While trying to update my ~amd64 system: app-text/poppler:0 (app-text/poppler-0.24.5::gentoo, installed) pulled in by (no parents that aren't satisfied by other packages in this slot) (app-text/poppler-0.24.3::gentoo, ebuild scheduled for merge) pulled in by app-text/poppler:0/43=[cxx,jpeg,lcms,tiff,xpdf-headers(+)] required by (net-print/cups-filters-1.0.43-r1::gentoo, installed) As a workaround, I may mask the 0.24.5 version, but it's not really the best way I think ;). Thank you for your time/correction. Cheers, C. Reproducible: Always Steps to Reproduce: 1. have some ~amd64 gentoo with cups and poppler (about 1 week old) 2. try an update Actual Results: Multiple packages instances within a single package slot are pulled Expected Results: either a new slot should be created or, for sure better, cups-filter shouldn't depend on an old poppler version (minor number, shouldn't create any problem I guess?)
Experiencing this as well.
emerge --info please! (need the portage version!)
Created attachment 367884 [details] emerge --info output emerge --info — sorry for not adding it before
Hi portage team, could you please have a quick look at this? TIA
Please try to increase the --backtrack value, i.e. --backtrack=30. If this doesn't help, attach the output with --debug added.
same with =net-print/cups-filters-1.0.36-r1. There are many other slot dependencies with the same problem. Sometimes ":=" results in a slot conflict instead of a recompile for no apparant reason, when the dependency is upgraded. Might be the removal of the original ebuild (in this case, poppler-0.24.3) triggers this behavior.
@Carsten: emerge --info and please try the things in comment #5. Otherwise we can't help you. (In reply to Sebastian Luther (few) from comment #5) > Please try to increase the --backtrack value, i.e. --backtrack=30. If this > doesn't help, attach the output with --debug added.
The --backtrack=30 option never helped so far. Unfortunately, I already wiped and reinstalled the two packages, Im working on another project and need a running system :/ However, next time I come across a slot conflict of the same kind I'll gather and post the --info / --debug outputs here.
I finally managed to reproduce it. Command run: emerge --with-bdeps=y -NuaD --backtrack=30 --debug --exclude=net-libs/telepathy-farstream @system @world|tee emerge-debug.txt Two slot conflicts are reported, although from the report it seems they would be safe to upgrade (unfortunately the color isnt visible, but the "older" version and all the packages wanting it are shown in blue, in both cases) and an upgrade of world and all deep dependencies was requested by the command. See attachments for outputs.
Created attachment 377830 [details] emerge --info (Carsten Milkau)
Sorry, the debug output is too large to attach, even in compressed form. Here's a link: http://test.verlorene-worte.de/upload/497964-emerge-debug.txt.gz
Created attachment 377832 [details] emerge --pretend ... (Carsten Milkau) Output without --debug
Use `xz -9` instead of `gzip` to compress this file. Then its size will be 414 KiB.
I uninstalled ffmpeg before running the command again. The result was quite telling: portage asked for a keyword unmask of ffmpeg. Doing so made the slot conflicts vanish. A little more research showed: the ffmpeg package I had installed previously has left portage, but couldn't be upgraded due to keywords (I had only unmasked the specific version installed). Sorry for the false alert. I will keep an eye open whether I can actually reproduce the original problem. PLEASE NOTE: please do NOT show packages in blue just because they have left portage. Only packages that are scheduled for uninstall should be shown in blue. It is very hard to resolve a slot conflict if you cannot see that a package pulling in a dependency will NOT be uninstalled (and thus needs attention for conflict resolution). If a package will remain installed, please show it in green or some other distinctive color. (Is this a bug?)
Another strange upgrade refusal I don't understand: # emerge -p '=virtual/perl-Getopt-Long-2.390.0-r1:0' !!! FEATURES=userpriv is enabled, but os.setgroups is not available. !!! FEATURES=userfetch is enabled, but os.setgroups is not available. !!! See https://bugs.pypy.org/issue833 for details. These are the packages that would be merged, in order: Calculating dependencies ... done! [ebuild U ] dev-lang/perl-5.18.2-r1 [5.16.3] [ebuild U ] virtual/perl-Getopt-Long-2.390.0-r1 [2.380.0-r2] !!! Multiple package instances within a single package slot have been pulled !!! into the dependency graph, resulting in a slot conflict: dev-lang/perl:0 (dev-lang/perl-5.16.3::gentoo, installed) pulled in by dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/PlRPC-0.202.0-r2::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/Net-SSLeay-1.550::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/IO-Socket-SSL-1.953.0::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-vcs/subversion-1.7.14::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/Net-Daemon-0.480.0-r1::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-db/mytop-1.6-r4::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/HTTP-Negotiate-6.0.1-r1::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/Algorithm-Diff-1.190.200-r1::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/Error-0.170.210::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-vcs/git-1.8.5.5::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/libwww-perl-6.50.0::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/HTTP-Cookies-6.0.1-r1::gentoo, installed) dev-lang/perl:0/5.16=[-build(-)] required by (dev-perl/WWW-RobotRules-6.20.0::gentoo, installed) (dev-lang/perl-5.18.2-r1::gentoo, ebuild scheduled for merge) pulled in by =dev-lang/perl-5.18* required by (virtual/perl-Getopt-Long-2.390.0-r1::gentoo, ebuild scheduled for merge) (and 1 more with the same problem) I don't get why it doesn't try to recompile the subslot dependencies. Isn't that what the subslot dependencies are meant to do? Will update emerge --info in a blink.
Created attachment 382036 [details] emerge --info
