The backport of JDK-8232102 to the 8u Shenandoah port - specifically the use of byte_size_in_proper_unit - broke the build on s390 (31-bit). This is the recurring problem that size_t and uint are not the same on this platform.

Hence we get several format warnings such as:

/builddir/build/BUILD/java-1.8.0-openjdk-1.8.0.242.b05-0.1.ea.el7.s390/openjdk/hotspot/src/share/vm/gc_implementation/shenandoah
/shenandoahHeapRegion.cpp: In static member function 'static void ShenandoahHeapRegion::setup_sizes(size_t)':
/builddir/build/BUILD/java-1.8.0-openjdk-1.8.0.242.b05-0.1.ea.el7.s390/openjdk/hotspot/src/share/vm/gc_implementation/shenandoah
/shenandoahHeapRegion.cpp:482:124: error: format '%zu' expects argument of type 'size_t', but argument 6 has type 'unsigned int'
 [-Werror=format=]
                       byte_size_in_proper_unit(ShenandoahMinRegionSize), proper_unit_for_byte_size(ShenandoahMinRegionSize));

Forcing byte_size_in_proper_unit to use size_t, as we did before in https://hg.openjdk.java.net/aarch64-port/jdk8u-shenandoah/hotspot/rev/ad280373e7cc, fixes the problem.