When running emerge on a list of several packages, more than one of which is masked (say by the ~x86 keyword on a x86-only system), portage exits after listing just the first masked package. When performing a revdep-rebuild, for instance, this can require dozens of false starts before getting the system updated, since the user can't issue a single "ACCEPT_KEYWORDS=~arch emerge <packages>" command. portage should list all masked packages, as it does with blocked packages. Reproducible: Always Steps to Reproduce: Try to emerge more than one masked package. Actual Results: portage reports the first masked package but not the other(s) Expected Results: portage should report all problems with masked packages
I think you have a valid request, but generally, it's better practice to use the autounmask program (from the app-portage/autounmask package) to unmask packages instead of overriding ACCEPT_KEYWORDS via the environment.
In this case, the actual underlying issue was not a masking problem, but rather treated as a masking problem by portage and resolved in the same way. Specifically, revdep-rebuild can try to emerge specific atoms that no longer exist in the portage tree, and instead of reporting all of the missing ebuilds (which portage treats as having a null keyword), it only reports the first. However, reporting all the masking problems at once will both fix the ~arch issue and the no-ebuild issue.
*** Bug 236225 has been marked as a duplicate of this bug. ***
*** Bug 253845 has been marked as a duplicate of this bug. ***
*** Bug 281185 has been marked as a duplicate of this bug. ***
Add depends on 280097 because Zac add it in duplicate bug 281185
Is this bug not resolved by the addition of --autounmask (portage-2.1.9)? Seems to work pretty well for me.
*** This bug has been marked as a duplicate of bug 280097 ***