#trinity-desktop < 2022/04/25 > | |
---|---|
[06:36] | agneli: anybody here willing to help me with my rebuild of trinity on gentoo :) |
[06:52] | agneli: i think it crashes on tdelibmid |
[06:52] | agneli: libtdemid ;) |
[06:52] | agneli: but i have no idea why... |
[07:00] | denk: you could start with a paste |
[07:01] | denk: just share the result with errors |
[07:24] | agneli: http://fileconvoy.com/index.php?Section=25678 |
[07:24] | tde-bot: Page title: Free file sharing without registration and size limits |
[07:24] | agneli: build.log and environment |
[07:24] | agneli: if that is OK for you |
[07:28] | agneli: https://paste.debian.net/plain/1238934 |
[07:28] | agneli: that is the message when it crashes |
[07:29] | agneli: in the mean time I am doing to compliation "manually" |
[07:29] | agneli: I went to portage directory and did cmake - it objected the last of arts |
[07:29] | agneli: which I emerged |
[07:30] | agneli: then cmake wen okay and now i did make and it seems it is compiling |
[07:30] | agneli: 76% already |
[07:30] | agneli: I believe it shoudl have already crashed... |
[07:30] | agneli: i also have a strace of the emerge :) |
[07:31] | denk: I don't see a cause in your paste, but I would change -j4 to -j1 as first step |
[07:34] | agneli: cpu is 2(4) and 16 GiB RAM, but OK I will do it, could you elaborate why, please? |
[07:37] | denk: when you find an error you should stop the process at the time it occured and not in some (long) time when all other processes finished |
[07:37] | agneli: of course |
[07:37] | agneli: i dhould have thought about it myself |
[07:37] | agneli: thank you |
[07:37] | agneli: but this "manual" make finished without crashing... |
[07:38] | denk: so, I'd like to see the output from the command "make" or whatever you run to the last message "error" |
[07:39] | agneli: i shared complete build.log and environment files above |
[07:39] | agneli: the merge erro is what I pasted |
[07:39] | agneli: I will run it now with -j1 and also strace |
[07:39] | agneli: it will take some time ;) |
[07:40] | denk: how to download from the site? I don't see any link to the file |
[07:40] | agneli: aaa |
[07:40] | agneli: my bad |
[07:41] | agneli: http://www.fileconvoy.com/dfl.php?id=gf7ffddd6ea7bc0221000425790bad11544c97ddc9d |
[07:41] | agneli: whong link |
[07:41] | tde-bot: Page title: Free file sharing without registration and size limits |
[07:41] | denk: there is the only form to upload a file |
[07:42] | agneli: the link i pasted now should give a list of files to download |
[07:42] | agneli: first link was wrong, yes I have given the main link, not the one to download |
[07:42] | denk: ok, downloaded |
[07:43] | agneli: js needs to be allowed on that site, all the linked google crap can be blocked |
[07:43] | agneli: ach OK, thank you |
[07:46] | denk: the only error I see in the log is: "standard input:524: Error: syntax error" |
[07:48] | agneli: i am running it with -j1 now |
[07:51] | agneli: in strace I have taken it started complaining around libtdemid compliation |
[07:51] | agneli: I will take another strace now with this -j1 should be much clearer I guess |
[07:51] | agneli: thank you denk for your time :) |
[07:52] | agneli: and as said when I complied manually it worked |
[07:52] | agneli: so maybe issue is actually with my portage and not with trinity |
[08:10] | SlavekB: In the buidl.log I see one FAILED, which looks like it relates to generated "keramik.h" (pixmaps.keramik). |
[08:50] | agneli: SlavekB: seems you won |
[08:51] | agneli: with -j1 this is exactly the last commakd |
[08:56] | agneli: https://paste.debian.net/plain/1238941 |
[08:56] | agneli: http://www.fileconvoy.com/dfl.php?id=ge96c430ca581db661000425813efd206f9ad6053b0 |
[08:56] | tde-bot: Page title: Free file sharing without registration and size limits |
[08:56] | agneli: those are current logs |
[09:00] | denk: try to run it manually |
[09:00] | denk: but do not redirect stderr to /dev/null: 2>/dev/null |
