Something isn't quite right with these packages currently. More investigation (potential toolchain problem? ) Needless to say, need to be looked into and fixed!
Imho gcc related. Does this happen to be I/O related?
exactly the problem is 3 fold:
1) some fine uninformed person hard wired ppc64 builds to be 32 bit ppc builds
2) sanza missing for 64 bit build settings
3) we've got a toolchain problem
the results fixing #1 and #2 are:
test 7 ok
error calculating RIPEMD160 on '12345678901234567890123456789012345678901234567890123456789012345678901234567890'
got 533600f779ddd1958449fbd3d7e38bac2f338621 instead of 9b752e45573d4b39f4dbd3323cab82bf63326bfb
die bug die!
removing -O2 fixes it for now ... changing to the 3.4 toolchain had no effect
I've mentioned this to alanm
things work great now!