[Debian-med-packaging] Endianness of libsis-base [Was: failed powerpc build of libsis-base-java 14.12.0-2]

Andreas Tille andreas at an3as.eu
Wed Dec 9 22:18:02 UTC 2015


Hi Bernd,

thanks a lot for your wuick and precise response.

I think I was able to fix this by the following patch against the helper
Makefile.am craftet by the other Debian developers:

$ git diff HEAD^ Makefile.am 
diff --git a/debian/build_native/Makefile.am b/debian/build_native/Makefile.am
index 494500a..6f00bb8 100644
--- a/debian/build_native/Makefile.am
+++ b/debian/build_native/Makefile.am
@@ -1,4 +1,5 @@
-AM_CFLAGS=-I/usr/lib/jvm/default-java/include -I/usr/lib/jvm/default-java/include/linux -DMACHINE_BYTE_ORDER=1
+MBO=$(shell if [ `dpkg-architecture -qDEB_BUILD_ARCH_ENDIAN` = little ] ; then echo 1 ; else echo 2; fi )
+AM_CFLAGS=-I/usr/lib/jvm/default-java/include -I/usr/lib/jvm/default-java/include/linux -DMACHINE_BYTE_ORDER=$(MBO)
 
 lib_LTLIBRARIES = libcisd_unix.la libcisd_nativedata.la
 libcisd_unix_la_SOURCES = unix.c


Kind regards

        Andreas.

On Wed, Dec 09, 2015 at 08:12:51PM +0100, Bernd Rinn wrote:
> Hi Andreas,
> 
> There is no automatic detection of the endiness of a platform in
> libsis-base, but this is a compile time setting for the C library. I
> assume that you will find somewhere in your build scripts for PowerPC
> 
>    -DMACHINE_BYTE_ORDER=1
> 
> which means: platform is 'little endian', when you actually should have
> there
> 
>    -DMACHINE_BYTE_ORDER=2
> 
> which stands for 'platform is big endian'. The particular test that you
> found is failing checks whether the platform really has the endiness
> that was specified in the build file.

I can confirm that the full build log

   https://buildd.debian.org/status/fetch.php?pkg=libsis-base-java&arch=powerpc&ver=14.12.0-2&stamp=1449660924&file=log

exactly shows -DMACHINE_BYTE_ORDER=1.  However, it is not really clear
to me how to tweak the

> All the best,
> 
> Bernd
> 
> On 2015-12-09 12:57, Andreas Tille wrote:
> > Hi,
> > 
> > the build problem is in the test suite:
> > 
> > Running testPlatformEndiness
> > Exception in thread "main" java.lang.AssertionError: expected:<LITTLE_ENDIAN> but was:<BIG_ENDIAN>
> > 	at org.testng.AssertJUnit.fail(AssertJUnit.java:59)
> > 	at org.testng.AssertJUnit.failNotEquals(AssertJUnit.java:364)
> > 	at org.testng.AssertJUnit.assertEquals(AssertJUnit.java:80)
> > 	at org.testng.AssertJUnit.assertEquals(AssertJUnit.java:88)
> > 	at ch.systemsx.cisd.base.convert.NativeDataTests.testPlatformEndiness(NativeDataTests.java:377)
> > 	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
> > 	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
> > 	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> > 	at java.lang.reflect.Method.invoke(Method.java:606)
> > 	at ch.systemsx.cisd.base.convert.NativeDataTests.main(NativeDataTests.java:446)
> > 	at ch.systemsx.cisd.base.AllTests.main(AllTests.java:39)
> > debian/rules:45: recipe for target 'override_dh_auto_test-arch' failed
> > 
> > I wonder whether there is any restriction to little endian architectures.
> > 
> > Kind regards
> > 
> >      Andreas.
> > 
> > ----- Forwarded message from Debian buildds <noreply at buildd.debian.org> -----
> > 
> > From: Debian buildds <noreply at buildd.debian.org>
> > To: dispatch at tracker.debian.org
> > Subject: failed powerpc build of libsis-base-java 14.12.0-2
> > 
> >  * Source package: libsis-base-java
> >  * Version: 14.12.0-2
> >  * Architecture: powerpc
> >  * State: failed
> >  * Suite: sid
> >  * Builder: powerpc-unicamp-01.debian.org
> >  * Build log: https://buildd.debian.org/status/fetch.php?pkg=libsis-base-java&arch=powerpc&ver=14.12.0-2&stamp=1449660924&file=log
> > 
> > Please note that these notifications do not necessarily mean bug reports
> > in your package but could also be caused by other packages, temporary
> > uninstallabilities and arch-specific breakages.  A look at the build log
> > despite this disclaimer would be appreciated however.
> > 
> > 
> > ----- End forwarded message -----
> 

-- 
http://fam-tille.de



More information about the Debian-med-packaging mailing list