;;; Loading src:lsp;evalmacros.lsp ;;; Loading src:lsp;cmuutil.lsp ;;; Loading src:lsp;setf.lsp/bin/sh: line 4: 22882 Segmentation fault ECLDIR=`pwd`/ ./ecl_min compile make[1]: *** [Makefile:241: bin/ecl] Error 139 make[1]: Leaving directory '/var/tmp/portage/dev-lisp/ecls-16.1.3/work/ecl-16.1.3/build' make: *** [Makefile:72: all] Error 2 ------------------------------------------------------------------- This is an unstable amd64 chroot image at a tinderbox (==build bot) name: 17.0-systemd_abi32+64_20181020-092847 ------------------------------------------------------------------- Please see the tracker bug for details. gcc-config -l: [1] x86_64-pc-linux-gnu-8.2.0 * Available Python interpreters, in order of preference: [1] python3.6 [2] python2.7 (fallback) Available Ruby profiles: [1] ruby23 (with Rubygems) * emerge -qpvO dev-lisp/ecls [ebuild N ] dev-lisp/ecls-16.1.3 USE="cxx gengc libatomic threads unicode -X -debug -emacs -precisegc" CPU_FLAGS_X86="sse"
Created attachment 552054 [details] emerge-info.txt
Created attachment 552056 [details] dev-lisp:ecls-16.1.3:20181021-070537.log
Created attachment 552058 [details] emerge-history.txt
Created attachment 552060 [details] environment
Created attachment 552062 [details] etc.portage.tbz2
Created attachment 552064 [details] logs.tbz2
Created attachment 552066 [details] temp.tbz2
@common-lisp: any idea what's going on here?
1) This is more likely a gcc-8 problem than a glibc problem, I suspect. 2) I can't reproduced it anywhere (not on amd64, not on ~amd64) So let's assume it has been fixed. Somewhere.
it is still an issue at unstable amd64 tinderbox image 17.1-no-multilib_libressl_20190401-161804
Created attachment 571810 [details] emerge-info.txt
Created attachment 571812 [details] dev-lisp:ecls-16.1.3:20190404-064804.log
Created attachment 571814 [details] emerge-history.txt
Created attachment 571816 [details] environment
Created attachment 571818 [details] etc.portage.tbz2
Created attachment 571820 [details] logs.tbz2
Created attachment 571822 [details] temp.tbz2
I am encountering this issue on a system using x86_64-pc-linux-gnu-8.2.0, and sys-libs/glibc-2.29-r2. Will attach the relevant files.
Created attachment 577908 [details] emerge pqv output
Created attachment 577910 [details] emerge --info output
Created attachment 577912 [details] emerge build.log
Created attachment 577914 [details] emerge environment
I do not encounter this issue when building from the latest git sources. https://gitlab.com/embeddable-common-lisp/ecl/ 7fd97549d66f3fdeefa0ce304ce925af2e1e0bd6
(In reply to Andreas K. Hüttel from comment #8) > @common-lisp: any idea what's going on here? After a few tests, the combination USE="gengc threads" makes the build to fail (USE="-libatomic" fails too). The pkg_setup phase says the gengc and preceisegc flags are both experimental.
*** Bug 735656 has been marked as a duplicate of this bug. ***
With ecls-21.2.1, the generational and precise garbage collectors are considered stable. ECL now builds with USE="gengc" although the result still crashes while running SageMath's maxima tests. USE=threads is irrelevant. USE=precisegc works fine. Suggestion: mask USE=gengc for now, or put back the pkg_setup() warning about it (and direct bug reports upstream).