Qt-interest Archive, August 2007
qt4/ppc64 linux: anyone build/use this successfully?
Message 1 in thread
I'm having a hard time getting a linux (fedora) qt4/ppc64 build to function
properly. Anyone out there have any experience (success or failure) with
this?
Short version of the story is that most qt4 binaries segfault immediately
(e.g. designer, lrelease, lupdate, etc...):
See also:
bug tracking this: http://bugzilla.redhat.com/246324
build.log of package tested:
http://koji.fedoraproject.org/koji/getfile?taskID=97931&name=build.log
cvs of specfile/patches used:
http://cvs.fedoraproject.org/viewcvs/devel/qt4/
-- Rex
--
[ signature omitted ]
Message 2 in thread
> I'm having a hard time getting a linux (fedora) qt4/ppc64 build to function
> properly. Anyone out there have any experience (success or failure) with
> this?
Exact same problems in Gentoo land as well.
http://bugs.gentoo.org/show_bug.cgi?id=178779
As best as I can tell from some debugging output, it looks like some qatomic stuff
is broken with regards to MSB/LSB stuff, but I'm far from a ppc expert.
Caleb
--
[ signature omitted ]
Message 3 in thread
Caleb Tennis wrote:
>> I'm having a hard time getting a linux (fedora) qt4/ppc64 build to
>> function
>> properly. Anyone out there have any experience (success or failure) with
>> this?
>
> Exact same problems in Gentoo land as well.
> http://bugs.gentoo.org/show_bug.cgi?id=178779
Thanks, glad to know I'm not crazy. :)
has trolltech been made aware, any reports to them yet? (if not, I can do
it).
-- Rex
--
[ signature omitted ]
Message 4 in thread
> Thanks, glad to know I'm not crazy. :)
>
> has trolltech been made aware, any reports to them yet? (if not, I can do
> it).
Not reported that I know of. We weren't sure if it was a fluke or not, but it
sounds like it isn't.
I'm guessing it's probably only going to get fixed with a patch, since it's taken
this long for it to manifest itself into another ppc64. There must just not be many
ppc64 Qt users. :)
Caleb
--
[ signature omitted ]
Message 5 in thread
Caleb Tennis wrote:
>> Thanks, glad to know I'm not crazy. :)
>>
>> has trolltech been made aware, any reports to them yet? (if not, I can
>> do it).
>
> Not reported that I know of. We weren't sure if it was a fluke or not,
> but it sounds like it isn't.
I just sent in a bug report to Trolltech... I'll followup when/if I hear
anything.
-- Rex
--
[ signature omitted ]
Message 6 in thread
>
> I just sent in a bug report to Trolltech... I'll followup when/if I
> hear
> anything.
>
Just received word that the following patch seems to fix the
problem. Not sure if it will be an official fix or not, but thought
you might want to try it out:
--- qt-x11-opensource-src-4.3.1.orig/src/corelib/arch/
qatomic_powerpc.h 2007-08-07 15:08:39.000000000 +0200
+++ qt-x11-opensource-src-4.3.1/src/corelib/arch/qatomic_powerpc.h
2007-08-22 21:20:24.153357369 +0200
@@ -44,7 +44,7 @@
QT_BEGIN_HEADER
#if defined(Q_CC_GNU)
-#ifdef __64BIT__
+#ifdef __powerpc64__
# define LPARX "ldarx"
# define CMPP "cmpd"
# define STPCX "stdcx."
Message 7 in thread
Rex Dieter wrote:
> Caleb Tennis wrote:
>
>>> Thanks, glad to know I'm not crazy. :)
>>>
>>> has trolltech been made aware, any reports to them yet? (if not, I can
>>> do it).
>>
>> Not reported that I know of. We weren't sure if it was a fluke or not,
>> but it sounds like it isn't.
>
> I just sent in a bug report to Trolltech... I'll followup when/if I hear
> anything.
Heard back from Trolltech, this has been assigned issue number #176132 (at
least for when I reported it).
-- Rex
--
[ signature omitted ]
Message 8 in thread
On Thu, Aug 23, 2007 at 06:53:00AM -0500, Rex Dieter wrote:
> Rex Dieter wrote:
>
> > Caleb Tennis wrote:
> >
> >>> Thanks, glad to know I'm not crazy. :)
> >>>
> >>> has trolltech been made aware, any reports to them yet? (if not, I can
> >>> do it).
> >>
> >> Not reported that I know of. We weren't sure if it was a fluke or not,
> >> but it sounds like it isn't.
> >
> > I just sent in a bug report to Trolltech... I'll followup when/if I hear
> > anything.
>
> Heard back from Trolltech, this has been assigned issue number #176132 (at
> least for when I reported it).
I have built Qt 4.3.1 for 64-bit AMD. It seems to be OK, but I have not tested it much.
Dave Feustel
--
[ signature omitted ]
Message 9 in thread
dave wrote:
> I have built Qt 4.3.1 for 64-bit AMD. It seems to be OK, but I have not
> tested it much.
x86_64 has always worked fine (for me), this is a powerpc64-specific issue.
-- Rex
--
[ signature omitted ]
Message 10 in thread
Rex Dieter wrote:
>> I just sent in a bug report to Trolltech... I'll followup when/if I hear
>> anything.
>
> Heard back from Trolltech, this has been assigned issue number #176132 (at
> least for when I reported it).
I also just found an almost identical older reported issue:
http://trolltech.com/developer/task-tracker/index_html?method=entry&id=133870
-- Rex
--
[ signature omitted ]
