Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
Wol wrote: > On 17/02/2022 07:42, Dale wrote: >> I also commented out as much as I could in >> package.use, the things I'd tried previously. Now it has a clean path >> to upgrade. > > Is package.use a file or a directory? If it's a file, convert it to a > directory (you can just put the existing package.use file in the new > directory). > > Then make sure every time you add new use flags they're either in a > file dedicated to a particular package, or they're in a dated file and > refer to particular package versions. That dated file, when etc-update > wants to update package.use, I just rename its temporary file into a > proper dated file, rather than let it mess with a file it's chosen at > random. > > That way, you can delete packages you no longer use, clear out old > package/use combos, and generally keep things tidy. > > And when you're trying to fix something, that all goes in one file you > can ditch when it's all gone wrong :-) > > Cheers, > Wol > > I have them all as directories. Some are specific to KDE, some are for other things. I organize them, sort of. I also have a catch all with the default name. As I continued to try to get everything up to date and a clean output from emerge world, I started running into the same problems as before. So, I tried setting backtrack to a insanely high number but that didn't help. I then went back to a tried and generally true method, emerge -e world. I did that in a chroot first and am currently doing it with emerge -ek world on my live system. That seems to have sorted out whatever was causing this USE flag circle. It compiled everything fine in the chroot. Maybe I will have a clean system in a few hours. Even emerging binaries seem to take longer than it used to for some reason. Oh well. Dale :-) :-)
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
On 17/02/2022 07:42, Dale wrote: I also commented out as much as I could in package.use, the things I'd tried previously. Now it has a clean path to upgrade. Is package.use a file or a directory? If it's a file, convert it to a directory (you can just put the existing package.use file in the new directory). Then make sure every time you add new use flags they're either in a file dedicated to a particular package, or they're in a dated file and refer to particular package versions. That dated file, when etc-update wants to update package.use, I just rename its temporary file into a proper dated file, rather than let it mess with a file it's chosen at random. That way, you can delete packages you no longer use, clear out old package/use combos, and generally keep things tidy. And when you're trying to fix something, that all goes in one file you can ditch when it's all gone wrong :-) Cheers, Wol
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
Hi Miles, This install is likely over 10 years old. I might add, I copied a lot of things over from my previous system which was a AMD 2500+ single core puter. In a way, some things could be dating back to around 2003 or 2004. I use eix-test-obsolete to clean up package.* files but there isn't much to clean up make.conf except to do it by hand. That said, I disabled threads globally. I also disabled a few other things and enabled in package.use for those that really need it. One thing that confuses this, not all flags of one name does just one thing. It may do one thing for one package and another thing for other package(s). It makes things interesting. I just got a nice clean emerge output while doing kde-plasma/plasma-meta. Once that is done, world is next but should be little change. I just hope old problems don't pop up again. I keep forgetting to try flaggie. I tend to edit by hand. Recently, I've started including notes on some entries. I tend to forget why something is there, sometimes even just a week later. :/ Thanks for the help. Dale :-) :-) Miles Malone wrote: > threads very rarely makes sense for anything, btw. > > On Thu, 17 Feb 2022 at 21:41, Miles Malone > wrote: >> Now for your own sanity you might consider stopping adding things >> globally constantly, and using app-portage/flaggie to sanely manage >> them per-package... Cause there's far more use flags that make sense >> per-package than make sense globally. I used to manage them all >> largely globally like ten years ago, but it's utterly unrealistic to >> do that today. >> >> On Thu, 17 Feb 2022 at 21:34, Dale wrote: >>> Neil Bothwick wrote: On Thu, 17 Feb 2022 01:42:39 -0600, Dale wrote: > P. S. Is there a tool to make the USE line in make.conf in alphabetical > order or something? When I add things, I try to put them in order so it > is easier