Created attachment 280565 [details] emerge --info Looks like bug #369163 has creeped up again. After upgrading to 0.14.1-r2 the those files are gone.
app-emulation/ganeti and possibly some other vm stuff need /usr/bin/qemu-system-x86_64 and /usr/bin/kvm and /usr/bin/qemu-kvm. qemu-kvm-0.14.1-r1 working fine drinking beer, waiting for fix ;)
I confirm!! it install 2 binaries: qemu-x86_64 & qemu-i386 I request old behavor !!! kvm bla-bla....
Created attachment 280601 [details, diff] Add i386 and x86_64 back to IUSE_SOFTMMU_TARGETS This patch (based on diffing the previous and new ebuild) might help. Only lightly tested, I might have broken something else in the process.
(In reply to comment #3) > Created attachment 280601 [details, diff] > Add i386 and x86_64 back to IUSE_SOFTMMU_TARGETS > > This patch (based on diffing the previous and new ebuild) might help. Only > lightly tested, I might have broken something else in the process. Not sure what's going on but suddenly gcc processes are using up to 1GB memory. Build is failing do to OOM. (with 4G RAM). When I reverted the patch it compiled fine. Is anyone else experiencing this?
(In reply to comment #4) > (In reply to comment #3) > > Created attachment 280601 [details, diff] > > Add i386 and x86_64 back to IUSE_SOFTMMU_TARGETS > > > > This patch (based on diffing the previous and new ebuild) might help. Only > > lightly tested, I might have broken something else in the process. > > Not sure what's going on but suddenly gcc processes are using up to 1GB memory. > Build is failing do to OOM. (with 4G RAM). > > When I reverted the patch it compiled fine. > > Is anyone else experiencing this? This was a GCC regression issue with GCC 4.5.x. I believe it may be fixed in 4.5.2. There is a bug about this out there and I don't remember the # off hand.
This issue is fixed in 0.14.1-r2 assuming you have QEMU_SOFTMMU_TARGETS=x86_64 enabled, which is on by default.
(In reply to comment #5) > This was a GCC regression issue with GCC 4.5.x. I believe it may be fixed in > 4.5.2. There is a bug about this out there and I don't remember the # off hand. Hmm.. 4.5.2 is what I'm using.
patch working))))) thanks)))) to avoid gcc regression, I set MAKEOPTS=-j1 && add swap 4G via losetup. compiled success))) Total: RAM 2G + SWAP 5G && 1 thread compile. amd64.
(In reply to comment #3) > Created attachment 280601 [details, diff] > Add i386 and x86_64 back to IUSE_SOFTMMU_TARGETS > > This patch (based on diffing the previous and new ebuild) might help. Only > lightly tested, I might have broken something else in the process. thanks))))) beer from me)))) working)
*** Bug 376011 has been marked as a duplicate of this bug. ***