Compound problem. Sandbox is unmerged in OLD 2.0.51 ebuilds. doebuild() is not aware of failure due to missing sandbox. Affected Platform: AMD64 Affected Requirement: FEATURES=sandbox Initial Problem: Removal of sandbox library and binary. Compounding Problems: Deletion of package files during update. Most users will not be affected by this. Initial installs may fail, and those having specifically enabled Sandbox on AMD64 will be affected. To determine if your machines are affected run emerge info and view the FEATURES setting. If sandbox is listed, please do the following: (It does no harm to do this anyway) All AMD64 users are advised to do the following: (Again, regardless of how recently this was done) emerge sync FEATURES=-sandbox emerge portage The potential for portage to misbehave when interacting with sandbox exists for 2.0.51_rc10 and 2.0.51 for anyone sync'ing PRIOR to Thu Oct 21 18:14:09 UTC 2004. An oversight on my part, followed by coincidence, left sandbox uninstalled on AMD64 with the original portage-2.0.51 ebuilds. The ebuild has not been bumped, as anyone experiencing the problem would not be able to update without at least some fore-knowledge anyway. If sandbox is not enabled, this will go unnoticed. It appears that due to profiles and other tree-affects, that sandbox is disabled on most boxes but the full extent of that default is not yet known. UPDATE PORTAGE __AGAIN__ after sync'ing now. The potential exists for portage to remove installed files without replacing them with updated files when doing updates. MERGE PORTAGE FIRST or disable sandbox.
OK, I'm one affected as I already switched to the 2004.3 cascading profile for AMD64 and portage 2.0.51, and it has sandbox enabled. The instructions are slightly unclear. Clear: 1. Verify whether affected. (emerge info|grep FEATURES, see if sandbox is on.) Clear: For those affected: 2. emerge sync 3. FEATURES=-sandbox emerge portage I've done that. Then the instructions go into an explanation. Thanks for that, as it can be vital information and skipping the explanation is all to easy to do. Unclear: At the bottom, it says to "UPDATE PORTAGE __AGAIN__ after sync'ing now." Unclear is whether this necessary AGAIN after step 3, making it step 4 (repeated sync) and step 5 (apparently normal update of portage, this time with sandbox enabled), or whether this is a RESTATEMENT of steps 2 and 3. IOW, assuming the problem, is it sync sandbox off re-emerge portage done, or sync sandbox off re-emerge portage sync re-emerge portage done. I've done the first set, altho I may have caught the new one anyway, but haven't done the second (steps 4 and 5 as outlined above) just yet, tho I probably will for safety.
Update the portage tree. Update portage with sandbox That is all. emerge sync FEATURES=-sandbox emerge portage (Disabling sandbox should not be required as of -r1)
portage-2.0.51-r2 is available.
The bug that deleted packages is gone but if I reenable my original FEATURES(ccache sandbox userpriv usersandbox) and try to emerge gpm it dies with the following error: >>> Test phase [not enabled]: sys-libs/gpm-1.20.1 touch: cannot touch `.tested': Permission denied or this in the case of portage: >>> Test phase [not enabled]: sys-apps/portage-2.0.51-r2 touch: cannot touch `.tested': Permission denied But it will compile without problem with only ccache set.
In response to #4: Have a look at http://bugs.gentoo.org/show_bug.cgi?id=66957
This should be able to be closed in the next round...
