Test fails on all big-endian architectures

[skitt]

Hi Claudio,

In 4.2.1 the tests fail on all big-endian architectures Debian builds for (mips, powerpc, s390x, sparc); see https://buildd.debian.org/status/package.php?p=libxmp for the build logs (not very detailed on the test side). I'll look into the failures some more as time permits!

Regards,

Stephen

[cmatsuoka]

Hi Stephen,

Thanks, this is probably caused by a change in the way the test is
performed (it now checks the MD5 digest of the generated data instead of
comparing it to a large raw PCM dump), to save about 200KB in the
distribution package. I'll set up a big-endian system emulator here to fix
that and provide a patch.

Best regards,
Claudio

On Tue, Dec 31, 2013 at 12:35 PM, Stephen Kitt notifications@github.comwrote:

Hi Claudio,

In 4.2.1 the tests fail on all big-endian architectures Debian builds for
(mips, powerpc, s390x, sparc); see
https://buildd.debian.org/status/package.php?p=libxmp for the build logs
(not very detailed on the test side). I'll look into the failures some more
as time permits!

Regards,

Stephen

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/9
.

[cmatsuoka]

Hi,

It was indeed a bug in the test validation. The attached patch should fix
it.

Best regards,
Claudio

On Tue, Dec 31, 2013 at 12:35 PM, Stephen Kitt notifications@github.comwrote:

Hi Claudio,

In 4.2.1 the tests fail on all big-endian architectures Debian builds for
(mips, powerpc, s390x, sparc); see
https://buildd.debian.org/status/package.php?p=libxmp for the build logs
(not very detailed on the test side). I'll look into the failures some more
as time permits!

Regards,

Stephen

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/9
.

[skitt]

Excellent, that does indeed fix the validation. Thanks!