Bummer, sandbox file does not exist: /var/tmp/portage/dev-java/gwt-2.8.2/temp/sandbox.log ------------------------------------------------------------------- This is an unstable amd64 chroot image at a tinderbox (==build bot) name: 17.1_hardened-20201219-112410 ------------------------------------------------------------------- gcc-config -l: [1] x86_64-pc-linux-gnu-7.3.1 [2] x86_64-pc-linux-gnu-8.3.1 [3] x86_64-pc-linux-gnu-10.2.0 * clang version 11.0.0 Target: x86_64-pc-linux-gnu Thread model: posix InstalledDir: /usr/lib/llvm/11/bin /usr/lib/llvm/11 11.0.0 Available Python interpreters, in order of preference: [1] python3.9 [2] python3.8 (fallback) [3] python3.6 (fallback) [4] python2.7 (fallback) [5] pypy3 (fallback) Available Ruby profiles: (none found) Available Rust versions: [1] rust-1.48.0 * The following VMs are available for generation-2: 1) JamVM JDK 2.0.0 [jamvm] *) AdoptOpenJDK 8.272_p10 [openjdk-bin-8] Available Java Virtual Machines: [1] jamvm [2] openjdk-bin-8 system-vm The Glorious Glasgow Haskell Compilation System, version 8.8.4 timestamp(s) of HEAD at this tinderbox image: /var/db/repos/gentoo Wed Dec 30 09:05:18 AM UTC 2020 emerge -qpvO dev-java/gwt [ebuild N ] dev-java/gwt-2.8.2 USE="-source"
Created attachment 680275 [details] emerge-info.txt
Created attachment 680278 [details] dev-java:gwt-2.8.2:20201230-093821.log
Created attachment 680281 [details] emerge-history.txt
Created attachment 680284 [details] etc.portage.tbz2
Created attachment 680287 [details] logs.tbz2
Any idea on what could fix this, at least until a fix is pushed? It's preventing Rstudio being installed. It would be *much* appreciated :)
I have similar sandboxing issue with different software, normally when you run emerge the HOME is set to "${PORTAGE_BUILDDIR}/homedir" (https://devmanual.gentoo.org/ebuild-writing/variables/#predefined-read-only-variables), but it seems that when you run java to build things with emerge, it is ignored and portage's normal home is used. There should be many ebuilds affected. I would think the problems are in the java virtual implementation, may be there is an eclass to inherit to set things right. I have trouble when using icedtea-bin and this bug started with openJDK-bin.
*** Bug 763639 has been marked as a duplicate of this bug. ***
CCing openjdk maintainers as maybe they know what variable we need to tweak :/
I found that symlinking /var/lib/portage/home/.java to ${PORTAGE_BUILDDIR}/homedir/.java works as a temporary workaround - I managed to build the package by doing that.
> I found that symlinking /var/lib/portage/home/.java to > ${PORTAGE_BUILDDIR}/homedir/.java works as a temporary workaround - I managed > to build the package by doing that. I can confirm, this works as a temporary workaround for me too.
Worked for me - full details for people like me from an answer in the forum by fedeliallalinea - https://forums.gentoo.org/viewtopic-t-1130784-highlight-portagebuilddir.html This sequence works Code: # ebuild $(portageq get_repo_path / gentoo)/dev-java/gwt/gwt-2.8.2.ebuild unpack Then set the symlink Code: # ln -s $(portageq get_repo_path / gentoo)/dev-java/gwt/homedir /var/lib/portage/hiome/.java Then finish the emerge Code: # ebuild $(portageq get_repo_path / gentoo)/dev-java/gwt/gwt-2.8.2.ebuild merge Thanks!
I don't know the variable, but looks it for some reason uses portage user homedir defined in /etc/passwd instead of $HOME value set at build time. maybe something trying to be way to smart and uses getpwnam to get the dir instead of env?
maybe try setting -Duser.home ? ANT_OPTS="-Duser.home=${T}" may help. some ebuilds do that or similar trick.
and sorry for delay =) ETOOMUCHMAIL
The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=cc67cd13dc32e350a03f875ce56dcd7b4e84510e commit cc67cd13dc32e350a03f875ce56dcd7b4e84510e Author: Wolfgang E. Sanyer <ezzieyguywuf@gmail.com> AuthorDate: 2021-04-01 20:51:37 +0000 Commit: Miroslav Šulc <fordfrog@gentoo.org> CommitDate: 2021-06-11 09:51:07 +0000 dev-java/gwt: bump to 2.9.0 This resolves the two bugs references below, but does not close them since they still exist in the older version of this ebuild. Bug: https://bugs.gentoo.org/762619 Bug: https://bugs.gentoo.org/716220 Signed-off-by: Wolfgang E. Sanyer <ezzieyguywuf@gmail.com> Closes: https://github.com/gentoo/gentoo/pull/20222 Signed-off-by: Miroslav Šulc <fordfrog@gentoo.org> dev-java/gwt/Manifest | 2 + .../gwt/files/gwt-2.9.0-jsinterop-classpath.patch | 20 +++++ .../gwt/files/gwt-2.9.0-remove-git-usage.patch | 26 ++++++ dev-java/gwt/gwt-2.9.0.ebuild | 96 ++++++++++++++++++++++ 4 files changed, 144 insertions(+)
Affected version was dropped with f33b3483a90cc9b84c7a92d933994669da218e30. Closing.