Bug 14418 - Qt 5.3 or programmed obsolescence?
Summary: Qt 5.3 or programmed obsolescence?
Status: RESOLVED DUPLICATE of bug 13991
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: Cauldron
Hardware: i586 Linux
Priority: Normal critical
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks: 13991 14229
  Show dependency tree / graph
 
Reported: 2014-10-30 10:53 CET by Georges Eckenschwiller
Modified: 2015-01-05 15:26 CET (History)
8 users (show)

See Also:
Source RPM:
CVE:
Status comment:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Description Georges Eckenschwiller 2014-10-30 10:53:35 CET 
Description of problem:

Qt 5.3 or programmed obsolescence?
===================================
By default, Qt 5.3 is compiled by activating the use of the instructions SSE2 of the processor.
References:
http://lists.qt-project.org/pipermail/development/2013-November/014085.html
https://lists.debian.org/debian-qt-kde/2014/07/msg00169.html

The consequences:
------------------
A lot of programs do not comply any more on computers equipped with a little bit former processors, devoid of instructions SSE2.
During my some tests, I have already listed sddm, AdminPanel, Baloo de KDE, Qt Creator. There is certainly the others else.
Bug #13991 and #14229
Graver still: LXQt, an desktop foreseen for ligth computers, cannot work any more exactly on the computers for which it is foreseen.

However, there is a possibility:
Compile by clarifying not to activate SSE2. It seems that Mageia did not choose this solution.


Reproducible: 

Steps to Reproduce:
Rémi Verschelde 2014-10-30 11:06:42 CET

CC: (none) => lmenut, mageia, remi

Angelo Naselli 2014-10-30 12:24:44 CET

CC: (none) => anaselli
Blocks: (none) => 14229

Comment 1 José Jorge 2014-10-30 14:23:06 CET 
What is the build option for remove SSE2?

CC: (none) => lists.jjorge

Comment 2 Georges Eckenschwiller 2014-10-30 16:21:47 CET 
I am hopeless more or less in English (Reverso translates for me).
Maybe will be the answer here:
https://bugreports.qt-project.org/browse/QTBUG-35430
Georges Eckenschwiller 2014-10-30 16:23:50 CET

Blocks: (none) => 13991

Comment 3 Thomas Backlund 2014-10-30 18:28:34 CET 
Fedora has fixed their package to build a non-sse2 + a sse2 version.

I think we should do the same

CC: (none) => tmb

Felix Miata 2014-10-31 00:19:40 CET

CC: (none) => mrmazda

Comment 4 Angelo Naselli 2014-10-31 12:25:20 CET 
On Comment #3 (+1), just out of curiosity is that change made only for i586 arch?
Comment 5 Thomas Backlund 2014-10-31 12:49:51 CET 
(In reply to Angelo Naselli from comment #4)
> On Comment #3 (+1), just out of curiosity is that change made only for i586
> arch?

Yes,

all x86_64 hw supports atleast SSE2
Comment 6 Florian Hubold 2014-11-11 17:23:26 CET 
Why do we have another bugreport for this?
Relevant comments are already in https://bugs.mageia.org/show_bug.cgi?id=13991

This is the third bugreport for the same issue, I've already closed Bug 14491 due to this.

*** This bug has been marked as a duplicate of bug 13991 ***

Status: NEW => RESOLVED
CC: (none) => doktor5000
Resolution: (none) => DUPLICATE

Comment 7 Georges Eckenschwiller 2015-01-05 15:26:03 CET 
I was able to redo the tests on my old machine.
I want just to confirm that the recompilation of qtbase5 by doktor5000 solved several problems.
Note You need to log in before you can comment on or make changes to this bug.