[09:01] | agneli: you mean this whole giantic line? :) |
[09:01] | denk: cd /var/tmp/portage/trinity-base/tdelibs-9999/work/tdelibs-9999_build/tdestyles/keramik && ./genembed ... /vslider.png > pixmaps.keramik |
[09:01] | denk: copy it to a shell script :) |
[09:01] | agneli: heheh ;) |
[09:02] | agneli: so not this one |
[09:02] | agneli: FAILED: tdestyles/keramik/pixmaps.keramik tdestyles/keramik/keramikrc.h /var/tmp/portage/trinity-base/tdelibs-9999/work/tdelibs-9999_build/tdestyles/keramik/pixmaps.keramik /var/tmp/portage/trinity-base/tdelibs-9999/work/tdelibs-9999_build/tdestyles/keramik/keramikrc.h |
[09:02] | agneli: but next one that starts with cd |
[09:04] | denk: SlavekB spoted you an error, find the cause |
[09:25] | agneli: that is not very resorceful answer, ist? I ran them... no errors... :( i frankly have no idea where to go |
[09:26] | agneli: if I understand what I see correctly package prefers trinity libs |
[09:26] | agneli: it looks in the trinity subdirs for libraries |
[09:27] | agneli: my libraries there are broken for sure |
[09:28] | agneli: i have no idea how to check which one affects my compilation |
[09:32] | denk: genembed should generate some content, did it do it? |
[09:33] | denk: e.g. pixmaps.keramik must have some size greater than zero |
[09:35] | denk: if it works manually but the same command does not work in the stream, it means something wrong in the stream (or environment) |
[09:37] | agneli: seems it is ~55kiB |
[09:44] | agneli: and why it wants to load the libraries that come from that very package? |
[09:44] | agneli: how would it then build on a fresh system? |
[09:49] | denk: ldd genembed |
[09:49] | denk: or use the full path: ldd /var/tmp/portage/trinity-base/tdelibs-9999/work/tdelibs-9999_build/tdestyles/keramik//genembed |
[09:55] | agneli: i restarted the build |
[09:55] | agneli: I saw this somewhere in the strace output |
[09:55] | agneli: libtdefx.so.14.0.0 |
[09:55] | agneli: I removed it |
[09:56] | agneli: and yes build crashed citing lack of /usr/trinity/14/lib64libtdefx.so.14.0.0 |
[09:56] | agneli: equery b libtdefx.so.14.0.0 |
[09:56] | agneli: returns tdelibs package |
[09:57] | agneli: [16547.634611] traps: genembed[19853] trap invalid opcode ip:7f59f2ade254 sp:7ffc598fa090 error:0 in libtdefx.so.14.0.0[7f59f2acc000+1d000] |
[09:58] | denk: as asnwer I expected output of ldd |
[09:58] | denk: *answer |
[09:58] | agneli: so that is consistent with dmesg |
[09:58] | agneli: i cannot give you ldd and build is ongoing so the file was deleted |
[09:59] | agneli: :( |
[09:59] | denk shrugs |
[09:59] | agneli: anyway I supplemented that lib from a working system and restarted the build, cpu of a working system is older than this one |
[10:04] | agneli: it is already 11 minutes |
[10:04] | agneli: ooo |
[10:04] | agneli: installing |
[10:07] | agneli: done |
[10:07] | agneli: first of all thank you denk for your time and attention |
[10:07] | agneli: thank you SlavekB as well |
[10:07] | SlavekB: you're welcome |
[10:07] | agneli: root cause seem to be /usr/trinity/14/lib64/libtdefx.so.14.0.0 |
[10:08] | agneli: tdelibs package was using broken library to rebuild intsefl |
[10:08] | agneli: that should not be the cause I guess? |
[10:08] | agneli: this is a circular dependency |
[10:09] | agneli: and how it installs itself in first place, when no tde packages are present? |
[10:11] | SlavekB: tdelibs with certainty can be built without the presence of tdelibs in the system. |
[10:12] | agneli: that I know as I installed that on fresh gentoo many times :) |
[10:12] | agneli: i am just pointing out the situation, maybe useful? |
[10:12] | agneli: some taths in the ebuild? |
[10:12] | agneli: *paths |
[10:12] | agneli: maybe it just favours |
[10:13] | agneli: no not true |
[10:13] | agneli: when I deleted that library is stopped to work |
[10:13] | agneli: i had to provide a working one for the things to happen... |
[10:13] | agneli: so idk |
[10:13] | agneli: anyways thank you guys |
