We plan to release coreutils-8.31 in the coming week so any testing you can do on various different systems between now and then would be most welcome. -- You can download the coreutils snapshot in x
I was able to build it (and run the upstream test-suite) for Fedora 28+: https://copr.fedorainfracloud.org/coprs/kdudka/coreutils-next/build/864469/ Our internal static analysis service sees no newly
Hello Pádraig, Thanks for all the hard work, and for cleaning up and finalizing the 'env' patch. Some test results: No failures on the following: Debian unstable (sparc64) CentOS 7 (ppc64le) CentOS
Thanks for all the hard work, and for cleaning up and finalizing the 'env' patch. On AIX-7.2 compiling for 64bit with CC="xlc -q64", compilation fails with: == CCLD src/split CC src/stat.o "src/stat.
On NetBSD 7.1 / x86_64: first, the "tests/misc/tac-2-nonseekable" test seems to hang, and I had to kill the command "tac - -" to force the tests to continue. Then, the following failed: FAIL: tests/
On OpenSolaris, 5.10/5.11, both x86 and sparc, compilation fails with: == CCLD src/echo CC src/env.o src/env.c:62:25: error: variably modified 'signals' at file scope static enum SIGNAL_MODE signals
Thanks! Newer openSUSE:Leap-15 and SUSE:SLE-15 seem to be fine (some archs are still building), but the builds on all SUSE:SLE-12 versions and e.g. not so fresh openSUSE Leap_42.3 (released 2017-07-2
Hello Bernhard, Older makeinfo-4.13a seems to have problems with the addition of the macro '@{hashchar}' of commit v8.30-33-g2f438fa9f: I believe some time ago (ca. 2017) there was a change to requir
Cool, we should probably use Assaf's patch to avoid that. Though I also am surprised that makeinfo is needed at all. Before release I checked the extracted archive had appropriate timestamps. I used:
So SIGNUM_BOUND is not a constant. Interesting. For now we can deal with this using the attached. thanks! Pádraig Attachment: env-opensolaris.patch Description: Text Data
Thanks for all the testing! I think the wc-nbsp failure stems from FreeBSD's concept of printable, assuming nbsp is non printable. For example: https://mpc.lists.freebsd.questions.narkive.com/4m5jNqL
Interesting, AIX must have a separate statx. The attached should avoid this build issue. cheers, Pádraig Attachment: statx-aix.patch Description: Text Data
Worth spelling it "echo" instead of the more verbose "printf '\n'"? -- Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3226 Virtualization: qemu.org | libvirt.org Attachment: signatu
Hrm, I presume <&- is not working there to close stdin? The attached should detect and avoid this case. thanks, Pádraig Attachment: test-tac-hang.patch Description: Text Data
The tail-2/pipe-f.sh failure seems similarly due to the shell not supporting >&- The attached should avoid that. thanks, Pádraig Attachment: tail-pipe-netbsd-test.patch Description: Text Data
We plan to release coreutils-8.31 in the coming week so any testing you can do on various different systems between now and then would be most welcome. -- You can download the coreutils snapshot in x