onli

That's more a questions whether someone can confirm it's a bug and I should report it somewhere else. When emerging media-libs/sdl2-image I got this error: libtool: link: x86_64-pc-linux-gnu-gcc -m32 -I/usr/include/libpng16 -march=native -O1 -pipe -I/usr/include/SDL2 -D_REENTRANT -Wl,-O1 -Wl,--sort-common -Wl,--as-needed -o .libs/showimage showimage.o ./.libs/libSDL2_image.so -lpng16 -ljpeg -ltiff -lz -lwebp -lSDL2 ./.libs/libSDL2_image.so: error: undefined reference to 'sqrtf' ./.libs/libSDL2_image.so: error: undefined reference to 'sincosf' ./.libs/libSDL2_image.so: error: undefined reference to 'floorf' ./.libs/libSDL2_image.so: error: undefined reference to 'acosf' ./.libs/libSDL2_image.so: error: undefined reference to 'ceilf' ./.libs/libSDL2_image.so: error: undefined reference to 'atan2f' ./.libs/libSDL2_image.so: error: undefined reference to 'sqrt' ./.libs/libSDL2_image.so: error: undefined reference to 'pow' ./.libs/libSDL2_image.so: error: undefined reference to 'tanf' ./.libs/libSDL2_image.so: error: undefined reference to 'fmodf' collect2: error: ld returned 1 exit status make: *** [Makefile:550: showimage] Error 1 * ERROR: media-libs/sdl2-image-2.0.2::media-kit failed (compile phase): * emake failed After a lot of searching I realized the linker is missing the lm flag: sudo LDFLAGS="-lm" emerge media-libs/sdl2-image That way it compiled successfully. Is that somehow specific to my system?

I try to upgrade my system and run into an issue again. This time it is encfs: That relies on dev-libs/tinyxml2:0/3, but the upgrade schedules an upgrade for that: WARNING: One or more updates/rebuilds have been skipped due to a dependency conflict: dev-libs/tinyxml2:0 (dev-libs/tinyxml2-6.0.0:0/6::dev-kit, ebuild scheduled for merge) conflicts with dev-libs/tinyxml2:0/3 required by (sys-fs/encfs-1.9.2:0/0::core-kit, installed) ^^^^ Nothing to merge; quitting. What's the best solution here?

The issue still seems to exist?

Thanks. I'm getting this error on an upgrade after this: ERROR: Specified xorg-kit branch 1.20-release is missing! Is it included in this release? Exiting. Do I have to run anything else instead of `ego sync`? Edit: notized the error message looks like there is a tab or something before the version number, so I added the xorg-kit line again. Seems to have worked now, I will now upgrade the system. Let's see ?

Hi Yes, I think so. I did the lxd init, created /etc/subuid and subguid and executed the usermod commands.

Oh, so it's not my system that has a problem there? Thanks for the confirmation. I just bought a RX 580, hope the update arrives before the card.

