4alpha2 DVD 64 with default E17 install (Date.txt Thu Sep 5 16:22:20 CEST 2013) First seen when previously installed with lxde, windowmaker, xfce icewm & e17, reproduced when installed with just E17. On logging in for the first time, whilst going through the configuration steps there is an error message.. ============================================ Enlightenment Error. This is very bad. Enlightenment SEGV'd. This is not meant to happen and is likely a sign of a bug in enlightenment or the libraries it relies on. We were not able to generate a backtrace, check if your 'sysactions.conf' has a 'gdb' action line. Please compile latest svn E17 and EFL with -g and -ggdb3 in your CFLAGS. (F1) Recover (F12) Logout ============================================ I'll attach some logs. Reproducible: Steps to Reproduce:
CC: (none) => ennael1Whiteboard: (none) => 4alpha2
CC: (none) => doktor5000
Created attachment 4331 [details] xsession-errors /home/user/.xsession-errors The .e-crashdump.txt file does not exist
Created attachment 4332 [details] journal.txt
Created attachment 4333 [details] /var/log/lxdm.log
Created attachment 4334 [details] Xorg.0.log
I've reproduced a logging issue on cauldron, I suppose that it's the same. I've commited a patch from the enlighment repository : http://git.enlightenment.org/core/enlightenment.git/commit/?h=enlightenment-0.17&id=b17a9b9cc9438b6dfac4402ac4107f08e23a4373 It should fix the issue.
Still valid 4beta1 trem, sorry. It wasn't immediate, I clicked the 'root' icon on the desktop and opened the menu before it showed.
Whiteboard: 4alpha2 => 4beta1 4alpha2
Blocks: (none) => 11608
I've tried to reproduce it when my cauldron, but I don't reproduce this issue. I've remove the .e file to simulate a first connection, and then I've successfully connect on e17. If it's not too late, I propose to move to 1.7.9 (efl) and 0.17.5 (e17). What do you think about it ?
You know best as maintainer trem, you'd need to ask for a freeze push on dev ml. If you like access to the actual pre-release isos to debug (or help testing!) can you add yourself here please: https://wiki.mageia.org/en/QA_ISO_testers
CC: (none) => davidwhodgins
I've tried to reproduce this issue with mageia 4 alpha 3 on virtualbox. But I don't reproduce it, e17 seems to work fine.
I've requested the update of e17 to efl 1.7.9 and e17 0.17.5. It's only bugfix release. It should be better with those versions.
Pushed - please let us know about the general status of e17
looks ok
Status: NEW => RESOLVEDResolution: (none) => FIXED
Created attachment 4872 [details] Crash dump after segfaulted while filling menus This is probably a different bug, but in reference to Comment 11 from Anne about the current status, seemed like the right place to post. After installing all GEs from the second round of both 32 and 64 Mageia4 install iso images and starting e17, trying to run a application from the menu resulted in a segfault and the same dialog as Claire's. The attached file is the crash dump from that segfault for the 32 bit version. The crash dump from the 64 bit version is identical. After hitting F1 to recover, E17 restarted without complaint and the menus were populated. On second boot, everything was fine. The logs show almost the same errors as Claire's and seem unrelated to this. The only other problem encountered was that MCC failed to launch from the application menu, giving an error dialog saying the application failed to start.
CC: (none) => warrendiogenese
reopening for William
Status: RESOLVED => REOPENEDResolution: FIXED => (none)
I've reproduced the "crash" on first login. It could be reproduced if you remove some directory on the home : rm -rf .e .cache .dbus .config If your log after having removing those directory, the crash happen when you try to navigate on menu. I suppose that it occurs when e17 populate menu. I have no idea on how to fix it. Any help would be very appreciated.
Please email dev list trem, other DE maintainers are there and may have some ideas.
As proposed by claire, I've send a mail on the ML.
I prepare E18 on cauldron, and I don't reproduce this issue with E18. May be, we could push E18 on backport.
Created attachment 7341 [details] crash dump with E18 and E17 Themes
CC: (none) => nic
Comment on attachment 7341 [details] crash dump with E18 and E17 Themes I installed E18 with E17 Themes on MGA5 and can reproduce the bug.
(In reply to Nic Baxter from comment #20) > I installed E18 with E17 Themes on MGA5 and can reproduce the bug. Are there any special steps needed, like did you try this under a newly created user, where some folders are missing like mentioned in comment 15 ? I cannot reproduce, first I deleted existing e configs (rm -rf ~/.e/ ~/.cache/efreet/ ~/.cache/evas_gl_common_caches/) and then logged in, went through the first start routine. But what is apparent is that no themes apart from the default are selectable, probably because the themes are binary and have been made for e17 and not updated since. I could try to ask Agust again, who is the creator of the themes in our e17-themes-mageia package and for the themes in e17-themes-extra we will either have to update them or drop them.
(In reply to Nic Baxter from comment #19) > Created attachment 7341 [details] > crash dump with E18 and E17 Themes You tested in virtualbox right? Could be an issue with the openGL driver which is known to cause issues. Although William's crashdump in comment 13 is not with virtualbox ... :/
Yes I did use virtualbox. I have now installed it on my AMD box (Phenom II X6 1055T & Radeon HD 6770) and it just works. So possibly openGL is the issue.
I'd say we close this one then. It was reopened one year ago, and can't be reproduced anymore on physical installations. I'll try to open another bug for the non-selectable themes.
Status: REOPENED => RESOLVEDResolution: (none) => WORKSFORME