I've received a large set of build and test results from a user who wishes to remain anonymous. That set includes configure/make/make-check results for 163 system/compiler/env combinations. There's no way I can triage all of that before release. Maybe others who care about some of the affected platforms will lend a hand.
However, there are a few things to be aware of: * there are some openbsd failures in these logs, but they should be fixed, now. * Compilation errors due to remove.c's declaration-after-code should be ignored. The c99-to-c89 patch mentioned in README was not applied. * some failures are due to ZFS ACLs via Linux NFS and cp/mv reporting failure to set permissions. This is a known problem. * I'm sure there are more, ... The lzma-compressed tarball is under 400KB, but the uncompressed logs are over 30MB. http://meyering.net/cu/build-logs/coreutils-build-log-6.10.183-512c9.tar.lzma I'm publishing this mainly so that people who care about the affected platforms can investigate further. If there's something really important hiding in all of those logs, I might delay the coreutils release further, but it's not likely. Here's what I hope will be the final snapshot before 6.11: ================================================= coreutils snapshot: http://meyering.net/cu/coreutils-ss.tar.gz 8.9 MB http://meyering.net/cu/coreutils-ss.tar.lzma 3.6 MB http://meyering.net/cu/coreutils-ss.tar.gz.sig http://meyering.net/cu/coreutils-ss.tar.lzma.sig aka http://meyering.net/cu/coreutils-6.10.194-0d03b.tar.gz http://meyering.net/cu/coreutils-6.10.194-0d03b.tar.lzma Changes since 6.10.183-512c9: Jim Meyering (10): doc: fix typo seq: work around floating point inaccuracies on more systems tests: accommodate a different errno string on Irix 6.5 tests: don't fail on systems without a "stat" syscall md5sum, sha1sum, etc: handle invalid input (i.e., don't segfault) tests: avoid mkdir/selinux failure when mknod is a shell built-in tests: add a comment explaining the potential failure avoid "may be used uninitialized" warning from newer gcc mknod --help: note that this command may be a shell built-in tests: accommodate built-in mknod more cleanly Matthew Woehlke (1): tests: accommodate built-in mknod more cleanly still