to find them. For some reason, they are out of order, a lot. > Something at some point messed up my organizing, badly. emerge --info shows the flags in alphabetical order, whatever their order in make.conf. It lso shows all flags, including those set b the profile, so you can see exactly what portage is using. -- Neil Bothwick Deja Foobar: A feeling of having made the same mistake before. >>> >>> Now that is cheating big time. Why didn't I think of that? Now they >>> back in order. I been cleaning up my USE line in make.conf. I think >>> some flags I added ages ago were only used by a few packages but are in >>> wide use now, and I don't always need them. So, I edit, run emerge >>> -auDN world to see what blows up. Edit again, run emerge and repeat. I >>> think I'm on about the 10th repeat now. Slowly cleaning things up. >>> >>> I love the sig on this one. How is it that thing knows which to pick? >>> ROFL >>> >>> Thanks for the cheat, I mean tip. Hit the nail on the head. >>> >>> Dale >>> >>> :-) :-) >>> >
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
threads very rarely makes sense for anything, btw. On Thu, 17 Feb 2022 at 21:41, Miles Malone wrote: > > Now for your own sanity you might consider stopping adding things > globally constantly, and using app-portage/flaggie to sanely manage > them per-package... Cause there's far more use flags that make sense > per-package than make sense globally. I used to manage them all > largely globally like ten years ago, but it's utterly unrealistic to > do that today. > > On Thu, 17 Feb 2022 at 21:34, Dale wrote: > > > > Neil Bothwick wrote: > > > On Thu, 17 Feb 2022 01:42:39 -0600, Dale wrote: > > > > > >> P. S. Is there a tool to make the USE line in make.conf in alphabetical > > >> order or something? When I add things, I try to put them in order so it > > >> is easier to find them. For some reason, they are out of order, a lot. > > >> Something at some point messed up my organizing, badly. > > > emerge --info shows the flags in alphabetical order, whatever their order > > > in make.conf. It lso shows all flags, including those set b the profile, > > > so you can see exactly what portage is using. > > > > > > > > > -- Neil Bothwick Deja Foobar: A feeling of having made the same > > > mistake before. > > > > > > Now that is cheating big time. Why didn't I think of that? Now they > > back in order. I been cleaning up my USE line in make.conf. I think > > some flags I added ages ago were only used by a few packages but are in > > wide use now, and I don't always need them. So, I edit, run emerge > > -auDN world to see what blows up. Edit again, run emerge and repeat. I > > think I'm on about the 10th repeat now. Slowly cleaning things up. > > > > I love the sig on this one. How is it that thing knows which to pick? > > ROFL > > > > Thanks for the cheat, I mean tip. Hit the nail on the head. > > > > Dale > > > > :-) :-) > >
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
Now for your own sanity you might consider stopping adding things globally constantly, and using app-portage/flaggie to sanely manage them per-package... Cause there's far more use flags that make sense per-package than make sense globally. I used to manage them all largely globally like ten years ago, but it's utterly unrealistic to do that today. On Thu, 17 Feb 2022 at 21:34, Dale wrote: > > Neil Bothwick wrote: > > On Thu, 17 Feb 2022 01:42:39 -0600, Dale wrote: > > > >> P. S. Is there a tool to make the USE line in make.conf in alphabetical > >> order or something? When I add things, I try to put them in order so it > >> is easier to find them. For some reason, they are out of order, a lot. > >> Something at some point messed up my organizing, badly. > > emerge --info shows the flags in alphabetical order, whatever their order > > in make.conf. It lso shows all flags, including those set b the profile, > > so you can see exactly what portage is using. > > > > > > -- Neil Bothwick Deja Foobar: A feeling of having made the same > > mistake before. > > > Now that is cheating big time. Why didn't I think of that? Now they > back in order. I been cleaning up my USE line in make.conf. I think > some flags I added ages ago were only used by a few packages but are in > wide use now, and I don't always need them. So, I edit, run emerge > -auDN world to see what blows up. Edit again, run emerge and repeat. I > think I'm on about the 10th repeat now. Slowly cleaning things up. > > I love the sig on this one. How is it that thing knows which to pick? > ROFL > > Thanks for the cheat, I mean tip. Hit the nail on the head. > > Dale > > :-) :-) >
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
