new file. Prevent that by adding --follow-symlinks
option to sed
Signed-off-by: Joakim Tjernlund
---
- v2: Use --follow-symlinks rather than realpath
meta/classes/rootfs-postcommands.bbclass | 4 ++--
Add support for creating passwd files in a /etc subdir
Set PW_SUBIR to pwdb to get passwd files in /etc/pwdb
Signed-off-by: Joakim Tjernlund
---
.../0001-Define-SUBUID_FILE-SUBGID_FILE.patch | 92 +++
meta/recipes-extended/shadow/shadow.inc | 30 +-
2 files changed,
Add support for creating passwd files in a /etc subdir
Set PW_SUBIR to pwdb to get passwd files in /etc/pwdb
Signed-off-by: Joakim Tjernlund
---
.../base-passwd/base-passwd_3.5.29.bb | 24 ---
1 file changed, 16 insertions(+), 8 deletions(-)
diff --git
These patches adds the possibility to store passwd/shadow files
in a sub dir, like /etc/pwdb
In a RO Root FS one can bind mount a writeable dir on /etc/pwdb
to support password changes etc.
This patchset is to probe wether OE would be interested in such feature
Joakim Tjernlund (3):
Add support for creating passwd files in a /etc subdir
Set PW_SUBIR to pwdb to get passwd files in /etc/pwdb
Signed-off-by: Joakim Tjernlund
---
meta/recipes-devtools/pseudo/pseudo.inc | 11 ---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git
If /etc/passwd is a symlink, sed -i on same file will replace the
symlink with a new file. Prevent that by expanding the file with realpath
before giving it to sed.
Signed-off-by: Joakim Tjernlund
---
- I have patches which moves shadow/passwd files to an subdir in order to
have passwd mgmt
On Wed, 2023-11-08 at 11:22 +, Jose Quaresma wrote:
Hi Joakim,
Joakim Tjernlund via lists.openembedded.org<http://lists.openembedded.org/>
mailto:infinera@lists.openembedded.org>>
escreveu no dia quarta, 8/11/2023 à(s) 10:56:
If /etc/passwd is a symlink, sed -i on same file
36 matches
Mail list logo