Hi I tried to get LXD working. I noticed the wiki page https://www.funtoo.org/LXD but it does not work for me (and I'm surprised to see calls to "service", is that usually installed?) I can execute launch, but the image then does not start. I get this error message: onli@Fallout:~$ lxc start pipeblock Error: Failed to run: /usr/libexec/lxd forkstart pipeblock /var/lib/lxd/containers /var/log/lxd/pipeblock/lxc.conf: Try `lxc info --show-log pipeblock` for more info onli@Fallout:~$ lxc info --show-log pipeblock Name: pipeblock Location: none Remote: unix:// Architecture: x86_64 Created: 2018/10/21 14:32 UTC Status: Stopped Type: persistent Profiles: default Log: lxc pipeblock 20181021150139.557 ERROR cgfsng - cgroups/cgfsng.c:create_path_for_hierarchy:1198 - The cgroup "/sys/fs/cgroup/openrc//lxc/pipeblock" already existed lxc pipeblock 20181021150139.558 ERROR cgfsng - cgroups/cgfsng.c:cgfsng_create:1287 - Failed to create cgroup "/sys/fs/cgroup/openrc//lxc/pipeblock" lxc pipeblock 20181021150139.558 ERROR cgfsng - cgroups/cgfsng.c:create_path_for_hierarchy:1198 - The cgroup "/sys/fs/cgroup/openrc//lxc/pipeblock-1" already existed lxc pipeblock 20181021150139.558 ERROR cgfsng - cgroups/cgfsng.c:cgfsng_create:1287 - Failed to create cgroup "/sys/fs/cgroup/openrc//lxc/pipeblock-1" lxc pipeblock 20181021150139.558 ERROR cgfsng - cgroups/cgfsng.c:create_path_for_hierarchy:1198 - The cgroup "/sys/fs/cgroup/openrc//lxc/pipeblock-2" already existed lxc pipeblock 20181021150139.558 ERROR cgfsng - cgroups/cgfsng.c:cgfsng_create:1287 - Failed to create cgroup "/sys/fs/cgroup/openrc//lxc/pipeblock-2" lxc pipeblock 20181021150139.563 ERROR conf - conf.c:lxc_map_ids:2999 - newuidmap failed to write mapping "newuidmap: uid range [0-1000000000) -> [1000000-1001000000) not allowed": newuidmap 2186 0 1000000 1000000000 lxc pipeblock 20181021150139.563 ERROR start - start.c:lxc_spawn:1677 - Failed to set up id mapping. lxc pipeblock 20181021150139.627 WARN network - network.c:lxc_delete_network_priv:2597 - Invalid argument - Failed to remove interface "vethB76GK4" from "lxdbr0" lxc pipeblock 20181021150139.627 ERROR lxccontainer - lxccontainer.c:wait_on_daemonized_start:840 - Received container state "ABORTING" instead of "RUNNING" lxc pipeblock 20181021150139.628 ERROR start - start.c:__lxc_start:1910 - Failed to spawn container "pipeblock" lxc pipeblock 20181021150139.633 ERROR conf - conf.c:lxc_map_ids:2999 - newuidmap failed to write mapping "newuidmap: uid range [0-1000000000) -> [1000000-1001000000) not allowed": newuidmap 2197 0 1000000 1000000000 1000000000 0 1 lxc pipeblock 20181021150139.633 ERROR conf - conf.c:userns_exec_1:4358 - Error setting up {g,u}id mappings for child process "2197" lxc pipeblock 20181021150139.633 WARN cgfsng - cgroups/cgfsng.c:cgfsng_destroy:1119 - Failed to destroy cgroups lxc 20181021150139.633 WARN commands - commands.c:lxc_cmd_rsp_recv:130 - Connection reset by peer - Failed to receive response for command "get_state" Is there a likely reason for the cgroup already existing, and for newuidmpa to be failing?

Hi A new mesa version would be important to me, I play quite often right now, also thanks to Steams Proton thing. I just stumbled over the news that the new version will contain game specific patches. There was already a post in this forum some time ago that mesa should be update soon, but that did not seem to happen. Or is there something wrong with my system?

@Oleg Vinichenko Okay, thank you! I got a bunch of different related upgrades that worked. But mesa is still on 17.2.8, was that the idea? I guess that 18.2.0-rc2 was a mistake, but 18.1.5 should be the target now, right?

I tried to fix this by removing the xa useflag I had in my /etc/portage/make.conf. That specific error vanished for me. However, now I'm getting a different error. I'm not even sure what exactly goes wrong there. Is it stopping because of the warning? {anonymous}::get_symbol_offsets(const llvm::Module&)’ defined but not used [-Wunused-function] get_symbol_offsets(const ::llvm::Module &mod) { ^~~~~~~~~~~~~~~~~ After all that is the last compilation step done. Or is the problem the llvm lib issues above? src/gallium/auxiliary/libgallium.a(gallivm_lp_bld_tgsi_action.c.o):lp_bld_tgsi_action.c:function d2f_emit: error: undefined reference to 'LLVMBuildFPTrunc' src/gallium/auxiliary/libgallium.a(gallivm_lp_bld_tgsi_action.c.o):lp_bld_tgsi_action.c:function up2h_emit: error: undefined reference to 'LLVMInt16TypeInContext' src/gallium/auxiliary/libgallium.a(gallivm_lp_bld_tgsi_action.c.o):lp_bld_tgsi_action.c:function i64div_emit_cpu: error: undefined reference to 'LLVMBuildSDiv' src/gallium/auxiliary/libgallium.a(gallivm_lp_bld_tgsi_action.c.o):lp_bld_tgsi_action.c:function u64div_emit_cpu: error: undefined reference to 'LLVMBuildUDiv' src/gallium/auxiliary/libgallium.a(gallivm_lp_bld_format_aos_array.c.o):lp_bld_format_aos_array.c:function lp_build_fetch_rgba_aos_array: error: undefined reference to 'LLVMBuildFPTrunc' I already recompiled gcc and llvm, so I'm stuck there. But maybe removing xa makes it work for you and the llvm stuff is something specific to my system?

I just tried to do a system upgrade. I get this error message for the mesa build: The log of this is here. It mentions the meson log, I uploaded that one here. The initial portage output is here I will now dig into the libva problem, maybe the issue is that requirement? I'm using the radeon driver and do not see how any of those options mentioned in the error message could be required for me.