Neil Bothwick wrote: > On Thu, 17 Feb 2022 01:42:39 -0600, Dale wrote: > >> P. S. Is there a tool to make the USE line in make.conf in alphabetical >> order or something? When I add things, I try to put them in order so it >> is easier to find them. For some reason, they are out of order, a lot. >> Something at some point messed up my organizing, badly. > emerge --info shows the flags in alphabetical order, whatever their order > in make.conf. It lso shows all flags, including those set b the profile, > so you can see exactly what portage is using. > > > -- Neil Bothwick Deja Foobar: A feeling of having made the same > mistake before. Now that is cheating big time. Why didn't I think of that? Now they back in order. I been cleaning up my USE line in make.conf. I think some flags I added ages ago were only used by a few packages but are in wide use now, and I don't always need them. So, I edit, run emerge -auDN world to see what blows up. Edit again, run emerge and repeat. I think I'm on about the 10th repeat now. Slowly cleaning things up. I love the sig on this one. How is it that thing knows which to pick? ROFL Thanks for the cheat, I mean tip. Hit the nail on the head. Dale :-) :-)
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
On Thu, 17 Feb 2022 01:42:39 -0600, Dale wrote: > P. S. Is there a tool to make the USE line in make.conf in alphabetical > order or something? When I add things, I try to put them in order so it > is easier to find them. For some reason, they are out of order, a lot. > Something at some point messed up my organizing, badly. emerge --info shows the flags in alphabetical order, whatever their order in make.conf. It lso shows all flags, including those set b the profile, so you can see exactly what portage is using. -- Neil Bothwick Deja Foobar: A feeling of having made the same mistake before. pgpd6qxSi84xk.pgp Description: OpenPGP digital signature
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
Arve Barsnes wrote: > On Thu, 17 Feb 2022 at 08:02, Dale wrote: >> !!! The ebuild selected to satisfy "> unmet requirements. >> - sci-libs/hdf5-1.10.5-r1::gentoo USE="cxx hl threads zlib -debug >> -examples -fortran -mpi -szip -unsupported" ABI_X86="(64)" >> >> The following REQUIRED_USE flag constraints are unsatisfied: >> !unsupported? ( threads? ( !cxx !hl ) ) >> >> The above constraints are a subset of the following complete expression: >> !unsupported? ( at-most-one-of ( cxx mpi ) threads? ( !cxx !mpi >> !fortran !hl ) ) > The mpi flag now seeds to be a distraction, as the actual problem now > is that hdf5 can't have both the threads and the cxx USE flag enabled > at the same time. Maybe you have these in global, or set somewhere in > your config? > > Regards, > Arve > > Ah ha. Success. I took a look at my USE line in make.conf. I did have threads in there so I moved it to a commented line to remove its consideration. I also noticed -cxx was in there as well and moved it to a commented line. I also commented out as much as I could in package.use, the things I'd tried previously. Now it has a clean path to upgrade. Thanks much. Once those get done updating, I'll apply my hammer to libreoffice. :-D Dale :-) :-) P. S. Is there a tool to make the USE line in make.conf in alphabetical order or something? When I add things, I try to put them in order so it is easier to find them. For some reason, they are out of order, a lot. Something at some point messed up my organizing, badly.
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
On Thu, 17 Feb 2022 at 08:02, Dale wrote: > !!! The ebuild selected to satisfy " unmet requirements. > - sci-libs/hdf5-1.10.5-r1::gentoo USE="cxx hl threads zlib -debug > -examples -fortran -mpi -szip -unsupported" ABI_X86="(64)" > > The following REQUIRED_USE flag constraints are unsatisfied: > !unsupported? ( threads? ( !cxx !hl ) ) > > The above constraints are a subset of the following complete expression: > !unsupported? ( at-most-one-of ( cxx mpi ) threads? ( !cxx !mpi > !fortran !hl ) ) The mpi flag now seeds to be a distraction, as the actual problem now is that hdf5 can't have both the threads and the cxx USE flag enabled at the same time. Maybe you have these in global, or set somewhere in your config? Regards, Arve
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
