Hi Richard,
I have verified that it fails the reproducer also on master (poky: 23212c8c44
(HEAD, mirror/master) rpm: fix RPM_ETCCONFIGDIR value in SDK)
I don't think it hides anything.
The content of ${S} is completely irrelevant for the task except that it is
used to create paths to files based on "${S}/..".
Best Regards,
Peter
-----Original Message-----
From: Richard Purdie <richard.pur...@linuxfoundation.org>
Sent: Thursday, March 23, 2023 12:40
To: Marko, Peter (ADV D EU SK BFS1) <peter.ma...@siemens.com>;
openembedded-core@lists.openembedded.org
Subject: Re: [OE-core][master][langdale][kirkstone][PATCH] gcc-shared-source:
ensure that S exists in do_deploy_source_date_epoch
> On Thu, 2023-03-23 at 12:01 +0100, Peter Marko wrote:
> > From: Peter Marko <peter.ma...@siemens.com>
> >
> > This include overrides do_deploy_source_date_epoch function and is
> > additionally referencing ${S}.
> > If ${S} does not exist yet, the function will fail because it cannot
> > evaluate path .. from non-existing directory.
> >
> > Reproducer (verified in kirkstone):
> > bitbake gcc -c deploy_source_date_epoch
> > bitbake gcc -c cleansstate
> > rm -rf build/tmp
> > bitbake gcc -c deploy_source_date_epoch
> >
> > Signed-off-by: Peter Marko <peter.ma...@siemens.com>
> > ---
> > meta/recipes-devtools/gcc/gcc-shared-source.inc | 1 +
> > 1 file changed, 1 insertion(+)
> >
> > diff --git a/meta/recipes-devtools/gcc/gcc-shared-source.inc
> > b/meta/recipes-devtools/gcc/gcc-shared-source.inc
> > index cd2e341099..7c84ea088d 100644
> > --- a/meta/recipes-devtools/gcc/gcc-shared-source.inc
> > +++ b/meta/recipes-devtools/gcc/gcc-shared-source.inc
> > @@ -10,6 +10,7 @@ SRC_URI = ""
> > do_configure[depends] += "gcc-source-${PV}:do_preconfigure"
> > do_populate_lic[depends] += "gcc-source-${PV}:do_unpack"
> > do_deploy_source_date_epoch[depends] +=
> > "gcc-source-${PV}:do_deploy_source_date_epoch"
> > +do_deploy_source_date_epoch[dirs] += "${S}"
> >
> > # Copy the SDE from the shared workdir to the recipe workdir
> > do_deploy_source_date_epoch () {
>
> I have a suspicion this was fixed differently in master.
>
> The fix looks incorrect since it will just make the directory exist, which
> will hide the error but won't fix whatever the race is causing the underlying
> determinism problem.
>
> Can you confirm the problem exists on master to start with please?
> >
> Cheers,
>
> Richard
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#178979):
https://lists.openembedded.org/g/openembedded-core/message/178979
Mute This Topic: https://lists.openembedded.org/mt/97798316/21656
Group Owner: openembedded-core+ow...@lists.openembedded.org
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-
