| Summary: | app-emulation/qemu[-python] should not require PYTHON_TARGETS | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Maciej Piechotka <uzytkownik2> |
| Component: | Current packages | Assignee: | Doug Goldstein (RETIRED) <cardoe> |
| Status: | RESOLVED DUPLICATE | ||
| Severity: | normal | CC: | python, qemu+disabled |
| Priority: | Normal | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
|
Description
Maciej Piechotka
2013-08-09 18:31:13 UTC
@python: Alright guys. What'd I do wrong in the ebuild? We still need Python 2.5, 2.6 or 2.7 to compile QEMU but when the user has USE=-python, we don't need a RDEPEND on it for the installed scripts. (In reply to Doug Goldstein from comment #1) python-r1 doesn't really deal with this situation optimally. We would need to borrow some functionality from python-any-r1 to make this work the way the requester wants. That would be tricky to implement, so I suggest the following patch as an alternative. This will force the user to set PYTHON_TARGETS via REQUIRED_USE, and will utilize that value in DEPEND. Runtime deps remain unchanged. Index: qemu-1.5.2-r1.ebuild =================================================================== RCS file: /var/cvsroot/gentoo-x86/app-emulation/qemu/qemu-1.5.2-r1.ebuild,v retrieving revision 1.1 diff -u -r1.1 qemu-1.5.2-r1.ebuild --- qemu-1.5.2-r1.ebuild 3 Aug 2013 19:47:09 -0000 1.1 +++ qemu-1.5.2-r1.ebuild 12 Aug 2013 02:48:25 -0000 @@ -58,7 +58,7 @@ # Block USE flag configurations known to not work REQUIRED_USE="${REQUIRED_USE} - python? ( ${PYTHON_REQUIRED_USE} ) + ${PYTHON_REQUIRED_USE} static? ( static-softmmu static-user ) static-softmmu? ( !alsa !pulseaudio !bluetooth !opengl !gtk ) virtfs? ( xattr )" @@ -122,7 +122,7 @@ DEPEND="${RDEPEND} dev-lang/perl - =dev-lang/python-2* + ${PYTHON_DEPS} sys-apps/texinfo virtual/pkgconfig kernel_linux? ( >=sys-kernel/linux-headers-2.6.35 ) What Mike said. Though I'd disagree that it's not optimal. It nearly boils down to the same result on most systems. duping to newer version as we've already sorted out a way forward & committed things there *** This bug has been marked as a duplicate of bug 496554 *** |