new kdelibs3 and issues stemming from it

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



I noticed something interesting when rpmbuild --rebuild kdelibs3 this
afternoon

checking whether g++ accepts -g... yes
checking dependency style of g++... none
checking whether gcc is blacklisted... yes
configure: error:

This particular compiler version is blacklisted because it
is known to miscompile KDE. Please use a newer version, or
if that is not yet available, choose an older version.

Please do not report a bug or bother us reporting this
configure error. We know about it, and we introduced
it by intention to avoid untraceable bugs or crashes in KDE.


error: Bad exit status from /var/tmp/rpm-tmp.51024 (%build)

So last week I was in some conversions about procedures for rpmbuild-ing
srpms and how the rpms were built was the very topic of discussion.   Also
some quotes from John Summerfield about "non standard modifications" of
packages to be distributed is an issue to all Linux Users in a discussion
about building rpms for my own personal use :-) 
Anyway I have some questions now about this myself
1.  What compiler was used to build these packages 
http://koji.fedoraproject.org/koji/buildinfo?buildID=51515
if the latest compiler from Fedora has been blacklisted?
2.  If the check in acinclude.m4 was removed how does software run?  Well I
can do that and report back....
3.  Why does rpm or Fedora's implementation or RPM not allow for multiple
installations of gcc
gjohnson@loser SOURCES]$ rpm -qa gcc
gcc-4.3.0-8.x86_64
[gjohnson@loser SOURCES]$ ls -al /usr/bin/gcc
-rwxr-xr-x 2 root root 241704 2008-04-28 17:06 /usr/bin/gcc
[gjohnson@loser SOURCES]$

Excerpt from from the acinclude.m4
  AC_MSG_CHECKING([whether $CC is blacklisted])

  dnl In theory we have tu run this test against $CC and $CXX
  dnl in C and in C++ mode, because its perfectly legal for
  dnl the user to mix compiler versions, since C has a defined
  dnl ABI.
  dnl
  dnl For now, we assume the user is not on crack.

  AC_TRY_COMPILE([
#ifdef __GNUC__
#if __GNUC__ == 4 && __GNUC_MINOR__ == 0 && __GNUC_PATCHLEVEL__ == 0
choke me
#endif
#endif
], ,
  kde_bad_compiler=no,
  kde_bad_compiler=yes
)

  AC_MSG_RESULT($kde_bad_compiler)

4.  Wow , someone thinks my question number 3 is like smokin crack...  Well
youd need to smoke something, because this doesn't work with the latest gcc
from Fedora...
Also if I'm reading this check correctly, it seems to be saying that if your
compiler is gcc-4.0.0 then configure should fail.  But I'm using
gcc-4.3.0...

Anway I changed the __GNUC_MINOR__ == 3 and configure did not throw the
error.  So I'm confused on that one 

Anyway I tend to build KDE from SRPM since KDE is a resource hog and it does
seem to run better (atleast from my test) when optimized for the hardware
the software is running on.  So I like to get the latest SRPMS and rebuild
them.


Anyway thanks again,
Tony

-- 
View this message in context: http://www.nabble.com/new-kdelibs3-and-issues-stemming-from-it-tp17654258p17654258.html
Sent from the Fedora Test List mailing list archive at Nabble.com.

-- 
fedora-test-list mailing list
fedora-test-list@xxxxxxxxxx
To unsubscribe: 
https://www.redhat.com/mailman/listinfo/fedora-test-list

[Index of Archives]     [Fedora Desktop]     [Fedora SELinux]     [Photo Sharing]     [Yosemite Forum]     [KDE Users]

  Powered by Linux