openjdk team mailing list archive

[Julian Gilbey <jdg@xxxxxxxxxx>]

On Sun, Sep 20, 2020 at 09:59:22PM -0700, tony mancill wrote:
> Hi Tiago, hi Julian,
> [...]
> Julian, I applied the patch and built the package successfully, but
> jlink still fails with the "expected hash" error.  It's (perhaps)
> interesting that the expected hash does differ between the patched build
> and the unpatched build.

That's really bizarre.  I checked running strip-nondeterminism on the
resulting jmods and it seemed to be OK.  But if so, then I've clearly
overlooked something.

> Disabling the invocation of 'dh_strip_nondeterminism -a' in
> debian/rules *does* address the problem with jlink, either with or
> without the jmod patch applied.  Also, I had to add strip-determinism to
> the build-deps to build with the patch.

You should be able to make do with just adding -Xjmods to the
dh_strip_nondeterminism invocation.  Removing it entirely also
resolves this bug, but moves further away from the reproducible builds
target.

So I think there may be something weird going on in your build.  If
you go into the build directory with the patch applied and
dh_strip_nondeterminism -a still present, what is the result of
running the following command after doing a build (perhaps with
DEB_BUILD_OPTIONS=notest to save running the whole test suite)?

sha256sum `find . -name jdk.management.jfr.jmod`

You should find that all of the files found have identical SHA256
sums.

And yes, I did say that you needed to add strip-determinism to the
build-deps ;-)

> Given that the package doesn't build reproducibly anyway, my inclination
> is to make the change in debian/rules to address this long-standing bug.

Could we see if we can figure out why my patch is not working first?

Best wishes,

   Julian