Hm, an incompatibility with python 3.4 would be bad. But it seems it works now. So, what I did: sudo eselect python set 3 to select python 3.5, and then I re-added the missing # to the first line of /var/git/meta-repo/kits/core-kit/profiles/base/make.defaults. Now the upgrade commands reacht the step of calculating dependencies, unlike before. Thanks.

@kris: onli@Fallout:~$ sudo eselect python list Password: Available Python interpreters, in order of preference: [1] python3.4 [2] python2.7 [3] python3.5 (fallback) You mean portage/ego relies on python 2 and I should select that?

I think so, doesn't that follow from the log file output? Where do I check that?

Otakku, I hadn't seen your. Though your thread is marked [solved] without there being a fix in the thread. How did you solve it? I can of course add the # to the file, but when I try to upgrade it gets removed again.

I'm running this script for upgrades: #!/bin/sh ego sync emerge --sync layman -S emerge -uavDN @world But today I'm getting the error message from the title. The full message is: That means one of those scripts removes the # sign from the first line of that file. There is an other error before: Already up to date. Sync successful and kits in alignment! :) Updating /etc/portage/repos.conf... Updating profiles at /etc/portage/make.profile/parent... Updating non-funtoo repositories... Traceback (most recent call last): File "/usr/bin/ego", line 120, in <module> EgoModule.run_ego_module(action, econfig, args, VERSION) File "/usr/share/ego/python/ego/module.py", line 102, in run_ego_module ego_module(*args) File "/usr/share/ego/python/ego/module.py", line 87, in __call__ self.handle() File "/usr/share/ego/modules/sync.ego", line 337, in handle self.sync_meta_repo() File "/usr/share/ego/modules/sync.ego", line 325, in sync_meta_repo self._do_package_moves() File "/usr/share/ego/modules/sync.ego", line 46, in _do_package_moves import portage ImportError: No module named 'portage' Full output here. I'm not aware of having changed anything related to this. How can I fix this?

Hi, just checking in to see if I'm doing something wrong. I'm running the current funtoo, I switched to 1.2 pretty early. Now my steam package (coming from the overlay) is complaining: emerge: there are no ebuilds to satisfy ">=media-libs/mesa-17.3.9[-bindist(-)]". (dependency required by "games-util/steam-games-meta-0-r20180619::steam-overlay" [ebuild]) And if I understand this right, https://github.com/funtoo/xorg-kit/tree/master/media-libs/mesa got his last upgrade indeed 9 month ago, while https://gentoobrowse.randomdan.homeip.net/packages/media-libs/mesa and https://mesa.freedesktop.org/ look like there were some updates. Am I looking into the wrong place, or is that kit just a bit behind (which would not be a big problem yet, it's just that mesa upgrades are often very useful for getting better performance).

Hm, it seemed to work just fine with Python 2.7 as the first, but I changed it around. Thanks :) That I double-posted to describe the solution was not really easy to see with the re-ordering going on here, sorry about that. Everything is good now (or rather, the QT issues I have now seem to be unrelated and already in the bugtracker).

Okay, I got it. A configuration thing/possible conflict. I had added this to package.use: Because I had gotten this error before: Now I removed that line and unmerged libreoffice, emerge could calculate the upgrade and found 72. Yeah :) Thanks for the help!

I did an ego sync now, but that did not help. There is no error when updating ego :) That works just fine, it gets compiled. Portage might be outdated, but I tried to update it like ego, did not change.

@Sandro: (Sorry, the Code Box does not work). @dhudson: I did that already (and now tried again of course), does not help. Though I don't actually know what ego does. Do I have to call that before using that emerge command, or is that used internally by emerge?

I run the kit system for a while now. Was gone for the month, now when I try to upgrade I get this: I already deleted /var/git/meta-repo and followed https://github.com/funtoo/meta-repo#using-meta-repo to recreate it, but now I don't have any ideas left.

Hi jhan. I think there was an update to the ebuilds by now. I am not aware of having changed my system (and certainly not the USE flags), but meld compiles now. The dbus USE flag is disabled on my system, but I had to enable it for some specific packages. But I don't see it activated for meld.

Emerge is failing to upgrade: emerge: there are no ebuilds to satisfy ">=dev-libs/glib-2.50:2[dbus]". (dependency required by "dev-util/meld-3.17.4::dev-kit" [ebuild]) (dependency required by "@selected" [set]) (dependency required by "@world" [argument]) And indeed, the ebuild of meld, https://github.com/funtoo/dev-kit/blob/master/dev-util/meld/meld-3.17.4.ebuild, got an update 6 hours ago that added a new ebuild for 3.17.4 that requires glib-2.50:2, but https://github.com/funtoo/gnome-kit/tree/3.20-prime/dev-libs/glib contains no ebuild for that version. Should I report this in the bug tracker?

Same for me, btw. Problem began when moving to the kits, setting force_xinerama = true made it work again.