View previous topic :: View next topic |
Author |
Message |
sno35 Guru
Joined: 15 May 2004 Posts: 334 Location: Paris, France
|
Posted: Sat Jan 21, 2012 1:01 pm Post subject: cairo configure script that detects 'Intel' on my G4 Mini |
|
|
Hi
I have hard time with cairo : firefox and thunderbird no longer work here :/
I had a kinda workaround with cairo-9999 build from beginning of may but I had to try newer for qt evolution but got into a situation I can not get back to the cairo-9999 version I quickpkg-ed then.
So now, I try to fight with cairo's master HEAD and I could see next line when running configure :
checking for native atomic primitives... Intel
does it make sens, after all ?
Thanks in advance.
[EDIT] corrected title
Last edited by sno35 on Tue Feb 07, 2012 7:04 pm; edited 1 time in total |
|
Back to top |
|
|
mattst88 Developer
Joined: 28 Oct 2004 Posts: 422
|
Posted: Mon Jan 30, 2012 6:25 pm Post subject: |
|
|
I think it selects Intel atomic primitives when it cannot find something better.
What's your actual error message? _________________ My Wiki page |
|
Back to top |
|
|
sno35 Guru
Joined: 15 May 2004 Posts: 334 Location: Paris, France
|
Posted: Thu Feb 02, 2012 8:21 pm Post subject: |
|
|
In fact I get no _compile_ error. Don't know how it manages to go on.
I quite got back to the beginning of may version (using git) of cairo that used to function OK for mozilla things and others but this time mozilla things do not work and gvim almost work (I have quirks on last entries of some menus).
I'd like to help get the 1.11.x version of cairo OK on my machine but I guess I am not smart enough. And I don't have much time, either.
Thank you for trying to help. |
|
Back to top |
|
|
mattst88 Developer
Joined: 28 Oct 2004 Posts: 422
|
Posted: Thu Feb 02, 2012 8:28 pm Post subject: |
|
|
I know that Firefox and Thunderbird aren't working, but you haven't told me what error message you're seeing. _________________ My Wiki page |
|
Back to top |
|
|
sno35 Guru
Joined: 15 May 2004 Posts: 334 Location: Paris, France
|
Posted: Sun Feb 05, 2012 2:27 pm Post subject: |
|
|
checking whether float word ordering is bigendian... yes
-> checking for native atomic primitives... Intel
checking whether atomic ops require a memory barrier... yes
I was hopelessly searching for some warning that would be a start to find something even remote. It seems it is a false signal.
But now I clarified something : my main problem was with having gcc given -g -O and cairo build having Valgrind checkings enabled at the same time.
Cf https://bugs.gentoo.org/show_bug.cgi?id=336687#c16
The way I found a decisive crash was by running the cairo tests under gdb.
As I know quite nothing about ppc assembly, I don't think I can tell the error is on the cairo side / valgrind side / gcc side. |
|
Back to top |
|
|
|