[kde-linux] KDE-4.14.0 doesn't work

[James Tyrer]

On 09/13/2014 03:04 AM, Ralf Mardorf wrote:
> On Sat, 2014-09-13 at 05:54 +0000, Duncan wrote:
>> James Tyrer posted on Fri, 12 Sep 2014 22:21:04 -0700 as excerpted:
>>
>>> I usually go ahead and update GCC because if there is a building problem
>>> with anything on the LFS/BLFS sites, there is a fix for it.
>>>
>>> Now I see that GCC 4.9.1 has been released so there might have been some
>>> problems with the ".0" release.  So, since LFS is going to upgrade to
>>> that, I will try building Qt-4.8.6 with that and see if it works.  That
>>> will take a while even with a 4400+ processor.
>>
>> In general that has been my experience with gcc on gentoo as well;
>> there's usually fixes available, most often already found on gentoo's
>> bugzilla so I don't have to go far. =:^) [snip]
>
> I didn't experience issues and didn't read that others experienced
> issues, when using 4.9.1. It seems that Arch Linux does use one 4.9
> patch and two 4.8 patches to build 4.9 [1].
>
> For release model distros such as *buntu, the only problem should be,
> that a new release from upstream doesn't fit to the outdated sources of
> the other packages provided by official repositories, but I'm not aware
> that 4.9 has any serious bugs. If I would have experienced an issue, I
> would have downgraded gcc, but I didn't.
>
> Please, don't misinform users about alleged issues, without evidences.
> Spreading fear, uncertainty and doubt about a stable release from
> upstream for a Linux core component is something very bad.
>
There isn't something very bad.  Just an obscure bug.  A specific 
compiler (an ".0" release) and a specific set of optimization options 
(the ones for AMD 64 bit processors) appear to have resulted in a 
strange problem.  This is clearly a bug that is not going to occur for a 
large number of people and would probably be fixed by not using the AMD 
optimization and I hope that it will be fixed by the next release of GCC.

However, since the same thing but has occurred before, I presume that it 
is real.

It might even be hard to reproduce with the millions of perturbations 
possible with from scratch systems even with people that, more or less, 
follow a DIY distribution.

People should presume that this problem will not occur if they install 
binaries from a distro.

-- 
James Tyrer

Linux (mostly) From Scratch

-- 
James Tyrer

Linux (mostly) From Scratch