Miles Malone wrote: > Well I just installed kicad-meta (6.0.1), and neither vtk nor hdf5 are > pulled into the dep tree as a result... Given opencascade[vtk] is > what's pulling in vtk in your emerge above, disabling the vtk > useflag's probably going to help. > > In addition, I cant imagine why in a million years MPI should be > getting pulled in unless you've explicitly enabled it somewhere else, > MPI's hardly something you'd be using if you werent USING, i.e. on a > distributed memory system. Looking at the dependency tree, the most > likely place for an MPI use flag to be causing all kinds of mpi > related shenanigans with other packages is Boost. If you've enabled > MPI for Boost without actually needing it, that'll cause all kinds of > tree problems. In my experience if you're using MPI you need to be > really really careful about enabling it globally, it's incredibly hard > for the devs to test given test systems are rarely clusters, so it's > often got significant breakage. Of course if you're developing for a > cluster, which I do from time to time, this can be a pain. > > So assuming you've got *absolutely no libraries* in your world file > (there really shouldnt be), make sure you've got mpi and vtk disabled > globally, and not per-package enabled for boost and opencascade. Get > rid of vtk and hdf5, and try again? (Assuming you dont need hdf5 for > something other than kicad that is, of course. It's the main format I > use for data storage as it plays nicely with Matlab and okay with c++ > and python) > > That's my best guess as to what's causing your blockers, anyways. > Give it a go and see where the error moves to > > I'm fine with disabling it globally so I added it to make.conf. I also made sure any entries in package.use were commented out. Now I get this: root@fireball / # emerge -auDN world These are the packages that would be merged, in order: Calculating dependencies... done! !!! The ebuild selected to satisfy "
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
Well I just installed kicad-meta (6.0.1), and neither vtk nor hdf5 are pulled into the dep tree as a result... Given opencascade[vtk] is what's pulling in vtk in your emerge above, disabling the vtk useflag's probably going to help. In addition, I cant imagine why in a million years MPI should be getting pulled in unless you've explicitly enabled it somewhere else, MPI's hardly something you'd be using if you werent USING, i.e. on a distributed memory system. Looking at the dependency tree, the most likely place for an MPI use flag to be causing all kinds of mpi related shenanigans with other packages is Boost. If you've enabled MPI for Boost without actually needing it, that'll cause all kinds of tree problems. In my experience if you're using MPI you need to be really really careful about enabling it globally, it's incredibly hard for the devs to test given test systems are rarely clusters, so it's often got significant breakage. Of course if you're developing for a cluster, which I do from time to time, this can be a pain. So assuming you've got *absolutely no libraries* in your world file (there really shouldnt be), make sure you've got mpi and vtk disabled globally, and not per-package enabled for boost and opencascade. Get rid of vtk and hdf5, and try again? (Assuming you dont need hdf5 for something other than kicad that is, of course. It's the main format I use for data storage as it plays nicely with Matlab and okay with c++ and python) That's my best guess as to what's causing your blockers, anyways. Give it a go and see where the error moves to On Thu, 17 Feb 2022 at 16:09, Dale wrote: > > Miles Malone wrote: > > Hi Dale, > > > > Try removing the vtk useflag from opencascade. Also, add verbose to > > your emerge arguments and it may show you the full RDEPEND > > > > Regards, > > > > Miles > > > > > > I already have -v in my make.conf defaults. So it is already there I > just don't have to type it. I tried a huge number of options before > getting this close. Sometimes I'd have a dozen packages complaining > about USE flags or other issues. This is the new output but it looks > the same to me. > > > root@fireball / # emerge -auDN world > > These are the packages that would be merged, in order: > > Calculating dependencies... done! > > The following USE changes are necessary to proceed: > (see "package.use" in the portage(5) man page for more details) > # required by sci-libs/vtk-9.0.3-r4::gentoo > # required by sci-libs/opencascade-7.5.2-r5::gentoo > # required by sci-electronics/kicad-5.1.12-r2::gentoo > # required by sci-electronics/kicad-symbols-5.1.12::gentoo > # required by sci-electronics/kicad-meta-5.1.12::gentoo > # required by @selected > # required by @world (argument) > =sci-libs/hdf5-1.10.5-r1 mpi > # > >=sci-libs/vtk-9.0.3-r4 -mpi > > Would you like to add these changes to your config files? [Yes/No] n > root@fireball / # > > > Open to ideas? Maybe I'm catching the tree in a bad state or > something. I don't recall ever seeing something like this. > > This is the complete emerge command from emerge.log. > > > emerge --newuse --oneshot --unordered-display --update --ask > --backtrack=100 --deep --keep-going --with-bdeps=y --quiet-build=n > --regex-search-auto=y --verbose world > > > Dale > > :-) :-) >
