From kde-bugs-dist Sat Nov 08 19:35:44 2003 From: Torsten Kasch Date: Sat, 08 Nov 2003 19:35:44 +0000 To: kde-bugs-dist Subject: [Bug 65945] kdebase HEAD fails to compile on Solaris 8 due to X-MARC-Message: https://marc.info/?l=kde-bugs-dist&m=106832015502651 ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is. http://bugs.kde.org/show_bug.cgi?id=65945 tk@Genetik.Uni-Bielefeld.DE changed: What |Removed |Added ---------------------------------------------------------------------------- Attachment #2885 is|0 |1 obsolete| | ------- Additional Comments From tk@Genetik.Uni-Bielefeld.DE 2003-11-08 20:35 ------- Created an attachment (id=3085) --> (http://bugs.kde.org/attachment.cgi?id=3085&action=view) updated patch for kdebase Sorry again for the delay; this issue turned out to be trickier than I thought... Here are some notes about this updated patch: * Initially, I tried to use KDE_CHECK_FUNC_EXT, but it turned out that this would write the missing prototypes directly to config.h, including the #include statements for the required X11 headers. The latter broke the compilation for almost every *.cpp file that included qt headers -- I'm not sure if this is only the case for Solaris. Anywar, I had to give up this idea, obviously. * I think I've found a bug in kde-common/admin/acinclude.m4.in which turned up while trying to get AC_CHECK_DECL working: $ac_compile is set in two places without restoring the previous value. Especially setting "-o conftest" breaks AC_CHECK_DECL as can be seen in the following excerpt from "config.log": [...] configure:35923: checking whether DPMSCapable is declared configure:35945: gcc -g -O2 -I/usr/openwin/include conftest.c -o conftest -I/v ol/openssl/include configure:35948: $? = 0 configure:35951: test -s conftest.o [...] The attached patch fixes this issue. * I've not (yet) tried to actually compile kdebase with this patch (this will take some time on the machines I've currently available), but I made sure that I get the correct settings for the HAVE_*_PROTO macros correctly (i.e. they get defined on Solaris 9 and undefined on Solaris 8). cheers, Torsten