Runner 18522 exiting Runner 18520 exiting Runner 18515 exiting make[1]: *** [Makefile:874: quiet-test] Error 1 make[1]: Leaving directory '/var/tmp/portage/net-misc/curl-8.8.0/work/curl-8.8.0-abi_x86_64.amd64/tests' make: *** [Makefile:1678: test] Error 2 * ERROR: net-misc/curl-8.8.0::gentoo failed (test phase): * emake failed ------------------------------------------------------------------- This is an unstable amd64 chroot image at a tinderbox (==build bot) name: 23.0_desktop_systemd_test-20240518-162232 The attached etc.portage.tar.xz has all details. ------------------------------------------------------------------- gcc-config -l: [1] x86_64-pc-linux-gnu-14 * clang/llvm (if any): clang version 18.1.5 Target: x86_64-pc-linux-gnu Thread model: posix InstalledDir: /usr/lib/llvm/18/bin Configuration file: /etc/clang/x86_64-pc-linux-gnu-clang.cfg /usr/lib/llvm/18 18.1.6 Python 3.11.9 Available Ruby profiles: [1] ruby31 (with Rubygems) [2] ruby32 (with Rubygems) [3] ruby33 (with Rubygems) * Available Rust versions: [1] rust-bin-1.77.1 * php cli (if any): go version go1.22.3 linux/amd64 HEAD of ::gentoo commit 7ec4d51c1b449743ff606a27d27e3e623d092c3d Author: Repository mirror & CI <repomirrorci@gentoo.org> Date: Fri May 24 15:48:49 2024 +0000 2024-05-24 15:48:48 UTC emerge -qpvO net-misc/curl [ebuild U ] net-misc/curl-8.8.0 [8.7.1-r4] USE="adns alt-svc ftp gnutls* hsts http2* imap mbedtls* openssl pop3 psl rtmp* rustls* samba* smtp ssh* ssl test* tftp websockets* -brotli -gopher -idn -kerberos -ldap -nghttp3 -progress-meter* (-sslv3) -static-libs -telnet -verify-sig -zstd" ABI_X86="(64) -32 (-x32)" CURL_SSL="openssl -gnutls -mbedtls -rustls"
Created attachment 894287 [details] emerge-info.txt
Created attachment 894288 [details] emerge-history.txt
Created attachment 894289 [details] environment
Created attachment 894290 [details] etc.clang.tar.xz
Created attachment 894291 [details] etc.portage.tar.xz
Created attachment 894292 [details] logs.tar.xz
Created attachment 894293 [details] net-misc:curl-8.8.0:20240524-191722.log.xz
Created attachment 894294 [details] qlist-info.txt
Created attachment 894295 [details] temp.tar.xz
Created attachment 894296 [details] tests.tar.xz
I'm seeing this too.
I saw this when doing mbedtls patching on one machine, but _not_ on the machine where I first packaged this. Daniel and Jay over at cURL, and their CI aren't seeing it either. iirc this is an IMAP test. What are _we_ doing that's different here?
I'm not quite sure what's happening here. The test seems to be hanging. If I run it with -network-sandbox it consistently passes (can anyone validate?), but I also have one machine where it just works all the time. I've raised the issue upstream.
Just to dump how I bisected in case someone finds it useful: """ [23:03] <sam_> for bisecting, I'm doing https://wiki.gentoo.org/wiki/Bisecting_with_live_ebuilds [23:03] <sam_> i.e. [23:03] <sam_> cloned curl [23:03] <sam_> [23:03] <sam_> /tmp/bisect.sh [23:03] <sam_> #!/bin/bash [23:03] <sam_> export EGIT_OVERRIDE_COMMIT_CURL_CURL=$(git rev-parse HEAD) [23:03] <sam_> export EGIT_OVERRIDE_BRANCH_CURL_CURL=master [23:03] <sam_> export ACCEPT_KEYWORDS="**" [23:03] <sam_> FEATURES="network-sandbox" CFLAGS="-O2" LDFLAGS="-Wl,-O1" CC=gcc-13 e /home/sam/git/gentoo/net-misc/curl/curl-9999.ebuild clean compile || exit 125 [23:03] <sam_> FEATURES="network-sandbox" CFLAGS="-O2" LDFLAGS="-Wl,-O1" CC=gcc-13 e /home/sam/git/gentoo/net-misc/curl/curl-9999.ebuild test [23:03] <sam_> if grep -q 'TESTFAIL: These test cases failed: 1553' /var/tmp/portage/net-misc/curl-9999/temp/build.log ; then [23:03] <sam_> exit 1 [23:03] <sam_> fi [23:03] <sam_> exit 0 [23:03] <sam_> [23:03] <sam_> git bisect bad master [23:03] <sam_> git bisect bad curl-8_8_0 [23:03] <sam_> git bisect good curl-8_7_1 [23:04] <sam_> chmod +x /tmp/bisect.sh [23:04] <sam_> git bisect run /tmp/bisect.sh [23:04] <sam_> with the ebuild modified as: [23:04] <sam_> - multilib_is_native_abi && emake test TFLAGS="-n -v -a -k -am -p -j$((2*$(makeopts_jobs))) !241 !1083" [23:04] <sam_> + multilib_is_native_abi && emake test TFLAGS="-n -v -a -k -am -p -j1 !241 !1083 1553" [23:04] <sam_> (to run only 1553) [23:05] <sam_> """ which then gave: """ [23:08] <sam_> 2d2c27e5a319665fdfbe75422dac55ff3a2218b6 is the first bad commit [23:08] <sam_> commit 2d2c27e5a319665fdfbe75422dac55ff3a2218b6 [23:08] <sam_> Author: Stefan Eissing <stefan@eissing.org> [23:08] <sam_> Date: Wed Mar 20 08:08:43 2024 +0100 [23:08] <sam_> multi: multi_wait improvements """ It indeed passes for me with FEATURES="-network-sandbox" or running the test manually outside of the ebuild. It consistently fails for me with network-sandbox on.
The bug has been closed via the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=b4a094f4165eba939dc23b9af9c660ab3b45f449 commit b4a094f4165eba939dc23b9af9c660ab3b45f449 Author: Matt Jolly <kangie@gentoo.org> AuthorDate: 2024-05-30 00:30:56 +0000 Commit: Matt Jolly <kangie@gentoo.org> CommitDate: 2024-05-30 02:02:10 +0000 net-misc/curl: add 8.8.0-r1 This release fixes an issue where under certain circumstances (in particular when invoked within the Gentoo sandbox), `multi_wait()` would incorrectly calculate the timeout, resulting in test failures. Allow setting `--enable-debug` via USE=debug with an appropriate warning to the ebuild and live ebuild. Closes: https://bugs.gentoo.org/932660 See-also: https://github.com/curl/curl/issues/13782 Signed-off-by: Matt Jolly <kangie@gentoo.org> net-misc/curl/curl-8.8.0-r1.ebuild | 387 +++++++++++++++++++++ net-misc/curl/curl-9999.ebuild | 16 +- .../curl/files/curl-8.8.0-multi_wait-timeout.patch | 75 ++++ 3 files changed, 477 insertions(+), 1 deletion(-)