I have tried this fix, and it has not helped. I first ran across this in Bug #72379. I have also run across this when I "emerge tetex jadetex". In the mentioned bug, they talk about multilib support and sandbox. I originally did not have multilib listed in my use variable, but it worked for gcc. # emerge info Portage 2.0.51-r15 (default-linux/amd64/2004.3, gcc-3.4.3, glibc-2.3.4.20041102-r0, 2.6.9-gentoo-r14 x86_64) ================================================================= System uname: 2.6.9-gentoo-r14 x86_64 AMD Athlon(tm) 64 Processor 3200+ Gentoo Base System version 1.4.16 Python: dev-lang/python-2.3.4-r1 [2.3.4 (#1, Feb 16 2005, 00:17:11)] dev-lang/python: 2.3.4-r1 sys-devel/autoconf: 2.13, 2.59-r6 sys-devel/automake: 1.5, 1.9.4, 1.6.3, 1.7.9-r1, 1.4_p6, 1.8.5-r3 sys-devel/binutils: 2.15.92.0.2-r1 sys-devel/libtool: 1.5.10-r4 virtual/os-headers: 2.6.8.1-r2 ACCEPT_KEYWORDS="amd64" AUTOCLEAN="yes" CFLAGS="-O2 -march=athlon64 -pipe -fomit-frame-pointer" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/2/share/config /usr/kde/3/share/config /usr/lib/X11/xkb /usr/share/config /usr/share/texmf/dvipdfm/config/ /usr/share/texmf/dvips/config/ /usr/share/texmf/tex/generic/config/ /usr/share/texmf/tex/platex/config/ /usr/share/texmf/xdvi/ /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d" CXXFLAGS="-O2 -march=athlon64 -pipe -fomit-frame-pointer" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs autoconfig ccache distlocks sandbox" GENTOO_MIRRORS="http://gentoo.osuosl.org http://mirror.datapipe.net/gentoo http://gentoo.ccccom.com http://www.ibiblio.org/pub/Linux/distributions/gentoo" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" SYNC="rsync://rsync.namerica.gentoo.org/gentoo-portage" USE="amd64 X Xaw3d acpi aim alsa apache2 bash-completion berkdb bitmap-fonts bzlib cdr crypt cups dbm dbx doc dvd fastcgi fbcon flatfile font-server foomaticdb ftp gd gdbm gif gnome gphoto2 gpm gps gtk gtk2 icq imap inifile innodb iodbc ipv6 java javascript jikes joystick jp2 jpeg junit kde kerberos ldap libg++ libwww lm_sensors lzw lzw-tiff mbox milter mime mmap motif mozilla mp3 mpeg msn mssql multilib mysql mysqli ncurses nls ntpl odbc opengl oscar oss pam pcre pdflib perl php png posix postgres ppds python qt quicktime readline samba sasl scanner session sharedmem simplexml skey slang snmp soap sockets spell spl ssl symlink tcltk tcpd tetex threads tidy tiff tokenizer truetype truetype-fonts type1-fonts usb userlocales vhosts xml xml2 xmlrpc xmms xpm xrandr xsl xv zeo zlib" Unset: ASFLAGS, CBUILD, CTARGET, LANG, LC_ALL, LDFLAGS, PORTDIR_OVERLAY
You guys do realize that no 32-bit sandbox is being built on amd64 systems? The absence of this library is causing problems, especially when it comes to building a multilib gcc. Oddly enough a 32bit version exists in the amd64 2004.3 stage2/3 tarballs and most people are remedying this problem by copying it from those tarballs.
make[1]: *** [sandbox] Error 1 make[1]: *** Waiting for unfinished jobs.... make[1]: Leaving directory `/var/tmp/portage/sandbox-1.2.9/work/sandbox-1.2.9- x86' make: *** [all] Error 2 * Building failed for ABI=x86!. This usually means a broken * multilib setup. Please fix that before filling a bugreport * against sandbox. !!! ERROR: sys-apps/sandbox-1.2.9 failed. !!! Function src_compile, Line 79, Exitcode 0 !!! emake failed for x86 !!! If you need support, post the topmost build error, NOT this status message. Broken as in doesn't exist
Fixed on or before 2.0.51.22-r1
Looking through the batch of bugs, I'm not sure that some of these are actually fixed in stable. Others, the requirements have possibly changed after the initial fix was committed. If you think this bug has been closed incorrectly, please reopen or ask that it be reopened.
