When building latest e2fsprogs on big endian platforms (ppc64, s390x) 2 tests are failing ... u_mke2fs_opt_offset: e2undo and mke2fs with offset option: failed u_offset: e2undo with the offset option (-o offset): failed ... for full logs please see eg. http://s390.koji.fedoraproject.org/koji/taskinfo?taskID=2255925 Version-Release number of selected component (if applicable): e2fsprogs-1.43.1-1.fc25
Created attachment 1166559 [details] u_mke2fs_opt_offset.failed
Created attachment 1166566 [details] u_mke2fs_opt_offset.log
Created attachment 1166585 [details] u_offset.failed
Created attachment 1166587 [details] u_offset.log
It's quite possible that these are not regressions; they are new tests: # git show cc090a3634b6a43979c7242a80c4d457b7a08ae2 | lsdiff b/tests/u_mke2fs_opt_offset/script b/tests/u_offset/script # git describe --contains cc090a3634b6a43979c7242a80c4d457b7a08ae2 v1.43~9
I haven't yet found the root cause; I thought I found an endian conversion problem in the utility, but it did not resolve this issue.
Should be fixed in rawhide now.
Thanks, successfully built in s390 and ppc kojis.