Created attachment 381034 [details] ebuild with 2 patches for removing -pg and slotted to "3" Thanks for committing the iperf-3.0.5 to the tree in #466508 Here i attach the needed patches for removing "-pg" as this conflicts with "-fomit-frame-pointer". Also i attached a ebuild, which does iperf3 put in slot3 (as iperf2 and iperf3 are NOT compatible and should be usable concurrent (the get named correctly, iperf-2 is named "iperf" and iperf-3.X is named "iperf3" as binary. For me this works at x86 arch without any issues now, hopfefully, we can commit this to tree to help spreed the newer iperf3 :) Ronny
Created attachment 381036 [details, diff] 1st patch for removing -pg
Created attachment 381038 [details, diff] 2nd patch for removing -pg
What would be the point of having a SLOTted ebuild? Just because you can?
Comment on attachment 381036 [details, diff] 1st patch for removing -pg That leaves -g in place.
Comment on attachment 381038 [details, diff] 2nd patch for removing -pg If we're going to patch Makefile.am, then it makes more sense to run automake than to patch Makefile.in as well.
Fixed without revision bump except for the SLOTting.
(In reply to Jeroen Roovers from comment #3) > What would be the point of having a SLOTted ebuild? Just because you can? Thanks for cleanup and integrating. For the slotted thing, iperf3 is NOT compatible to iperf2, so it makes sense to slot the version 3, so we can have iperf v2 and iperf v3 concurrently installed. For me, i have clients, which are not aware of the iperf3 protocol and with non-slotted ebuilds here, if have the option to install iperf OR iperf3. It could also be possible to create a new package named net-misc/iperf3 to seperate this. But I thought, the slotted way is the best option.