[10:14] | agneli: I was battling this for 2 days already |
[11:51] | Space_Man has quit (Remote host closed the connection) |
[18:19] | agneli: and I ahve another topic :) |
[18:20] | agneli: when I use sftp to log onto some machine |
[18:20] | agneli: in terminal it works |
[18:20] | agneli: but in Konqueror all I am getting is "protocol died unexpectedly" |
[18:21] | agneli: that is true for any sftp connection |
[18:21] | agneli: and it is so on all my machines |
[18:21] | agneli: it used to work well not so long ago |
[18:21] | agneli: when I run konqueror from the terminal there is no error message |
[18:22] | agneli: any hint where I need to look please? |
[18:24] | agneli: there was some ssh update lately that removed some ciphers, might that be related? |
[18:35] | denk: in this case dtrace could be useful, but you don't have it, try to use what you have |
[18:42] | agneli: I do not: https://wiki.gentoo.org/wiki/DTrace |
[18:43] | agneli: so no magic setting that would just make konqueror more talkative in the konsole? |
[18:43] | agneli: thank you denk, I will use strace then |
[18:43] | agneli: if you could just confirm sftp in konqueror works for you? |
[18:43] | SlavekB: you can try to check ~/.xsession-errors file. |
[18:45] | agneli: SlavekB: thank you |
[18:46] | denk: agneli: I can't confirm anything because my tde build is too old |
[18:46] | agneli: [2022/04/25 20:46:34.011] [tdeio_sftp] [6732] ERROR: KSshProcess::version(): pclose failed. |
[18:47] | agneli: denk - thank you anyway :) |
[18:47] | denk: not at all |
[18:47] | agneli: so sth wrong is wth me... |
[18:48] | agneli: https://paste.debian.net/plain/1239000 |
[18:48] | denk: trace the process with all its children |
[18:49] | denk: assert |
[18:49] | denk: it reads unexpected data |
[18:50] | agneli: and i have 3 segfaults in the dmesg... |
[18:50] | SlavekB: you can install debug symbols to get more accurate backtrace. |
[18:51] | agneli: SlavekB: this is done via a separate package? or full rebuild of TDE would be needed? |
[18:52] | denk: rebuilding everything is the gentoo way :) |
[18:53] | SlavekB: It is possible that you have to do a rebuild with enabled debug symbols - or disable the "strip" of debug symbolsy if it is carried out at the level of the packaging system. |
[18:55] | agneli: ugh... then I will give a try with strace first... then I will look into those debug symbols |
[18:56] | agneli: yes denk you are right and when you do not have the glorious threadripper as Mr. Torvalds does then well.... it does take time and money (electricity bill) |
[18:56] | agneli: :) |
[19:10] | denk: agneli: you broke an unbreakable function |
[19:11] | denk: https://paste.dilos.org/?edb3e06566e3d24d#XFGSSxfNQse/SXuHRvOf7MYPSzBsWB4UgHdD5w7245g= guys look at that |
[19:11] | denk: line 23 |
[19:11] | denk: KSshProcess::version(): pclose failed. |
[19:11] | denk: what he found in the log file |
[19:12] | agneli: in this file: 20:43 < SlavekB> you can try to check ~/.xsession-errors file. |
[19:13] | denk: at this point I'd like to see errno |
[19:14] | denk: in my manual page I see the only error for pclose() |
[19:14] | denk: The pclose() function will fail if: |
[19:14] | denk: ECHILD |
[19:14] | denk: The status of the child process could not be obtained, as described in |
[19:14] | denk: the DESCRIPTION. |
[19:15] | agneli: how to I provide you the errno denk, please? |
[19:15] | agneli: do I |
[19:15] | agneli: :) |
[19:15] | denk: no, thanks, I'm going to go to sleep |
[19:15] | denk: and good luck to debug it |
[19:16] | agneli: not today... |
[19:16] | agneli: :) |
[19:26] | agneli: for the moment I will just blindly rebuild trinity-base/tdebase-tdeioslaves |
[19:26] | agneli: not sure if that is relevant :) |
[19:26] | SlavekB: ...and tqt3 itself. |
[19:27] | agneli: ach then I will add the symbols then... and rebuild everythng.... |
[19:29] | agneli: ioslaves did not help...time to sleep |
[19:29] | agneli: thanks guys again |
#trinity-desktop < 2022/04/25 > |
© 2010-2024 Trinity Desktop Project