Re: [OE-core] [PATCH] abi-version/ssate: Bump to avoid systemd hash corruption issue

[Richard Purdie <richard.purdie@linuxfoundation.org>]

On Sun, 2024-09-01 at 10:36 +0000, Peter Kjellerstedt wrote:
> > -----Original Message-----
> > From:
> > openembedded-core@lists.openembedded.org <openembedded-core@lists.openembedded.org
> > > On Behalf Of Richard Purdie
> > Sent: den 1 september 2024 09:19
> > To: openembedded-core@lists.openembedded.org
> > Subject: [OE-core] [PATCH] abi-version/ssate: Bump to avoid systemd
> > hash corruption issue
> > 
> > Unfortunately some recent patches caused non-deterministic output.
> > One input hash lead to both good and bad output and whilst that
> > patch
> > has been fixed, the problematic hash 'cross' linkage remains. Bump
> > to
> > a new sstate and hash equivalence version to avoid this and work
> > from
> > a clean slate.
> > 
> > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> > ---
> >  meta/classes-global/sstate.bbclass | 2 +-
> >  meta/conf/abi_version.conf         | 2 +-
> >  2 files changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/meta/classes-global/sstate.bbclass b/meta/classes-
> > global/sstate.bbclass
> > index beb22f424e8..fdd529ee4e8 100644
> > --- a/meta/classes-global/sstate.bbclass
> > +++ b/meta/classes-global/sstate.bbclass
> > @@ -4,7 +4,7 @@
> >  # SPDX-License-Identifier: MIT
> >  #
> > 
> > -SSTATE_VERSION = "12"
> > +SSTATE_VERSION = "14"
> > 
> >  SSTATE_ZSTD_CLEVEL ??= "8"
> > 
> > diff --git a/meta/conf/abi_version.conf
> > b/meta/conf/abi_version.conf
> > index 0fe91b5ddf3..ac97752c976 100644
> > --- a/meta/conf/abi_version.conf
> > +++ b/meta/conf/abi_version.conf
> > @@ -12,4 +12,4 @@ OELAYOUT_ABI = "15"
> >  # a reset of the equivalence, for example when reproducibility
> > issues break the
> >  # existing match data. Distros can also append to this value for
> > the same effect.
> >  #
> > -HASHEQUIV_HASH_VERSION  = "17"
> > +HASHEQUIV_HASH_VERSION  = "18"
> 
> Is this actually needed now after the changes to insane.bbclass?
> Won't those changes effectively invalidate the whole sstate anyway?

Firstly, those changes don't change the output of the tasks so no, the
mismatched output hashes would have remained a problem.

Secondly, those had already merged and been tested alongside the
systemd changes so it was already too late anyway.

So sadly I just can't see an alternative to this due to where the
corruption is.

Cheers,

Richard