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
Miles Malone wrote: > Hi Dale, > > Try removing the vtk useflag from opencascade. Also, add verbose to > your emerge arguments and it may show you the full RDEPEND > > Regards, > > Miles > > I already have -v in my make.conf defaults. So it is already there I just don't have to type it. I tried a huge number of options before getting this close. Sometimes I'd have a dozen packages complaining about USE flags or other issues. This is the new output but it looks the same to me. root@fireball / # emerge -auDN world These are the packages that would be merged, in order: Calculating dependencies... done! The following USE changes are necessary to proceed: (see "package.use" in the portage(5) man page for more details) # required by sci-libs/vtk-9.0.3-r4::gentoo # required by sci-libs/opencascade-7.5.2-r5::gentoo # required by sci-electronics/kicad-5.1.12-r2::gentoo # required by sci-electronics/kicad-symbols-5.1.12::gentoo # required by sci-electronics/kicad-meta-5.1.12::gentoo # required by @selected # required by @world (argument) =sci-libs/hdf5-1.10.5-r1 mpi # >=sci-libs/vtk-9.0.3-r4 -mpi Would you like to add these changes to your config files? [Yes/No] n root@fireball / # Open to ideas? Maybe I'm catching the tree in a bad state or something. I don't recall ever seeing something like this. This is the complete emerge command from emerge.log. emerge --newuse --oneshot --unordered-display --update --ask --backtrack=100 --deep --keep-going --with-bdeps=y --quiet-build=n --regex-search-auto=y --verbose world Dale :-) :-)
Re: [gentoo-user] Kicad and complications from hdf5 and vtk USE flags. Not package specific tho.
Hi Dale, Try removing the vtk useflag from opencascade. Also, add verbose to your emerge arguments and it may show you the full RDEPEND Regards, Miles On Thu, 17 Feb 2022 at 11:07, Dale wrote: > > Howdy, > > I seem to have ran into a road block. To get a clean path for emerge > recently, I had to unmerge a few packages. I removed Kicad, libreoffice > and a few others. Some of them I could add back with no problem. I > just emerged them and off it went. I'm having trouble with libreoffice > but I think it is a compile failure so I may just go back to a older > version. The one I can't get around is Kicad. When I run emerge the > first time, I get this: > > > root@fireball / # emerge -auDN world > > These are the packages that would be merged, in order: > > Calculating dependencies... done! > > The following USE changes are necessary to proceed: > (see "package.use" in the portage(5) man page for more details) > # required by sci-libs/vtk-9.0.3-r4::gentoo > # required by sci-libs/opencascade-7.5.2-r5::gentoo[vtk] > # required by sci-electronics/kicad-5.1.12-r2::gentoo > # required by sci-electronics/kicad-symbols-5.1.12::gentoo > # required by sci-electronics/kicad-meta-5.1.12::gentoo > # required by @selected > # required by @world (argument) > =sci-libs/hdf5-1.10.5-r1 -mpi > # > >=sci-libs/vtk-9.0.3-r4 mpi > > Would you like to add these changes to your config files? [Yes/No] n > root@fireball / # > > > I do my file edits manually so I go make the change it wants and try > again with the same command. > > > root@fireball / # emerge -auDN world > > These are the packages that would be merged, in order: > > Calculating dependencies... done! > > The following USE changes are necessary to proceed: > (see "package.use" in the portage(5) man page for more details) > # required by sci-libs/vtk-9.0.3-r4::gentoo > # required by sci-libs/opencascade-7.5.2-r5::gentoo[vtk] > # required by sci-electronics/kicad-5.1.12-r2::gentoo > # required by sci-electronics/kicad-symbols-5.1.12::gentoo > # required by sci-electronics/kicad-meta-5.1.12::gentoo > # required by @selected > # required by @world (argument) > =sci-libs/hdf5-1.10.5-r1 mpi > # > >=sci-libs/vtk-9.0.3-r4 -mpi > > Would you like to add these changes to your config files? [Yes/No] n > root@fireball / # > > > Each time it just reverses the recommended settings. I've been dealing > with this for a couple days now. I've tried changing settings on the > packages that are pulling it in and anything else I can think of but so > far, this is as close I've got to it being able to update. > > Does someone have their emerge decoder ring handy? I just can't figure > out what changes to make to get past this. Once this is done, I'll > start hammering on libreoffice. > > Thanks. > > Dale > > :-) :-) >
